An Odyssey in Learning and Perception (Learning, Development, and Conceptual Change)

An Odyssey in Learning and Perception Eleanor J. Gibson In the field of psychology, beginning in the 1950s, Eleanor J. ...

Author: Eleanor J. Gibson

13 downloads 783 Views 30MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

An Odyssey in Learning and Perception Eleanor J. Gibson

In the field of psychology, beginning in the 1950s, Eleanor J. Gibson nearly single-handedly developed the field of perceptual learning with a series of brilliant studies that culminated in the seminal work, Perceptual Learning and Development. An Odyssey in Learning and Perception brings together Gibson's scientific papers, including difficult-to- find or previously unpublished work, along with classic studies in perception and action. Gibson introduces each paper to show why the research was undertaken and concludes each section with comments linking the findings to later developments. A personal essay touches on the questions and concerns that guided her research.

TABLE OF CONTENTS SERIES FOREWORD FOREWORD BY ELIZABETH S. SPELKE INTRODUCTION I EXPERIMENTAL PSYCHOLOGY IN THE THIRTIES (1932-1942) 1 BILATERAL TRANSFER OF THE CONDITIONED RESPONSE IN THE HUMAN SUBJECT, J. J. GIBSON AND G. RAFFEL. J. EXP. PSYCHOL., 1932, 15, 416-421. 2 RETENTION AND THE INTERPOLATED TASK, WITH JAMES J. GIBSON. AMERI. J. PSYCHOL., 1934, 46, 603-610. 3 SENSORY GENERALIZATION WITH VOLUNTARY REACTIONS. J. EXP. PSYCHOL., 1939, 24, 237-253. 4 A SYSTEMATIC APPLICATION OF THE CONCEPTS OF GENERALIZATION AND DIFFERENTIATION TO VERBAL LEARNING. PSYCHOL. REVIEW, 1940, 47, 196-299. 5 RETROACTIVE INHIBITION AS A FUNCTION OF DEGREE OF GENERALIZATION BETWEEN TASKS. J. EXP. PSYCHOL., 1941, 28, 93-115. RETROSPECT AND PROSPECT: ARE THEORIES RECYCLED? II COMPARATIVE RESEARCH ON LEARNING AND DEVELOPMENT (1952-1970) 6 THE ROLE OF SHOCK IN REINFORCEMENT. J. COMP. PHYSIOL. PSYCHOL., 1952, 45, 18-30. 7 THE EFFECT OF PROLONGED EXPOSURE TO VISUALLY PRESENTED PATTERNS ON LEARNING TO DISCRIMINATE THEM, WITH R. D. WALK. J. COMP. PHYSIOL. PSYCHOL., 1956, 49, 239-242. 8 THE EFFECTIVENESS OF PROLONGED EXPOSURE TO CUTOUTS VS. PAINTED PATTERNS FOR FACILITATION OF DISCRIMINATION. J. COMP. PHYSIOL. PSYCHOL., 1959, 52, 519-521. 9 BEHAVIOR OF LIGHT- AND DARK-REARED RATS ON A VISUAL CLIFF, WITH R. D. WALK AND T. J. TIGHE. SCIENCE, 1957, 126, 80-81. 10 DEVELOPMENT OF PERCEPTION: DISCRIMINATION OF DEPTH COMPARED WITH DISCRIMINATION OF GRAPHIC SYMBOLS. REPRINTED FROM J. C. WRIGHT AND J. KAGAN (EDS.), BASIC COGNITIVE PROCESSES IN CHILDREN, MONOGR. SOC. RES. CHILD DEVELOPMENT, 1963, 28, NO. 2 (SERIAL NO. 86), 5-24. 11 THE DEVELOPMENT OF PERCEPTION AS AN ADAPTIVE PROCESS, ELEANOR J. GIBSON. AMERICAN SCIENTIST, 1970, 58, 98-107. RETROSPECT AND PROSPECT: COMPARATIVE PSYCHOLOGY AND ANIMAL COGNITION III PERCEPTION: PSYCHOPHYSICS AND TRANSFORMATIONS (1954-1959) 12 THE EFFECT OF TRAINING ON ABSOLUTE ESTIMATION OF DISTANCE OVER THE GROUND, WITH R. BERGMAN. J. EXP. PSYCHOL., 1954, 473-482. 13 THE EFFECT OF PRIOR TRAINING WITH A SCALE OF DISTANCE ON ABSOLUTE AND RELATIVE JUDGMENTS OF DISTANCE OVER GROUND, WITH R. BERGMAN AND J. PURDY. J. EXP. PSYCHOL., 1955, 50, 97-105. 14 DISTANCE JUDGEMENT BY THE METHOD OF FRACTIONATION, WITH J. PURDY. J. EXPER. PSYCHOL., 1955, 50, 374380. 15 CONTINUOUS PERSPECTIVE TRANSFORMATIONS AND THE PERCEPTION OF RIGID MOTION, WITH J. J. GIBSON. J. EXPER. PSYCHOL., 54, 129-138. 16 MOTION PARALLAX AS A DETERMINANT OF PERCEIVED DEPTH, WITH J. J. GIBSON, O. W. SMITH, AND H. FLOCK. J. EXPER. PSYCHOL., 1959, 58, 40-51. RETROSPECT AND PROSPECT: PSYCHOPYSICS TO COMPUTATION IV PERCEPTUAL LEARNING (1955-1969) 17 PERCEPTUAL LEARNING: DIFFERENTIATION OR ENRICHMENT?, WITH J. J. GIBSON. PSYCHOL. REV., 1955, 62, 3241. REPLY BY L. POSTMAN: ASSOCIATION THEORY AND PERCEPTUAL LEARNING. PSYCHOL. REV., 1955, 438-446. WHAT IS LEARNED IN PERCEPTUAL LEARNING? A REPLY TO PROFESSOR POSTMAN. PSYCHOL. REV., 1955, 62, 447-450. 18 PERCEPTUAL LEARNING. ANNUAL REVIEW OF PSYCHOLOGY, 1963, 14, 29-56. 19 PERCEPTUAL DEVELOPMENT AND THE REDUCTION OF UNCERTAINTY. IN PROCEEDINGS OF THE 18TH INTERNATIONAL CONGRESS OF PSYCHOLOGY, 7-17, MOSCOW, 1966 20 TRENDS IN PERCEPTUAL DEVELOPMENT. EXCERPTS FROM CHAPTER 20 OF PRINCIPLES OF PERCEPTUAL LEARNING AND DEVELOPMENT (PP. 450-472). ENGLEWOOD CLIFFS, NJ: PRENTICE-HALL, INC., 1969.

RETROSPECT AND PROSPECT: THE COMING OF AGE OF PERCEPTUAL DEVELOPMENT V YEARS OF SIGNIFICANCE: RESEARCH ON READING (1965-1977) 21 LEARNING TO READ. SCIENCE, 1965, 148, 1066-1072 22 A DEVELOPMENTAL STUDY OF VISUAL SEARCH BEHAVIOR, WITH A. YONAS. PERCEPTION AND PSYCHOPYSICS, 1966, 1, 169-171. 23 CONFUSION MATRICES FOR GRAPHIC PATTERNS OBTAINED WITH A LATENCY MEASURE, WITH F. SCHAPIRO AND A. YONAS. CORNELL UNIVERSITY, 1968. 24 THE ONTOGENY OF READING. AMERICAN PSYCHOLOGIST, 1970, 25, 136-143. 25 PERCEPTUAL LEARNING AND THE THEORY OF WORD PERCEPTION. COGNITIVE PSYCHOLOGY, 1971, 2, 351-368. 26 HOW PERCEPTION REALLY DEVELOPS: A VIEW FROM OUTSIDE THE NETWORK. IN D. LABERGE AND S. J. SAMUELS (EDS.), BASIC PROCESSES IN READING: PERCEPTION AND COMPREHENSION. HILLSDALE, N. J.: ERLBAUM, 1977, 155-173. RETROSPECT AND PROSPECT: PERCEPTION, COGNITION, OR BOTH? VI PERCEPTUAL DEVELOPMENT FROM THE ECOLOGICAL APPROACH (1972 TO THE PRESENT) 27 THE SENSES AS INFORMATION-SEEKING SYSTEMS, WITH J. J. GIBSON. (LONDON) TIMES LITERARY SUPPLEMENT, 1972, JUNE 23, 711-712. SEEING AS THINKING: AN ACTIVE THEORY OF PERCEPTION. R. GREGORY. (LONDON) TIMES LITERARY SUPPLEMENT, 1972, JUNE 23, 707-708. 28 PERCEPTION AS A FOUNDATION FOR KNOWLEDGE: THOUGHTS INSPIRED BY PAPERS OF FRAILBERG AND BELLUGI. DISCUSSION PREPARED FOR THE LENNEBERG SYMPOSIUM, CORNELL UNIVERSITY, ITHACA, N.Y., MAY 1976. 29 PERCEPTION OF INVARIANTS BY FIVE-MONTH-OLD INFANTS: DIFFERENTIATION OF TWO TYPES OF MOTION, WITH C. J. OWSLEY AND J. JOHNSTON. 30 DEVELOPMENT OF KNOWLEDGE OF VISUAL-TACTUAL AFFORDANCES OF SUBSTANCE, WITH A. WALKER. CHILD DEVELOPMENT, 1984, 55, 453-460. 31 EXCERPTS FROM THE CONCEPT OF AFFORDANCES IN DEVELOPMENT: THE RENASCENCE OF FUNCTIONALISM. IN W. A. COLLINS (ED.), THE CONCEPT OF DEVELOPMENT. THE MINNESOTA SYMPOSIUM ON CHILD PSYCHOLOGY, VOL. 15. HILLSDALE, N.J.: L. ERLBAUM ASSOC., 1982, 55-81. 32 DETECTION OF THE TRAVERSABILITY OF SURFACES BY CRAWLING AND WALKING INFANTS, WITH G. RICCIO, M. A. SCHMUCKLER, T. A. STOFFREGEN, D. ROSENBERG, AND J. TAORMINA. JOURNAL OF EXPERIMENTAL PSYCHOLOGY: HUMAN PERCEPTION AND PERFORMANCE, 1987, 13, 533-544. 33 EXPLORATORY BEHAVIOR IN THE DEVELOPMENT OF PERCEIVING, ACTING, AND THE ACQUIRING OF KNOWLEDGE. EXCERPTS FROM ANNUAL REVIEW OF PSYCHOLOGY, 1988, 39, 1-41. EPILOGUE: PROSPECTS FOR A NEW APPROACH TO PERCEPTUAL LEARNING REFERENCES AUTHOR INDEX SUBJECT INDEX

Series Foreword

Lila Gleitman SusanCarey ElissaNewport ElizabethSpelke

Forewordby ElizabethS. Spelke

:;:1. 1 This volume portrays ascientist and her science .We learn about Eleanor J. Gibson and her field ,experimental psychology ,both through the scientific papers that she has collected here and through the personal essay that weaves these papers together by highlighting the questions and concerns that guided and animated her work . There are many things togain from this book .First ,the reader will learn agood deal about the nature and development ofperception and action from Gibson 's classic papers . Some of these papers have influenced experi mental psychology so profoundly that they may seem familiar ,even com -monplace , tofirst -time readers .What psychologist inthe 1990s could doubt ,for example ,that learning can bring changes inwhat we perceive ,not just changes inhow we respond towhat we perceive ?Yet psychology owes this insight ,and the defeat of arguments to the contrary toEleanor and James Gibson 'svigorous studies of perceptual learning . ,largely Other papers describe findings that goagainst deeply ingrained beliefs about the nature and development ofpsychological capacities .For exam ple ,on consider Gibson 'sdemonstration ,as an undergraduate that learning to act an object with one hand immediately transfers to,the other hand . This experiment shows quite clearly that what one learns ,when one learns anew and appropriate action ,tone isnot aresponse contingency ("affordance such as "flex this finger when you hear this ")but what Gibson calls an ": aproperty ofthe environment with consequences forthe perceiver /actor (should such as "after the tone , this plate gives off a shock ") . Gibson 's finding give pause toof those engaged inthe long -standing effort tounder stand action interms elementary responses and their concatenation . Another finding that experimental psychology cannot quite digest is Gibson ,Walk ,and Tighe 'sdemonstration that young animals of awide range of species ,capable including humans ,avoid visually specified drop -offs as soon as they are oflocomotion .Elegant experiments with animals

.. Xll

E. S. Spelke

that locomote atbirth ,and also with animals reared inthe dark ,reveal that the capacity toperceive athree -dimensional surface layout can develop in the absence ofany visual experience ortrial and error learning .Perceptu systems appear to,be built to bring animals information about the layout and itsaffordances contrary to empiricist theories ofspace perception . The reader ofthis book also willlearn something about the flourishin ofexperimental psychology intwentieth -century America .Gibson 'spapers portray adiscipline coming into itsown , focusing onfundamenta ques tions about perception and action ,gathering insights into the nature and development ofthese functions , and raising new questions forthe next century ofpsychology topursue . Consider ,forexample ,the progress that has been made inthe study of perception action ininfancy .field With her studies ofthe the visual cliff Gibson was one ofand the founders ofthat of research in late 19505 .,Lackin her own laboratory ,she was not able tocontinue her studies ofinfants at that time .Research oninfant perception continued inother laboratorie in the 1960s .Nevertheless ,the fundamental questions about perception and perceptual development - how perceivers apprehend alayout ofsurface , objects ,and events ;how the various perceptual modes work together to produce an experience ofaunitary world ;how perceptual systems serve to guide effective and adaptive action were raised consistently only by Gibson 'sunofficial students , T.G.R.Bower and Albert Yonas .Then ,in 1975 ,Eleanor Gibson was forthe first time inaposition toestablish her own infant study laboratory . The papers collected here show what has happened tothe field ofinfant perception since that time :Gibson 'sstudie ofsurface perception ,event perception ,perception ofintermodal relation , and perceptual guidance ofaction have helped bring the study ofinfant perception itsown contribution ,faculty during the four years that was both ainto member of.Her Cornell 'sactive and the director ofshe her infant laboratory ,isincalculable .Those years ofwork provide asgood an example asany ofthe progress that can bemade inexperiment psy chology when agifted scientist isgiven the freedom and the resource to pursue her science . During much ofher career , Gibson didnotenjoy this freedom . Her first goal was tobecome acomparative psychologist .With her training in perception , and with her theoretical disposition to view perception and action asinterconnected functions ,each adapted tothe environme ofa particular species , Gibson could have undertaken alifelong compara and ecological study ofperception and perceptually guided ackion .She was prevented from achieving this goal each time she tried : as a new gradua student , rejected from her psychology department 'sonly laborator of comparative psychology because of her sex ; as a technical assistan in a different laboratory ,charged with incompetence after making adiscove

Foreword

xiii

thatthedirector ofthelaboratory later published ashisOwn; asamature scientist, butwithout anacademic appointment, herattempts toconduct research atheruniversitys laboratory thwarted bythesudden removal of

theanimals she had been rearing and observing inanextended longitudinal

project.

Eleanor Gibsons career hasnotgreatly suffered from these reverses,

thanks toherindomitable spirit, butscience has. The most basic questions about surface, object, andevent perception inmost nonhuman species remain unanswered, even unasked. Thelack ofavigorous, comparative study ofperception andaction hashadserious consequences notonly for psychology but fortwoofitssurrounding disciplines: neurobiology and evolutionary biology. Eleanor Gibsons career hascoincided with a tremendous growth of research inneuroscience, most ofitconducted with laboratory animals and much ofitexploring theneural structures andprocesses subserving per-

ception andaction. Hercareer hasalsocoincided witha blossoming of zoological studies ofbehavior initsadaptive, evolutionary context. Both

these fields areinneed ofacomparative psychology ofperception and only occur ifinvestigations aregrounded onasolid understanding ofthe functions thatneural structures serve. Moreover, theright ecological and evolutionary analysis ofspecies-typical behavior requires anappreciation action. Reason andexperience suggest thatprogress inneuroscience will

notjustofananimals physical environment butofitsperceived environThecomparative, experimental psychology ofperception hasbeen so little developed, however, thatsome neuroscientists andevolutionary biologists appear tohaveconvinced themselves thatthefield isnotneeded: thatstudies ofneurobiology atoneend, andofevolution attheother, will suffice to unravel thesecrets ofmind, brain, andbehavior. Since this conclusion threatens tohamper progress inthebrain andbehavioral sciment as well.

encesforyearsto come, onecanonlyhopethatfuture scientists and

teachers with Eleanor Gibsons gifts, education, andgoals will come along and that they will beallowed tobuild theresearch discipline sheenvisaged. A thirdreason to readthisbookis to learnaboutGibson herself. Reading thispersonal account ofherlifeandwork isanexhilarating

experience. Despite theobstacles inherpath, hersisastoryofsuccess. Her

professional lifewasdevoted towork onproblems thatengaged herfrom thebeginning. Her work afforded insight into many ofthese problems and

raised newproblems tochallenge her.Welearn fromGibsons lifethat professional status andmaterial resources donotmake ascientist, however necessary theymaybetoa scientists work. Themostimportant in-

gredients ofagood scientist arethepersonal qualities Gibson exemplifies.

xiv

E. S. Spelke

Whatarethesequalities? Some aregiftsatwhich mostofuscanonly marvel: theprodigious intelligence thatenabled herto complete allher

graduate training short ofthethesis inoneyear; theextraordinary eyefor phenomena thatallowed hertorecognize ina common observation (for

example, thatnewborn goats dontfalloffstools) thegerm ofexperiments Othersare qualitiesthat we mayemulate:the of major consequence. courage toespouse anddefend original views inthefaceofopposition or indifference; thejudicious combination offlexibility, ingenuity, anddetermination thatallowedherto faceobstacles andprevail; thegenerosity

andpatience shedisplayed withstudents, giving ofhertimeandwisdom whenever it wasneededto anyonewithanycuriosity andwillingness to learn.

Forme,Gibsons moststriking quality ishernearly magical blendof firmness andopenness. Thefirmness hasbeenfeltbyeveryone whohas

ever disagreed withher,anditisevident throughout thisbook: Shehashad thestrength, conviction, andintegrity todevelop andmaintain herown enduring conception ofmind andaction. Atthesame time, Gibson is always wholly open tochange. Heropenness isexemplified atmany points

in thisbook.Note,forexample, the readinessalmost delightwith which shenowcriticizes someofherearlier theories ofreading andmany ofherearlyclaims aboutinfant perception. Thisopenness ischaracteristic

anddeep. Gibson hasalways triedto findthings out,nottovindicate previously heldopinions. Shehaschanged herideasagain andagain in response to newexperimental findings. A major partofhergenius, I believe, liesinherability to remain opento further development and

change while maintaining a conception ofhersubject thatisstrong and systematic andrichenough toserve asthesource andframework for all her work.

Whatisthisconception? Thisbookisanextended answer tothatquestion.Gibsons conception canbefeltintheearliest papers inthiscollection; anddeepened, butnotfundamentally changed, inthe it emerges enriched

papers thatfollow. Herconception cannot beidentified with anyparticula period inthehistory ofpsychology: itwasnever fashionable inthepast, anditisnotantiquated now.I tendtodoubtthatGibsons approach will solve allthemysteries thatpsychology seeks tounravel. When weareface toface, however, Icannot helpwondering whether shemight notberight

abouteverything. Eleanor Gibson seestheproblems ofperceiving, acting

organisms soclearly, andshehasalways hadsuch afirm sense ofhowto practicepsychology as a science.

In tr ad ucti an

xvi

Introduction

force ordynamism wasresponsible forthechanges? Wearereaching that

point justnowinperceptual development, andtheory should takea new

spurt. There wasa strong developmental approach inpsychology before

1950(forexample, Gesell andPiaget). Intheupsurge ofinfant psychology, itwaslostasweadmired thefeatsofthecompetent infantdisclosed with ournewtechnology; andnowthetimehascomefora renewed andstrong developmental approach leading toanunderstanding ofwhatmoves perceptual development.

Inputting thisbooktogether, I chose papers thatseemed significant in theirtime,andinaccompanying textI haveexplained theoccasion for

writing them andconsidered what significance they have now. Aretheold questions stillwithusinnewclothing, orhave weformulated newones? Is

thereprogress andcontinuity inascientific history, oreveninonepersons living of it?

I havedivided thepapersintosixsections andarranged themin a

roughly chronological order. Aconsistent chronological order wontdo, because accidents intervene ina lifetime andupsettheconsistency ofones advance toward a goal.Yet,thereisanunderlying focus. Inmycase,it

began withthese concernsan interest inanimal behavior anddevelopment,inlearning, andgradually inperceptionand thesewovethemselves together overtime.Myaimforthisbook,sinceI believe thata developmental approach isthemostfruitful one,istoshowhowchanges takeplace, where thetransition points lie,andwhere thecurrent isheading. Id liketo thinkId shown,too,howmuchfunit canallbe.

Towhomis thisbookaddressed? Principally to todaysgraduate stu-

dents and young

researchers. I had themin mind,becauseI thoughtthey

might findencouragement intheearlyreverses thatbefell mewhich, however discouraging at thetime,leftchannels thatledonto interesting work. Myowngraduate students havereadit,theysaidwithprofit. Naturally, I hopeanyone insearch ofatheory ofperceptual learning, ora wayofthinking about cognitive development andtheorigins ofknow!edge, willreadit.AndI hope itfills ina fewdetails ofthehistory of psychology.

I ExperimentalPsychologyin the Thirties (1932- 1942)

Introduction to Part I

- In the 1930spsychology was a lusty young science. The American tradition of functionalism, madememorableby Jamesand Dewey, flourished as a strong conviction of evolutionary theory and a lot of researchon animals; the shadeof Titchener survived in the thriving enterpriseof psychophysics; behaviorismwas not just a legacy but a fact. Psychology laboratories(all universities had one and so did any self-respecting college) were full of activity and an air of excitement. Here was a new science, bursting with discoveriesto be made, fields to be mined, theories to be destroyed or constructed. The brightest young academicswere attracted to the scene, so the company was good and the competition challenging. Learningwas the big topic of the day, as befit the functional tradition, but there were plenty of other tempting ones, such as motivation and psychoanalysisfor the daring, and psychophysicalresearchpursuedunder the most stringent rules for the less imaginative. University faculties acceptedpsychology as a biological scienceand even allowed sciencecredit to introductory courses with laboratory sections, which nearly every department of psychology

offered .

This was where I camein. The hardheadedatmospherepromising new findings from laboratories focused on behavior was incredibly appealing, and I was won over before I was twenty years old. Clinical psychology and testing were already looming large, but it was the laboratory scenethat dazzled. Nearly sixty years and more than sixty experiments later I can

easily capturethe stirring moments. It seemsworth describingthe psychology of that era as it struck a student of the times, illustrating it with papers of the era. They are my own (often with a collaborator to make things more interesting). I explain what motivated me at the time, and how times changedas psychology (and I) grew up through the rest of the twentieth century. I begin with a sketchof the psychology of the 1930s, as it existed in the laboratory, and follow with a brief sketchof my introduction to it.

PartI

4 The

Functional

The

Tradition

functional

tradition

universities

,

was

a

the

of

call

influenced

by

malized

by

extends

.

colonies

had

a

were

1925

)

fanned

country

the

,

and

,

,

and

The

-

but

finding

the

on

how

line

he

major

of

an

descent

Yerkes

and

,

Yale

research

never

.

.

had

interest

in

other

psychologically

min

and

course

for

but

a

Hobhouse

Park

topics

flames

with

from

of

Orange

favored

for

resurfaced

,

.

,

and

emphasis

,

and

it

in

rewarded

expanded

action

stronger

would

couched

have

descended

Romanes

in

of

were

be

of

too

was

at

to

niche

animals

hand

learning

we

were

problems

a

-

what

they

soon

Chicago

maze

.

other

chimpanzees

to

how

environmental

research

Lashley

related

theories

and

these

however

of

investigations

learning

putting

its

like

(

control

,

Darwin

communication

(

years

to

"

. )

perceptual

homing

primate

of

-

etc

and

"

numerous

cues

,

recent

with

of

were

Maze

Columbia

control

,

functionalisffil

The

.

at

Sensory

connectionism

behavior

to

"

floor

right

for

research

followers

oriented

In

and

was

.

.

,

s

"

on

left

.

There

,

so

,

Hull

back

primates

and

'

its

There

,

left

information

adapts

ded

(

foraging

animal

.

cues

Thorndike

of

in

research

)

turns

Clark

speak

too

olfactory

echolocation

of

way

,

strong

colleges

for

vision

particularly

many

topic

sequences

we

was

at

popular

role

now

and

who

superintended

solving

'

-

cognitively

Problem

Kohler

an

was

s

and

Mentality

animal

of

Apes

laboratory

in

this

.

Functional

ural

psychology

outcome

infant

Darwin

biography

Man

(

)

,

thrived

than

theory

were

put

the

The

his

,

.

It

sixties

,

'

a

tale

Psychophysical

hold

Developmental

was

.

to

be

.

but

it

Boring

and

' s

was

psychology

phenomenological

his

remains

but

it

book

later

,

in

,

.

-

an

in

"

Child

Psycho

with

laboratory

-

norms

where

really

nat

ontogenetic

concerned

a

a

wrote

Evolution

and

more

of

,

fact

Mental

fascinating

was

excitement

told

well

theories

came

into

its

own

.

I

what

admired

for

of

the

'

t

being

call

the

to

on

in

1933

do

.

)

G

but

America

was

a

to

was

.

As

the

bible

for

percep

-

perception

days

the

considered

term

at

in

those

opposed

-

Boring

research

in

which

psycho

.

did

as

,

measurement

,

E

psychophysicists

scientific

psychologists

weak

1930

by

(

wanted

allowed

Gestalt

and

by

alight

Consciousness

who

what

rigorously

the

kept

of

else

Establishment

,

absent

flame

Dimensions

anyone

wouldn

the

conspiculously

,

Physical

and

Nowadays

was

in

,

phylogenetic

,

,

was

alive

students

tion

,

Darwin

psychology

introspection

physics

graduate

Romanes

Yale

the

.

approach

.

Enterprise

Titchenerian

Harvard

.

fanciful

at

t

test

developmental

comparing

however

didn

a

theory

chart

Gesell

the

supported

evolutionary

son

a

with

to

s

own

provided

that

logyll

'

of

1889

development

in

also

of

,

,

and

perception

II

psychophysics

it

soft

, "

Experimental Psychology(19321942)

5

indicates, it wasdualistic, aiming to matchup physical dimensions of frequency, intensity, andsoonwithmental scales ofjudgment. It was nonfunctional, dry,eveninhuman, buta strong technology wasspawned andit haditsuses,making possible applications to industry andother sciences. Graduate students hadto learnallthepsychophysical methods

andthemathematics thatevolved tosupport them.Ina latersection I turn

topsychophysics withoutdoor studies ofdistance perception. Themethodswerealsouseful instudies ofperceptual learning, laterona major

interestof mine.To become a psychophysicist, onewentto Harvard,

whereBoringreigned. Learning in the 1930s

Learning wasthe greattopicof the day.It mightbe studiedas verbal

learning orasmemory,wherethebignames wereE.S.Robinson, author

ofAssociation Theory Today (1932), andfollowers ofhissuchasJohn

McGeoch andArthurMelton; or it mightbecentered aroundthecondi-

tioned response. Clark Hull, atYale, epitomized thehardheaded learning tionalist, however. I wasinspired asa young student byhisPsychological Review papers onconditioning, especially A Functional Interpretation of theConditioned Reflex(1929), inwhich hestrove to interpret various manifestations ofconditioning asadaptive behavior. Thefourpapers inthis theorist, basinghistheoryonconditioning. Heconsidered himself a func-

series areclassics, clear andinsightful. Theypredated hispapers composed stress wasplaced onmechanism anddeductive rigor. Whether Hullreally asminiature logical systems andthebooksonlearning theory,wherethe

achieved rigorornot,hehadvigorandhislaboratory wasanexciting

place,seethingwithactivity.

Hullwasnottheonlyfunctionalist concerned withlearning theory.

Therewashisarch-rival, Tolman, at Berkeley. Howa rat learneda maze wasasinteresting to Tolman asit wasto Hull,butalthough heeschewed

thewords andonlydescribed behavior, Tolman wasathearta cognitive

psychologist. Therat learnedthemazebecause it discovered whatledto whatandformeda cognitive mapofthelayoutandwhatit afforded. One

of thejoysof a meeting of theAmerican Psychological Association in thosedayswasseeingHullandTolman withtheirrespective followers

holding forthinthesameroom, everyone arguing excitedly (andgood

naturedly).

TheImpactof the Thirtieson a Student

I attended Smith College asanundergraduate, a greatplace fora young

womanto be introduced to science. Asa womenscollege, Smithhada

6

PartI

strongfemale faculty inscience andthearts,because thatwaswhere a

womanscholarcouldfindhonorand promotion.Thereweretruly great womenat Smith,for exampleMarjorieNicholson, the worldsexperton

scienceand imagination. I firstencountered psychology in a year-long introductory coursethatincluded a weeklylaboratory (I latertaughtit).

MysecondcoursewascalledAnimal Psychology, taughtby Margaret Curti,a Chicago-trained functionalist. It wasentrancing, becausewe studentswerepermittedto runan experiment ourselves. Theanimalswere rats, of course, and they learned a maze.

The realexcitementcamemy senioryearwhenI had a year-longcourse

in advancedexperimental psychology withJamesGibson,a youngassis-

tantprofessor freshfromPrinceton. Therewereeightstudents in the

class(fourof themlaterobtainedPh.D.s). Weworkedin pairsandperformedan experiment per month,settingup apparatus, obtaining and

running ourownsubjects, researching thebackground, andwriting it all up.Every experiment wasa novelproblem andwecovered a widefield, withexperiments onreaction time,learning, memory, perception, adaptation to prismsthe gamut of the fieldat the time.It was one way

thata youngprofessor whowanted to doresearch couldkeephishandin, despite teaching twoorthreecourses aterm.Manypapers came outofthe workbeguninthatcourse. 1wasproudthatanexperiment ofminewith GertrudeRaffelSchmeidler (latera ColumbiaPh.D.)waspublishedin the

Journal ofExperimental Psychology. Theexperiment wasonbilateral transfer ofa conditioned reflex.Wehada functional interpretation oftheCRofour own (see followingpaper),ahead of its time.

Thatcoursetaughtme the thrillof doingexperiments andthe great satisfaction of discussing problems dailywithcolleagues. Smithhadan excellent courseon historyandsystemsofpsychology, taughtby a Ph.D. ofBorings,andby thetimeI wasa senior,KurtKoffka wasgivinga course on GestaltPsychology, whichI attended.Later,whenI wasa graduate student, I regularly attended Koffkas seminar asmostoftheSmithpsycho-

logyfaculty did.Itwasthelocalforum fordebate, despite Koffkas rather authoritarian style.I wasnevermuchattractedto Gestaltpsychology, as I haveeverbeenuneasywitha phenomenological approach,a factthat undoubtedly influenced mychoiceofYaleforgraduateschool. SmithCollegewasverysupportive andencouraged its studentsto go

onto graduate work.I stayedonat Smith fortwoyearsaftergraduation asa teaching assistant andobtained a masters degree there.Mythesis (donewithJames Gibson) wasonretroactive inhibition inverballearning,

verymuch inafunctional tradition. Itwasdeepdepression times andthere werefewfellowships for women.WhenI wentto graduateschoolat Yale,Smithgavemea smallfellowship, butYalecontributed nothing.

ExperimentalPsychology (1932- 1942)

7

At Yale, I intended to specializein comparativepsychology and work in Yerkes's laboratory, not so much becauseof Yerkes's ideas but for the glamor of working with chimpanzees(a poor reason, I fear). That expectation was not fulfilled . Yerkes informed me, none too graciously , that he allowed no women in his lab- a real shock after the nurturing atmosphere

of Smith. Clark Hull acceptedme as a graduate student, however. He was deep in his enterpriseof building logicodeductive systems. He allowed me to choosemy own topic, but I was required to back it up by a theoretical presentationwith formally stated premises, deductions, etc. I managedto

do it, but with a bit of tonguein cheek . I wantedto work on verballeaming andforgetting, andto showthat remembering andforgettingdepended on differentiatingcontentinto meaningfulunitsthat did not confusewith one another in an incoherent melange. I could borrow my main concepts, differentiation and generalization , from the conditioned response literature , which made them very respectable at the time . But the terms held com-

monsensepsychological meanings for me, perceptual meanings I would have said later . Noone at Yale worked on perception or was in the least interested

in it .

I performed my thesis experiments , except the one on generalization in a reaction time experiment , at Smith, where I returned as an instructor

after a year at Yale. I had a very busy year at Yale getting through all the many requirements, including a proseminar for all the new students. It was a showcasefor the professors, each of whom had a few weeks to display his ideas (all were men, of course). Yale was generous in letting me take a degreewith only one year's residencecredit. I took Clark Hull's enormous

memory

drum back to Smith and worked

there on my disserta -

tion. Smith retainedme on its faculty and kept me so busy that, except for directing a few master's theses, I got little further researchdone. It was after World War II before the scenereally changed for me, as it did in many ways for psychology in America.

Bilateral Transfer oftheConditioned Response

in the Human Subject

JamesJ. Gibson, Eleanor G.Jack,Gertrude Raffel

This paper isafunctional treatment oftheconditioned response, emphasizing its

adaptiveness asa response ofthewhole organism rather thananisolated reflex as Paviovians would haveit.Thediscussion emphasizes thispoint: Ourhypothesis would haveit thatintheadult,organized habits ofwith-

drawing orpulling awayhavebeen developed forallpartsofthebodywhich havecome incontact withpainful stimuli. Wecansuppose thenthatinour experiment, theconditioned stimulus hasbecome effective notonlyforthe

particular withdrawal response inconnection withwhichit waslearned, but alsoforthisentirerepertory ofavoiding reactions.

Thispointwassupported inlaterexperiments byWickens (Wickens 1939a andb)in whichfingerwithdrawal wasconditioned withthehandin normal position andthenreversed sothata flexorratherthanextensor movement hadto

beperformed iftransfer occurred. Itdid.InpartII,anexperiment thatIperformed much later with goats assubjects leads toasimilar conclusion, soitwasnotonly human adultsubjects forwhom avoidance toshock isnota simple reflex. When

wecometopartVI,I showhowthemuchlaterconcept ofaffordance fitsthese

cases, which arebasically similar totheresponse tolooming presented there.

It is interesting to findHilgardandMarquisin theirclassic work,Con-

ditioning andLearning, referring toourresults andWickenss as insightful behavior infypical conditioning experiments (Hilgard andMarquis 1940, 243). Theyalsosayin thesecases theequivalence isobviously learned. I wouldnt be

so sureaboutthat.I thinkwe havediscovered sincethenthat behavior is

organized originally inlarger synergies thatinvolve patterns ofpostural activities

thatcanbe broken downor individuated intosmaller ones,ratherthanthe

oppositenot a learned equivalence ofsmaller pieces asHilgard andMarquis supposed in1940. Theelementarism ofearlybehaviorism diedhard(ifit has). JournalofExperimental Psychology, 1932,41O421.

10

J. J. Gibson , E. G. Jack , & G. Raffel

However , thesetwo importantcriticsof theirtimedid raisemanydoubtsabout buildinga theoryof behavior on reflexes that couldonly widenthedomainof actingby being"conditioned " to whatever stimulichance andopportunity happened t-""" .""... ..., toJjuxtapose rr~rr-r --- with -- -- them. During the courseof an experimenton the conditionedwithdrawalresponseof the handto an electricshock,the opportunitypresented itselfto testfor transferof this conditionedresponse from the trainedright handto the untrainedleft hand. The subjectsat at a table with the palm of the handrestingon onelargebrasselectrodeandthe middlefingeron another smallerone. The handwasnot strappeddown or constrained in any way. Theapparatus andelectricalcircuitwereessentially the sameasin Watson's research ,l exceptthat the arrangemen ~ for graphicallyrecordingthe responseson a kymographdrumwas not employed . One of the Es simply satat the sideof S andrecordedthefingermovements asthey occurred . In order to keepthe attentionof S off the experiment , and to lessenany anticipatoryfearof the electricshock,S wasrequiredto readaloudduring an experimentalsitting. The conditionedstimuluswas the soundof an electricbuzzerlastingfor one secondand startingapproximately~ sec. beforethe shock.This timing wasaccomplished by usinga doublecontact key which, whendepressed , first completedthebuzzercircuitandthenthe shockcircuit. Theprocedurewasasfollows. After Swasseatedwith the right handon theelectrodes , andhadbegunto readaloud, thebuzzerwassoundedalone. Forseveralsubjectswho weresuspicious or fearful, the soundby itselfwas adequateto producea withdrawalmovement . Although this response alwaysdisappeared afteroneor two repetitionsof the buzzerstimulus , the recordsof suchsubjectshavebeenomittedfrom the dataon transfer . After the abovetest, combinedstimulationsof buzzerand shockweregiven at irregularintervalsvaryingfrom 3 to 20 seconds . When 10 repetitionshad beencompletedoneof two procedures wasused.(1) In thefirst, thebuzzer wassoundedaloneandany conditionedwithdrawalmovementwasnoted. If noneappeared , anotherseriesof reinforcements wasgivenanda second test for the conditionedresponsemade.Whenthe latter madeits appear ance, a final seriesof reinforcements was given and then S was told to shift over and placethe left hand on the electrodes , with the middle fingeron the far electrodeasbefore. The buzzerwasnow soundedalone andany withdrawalmovementnoted. (2) In the secondprocedure , instead of waitingfor theconditionedresponse to appearin theright hand, E made a testfor thetransferred response in theleft handimmediatelyafterthefirst 10 combinedstimulations . If noneappeared , Swasrequiredto shiftbackto the right handand more reinforcements were given, after which another

Bilateral Transfer of Conditioned Response test for transfer was made . As a check on the transferred

11

response when it

occurred, a final test for the conditioned movement in the right hand was

subsequentlymade. Results

Twenty subjects were used in the experiment . Of these, 19, or 95 percent,

were successfullyconditioned to withdraw the finger of the right hand when the buzzer was sounded alone . From 10 to 52 repetitions of the combined stimuli were necessary to establish the response . (The one sub -

ject who failed to becomeconditioned had undergoneelectricaltreatments for a year , and showed a very weak withdrawal

response even when the

shockadministeredwas much stronger than that usually employed. At the end of 200 repetitions no conditioned response had occurred and the

experimentwas stopped.) Of the 19, 6 showed someindication of responding to the buzzeralone before the experimentwas begun, and their results have accordingly not been considered. Thirteen subjectsremain for whom the data on transfer of the conditioned responseseemvalid. Although at no time did any subject in this group receive a shock with the left hand, 8,

or 62 percent made definite withdrawal responsesof the finger when the left hand was placed on the electrodesand the buzzer alone was sounded. Three of this group had given evidence of a conditioned response of the

right hand before the left hand was tested, whereas5 showed the response in the left or transferhand first and subsequentlyin the right . It is possible that a larger number of subjects might have manifested transfer if the experiment had been continued longer. No more than 100 combined presentations of buzzer and shock were given to anyone subject.

If by that time no indication of transfer had shown itself the attempt was abandoned. This was done in part out of consideration for the subject for whom the experiment was naturally somewhat uncomfortable , and in part

out of a desire to investigate other problems in addition to the transfer phenomenonduring the experimentalsitting.2 The appearanceof the conditioned withdrawal responsein the right hand itself was found to be somewhat irregular and unstable, this finding being in accord with the results of previous experiments.3 In 2 subjects, only one definite conditioned response could be elicited , all attempts there-

after being negative. These subjectswere apparently surprisedand somewhat chagrined to find that they had reacted to the sound of the buzzer alone , and thereafter

set themselves

" not to be fooled

more than once ."

They constitute two of the group for whom transfer could not be demon strated. In general the reports of subjects indicate that various attitudes such as timidity towards shocks, a desire to appear indifferent , a desire not

to be tricked into responding, and the like, played a large part in the

12 J. J. Gibson , E. G. Jack , &: G. Raffel experiment , andwerethe probableexplanationof muchof the irregularity with

which

the conditioned

response

showed

itself . In view

of this in -

stability of the responsein the right hand it is not surprising that the transferred response in the left hand could not always be demonstrated . Discussion

In a significant number of cases(62 percent) the establishing of a conditioned withdrawal responseof the right middle finger is accompaniedby the formation of a similar conditioned responseof the correspondingfinger of the other hand . From this fact alone little can be concluded

with certain -

ty . Chiefly it suggests the need for further research on such questions as the

extent of transfer to other (unsymmetrical) fingers of the same and the opposite hand, the latent periods of transferredresponsesascomparedwith the latency of the primary conditioned and unconditioned response, the

possible occurrenceof implicit movements in the transfer hand, and the like. But it also suggests a possible explanation towards which further experiments may be oriented .

It is clear in the first place that the conditioned withdrawal movement can scarcelybe considereda single isolated reflex, or even a conditioned reflex, if by the term we mean a responselimited to a specificmuscle or muscle group. This latter criterion is frequently regarded as definitive for the true reflex. The hypothesis is suggestedthat the conditioned with drawal to shock involves, or perhapsis itself, a generalizedhabit of avoiding or withdrawing when the buzzer is heard, which may be evoked from another part of the body than that in which the response was learned. Unfortunately no systematic data were obtained as to transfer of the

responseto other fingers. It was, however, informally determined with several subjects that the conditioned response did readily transfer to the

index finger of the same hand. The extent to which a general avoidance response to the buzzer was set up can only be inferred . But from the above

evidence and from what is known of transfer of learning to symmetrical and unsymmetrical parts of the body4 it is possible to supposethat the withdrawal response could have been elicited in the other fingers and perhapseven in other parts of the body. It should be noted that the conditioned response which is concurrently

establishedin the untrained hand, is really a latent or potential response. It becomes actual and overt only when the finger of the untrained hand is for

the first time resting on the electrode. Our hypothesis would have it that in the adult, organized habits of withdrawing or pulling away have been developed for all the parts of the body which have frequently come in contact with painful stimuli. We can then supposethat in our experiment , the conditioned stimulus has become effective not only for the

Bilateral Transfer ofConditioned Response 13

particular withdrawal response inconnection withwhichifwaslearned, but

alsoforthisentire repertory ofavoiding reactions. Furthermore when any

part,saya finger, is ina potentially painful situation (e.g.,restingonan

electrode) wemayconceive thattheappropriate withdrawal response for

thatfingeris in a stateof readinessperhaps of sub-activation. It is this

particular finger, therefore, which iswithdrawn undersuchcircumstances,

ratherthananyother.Consequently thedifference between thelatentor

potential conditioned response andtheactual onedepends onthepresence ofa specific preparatory setarousedby thesituation. Ontheconscious side,insofarastheseactivities arerepresented in

consciousness, the aboveformulation wouldrun as follows.The condi-

tionedstimulus hasbecome during training a general signal to withdraw.

Together withthishabit, there ispresent (normally) a definite preparatory

set to withdraw the particular fingerwhichis restingon the electrode. Hence whenthefingeroftheuntrained handisplaced inthatsituation, the soundofthebuzzer setsofftheproperresponse.

Theintrospective reports givenbythesubjects indicated notonlythat thesetorattitude wasanimportant factor intheexperiment butalsothat

theresponses, bothconditioned andunconditioned, couldnotbeclassed as

definitely either voluntary orinvoluntary. Theywerenotcompletely vol-

untaryin thesenseofbeingintended, as,e.g.,a fingermovement in the

reaction timesituation isintended. Theconditioned responses wereusually accompanied bya feeling ofsurprise. Ontheotherhand,theywerenot wholly involuntary inthesense ofbeing automatic asisthekneejerk.The conditioned responses couldhavebeeninhibited hadthesubject wished

to inhibitthem.Thecasesof thetwosubjects whoset themselves not to

befooled morethanonceareinpoint. Thattheunconditioned response itselfcouldbeinhibited inlargepartwasshown byonesubject whotook

the instructions to forgetaboutthe fingeron the electrode so far as

possible, asmeaning thatheshould notrespond voluntarily. Thefinger

wassosuccessfully forgottenthattheonlyresponse to theshock wasa slighttrembling or twitching movement, evenwhenthe stimulus was painfully strong. Fromtheaboveevidence ifappears evenmoreclearthat

avoidance toshock isnotareflex. Itisrather aresponse highenough inthe scaleofcomplexity tobeinpartdependent ona voluntary set.

Sincea conditioned responsecanbe formedbetweena stimulus anda

response which wasnever overtly made during theconditioning process, theinadequacy ofthesimple diagram orphysiological schema usedfrequently inexplaining conditioning isoncemoredemonstrated. Anyreal

physiological explanation of the presentfactswouldhaveto be based onanadequate theoryofthephysiological basisofbilateral transfer. Such a theoryhasnotyetbeenformulated. Itispossible thatfurther research on

14

J. J. Gibson , E. G. Jack , &: G. Raffel

transferof the simple type of learning here describedwill contribute to this problem. Notes

1. For the original researchon this subject, seeJ. B. Watson, The Place of the Conditioned Reflex in Psychology, Psychol . Rev., 1916, 23, 89- 116. 2. These additional problems had to do with the fact of so-called 'sensory irradiation' (Pavlov) and the establishing of differential responsesto buzzers of different pitch. 3. J. B. Watson, The Placeof the Conditioned Reflex in Psychology, Psycho 1. Rev., 1916, 23, 89- 116. G. Humphrey, The Effect of Sequencesof Indifferent Stimuli on a Reaction of the Conditioned Reflex Type, J. Abn. and 50c. Psychol ., 1927, 22, 194- 212. 4. Seefor example C. W . Bray, Transfer of Learning, J. Exp. Psychol ., 1928, II, 443- 467.

2

Retentionand the InterpolatedTask Eleanor J. Gibson, James J. Gibson

-

.

This paper is a condensationof my master's thesisat Smith College. JamesGibson

wasmy thesissponsor(hebecame my husbandwhile I wasworkingon thethesis ). Verballearningand memory , in the Chicagotradition, was muchin vogue(note thereference to E. S. Robinson , oneof the Chicagogroup). Thematerialemployed in theseexperimentswas typically letters, digits, or nonsensesyllables. Retroactive inhibition was a popular notion for explaining forgetting, and similarity of

interpolatedmaterial to whateverwas originally presentedfor learning was considered critical. But thesimilarity wasgenerallycontentoriented , and referred to likeness between individual items, consistent with the elementarism of the

thirties. My functional-mindedness ledme to the hypothesis that the taskengaged in, which is oriented toward process rather than content, must be of equal importance. It makes me think now of the later information processingtermi-

nology, emphasizing "processing " of "alphanumeric characters ." But of coursethat would be twenty-five years later. The experiment varied both content and task with respect to original and interpolated task relations. " The similarity of two tasks, then, will be expressedin

termsof the significant featureswhich the two taskshave in common." The features chosenwere the "material" and the "operation" performed on it . As the

experiment demonstrated , bothso-calledfeatureshad the effectof reducingretention when they were presentedseparately in an intervening task, but detrimental effectswhen they were combinedweregreater than the sum of the injury done by the two separately. I supposethat one could say that "processing" numbers is not the same task as "processing" letters, whatever the operation. This experimentled me to think further about retroactive inhibition and consequentlywas the fore-

runner to my Ph .D . thesis.

American Journalof Psychology , 1934, 46, 603- 610.

16

E. J. Gibson & J. J. Gibson

Thesimilarity problem inretroactive inhibition arises fromthefactthatthe integrity ofretention ofmemorized material isa function ofthesimilarity between thelearning andtheinterpolated activity. According totheSkaggsRobinson theory,thefunction is of thefollowing nature.Thecurveof retentionstartsat a highlevelwithmaximum similarity, whichRobinson took to meanthat the interpolatedtaskis identicalwith,or a continuation

of,the original learning. Thecurvefallsas thesimilarity of interpolated materialis reducedfromidentityuntilretentionreachesa minimum at the

stagewhich, presumably, otherinvestigators havecalled a similarinterpolation. Thecurvethenrisesuntil,at the stageof dissimilar or rest interpolation, retention isat thelevelusually considered normal. Thelatterportionof thiscurveembodies thetypeof resultsusually obtained in experiments on thesimilarity problem. Thefirstpart,showingdecreasing retention withdecreasing similarity, hasbeenverified by Robinson 2 and Dreis 3 with two experimentswhereinthe interpolations

consisted of varying amounts ofrehearsal to whichwereaddedcomple-

mentary amounts ofnewmemorization. Thegreater theproportion of rehearsal, thegreater wasthesimilarity oftheinterpolated activity to the original learning. These experiments, along withsomeothers, 4 arosefrom Robinsons tentativeassumption thatimprovement through practice andthe disintegrating effect ofinterpolation arerelated phenomena; inotherwords,

thatpractice-effect passesoverintointerpolation-effect without break. It shouldbe notedthat,in thisattackon the similarity problem, the interpolated activity isnota singletask,asit canbeonlyifit is separate anddistinctfromtheoriginal learning task.In otherwords,theinterpolationsof Robinson andDreiswerenot activities havingan effect(either facilitatory or inhibitory) on the integrityof retention of the primary learning. Theywereinstead continuations oftheprimary learning invary-

ingamounts, combined withnewlearning. Likewise, theretention which wasmeasured intheseexperiments wasnotretention ofa primarylearning, but of that plussomecontinuationof it.

Theposition canbetaken, ofcourse, thattheretroaction problem need

not be definedin termsof discreteprimaryand secondary tasks.The

interpolated activity doesnothavetobea separate andunique task,the argument mightrun,sinceit isprecisely thefactor ofindistinctness and confusion betweentaskswhichis mostlikelyto be the explanation of

retroactive inhibition. Thishypothetical argument, however,begstheques-

tioninfavorofa radical formofthetransfer theoryofretroaction. There

isasyetnoproofthatit is thesingle andcomplete explanation ofthe phenomenon.

Thetheoretical approach oftheexperiment to bedescribed differs from thatexpressed by the Skaggs-Robinson function in that retroaction is

thought of as oneof theproblems of theinterrelations of tasksthat

Retention andtheInterpolated Task 17

problem dealing withtheeffect ofa second taskonthetestedretention ofa first.Thisemphasis ontask-characteristics carries withititsownmode of dealing withthesimilarity factorin retroaction. Twotasksaredesignated assimilar withreference to definite features which thetwohave

incommon. Such features would notbeindependent elements butaspects bymeans ofwhich twotasks could becompared. There areundoubtedly manysuchwhichcouldbe usedas a basisof comparison, butcertain features ofa taskwillstandoutasbeingmoreimportant thanothers foragenuine description ofitandconsequently fora significant statement ofsimilarity. Thesimilarity oftwotasks, then, willbeexpressed interms

of thesignificant features whichthetwotaskshaveincommon.

Thetwomostobvious features ofa taskaretheoperation which the

subjectis instructed to perform andthe material withwhichhe must

perform it.Thetermoperation willbeusedto mean thataspect ofa

taskbestindicated bythewordAufgabe, andthetermmaterial willbeused to indicate whatcouldbecalledtheperceptual data.Thesetwofeatures

arenotdistinct components. They areatleastpartly interdependent, since achange inthematerial ofa taskwould beaccompanied bysome change intheoperation, andviceversa. Forexample, theoperation inmemorizing

nonsense syllables is not the sameas in memorizing nonsense forms, althoughboth are memorial in nature.

Thisdistinction isnotnew. There have beenrelated onesranging from behavior asgoal-seeking andasinvolving behavior-supports. 5 Robinson himself, inanearlier experiment, distinguished between theprocessand thedistinction between actandcontent to Tolmans characterization of

thecontentofa taskandfound thatthe degree ofretroactive inhibition isa function ofsimilarity ofprocess aswellassimilarity ofcontent. 6 He

further demonstrated thatanother feature, which hecalled formofpresen-

tation,wasimportant forretroactive inhibition. Hisdata,however, didnot

show thatthese features were independently capable ofcausing retroaction. Asaconsequence, perhaps, hedidnotfollow upthislineofthought inhis subsequent research. With theexception ofthisexperiment ofRobinsons, allresearches onthesimilarity factor haveworked withsimilarity ofmaterial alone. When variations intheoperation oftheinterpolated taskhave been included inanexperiment, thisfactor hasnotbeenspecified norhas

its effectbeenisolated andmeasured. Littleis known, then,aboutthe influence ofoperational orfunctional similarity onretroaction. TheExperiment

Intheexperiment to bedescribed, ourproblem wasto decide upona setofinterpolated tasks which would enable ustoisolate andcompare the decrements inretention separately caused bysimilar operation andbysimi-

18

E. J. Gibson & J. J. Gibson

lar material.The relative importanceof the two factors for retroaction

couldthusbe determined. Specifically, we wishedto comparedegreesof

retentionof the primarylearningin the following cases:(1)whenthe secondary taskis likethe primarytaskin bothoperation andmaterial; (2)whenit is likein operationbut differentin material; (3)whenit is differentin operationbut likein material; (4)whenit is differentin both operation and material.

In order to makethe above comparisonsit was necessaryto selecta primary

learning taskandfourkindsofinterpolated tasksfulfilling theaboverequirements. On the basisof a preliminary experiment, learninga listof pairedconsonantswas

chosenfortheprimary taskTheinterpolated taskswere(1)learning anotherlist

ofpairedconsonants, (2)learning a listofpaired digits,(3)cancelling a specified pairofconsonaMs whenever it appeared ina sheetofpiedtype,and(4)cancelling a pairofdigitsinthesame manner. Afifthinterpolated task,ofthesortgenerally employed asa control orrest condition, wasalsoincluded. Itconsisted oflooking at pictures (moving-picture stills)andit wasbelieved to differfromtheprimary task even morethan did cancelling digitsin fact it was designedto be as differenta task as possible.Amongother features,it lackedthe motivational or work characteristics of the other four tasks.

For the sakeof simplicity, we shallcall the tasks,in order,Learning Letters,

Learning Numbers, Cancelling Letters, Cancelling Numbers, andPictures. Thefollowing schemadesignates by capitallettersthe arrangement of tasksin the fiveinter-

polated conditions. Therewere5 groups ofSs,onegroupcorresponding toeach condition.

Interpolated Condition

(Group)

Primary Learning

Secondary Task

V

LL

P

I II III IV

LL LL LL LL

LL LN CL CN

Experimental materials

Forprimary learning, allgroups studied alistof10pairs ofconsonants. Nopairs

wereincluded which made familiar abbreviations (such asPM); nopairconsisted ofconsecutive letters ofthealphabet; andnearly alltheconsonants ofthealphabet wereusedinonelist.Forthesecondary task,Group I learned another listof

consonants made upliketheonejustdescribed, Group IIlearned alistof10pairs

ofdigits. Nodigit appeared with disproportionate frequency, nordidconsecu tivepairs contain adigit more than once. Group IIIhadtoread through asheet ofpied consonants striking outthecombination KSwhenever itappeared. The combination occurred aboutthreetimesineverytwolines. Group IVcancelled adigit combination inthesame manner. Thepictures secured forGroup Vwere

Retention andtheInterpolated Task 19

photographs distributor. (SxIIin.) ofdramatic situations obtained from amotion picture Procedure

Five laboratory sections ofalarge course inpsychology served asthe5groups of intheassigning ofstudents tosections, thegroups were probably asnearly equal Ss.They averaged about 26Sseach. inview oftheabsence ofanyselective factor

inability asanythatcould befound. Theexperiment wasadministered inthe

manner ofagroup mental testwiththereading ofdefinite instructions bytheF andemphasis oncareful adherence tothem. Thematerial foreach group was bound inpamphlet formandpassed outtotheSs.Thefirstsheetcontained a warning nottoopen, thesecond sheet hadonit thelistforprimary learning, thethird wasblank, thefourth contained thematerial fortheinterpolated task, the fifth hadacolumn of10short lines forthewritten recall oftheprimary list,and forGroups I andTithesixthhadanother setoflines fortherecall oftheinterpolated list.Thepictures forGroup Vcould notbeincorporated inthebooklets, sothey were placed face down ontheSstables before theexperiment began. Theexperiment wasrunoffinthesame wayforallgroups, apart from the differences intheinterpolated activity. Two minutes were allowed forthestudy of

he primary listwithaviewtoreproducing it,incorrect order, later. Ssthenturned totheblank pagefor30sec.,during which theyweregiven instructions forthe

interpolated task. Three minutes were allotted totheinterpolated task; thentheSs turned tothefifth sheet and,after brief instructions, wrote asmany pairs of consonants fromthefirstlistastheyremembered during i mm.Finally, for Groups IandII,theSsturned tothesixth page andreproduced thesecondary list. Thetime-intervals allowed forprimary learning, forinterpolated task,andfor recall, were determined after preliminary experimentation. Scoring

Thefollowing system wasusedinscoring therecalls written bytheSs.Onecredit wasgiven ifaconsonant-pair wasreproduced correctly inthesame position which itoccupied intheprimary list,orifitwasreproduced correctly initsrelatively correct position (i.e.relative tothepairofconsonants which preceded it onthe list). One-half credit wasgiven ifapairwascorrectly reproduced inaposition one place removed fromitscorrect position (relative orabsolute). One-half credit was givenifone-half a pairwasreproduced initscorrect position. Nootherrecall received credit. Results

Table2.1isa summary oftheresults, showing thenumber ofSsin each

group, therange oftheindividual recall scores ineachgroup, theaverage

number ofconsonant-pairs recalled, andtheprobable errorsoftheaver-

ages.As onewouldexpect,Condition I, learninga secondlist of con-

sonants, results inpoorer retention thananyoftheotherinterpolated conditions. Conditions IIandIIIshowabout thesame average number of pairsrecalled. Thelossresulting fromeither common operation alone, or

20

E. J. Gibson & 1. J. Gibson

Table

2.1

Summary of Resuhs Group

No.Ss

I

II

III

IV

V

LL-LL

LL-LN

LL-CL

LL-CN

LL-P

22

25

30

28

27

Range Av.

P.E.av

09 3.7 –28

0.510 6.2

–36

0.510

6.3 –38

310

410

7.6 –.25

8.6 –21

commonmaterialalone,seemsto be about the same,i.e. similarityof

operation andsimilarity ofmaterial areapparently ofaboutequalimportance in determining decrements of retention.

In ConditionIV (LL-CN), whereneitheroperationnor materialis like

that in the first task, recallis better than it is in the two conditionsin whichone of these featuresis similar.This improvementis manifestin

boththeaverage recallandtherangeoftheindividual scores. Condition V (P),wheretheinterpolated taskwassimply looking at pictures, givesthe bestrecallof all5 conditions. Boththeaveragerecallandtherangeof the

individualscores show that there is an improvementeven over LL-CN,

whereneitheroperation normaterial werecommon to thefirstandsecond tasks.

Fig.2.1shows graphically thedifferent percentages ofthetotalpossible

recallwhichresultedfrom the 5 interpolatedtasks.It shouldbe remembered that the maximumpossiblerecallwas 10 consonant-pairs. Using

thesepercentages, wemaymakethecomparisons indicated asthepurpose of ourexperiment. Condition IV(LL-CN) wasdifferent fromtheprimary taskin both materialand operation,and 76%of the maximum recallwas

obtained by thisgroup.Condition III(LL-CL) waslikeIV,exceptthat the materialwas similarto that usedin the primarytask.Its recallwas 63%.Subtracting thisfrom76%wefinda difference of 13units.Hencewe

maysaythatsimilarity totheprimary taskinmaterial onlycauses a lossin

retention,in thissituation,of 13unitsof percentage. ConditionII (LL-LN) waslikeIV(LL-CN) in material,but in operationwassimilarto the primary

task.Theaveragerecallfor thisgroupwas62%.Subtracting from76% (recallforIV)wefinda difference of 14units.Hencein thiscondition we

maysaythatsimilarity totheprimary taskinoperation onlycauses a lossin retention of 14 units of percentage.

Now,if it wereassumedthat a taskis madeup of two actuallyin-

dependent andseparable components, material ontheonehand,andan operation ontheother,itmightthenbeexpected thatwithaninterpolated taskhaving bothoperation andmaterial similar to thefirsttask,theloss

Retention and the Interpolated Task

21

Figure2.1 Effectof different interpolated tasksonretention of a list of 10pairedconsonants . Reading fromleft to right, thefiveinterpolated tasksareLL( learning a second list of letters ), LN (learning a).listof numbers ), CL(cancelling letters ), CN(cancelling numbers ), andP (looking at pictures

03NIV .L80 llVJ3 ~ 3181SS0dlV .LO.L::fa .LN3J~3d

would be a sum of the lossescausedby these two independently. Since there is a loss of 13 with similarity of material and 14 with similarity of operation, the sum of the losses, in this case, might be supposedto be about 27. ActuallyI however, comparing Condition I (both material and operation in common) with Condition IV (neither in common) the loss is 39 units of percentage.The presumptionis that the effect of an interpolated task in which similarity of material and operation is combined is greater than the sum of the effectsof two taskseachsimilar in one of thesefeatures respectively.7 If this inferenceis correct, it would indicate that the material and the operation within a single task interact, each enhancing to some extent the effect of the other. Consequentlyone can be separatedout only with violence to the other, and in varying them independently, as we have done, the remaining common feature is no longer as similar as it would beif both remainedthesame.

In ConditionV (P) the interpolatedtask was differentfrom the first task in a variety of features . The percentagerecalled(86) was greater than in Condition IV (CN), where the interpolatedtask was different in operationandin material.Thegainin V ascomparedto IV suggests that

22

E. J. Gibson & J. J. Gibson

thereare otherfeaturesthan thesetwo with referenceto whichcompar-

isonsmaybemade, andthatoneormoreofthesefeatures, e.g.thework attitude, remained common totheprimary andinterpolated tasksinIV,but not

in V.

Theabovecomparisons canbemade, withatleastequallogic,byexam-

ining gains inretention asa function ofnon-resemblance oftheinterpolated task. Thegains are25unitswithadifference inmaterial only, 26unitswith a difference in operation only,and39unitswitha difference inboth.As before, thenon-additive character ofthelattergaincanbeexplained by supposing material andoperation to beinterdependent. Summary

Ourexperiment showsthatcomparison withreference to features of two

tasksis usefulin understanding therelationbetweensimilarity andretroactive

inhibition.

It hasbeendemonstrated that underthe conditionsof this experiment

theinterpolation ofa taskwhich issimilar to theprimary learning either inoperation orinmaterial results inpoorer retention thandoestheinterpolationof a tasksimilarin neither.In thissituation,the two featuresoperation andmaterial, seemto be aboutequallyimportantin theeffecton retention.

Operation andmaterial werechosen asthetwomostobvious features or characteristics by meansof whichtwotasksmightbe compared. That thesearenot the onlyfeatureson whicha comparison maybe basedis

suggested bythefactthatCondition V.which differed inotherrespects,

gavebetter retention thanCondition IVwhich differed inthetwofeatures of material and operation.

Finally, theoriginal suggestion thatfeatures ofa taskareinterdependent seems tobeupheld bytheprobability thatthesumofdecrements dueto operation andmaterial doesnotequal thelosscaused bysimilarity inboth. Notes

1.SeeE.S.Robinson, Thesimilarity factorinretroaction, thisJOURNAL, 39,1927,297 312.ForSkaggsearlier formulation, seeF. B.Skaggs. Further studies in retroactive inhibition,Psycho!. Monog.,34, 1925,(no. 161),160. 2. E. S. Robinson, op. cit.

3. T. A. Dreis,Two studiesin retroaction,J. Gen.Psycho!., 8, 1933,157172.

4.L.Harden, Aquantitative studyofthesimilarity factor inretroactive inhibition, 1.Gen.

Psycho!., 2,1929, 421432; andN.Y.Cheng, Retroactive effect anddegree ofsimilarity

J. Exper.Psycho!.,12, 1929, 444449.

5.F.C.Tolman, Purposive Behavior inAnimals andMen,1932,10f.

6.E.S.Robinson, Somefactors determining thedegreeofretroactive inhibition, Psychol. Monog.,28, 1920, (no. 128), 28.

Retention and the Interpolated Task

23

Sensory Generalization withVoluntary Reactions EleanorJ.Gibson

This experiment wasthefirst ofseveral undertaken formydissertation research at

Yale University. Mydissertation sponsor, Clark L.Hull, generously gave mehis notebooks toperuse over a long weekend, thinking I might find aproject that

appealed tomeamong anumber roughly outlined there. Hewas inthehabit of

writing a sortofdiary every Sunday morning, noting down events oftheweek justpast, especially ideas thathadoccurred tohim, including ones fornew experiments. Reading thenotebooks was fascinating, butI returned them with the conviction thatI hadtoformulate myown problem andthatI would continue in

thevein ofmymasters thesis, working onverbal learning and forgetting with human subjects. Hull greeted mydecision without much enthusiasm. IfIwanted towork with him, I hadtostay inhisterritory. Hetold metowrite upmy ideas asaproposal, andhewould consider them. Aprevious student ofHull, W.M.Lepley, had published amonograph onserial learning and forgetting based on conditioned reflex principles. Hull called this work to my attention thinking I mightdosomething similar. Lepley hadproposed thatrote learning might bethought ofasa series of conditioned responses, with each item intheseries serving asastimulus foralater

response, either a simultaneous conditioned response forthenearest item, oras intrusions, which would have tobesuppressed andwould give risetoinhibition ofdelay. Hull(1935) spunthisideaintoanelaborate deductive model and showed thatanumber oftypical phenomena ofrote learning could beaccounted for. Although theideas and reasoning were clever, they somehow didnotconvey afeeling ofwhat oneactually does ina learning situation. However, Hull was insistent ontheimportance ofextending analogies withconditioning toother learning situations. Two other conditioning phenomena, generalization anddifferentiation, appealed tomeasreasonable candidates, fortunately. Generalization

conditioned trace responses forlater ones. The latter would lead toanticipatory

Journal ofExperimental Psychology, 1939,24,237253.

26

E. J. Gibson

Theoretical contributio to the problem of learn hav rec b characterize by a new series of attemp at linkin the fac of co tionalleamin experime with the condit resp . The me o these systematic studies has been to use the empi law of the co -

Sensory Generalization withVoluntaryReactions 27

28

E. J. Gibson

This is a reaction time experime . That means that whe a ce signal occurs ,signal you are to react as quickly as you can apa response .The will be avibratory stimulu on the (lwith owe ,upp ,rig ,

Procedure The following instructions were given thesubject toread atthebegin ning oftheexperiment .

SensoryGeneralizationwith Voluntary Reactions

29

orleft) partofyourbackwhichwillbedemonstrated to youin aminute , and theresponse willconsist ofspeaking OUTintothisvoicekey.Whenever you feelthevibrator , sayOUTimmediately . Thenoiseof theapparatus starting will serve asa readysignal soyoucangetprepared to respond in a hurry . Youwillbegivensome practise trialsbefore yourreaction timeisrecorded . Youareto respond onlytothedemonstrated stimulus . Laterin theexperi menttherewill beotherextraneous stimuli . These stimuliwill be other vibrators located in different places onyourback . Youarenotto respond to anyexcept thedemonstrated stimulus onthe- partof yourback . Experiments onreaction timeshowthatyoucangofaster if youkeep your mindontheresponse youareto make , sokeepamotorattitude , thatis, be prepared to sayOUTtheinstant thisvibrator stimulates you. Whenthesubject finished reading theinstructions , thevibrator to whichhewas to respond wasdemonstrated to him.Thisvibrator wasalways atoneendof the series (extreme right,extreme left, top, orbottom ). Then50trialswiththestimulus to whichhehadbeen instructed to respond (the"practised " stimulus ) weregiven . Reactions weretaken atarateof aboutonein 20seconds . At thebeginning of a trial, theexperimenter started thedrum(thesoundof whichserved asa ready signal ), a variable interval of fromk to 21seconds intervened , thestimulus occurred , the subject responded , the drum was stopped , and the chronoscope reading taken bytheexperimenter . A practise series of 50 trialswasgivenfirstwith thedesignated stimulus . Then100moretrialsweregivenduringwhichthe3 "prohibited " stimuliwere interspersed atrandom intervals among trialswiththedesignated stimulus . Each of the3 prohibited stimulioccurred 9 timesduringthe100trials . Theorderof theprohibited stimuliwasvaried witheach subject according to a prearranged plansothatorderofoccurrence didnotfavoranyone stimulus when a groupof subjects wasconsidered together . Thisprecaution wastakenin case thereshould beatendency forfalseresponses to decrease withpractise .4A group of 9 subjects wasrequired to balance theorder . Eighteen subjects (GroupI) were runwiththevibrators placed vertically up anddowntheback(halfwith the practised stimulus atthetopoftheseries , theotherhalfwiththepractised stimulus atthebottom ); 18moresubjects (GroupII) wererunwiththevibrators arranged horizontally across theback(again halfpractised ateach extreme ). The subjects were all students at Yale University and were paidfor their serVIces .

Controls Following theexperiment proper , thesubjects weregiventwo briefseries oftests , onetocheck ontherelative intensity ofthestimuli , and theothertocheck ontheirlocalizability . Theintensity ofthevibrators was controlled mechanically in sofaraspossible , in thatthevibrators , their wiring , andmounting wereidentical ; asafurther precaution , each subject wasputthrough a series of judgments of comparative intensity of the stimuli , incase there should beatendency formore false responses tooccur to a stimulus of stronger intensity andthereby distortthecurve of fre-

Table .-Group 3 1 Perc ~ of Tim Jud Str tha St . ' A & ' . & . v ' .Nea Ne F Stim St ~ u lu ~ I~p A 22 % 49 % 3 % B 88 54 7 II %86 %6~

30

E. J. Gibson

a

Sensory Generalization with Voluntary Reactions

31

other point occu . Whe the vibr we dis up an d the back , more error of iden occ . Po 0 1 w confu 14 times out 108 cha ( 1 pe ) Ag , 3 0 n with point 2 and 3 . It is quit in kee wi th fa o s sensit that the vibra plac ver sho be ha to lo , since spatia thres are high whe stim are ap in a lo direct on skin than whe they are app tra . It is c howe , that any gene foun in Gro II , at lea , ca ha b due to mista local of stim . The main question of the experim was wheth or not gen was more apt to occur in proport as the prohi stim app imated spatially the practise one .This quest was ans by the

Figure 3 . 1 Frequency of response to vibrators as a function of proximity to practised point ( G I , roup vibrators placed vertically ) . Frequency of response in percentage is represe on the ordinate ,and stimulated points in order of proximity on the abscissa . .LN3:> ~ 3dNI3SNOdS3 ~ ::fOA :>N3nO3 ~ .::f

STIMULATEDPOINTS

Results

relative frequency of respons to prohib stimu .The res ar best presented graphica .In Fig .the 3.1 are prese com for

32

E. J. Gibson

l.N3:J 'd3dNI3SNOdS3'd ::IOA:JN3nO3'd::l

the 18 subjectswho had the seriesof vibrators placed up and down the back. The practised stimulus is designatedas 0, and the others are designated 1, 2, and 3 in order of proximity to it, regardlessof whether the 0 stimulus was at top or bottom. The 0 stimulus is almost invariably respondedto (98 percent). Falseresponsesare madeto the closestprohibited vibrator 25 percent of the time, to the next closest14 percent of the time, and to the farthest 9 percent of the time. In general, the curve shows a continuous gradient from point 0 to point 3, in accordancewith the expectation that a processanalogousto generalizationof conditioned responses occurswith voluntary responses .? The differencesbetween the frequencies at the four points have very satisfactory critical ratios (44.2 to 2.67), with the exception of the differencebetween point 2 and point 3 which has a critical ratio of only 1.48.8 The continuous downward trend, however, is unquestionable. Furthermore, when the data for the two sub-groups of 9 subjectseachare plotted separately, the resulting curvesareboth similar to the one above, showing a continuous slope from 0 to 3. The results for the 18 subjectswho had the vibrators placedacrossthe back are given in Fig. 3.2. Here the slope from point 0 to point 2 resembles 100

80

60

40

20

0

0

1 2 STIMULATED POINTS

3

Figure3.2 Frequency of responses to vibrators asa functionof proximityto practised point(GroupII, vibrators placed horizontally ).

Sensory Generalization with Voluntary Reactions

33

closelytheslopein Fig. 1, but theupturnat point3 is not typicalof the usualcurveof generalization . It mustbe remembered , however , thatthe vibratorshereweredistributedhorizontally , andpoints0 and 3 were consequently symmetrical . Perhaps thesymmetrical locationis responsible for thehigherdegree of generalization at point3 thanat point2. Anrep(1) reportsthatsymmetry is aneffective cause of generalization ; hefound,in fact(with dogsassubjects ), thatstimulation of thepointsymmetrical with the0 pointwasequallyaseffective asstimulation of the0 pointitself.The difference in thisgroupbetween points2 and3 hasa statistical reliability of only 1.16 timesits standarderror. But whenthe datafor the two subgroups areplottedseparately , they resemble the curvedrawnfrom combined data,showinganupturnat3. It seems unlikelythatthisisa mere chance resemblance . It isnotable , also,thatin Fig. 3.1, wheretherewereno symmetrical points , no suchupturnoccurred . It is interesting to compare percentages of generalization at point 1 in GroupsI and II keepingin mind the fact that stimuli0 and 1 were accurately localized by GroupII butnotinvariably by GroupI. It mightbe expected on common sense groundsthatlessaccurate localization should accompany a greaternumberof falseresponses , but suchis not thecase . ForGroupII, thefrequency at point1 is 36percent , whilein GroupI it is only 25 percent ; morefalseresponses occurred whenthevibratorswere easilyidentified . Generalization of responses , therefore , canhardlybeexplainedby theinabilityof thesubject to identifythecorrectstimulus . A seconddifference is that betweenresponse to stimulation at the 0 pointsthemselves in the two cases . Figure3.1 showsa slightlylower percentage of response at this point thandoesFig. 3.2 (98 percentas compared with 100percent ).9 Againfrequency of response waslowered whenabsolute localization waspoorer . Thisfactsuggests thatthedifficulty of discriminating points0 and1 in GroupI induceda stronginhibitory attitudein thesubjects - i.e., caused themto reactwith greater"caution ." If thiswasthecase , reactiontimeshouldbeincreased . An examination of thetimesreveals thatit was.Takingthefirst50practise trialsasa control , andcomparing with themall thetrialswith the0 stimulus aftertheprohi bitedstimulihadbeen introduced , anaverage increase in latencyis foundin thelattercaseof 63ms. in GroupI, and42ms. in GroupII. It is significant thatthegreateraverage increase in latencyoccurs in GroupI. Theinvariable increase in latencyafterintroduction of prohibited stimuli is interesting . It is related , for onething, to the difference alwaysfound betweensimpleanddiscrimination reactions in thereactiontimeexperi ment.Furthermore , it suggests theanalogous situation with Pavlov 's dogs whena previously differentiated stimulus is introduced amongtrialswith a conditioned stimulus . An inhibitoryprocess occurswhich'spreads ' or leaves anafter-effectservingto depress theensuing conditioned responses .

34

E. J. Gibson

It should bepointed outthatthepossible influence of fatigue hasnot beencontrolled in theabove comparison ofthefirst50reactions withlater reactions . It is mostimprobable , however , thattheincrease in latency is evenpartially dueto fatigue . Foronething,practise effects wouldnormally continue after50reactions andbalance anyfatigue ; furthermore , theeffects offatigue in thereaction timeexperiment areveryslight(see4, p. 46). Thereaction timeswereanalyzed to seewhether thegradient of frequency of response to thefourstimuliwasaccompanied by a correlative gradient of speed . Superficial consideration mightleadto theexpectation of a negative correlation between frequency andlatency , sincepositive evidence of suchacorrelation in thecase of simple conditioned responses isavailable (3, 9). Butsucha correlation hasneverbeendemonstrated for generalized responses in the conditioning situation . Furthermore , in the present situation two factors arepresent whichcutacross theexpected correlation - in fact,workin opposition to it. In thefirstplace , thereis voluntary inhibitionof long-latencyfalseresponses . Theeffectof such inhibitionwouldpresumably be to allowonlythefastest potential responses to occur . Sincethe subject doesnot inhibitresponses to the designated or practised stimulus whentheyareslow , noselection of fast responses wouldoccurforthisstimulus . Simply averaging thetimesof all responses madeto eachof thefourstimuliwould , therefore , showan artificially decreased latency of response to theprohibited stimuli ; whereas thesuperficial prediction wasthatlatency of theseresponses should be greater thanthoseto thedesignated stimulus , since frequency of response islower . Table3.2 shows theaverage latency of response to thefourpointsof stimulation in bothgroups . Theaverage latency at point0 is calculated fromall theresponses to thatstimulus aftertheintroduction of thefalse stimuli (it willberemembered thatreaction timesforthepreceding practise responses wereshorter ). Theresults showthattheaverage latencies byno means increase frompoint0 to point3, astheyshouldif a negative correlation between frequency andlatency exists ; onthecontrary , thereis a decrease in latency asthestimulus is displaced fromthepractised point. Ourassumption of voluntary inhibition of long-latency falseresponses is therefore probably correct .

Table 3 . 2 Average Latency in Ms ~vt "labc L.a~ll'-L 'yof VI .Response I.~ ""t'.,& .,-to -the --4 -Vibrators -~ - 1-Vibrator -,-. 2 -~ 7.1,--:3 Vibrator 0 Vibrator Vibra --Group I ( V ertical distribution ) 233 219 149 151 Group II ( H orizontal distribution ) 170 161 106 122 ~:uup 11 \1J.V ~I.LV.L\.U "'I-& ."..,_.& _& "/

SensoryGeneralizationwith Voluntary Reactions

35

The second complicating factor is the slight tendency for the subjects of Group I to confusepoints 0 and 1. It has already been pointed out that frequencyof responseat thesepoints was lower than in Group II, probably becauseof "caution" or a strong inhibitory attitude. It is as if the subject had to make a discrimination

reaction between

two confusable

stimuli ; and ,

asis well known to be the case,! 0 the reaction time lengthenedaccordingly. The average latency at points 0 and I for Group I may be seen to be abnormally high in comparisonwith the rest of the figures. The first of these two complicationsmay possibly be eliminated by the following procedure. For eachpoint, the times of only the fastest6 percent of the total number of possiblereactionsmay be averaged. The hypothesis is that the whole distribution of possible responsesshould be slightly faster, the greater the frequency of reaction produced by the stimulus (or the nearer the stimulus to the designated point). By taking the top 6 percent of the potential responses , we include only reactionswhich were fast enough to slip by (above the threshold of voluntary inhibition), and avoid the difficulty . There is no way of avoiding the influence of the secondcomplicating factor in presentingthe results.

8 o -

-

-

-

GROUP

I

GROUP

II

Figure 3.3

Average latency of the fastest 6 percent of responsesat the four stimulated points .

36

E. J. Gibson

Figure 3.3shows thecurves oflatency ofresponse to thefourstimuli

whenonlythefastest6 percent ofthepotential responses areincluded in

eachcase.CurvesforbothGroupsareplottedon the sameaxes.If there

werea perfectinverse relationship between frequency andlatency, the

curvesshouldshowa constantupwardslope.A glanceat themdoesnot indicatemuchtendencyfor an increasein latencyto accompany spatial

displacement fromthe0 point,asdoesdecreasing frequency. Butifpoints 0 and1forGroup I arediscounted (forthereasons givenabove), therewill be seento be a roughupwardtrend.A perfectslope,evenasidefrom the two discounted points,couldhardlybe expectedconsidering that6

percent ofthepotential reactions inthecaseofthethreeprohibited stimuli includes only10responses ateachpoint.Therefore, although it cannot be concluded that speedis correlated withfrequency underthe presentcircumstances, it maybe suggestedthat sucha trendis not entirelyabsent whenthecomplications introduced by difficult localization andvoluntary inhibition

are considered.

Discussion

Severalreferenceshavealreadybeenmadeto the reactiontimeliterature

pointing outparallels between thepresent experiment andthediscrimination reactiontime experiment.Someother factsfromthis sourceseemto

be relevant.Thefalse reaction,aneventgenerallyfoundbut so faraswe know never studied, would seem to be understandable in many cases as an

example of generalization. Withreference to somerecentattemptsto condition voluntaryresponses in a reactiontimeset-up,it hasbeensug-

gestedbyGibson (5)thatwhatweretakento beconditioned voluntary reactionsweremorelikelygeneralized reactions. Thefactthatgeneralized

responses havebeendemonstrated to occur ina voluntary situation lends plausibility to this suggestion.

Furthermore, phenomena reportedin studiesoftheeffectofa distracting

stimulus onspeedofreaction to a givenstimulus canbeinterpreted inthe lightof ourapproach. Evans(4)foundthatdistraction wasparticularly effective whenthedistracting stimulus belonged to thesamesensedepart-

mentas the stimulusto be respondedto. Thatis,if the distractingstimulus

wasanintermittent flashoflight,whilethedesignated stimulus wasa flash

oflightofdifferent intensity coming froma slightly different place, the

increasein reactiontimeswas greaterthan if the distractingstimulus had been a tone or a touch.Our data likewiseshow that reactiontime

increases significantly whenotherstimuliof the samedimension as the practised stimulus areintroduced, especially whenthenewstimulus isnot easilydiscriminated fromthe practised one.it maybe tentatively suggestedthatin suchsituations a typeof inhibition is setup comparable

SensoryGeneralization withVoluntaryReactions 37

totheinhibitory after-effects reported byPavlov (17,p.125) andbyAnrep (1)whena differentiated stimulus isintroduced amongtrialswitha condi-

tioned stimulus. Butsuchapossibility mustbeexplored inanexperiment

especially designedforthepurpose.

Theresultsof thisexperiment havea certainbearing on therelation

between conditioned andvoluntary responses. Incommon withtheexperi-

mentsspecifically directed atthestudyofthisrelationship, theresultsshow

bothsimilarities anddifferences. Generalization andthegradient ofgeneralization appearunderbothconditions. Buttheinstructions whichwere

necessarily introduced inthevoluntarysituation apparently produce a setorattitudeofcaution whichhastheeffectofshortening reaction times ofthefewresponses madeto thefalsestimuli. It is a question whether

generalized responses intheconditioning situation wouldshowanincrease

ora decrease. Itseems possible thatthelatencies ofgeneralized responses

wouldbelonger whilethediscrimination isbeingbuiltup,andshorter, asin thepresent situation, whenthefrequency of generalized responses has

beenloweredby differential reinforcement. Thatis,establishment ofdifferentiationwouldinhibitfalseresponsesin mostcases;the few whichoc-

curred mightbefasterthannormal responses, asinthepresent situation. Suchcasesmightbe producedby disinhibition, for instance.Butinves-

tigation, ratherthanspeculation, isobviously called forbythisproblem. Anewlineofexperimentation issuggested bytheprobability thatfalse reactions to a prohibited stimulus donotoccursimply because thesubject cannottellthedifference between thisstimulus anda designated one. Similarities of a superthreshold natureappearto playa part.Theexact natureoftherelationship between generalization andperceptual similarity andidentity isanunanswered problem, andonewhich should yieldto a technique combining psychophysical methods andmethods designed for the studyof conditioned responses. Summary

Subjects wereinstructed to respond verbally asquickly aspossible to a

designated vibratory stimulus, butnotto respond in thiswayto other stimuli. Practise wasgiveninresponding to thedesignated stimulus. Later othervibratory stimuli, distributed either longitudinally ortransversely on thesubjects back,wereintroduced, to seewhether generalized orfalse responses wouldoccurandwhether theywouldbe morefrequent, the nearerthe stimulus to thedesignated oneon the skin.Thefollowing conclusionsmay be drawn.

1. Frequency of response to theprohibited stimuli showed a gra-

dientofgeneralization analogous to thegradient foundin thecondi-

38

E. J. Gibson tioned response experiment, when the vibrators were distributed longitudinally. 2. When the vibrators were distributed transversely, an upturn at the end of the curve of response resulted. This upturn was probably causedby the symmetrical relations of the point stimulated and the designatedpoint. 3. Introduction of prohibited stimuli apparently producedan increase in latency of responseto the designatedstimulus. This increasewas particularly noticeable when one of the prohibited stimuli was not easily discriminablefrom the designatedstimulus. 4. The averagelatency of responseto the prohibited stimuli was less than that to the designated stimulus, probably becausethe longlatency falseresponseshad been inhibited. 5. When the differential effectsof voluntary inhibition were avoided, there was some indication that speedwas positively correlatedwith frequencyof response. 6. With referenceto the problem of the relation between conditioned and voluntary responses , the present results suggest that similarities exist, but that differencesdue to the instructions introduced in the voluntary situation are to be expected.

Notes

I.The writer wishes toexpress gratitude toProfessor Clark L.Hull forencouragemen and advice during theprogress ofthis research .The experiment isone ofaseries ofstudies presented totheFaculty oftheGraduate School ofYale University inpartial fulfillment oftherequirements forthedegree ofDoctor ofPhilosophy inPsychology . 2.This experiment was reported before theEastern Branch oftheAmerican Psychologic Association in1937 (6). 3.Generalization gradients have been demonstrated withthisstimulus dimension by Anrep (1)and byBass and Hull (2). 4.Such atendency didoccur .About 70percent ofallfalse responses came during thefirst half ofthese 100 trials . 5.Mvers gives thetwo -point intheregion ofthespine as5.4cmwhen the ." (16 ,)'-' - threshold stimuli are distributed longitudinally . 6.Thedesignation 0 isused torefer tothestimulated point which thesubject was instructed torespond to. The numbers 1, 2, and 3refer respectively tothevibrator nearest ,next ,and farthest from that point . 7.Forasimilar negatively accelerated curve ofgeneralization intheconditioned response situation see Hovland (10 ). 8.Reliabilities were calculated with theformula given byYule (22 ,p.269 )forthestandard error ofthedifference between twoproportions . 9.Although small , thedifference isstatistically reliable , itscritical ratio being 3.67. 10 .Henmon (8) found that thesmaller thedifference between twostimuli , thelonger the reaction time tothedifference .

SensoryGeneralization withVoluntaryReactions 39 References

1.Anrep, G.V.,Theirradiation ofconditioned reflexes, Proc. Roy. Soc., 1923, 94,Series B, 404425.

2.Bass, M.J.andHull, C.L.,Irradiation ofa tactile conditioned reflex inman, J.Comp. Psycho!.,1934, 17, 4765. 3.Campbell, A.A.andHilgard, E.R.,Individual differences inease ofconditioning,]. Exper. Psycho).,1936, 19, 561571. 4.Evans, J.E.,Theeffect ofdistraction onreaction-time, withspecial reference topractice andthetransfer oftraining, Arch. ofPsycho!., 1916, 37,pp.106. 5.Gibson, J.J.,Anoteontheconditioning ofvoluntary reactions, J.Exper. Psycho!., 1936, 19, 397399. 6.Gibson, E. J.,Sensory irradiation withvoluntary responses, Psycho!. Bull., 1937, 34, 5115 12. 7.Gibson, E.J.,ASystematic Application oftheConcepts ofGeneralization andDifferentiation to Verbal Learning, Dissertation, YaleUniversity, 1938.

8.Henmon, V.A.C.,TheTime ofPerception asa Measure ofDifferences inSensations, New York: Science Press, 1906.

9.Hilgard, E.R.andMarquis, D.G.,Acquisition, extinction, andretention ofconditioned lidresponses to lightindogs,J. Comp. Psycho!., 1935,19,2958.

10.Hovland, C.I.,Thegeneralization ofconditioned responses: I.Thesensory generalization ofconditioned responses withvarying frequencies of[one,J.Gen. Psycho!., 1937, 17, 125148.

11.Hull, C.L.,Thegoalgradient hypothesis andmazelearning, Psycho!. Rev., 1932, 39, 2543. 12.

, Theconcept ofthehabit-family hierarchy andmazelearning, Psycho!. Rev., 134152.

1934, 41, 3352; 13.

491516.

, Theconflicting psychologies oflearningawayout,Psycho!. Rev., 1935, 42,

14.Lepley, W.M.,Atheory ofserial learning andforgetting based uponconditioned reflex principles,Psycho!. Rev.,1932,39, 279288.

15.Marquis, D.G.andPorter, J.M.,Differential factors inconditioned voluntary and conditioned involuntary responses. Psycho!. Bull., 1937, 34,p.772. 16.Myers, C.S.,AText-Book ofExperimental Psychology, Cambridge: University Press, 1928, pp. xiv + 344.

17.Pavlov, I. P.,Conditioned Reflexes. Translated andedited byG.V.Anrep, Oxford UniversityPress,1927,pp. xv + 430. 18.Peak, H.andDeese, L.,Experimental extinction ofverbal material, J.Exper. Psycho!., 1937, 20, 244261. 19.Spence, K.W.,Thenature ofdiscrimination learning inanimals, Psychol. Rev., 1936, 43, 427449. 20.

, Thedifferential response inanimals tostimuli varying within a single dimen21.Stephens, J.M.,Theconditioned reflex astheexplanation ofhabitformation. 111. The sion, Psycho!.Rev.,1937, 44, 430444.

operation oftwohigher orderreactions inclose succession,]. Exper. Psycho!., 1936, 19, 7790.

22.Yule, G.Udny, AnIntroduction totheTheory ofStatistics, London: Charles Griffin andCo., Ltd.,1922,pp. xv + 415.

A Systematic Application oftheConcepts of Generalization andDifferentiation to Verbal

Learning EleanorJ. Gibson

The centerpiece ofmydissertation wasthisattempt tousetheconcepts ofgeneralization anddifferentiation todeduce anumber ofphenomena characteristically found inpaired associate learning studies ofthetime, such astransfer, interference, andretroactive inhibition. Toplease Hull, I used concepts indicating phenomena

found inconditioned response learning situations, butasI thought ofthem, the concepts actually referred toperceptual aspects ofthelearning process. However,

onedidnt speak much ofperception inpublic inthose days, sothere isa kind of

surreptitious meaning smuggled in.I wanted toemphasize structural features of items tobelearned, rather thansheer association. Perceptual differentiation, preliminary toassociation with another unit, whatever itmight be,was promising as theprocess thatmust beinvolved. Theterm predifferen ia ion evolved inthe process ofconstructing themodel andpulling outdeductions. Itbecame apopular term andwasthefocus ofconsiderable research (forexample, Arnoult 1953).

Inpreparing thispaper, I tried tofollow Hullslead(Hull 1935) presentingdefinitions, assumptions (called postulates, inemulation ofa geometrical proof), andthen deductions thatcould betested experimentally, with thecomplete argument given. Itnowseems tomesomewhat labored andpretentious, asdoes Hulls work. Spelling everything out, with explicit definitions andpredictions, has itsmerits. Even so,incomplete definitions thatallow unvoiced assumptions tocreep incanoccur, although itisremarkable how often they gounremarked. Ifind afew

nowinmyownmodel andwonder whynoonepointed themoutat thetime.

This miniature system, aswell asHulls miniature andmore major ones, are longsince passØ formostpsychologists, butit seems thatthatdoesnt include

everyone who isconcerned with models ofbehavior orcognition. Afewyears ago Igave a seminar attheUniversity ofPennsylvania whose participants included

promotors ofArtificial Intelligence androboticists. I askedthemto read,as

historical background, thepaper that follows andasample ofHulls work. Tomy

astonishment, they considered ita revelation, exciting andpromising ! Psychological Review, 1940,47, 196229.

42

E. J. Gibson

Introduction

Arecent trendinexperimental andtheoretical psychology hasbeenthe

attempt tosystematize thefacts ofagiven field onthebasis ofempirical principles derived fromstudyoftheconditioned response (14,15,21,39, 16).1 Itisreasonable toexpect thatthesame general lawsholdfromone learning tasktoanother, insofarasthesituations aresimilar, andit is possible thatthegreater thesimplicity ofthelearning, thebetter arethe chances oflaying baremechanisms which operate inalllearning. Itisnecessary, however, toguard against thechance thattheprinciples chosen forsystematic exploitation areartifacts of theexperimental situationfromwhich theywerederived, andhavenoimportance asgeneral

characteristics oflearning. Hullhassuggested a procedure fortesting the valueof a theorybuilton suchconcepts (14).Theessentials of such a testarethedevelopment ofa clear-cut hypothesis, exploration ofthe hypothesis forallitsderivatives, andexperimental testofsuchderivatives ashavereceived nodirectsubstantiation or disqualification. Thetheory muststandorfallwiththeexperimental evidence fororagainst thepredictions which it makes.

Thefollowing account isanattempt todevelop atheory relating various factsofverbal learning (such astransfer, interference, certain intra-list effects andretroactive inhibition), totwoexperimentally defined characteristics of theconditioned responsegeneralization anddifferential inhibition. The

theory willbeexplored fordeductions ofknown facts aswell asforimplications yettobetested, since ithasbeen undertaken principally inthehope

ofsystematizing many ofthefactsalready known. Thattransfer, interference, andretroactive inhibition arerelated haslongbeensuspected, asthe so-called transfertheoryofretroactive inhibition implies (2,p. 427if.).

Thehypothesis tobepresented islikewise classifiable asatransfertheory, butit aimstomake asexplicit aspossible themechanisms assumed tobe responsible forthetransfer, which isitself aphenomenon aspoorly understood as retroactive inhibition. Thehypothesis willfirstbestated inabrief, informal fashion. Then, for thebenefitofthereaderwhowishesto makea morescrupulous examina-

tion,thetermsusedwillbeexplained, assumptions stated, andarguments for variouspredictionspresentedin detail. Statementof the Hypothesis

Thehypothesis asserts thata major necessity ofverbal learning is the establishment ofdiscrimination among theitemstobelearned, andthatthis

process ofdiscriminating isactually a fundamental partofwhatiscalled generally thelearning process. If nodiscrimination between theitems

VerbalLearning 43

already exists, thentheearly partofthelearning process willseeanincrease discrimination. Learning timeshould beata maximum inthiscase. Ifsuch discrimination already exists, learning timeshould beata minimum. Posi-

inthetendency toconfuse theitems, followed bythedevelopment of

tivetransfer willoccurin situations wherethenatureofa second task

permits discrimination acquired inaprevious task tobebeneficial. Negative transfer willoccur when generalization witha previous taskoccurs, but where thesituation issuch thatdiscrimination between some aspect ofthe twotasksthemselves isrequired, aswellaslearning ofthesecond. Retroactive inhibition willoccur, similarly, ifa second taskgeneralizes withone

already learned, andifthesituation issuchthatdiscrimination between

someaspectof the twotasksmustbe produced beforethe first:canbe

recalledadequately.

Thereader familiar withthefacts ofconditioned responses willseeimtion anddifferentiation. 2 Thehypothesis will bestated asitapplies toverbal learning bythepaired associates method. Itisassumed thatwhen a listis being learned bythismethod, generalization mayoccur between thevariousstimulus items, sothata response learned toonetendstooccur, asa response toother stimulus items inthelistalso. Figure 4.1represents alist

mediately thepossibility ofstating these assumptions interms ofgeneraliza-

inwhich generalization isoccurring. Thedotted lines represent generalizationtendencies andthesolid arrows represent connections with theright responses, which aretobelearned. Insuch a listasthis,where thegeneralizing stimulus items allhave different responses, those responses occurringortending tooccur byvirtue ofgeneralization willblock theright

responses inproportion tothestrength ofthetendency togeneralization. Inordertoreduce thestrength ofthegeneralizing tendencies (toincrease

thedifferentiation), differential reinforcement through practice mustbe applied. Itisassumed, thatvarying degrees ofgeneralization mayoccur between oneitem andothers inalist, andconsequently thatlists may vary

asto theaverage strength ofgeneralization occurring in thelist.The Sa \ >-__ .,

\

Sb (

,/ /%.

\)(/ Rb /

Figure 4.1

K

\

\

44

F. J. Gibson

List1

List2

List1 R

Sa

Ra

Sc

* R

,.

Sa

Rb Sb

b

Sd .

_-ł .Rd

Sb ..-

łRb

Figure 4.2

necessity forgreater differentiation willthenserveto makethelearning moredifficult inlistswherea highdegreeofgeneralization occursthanin oneswherelessgeneralization occurs; andit willalsobe responsible for

poorer retention, since thedifferentiation, especially withlower degrees of

learning, decreases overa periodof time,allowing spontaneous recovery ofthegeneralization. Iftheabovehypothesis iscorrect, thegeneralization in eithercaseshouldbe apparentin actualconfusion of responses.

Furthermore, in anytwo-list situations suchas theonesusedto study

interferenceand retroactiveinhibition,it is assumedthat inter-listgen-

eralization mayoccur. Whena firstlistisfollowed bya second onewhich

includes stimulus itemssimilar to onesin the first,stimulus itemsin the secondlist willtendto produceresponsesfromthe first;andmembers likewise whenlist 1 is againpresentedforrecallandrelearning, response

fromlist2 willtendto recur.Figure4.2represents responses fromlist1

tending tooccur bygeneralization inlist2,andonesfromlist2 occurring bygeneralization when list1isagain tested. Ineachcase, thegeneralizin tendencies willblocktherightexcitatory tendencies inproportion to the

strength ofthegeneralization. Again, thegeneralization should beevident inactualconfusion ofresponsesinthiscase,errorsofreversionto the wrong list.

Thedegree oflearning ofanylistcannot bepredicted fromthenumber ofpractice repetitions alone, according tothishypothesis, since theamount ofdifferential reinforcement orpractice required to reducegeneralization to

a givenstrength increases withthedegree ofgeneralization between stimulusitems.Theextentof differentiation achieved as a resultof a given amountofdifferential reinforcement willvary,then,withthisfactor.More-

over,theextentofdifferentiation mustbecalculated alsointermsoftime elapsing sincedifferential reinforcement hasbeendiscontinued, sincespon-

taneousrecovery willeventually occur.Pavlovhasshownthatsponta-

neousrecovery occurs moreslowly, thefarther differential inhibition has beencarried 33,p. 123).Temporal intervals between learning andrecall thusassume importance inthelightofthehypothesis, andtogether with

VerbalLearning 45 Sa

St\V1 ,4 a P\ Rb St i I, I \

t \RC Rd

Figure 4.3

degree oflearning, mustbetaken intoaccount inpredicting inter-list Thehypothesis hasbeenstated interms oflistslearned bythemethod ofpaired associates, butthesame logic willapply throughout to lists learned byanymethod which allows astimulus-response analysis. Figure 4.3represents a listbeing learned bytheanticipation method. Itema functions onlyasa stimulus andmust produce thenextitem, Ra,asits response. Butwhenthisitemactually appears, it functions as a stimulus. Thedottedlinesrepresent generalizing tendencies within thelist.The hypothesis cannot besoeasily applied toother verbal learning situations, effects.

sinceinmostofthemthesubjects performance isnotsocontrolled asto

allow a detailed S-Ranalysis. Such would bethecasewiththemethod of

complete presentation, forinstance. However, there seems nogood reason

to suppose thatthehypothesis cannot holdwithsucha method. In the sections whichfollow, evidence willbequoted without comment whenit

hasbeenobtained bythepaired associates oranticipation method, but

where any method hasbeen employed, thisfactwill bespecifically referred to. other Theroleoftheresponses indetermining easeoflearning a single list orintransfer situations isnotspecifically delineated bythepresent hypothesis. Itisconceivable thatsomething analogous togeneralization of stimulus items might occur among response items, especially since every

response isina sense a stimulus aswell. Mullers substitution hypothesis develops thispossibility (30). According toMuller, activesubstitution wassaidtooccur when ideaawasconnected withideab,andideaA,which

wassimilar toa,alsoproduced b.This description follows thepattern of

sensory generalization whichhasbeendescribed here.ButMulleralso

described passive substitution; Bmight besubstituted forb,when Bwas

similar tobbutnotassociated with a.This, interms ofthepresent hypoth-

46

E. J. Gibson

esis,wouldamount to generalization of response items,if it occurred. Confusion errorswouldresultineithercase,anda priorimightbe dueto

eithertypeofgeneralization. ButrecentworkofThorndike (41)suggests thatthestimulus itemsof a listarechiefly effective in producing such errors.Hefoundthatstimulus members of a verylonglistevokedre-

sponses made toother stimuli like them nearly twice asoften astheyevoked connectedwithunlikestimuli.Buta stimulusmemberevokeda responses

word which was like the responseconnected with it little or no oftener than it evoked unlike responses.

Anexperiment byMcGeoch andMcGeoch (28)seems alsosignificant in

thisconnection. In a retroactive inhibition experiment, stimulus members ofcorresponding pairedassociates ina listI anda list2 weresynonymous; or response members weresynonymous; or bothmembers weresynonymous;or neitherwere.Whenthe stimulusmembers weresynonymous, therewasa consistent increase inretroactive inhibition overthecondition whereneithermember wassynonymous, butthiswasnotthecasewhen response

membersweresynonymous. Furthermore, whenboth members

weresynonymous, therewasnoconsistent increase inretroactive inhibitionascompared withthecondition wherestimulus members onlywere synonymous. Onthebasisoftheseresults, thepresent hypothesis will confineitselfto theexploitation oftheconceptof sensorygeneralization as originally stated. Relation to Other Systems

Therehasbeeninthepastat leastoneratherambitious attemptto systematizesomeof the factswhicharebeingconsidered, andtherearecon-

temporary theories which applyto thesamerealm. Theearlier attempt referred to is thatof G.E.Muller andhisstudents. MullerandPilzecker (30), besides presenting anenormous massofexperimental databased on a thorough-going exploitation oftheTreffermethode, posited a number of kindsof inhibition (e.g.,associative inhibition, retroactive inhibition) to

explain theirresults. Their substitution hypothesis, which rosefroman

exhaustive analysis oferrors, ina senseforeshadows thepresent theory. Mullerstheories, naturally, roseoutof hisresults; andthoughhistermi-

nology haslefta lasting impression ontheliterature, it seems antiquate today, because thefoundation laidbytheseearlier investigations makes itpossible todevelop a morecomprehensive theory which mayevensug-

gestthepsychological processes ormechanisms underlying histypes of inhibition.

Asystematic attempt toexplain various factsofmemory andlearning is

being made at present byGestalt psychologists, andhasgiven riseto experimental work, notably byKhlerandhisstudents (20, 33,32,35). The

VerbalLearning 47

essence ofthetheory isthattheGestalt laws ofspatial organization hold organization, theGestalt psychologists predict such effects asdifficulty in learning homogeneous series, retroactive inhibition, andproactive inhibition.Forinstance, thelawofsimilarity inspatial organization isassumed toholdformemory, andisdemonstrated ina series ofexperiments by

alsointhefield ofmemory; byanalogy withtheir laws ofperceptual

vonRestorff (35),inwhich retention ofhomogeneous seriesofitemsis

compared with retention ofheterogeneous series. Anitem inaheteroge-

neousseriesis saidto beina betterposition to beretained thanthesame

iteminanentirely homogeneous series, since bythelawofsimilarity there

isaggregation ofthetraces ofthehomogeneous items, thereby causing any single item toloseitsidentity. Also, anisolated item inoneseries may loseitssuperiority ifa second series follows which iscomposed ofitems similar toit(retroactive inhibition), while anisolated iteminasecond series willhavenoadvantage iftheseries preceding it ismade upofsimilar

items(proactive inhibition). Thetheoryis characterized by Koffka as a dynamic tracetheoryandhasbeenelaborated inhisbook(19).It could probably berestated intermsofdiscrimination, andinthissenseissimilar to the theorypresented in thispaper.Thechiefdifference liesin the concepts on which the two theories are based, and the methods bywhich they are developed. Athirdtheory, advanced byJames (17)toexplain interference, restslike

thepresent theory onPavlovian concepts. Histheory differs considerably

fromthepresent one,however, inthatthekeyconcept isexternal inhibi-

tion,ratherthandifferential inhibition. Thetheoryis extended to cover retroactive inhibition, butitisunfortunately notdeveloped totheextent of

predicting experimental results orspecific tests. Noexperimental work seems tohavefollowed it.Thetheory could consistently supplement the onesuggestedin thispaper. Thepresent-day associationists haveassembled a number of laws

which bear some relation tothecentral notions ofthepresent hypothesis. InRobinsons book (36), a Law ofAssimilation isincluded which isvery similar tothenotion ofgeneralization. Itisstated asfollows: Whenever an associative connection issoestablished thatanactivity, A,becomes capable ofinstigating anactivity, B,activities other than Aalsoundergo anincrease ordecrease intheircapacity toinstigate B. Thelawisextended tocover assimilation ofinstigated processes aswell. TheLaw ofAcquaintance, suggested inthesame work, issuperficially similar toaprinciple ofdiscrimination.Butthelawapparently means byacquaintance mere isolated repetitionofagiven item; acquaintance isrelevant tothepresent theory only insofarasitproduces discrimination from other items. Mere acquaintance doesnotseem toincrease efficiency oflearning (Waters, 42).Much closer tothenotion ofdiscriminability ofstimulus items isThorndikes Law of

48

E. J. Gibson

Identifiability , which states that IIconnections are easy to form in proportion as the situation is identifiable , distinguishable from others . . . " (40 , p . 87 ). In general , these laws seem to name results , rather than to form a coherent and interrelated system .

Explanation of Terms The following explained

terms are crucial to the hypothesis

in some detail . Illustrations

and are consequently

will be presented when necessary .

Generalization : thetendency fora response Ralearned to 5ato occur when Sb(withwhich it hasnotbeen previously associated ) is presented .3 A generalization gradient is saidto beformed when a number ofstimulus i~ems show decreasing degrees ofgeneralizatio witha given standard stimulus . Thehypothesis need make noassumption asto thetypeof stimulus continuum which willyielda generalization gradient , butit isconsistent withit to suppose that such agradient willbeyielded byagroup ofstimuli which canbe arranged along anydimension orscale withrespect tothepresence of some discriminable quality oraspect - inother words , stimuli which would beconsidered to varyin degree of similarity . Experiment evidence proves thatnotonlysimple dimensions such aspitchand intensity (11,12)yieldgeneralization gradients . Yum(44)found that when nonsense syllables , words , orvisual patterns were stimuli ina learning series , different butsimilar stimulus items would in a test series elicittheresponses learned , thestrength ofthetendency vary ingwiththedegree of similarity between testitemandoriginal stimulus item ; andGulliksen (10)found thatsimilarity oftestfigures to a training figure correlated significantly withtendency to give thetraining figure 'sresponse .Razran (34)has recently demonstrat generalization onthebasis ofsynonymity ofwords . 2. Multiple generalization : thetendency forresponses which arebeing learned to members ofa series of stimulus items to occur asgen eralized responses tomembers other than their"right " associate . The situation employed inYum 'sexperiment (44) wassuch astoyield multiple generalization . Gulliksen 'sexperiment (10)provides atransi tionbetween simple generalization andmultiple generalization , since a choice between twoantagonistic responses , each of which had been learned toastandard stimulus , was possible . 3. Right response : theresponse paired withaparticular stimulus item inalistpresented bythepaired associates method , ortherespons immediately following aparticular stimulus iteminalistpresented by theanticipation method , 1.

Verbal Learning

49

4. Learning : a list is said to be completely learned when the right response item is given in each case upon presentation of the stimulus

item. A given degreeof learningor criterion of learning meansthat a certainpercentageof right responsesis reproducedupon presentation

of the stimulus items .

5. Differentiation : a progressive decreasein generalizationas a result of reinforced practice with Sa-... Ra and unreinforced presentation of Sb.

6. Reinforcement : a processwhich occursduring verbal learning when a subjectseesa responseas he anticipatedit and thinks "that's right ." Differential reinforcement designatesthe situation wherein a right response is reinforced and a generalized response is not reinforced .

7. Excitatorytendency : tendency for a particular stimulus to evoke a particular responsein a degreegreater than zero. 8. Meaning : a characteristic of a verbal or visual item which serves to differentiate

it from other items .

Statementof Postulates

A number of postulates are basic to the hypothesis and should be made evident at the outset. In some casestheseprinciples have been empirically demonstratedin the conditioned responsesituation; in other casesthey are incapable of empirical demonstration, and must rest on confirmation of the relationshipswhich they predict in combination. 1. When lists are learned by a paired associates or anticipation meth od , S-R connections

are set up between

certain of the items (between

members of a pair in paired associates, and one and the next in the anticipation method ).

2. If there is a right excitatory tendency and also a generalizedone between the same5 and R, the resultant excitatory tendency will be stronger than either one alone; if a right excitatory tendency and a generalized one are aroused by the same 5 but lead to different R's,

they will interfere, the weaker tending to block the stronger in proportion to the strength of the weaker. 3. Stimulus items which generalizewhen presentedfor learning in a single list will also do so when they are presented in the form of two

lists, and with the samerelativedegreeof generalization. 4. Generalizationwill increaseto a maximum or peakduring the early stages of practice with a list , after which it will decrease as practice is continued . (This assumption has been confirmed by the author in an experiment to be reported , 7.)

50

E. J. Gibson 5 . In list -learning , all decrease in generalization reinforcement .4 6 . The amount tendency

of reinforcement

to a given strength

required

is due to differential

to reduce a generalizing

will increase with increasing

strength

of

the generalizing tendency . (This postulate is an " empirical " one , in the sense that it has often been demonstrated in the conditioned response situation , 33 .) 7. After the cessation of practice , differentiation

will decrease over a

period of time , leading to an increase (" spontaneous recovery " ) of generalization . (This , again , is an " empirical " postulate , 13 , 33 .) 8 . Spontaneous recovery will occur more slowly , the more differ ential reinforcement has been applied . (Demonstrated in the condi tioned response situation , 33 .) 9 . If differentiation has been set up among items , it will be easier to differentiate

a number

of stimulus

them again later , even though

they are paired with different overt responses than those learned when the original differentiation was set up . (This assumption might be called " transfer of differentiation ." ) 10 . If differentiation has been set up among a number of stimulus items , there will be less tendency for them to generalize with new stimulus items or for new stimulus items to generalize with them , the decrease in generalization being proportional to the amount of differ ential reinforcement given . (Analogous to this are the cases reported in Pavlov (33 ), of dogs who , having once learned to differentiate a circle from an ellipse , found it easier to differentiate

between

the circle

and other ellipses of different ratio . A common sense analogy would be the case of a child who , after at first generalizing all the animals he sees, learns that cats are not dogs ; after this , he does not have equal trouble

in learning

that rabbits are not dogs . He would

see every animal in the world animals .)

in order to differentiate

not have to

dogs from other

Some Propositions that Follow from the Hypothesis A number of propositions

which may be predicted

from the hypothesis

just

presented will be stated . These will not be formally derived in the strictest logical sense, but for each one the argument will be indicated as clearly as possible , with reference to the definitions and postulates on which it de pends . Examples will follow the argument when any clarification seems necessary , and evidence for or against the proposition will be reported . In some cases, the propositions are familiar facts of verbal learning . For the convenience of the reader , the propositions will be given titles and grouped under traditional

headings .

VerbalLearning

51

Similarity

Similarity willbeconsidered, inthepresent instance, asthatrelationship

betweenstimulusitemswhichcanbe indicatedandmeasuredin termsof theirtendency togeneralize. A priori,thisis a reasonable criterion of similarity, sincea groupof stimuliamongwhichgeneralization occurswould,

according to usualstandards, beconsidered similar (e.g.,a seriesoftones,

or ofspotsalongthefore-leg). Yum(44)foundthatjudgesratings of perceptual similarity ofvisual patterns andsynonyms to theirrespective standards coincided withthedegree towhich theforms orsynonyms were

capable ofproducing a response originally learnedto thestandard. Furthermore,it seemsobviousthata common wayofspeaking ofsimilarity is in termsofdiscriminability, or tendency notto generalize.

Acomplication interminology isintroduced bythefactthatgeneralizationwillvarywithotherfactors thanoriginalgeneralization tendency. Ashasalready beenpointed out,itwillvaryalsowithdegree oflearning andwiththetimeintervalsincelearning wasleftoff.Thesetwofactorsare

ina sense secondary determiners ofdegree ofgeneralization, andoriginal generalization tendency willbeusedtherefore to denote thedegree of generalization apparent before differentiation hasbeensetupthrough

differential reinforcement. Undertheheading ofsimilarity,then,willbe

included propositions which haveastheirmajorvariable thedegree of

originalgeneralizationof stimulusitems.

I. Intra-ListTransferand ConceptFormation

Ifmultiple generalization occurs during thelearning ofa list,andifthelist isconstituted sothatthegeneralizing stimulus items arepaired withthe

sameresponses, learning shouldbeeasierin proportion to thedegree of

generalization.

Theargument forthisproposition would beasfollows: Suppose thata lististobelearned inwhich multiple generalization mayoccur, andsupposethatthosestimulus itemswhich aresimilar (i.e.,generalize toa high

degree)actuallyhavethe sameresponses to be learnedto them.The

generalizing excitatory tendency andtherightexcitatory tendency, inthis case,willcoincide. According topostulate 2,ifthereisa rightexcitatory tendency anda generalized onebetweenthesameS andR,theresultant

excitatory tendency willbestronger thaneither onealone. Now,according

to the definition of generalization, generalization mayexistin different

degrees, sothattheaggregate generalization ina listmayvaryfromlow tohigh.Then, inthepresent case,thehigher thedegree ofgeneralization

inthelist,thestronger willbetheresultant excitatory tendencies afterthe sameamountofpractice, andtheeasierwillbethelearning.

52

E. J. Gibson

Sc"' .... ~ Rc ....... ............... ><.. ....""'" .... Rc Sd ,""'" Figure 4.4 An

example

arrows and

of

indicate both

items

of

generalization easier the

list

,

related

to

are has

II . Inter

items

4 .4 . Sc

,

same

about

employed

learning

generalize

the

degree

; of

response

, the

situation

studying

is

concept

concept

formation

since

both

depend

lists

where

little

concept

broken

Sd

this

in

called

control

the

higher the

thing

process

As

and

is

upon

differ

-

predic

-

generaliz

-

generalization

experiments

, the

a second

list , and

above

.

3 )

1

a first

that

of

the

and

with that

. The

items

the

greater

the

same

do

degree

of

the

to

be 1

they

in

the as

2 . It

the

degree

a list

2

of

the

is

in

for the

generalization

two

lists

form . Now

have will

easier

so

assumed

presented

presented

proposition ,

by

list when

the

above

generalization

if

learn

followed to

are

degree

items for

to

.

generalize

relative

given

easier

list

when

generalizing

reasoning

is from

which so

be

increases

associates

also

the

2 will

lists

occurs

will

to

, list

two

paired

stimulus list

list

responses

generalization

a single ,

from

same

the

same

then

list

2

, is

given

apply

will

be

, to

. An

4 . 5 . The 1 .

the

a list

further

learn

the

.

. The

having

important

in

Fig items

response items

commonly

verbal

between

that

lists

the

. The

that

occurs

have

that

in

responses

of

included

generalization

( postulate

suppose

same

one

checked

in , and

Transfer

Suppose

two

the

. Since

usually been

constituted

learning

the

of

generalization

of

and

is

presented

generalize

groups

learn

suggests

not

- List

ing

of

to

generalization

not

If

have

process

and

is Sb

the

be

which the

entiation

tion

will

list

and

a pair

arrangement

formation

occurs

a Sa

between

the

that

such

, items

But

transfer

example

of

broken since should

this

situation

arrows the

in

list

generalizing result

. It

, obviously 2

represent and

must

not

the be

a

transfer

generalizing right forgotten

one

responses that

in

tendencies coincide internal

from ,

positive

generalization

Fig list

.

Verbal Learning List 1

Sa

Sb

53

List 2 "" ""

, Ra "" ""

. Ra

Sc -'

... Rb

""~ Rb ,..,."", "" Sd""" ~ Rb

~ Ra

Figure 4.5

will also occur to some extent in both lists ; but it should not interfere

with

the transfer, since a control group learning merely list 2 would have the same internal generalization to contend with , and would only lack the beneficial generalization from list 1. No evidence obtained under these precise conditions is available, but an experiment of Bruce's (3) is relevant . He found positive transfer when the responses of two lists were the same , and the stimulus items very similar (two letters in common ). Less transfer occurred when stimulus items had no letters

in common

.

III . Retraactive Facilitation

If generalization occursbetweentwo lists, and if the generalizing items have

the sameresponses , recallar relearningaf list 1 will befacilitatedas the degreeafgeneralization betweenthe two lists increases . The situation is essentially that described in the above proposition ,

except that retention of list I is to be tested following the learning of list 2. The argument above will apply. This situation might be dubbed retroactive facilitation (Fig. 4.6). Retention of list I should be facilitated increasingly as generalization between the two lists increases , since learning list 2 should strengthen proportionally the excitatory tendenciesalready developedin learning list I . A retroaction experiment by Bunch and Winston (4) contained one condition in which response syllables from the first list were retained intact in the second, while the stimulus syllables were different . Inhibition , rather

than facilitation, resulted during recall and relearning of list 1. This evidenceis apparently not in harmony with the presenttheory. But degreeof generalization of stimulus members was not varied or specified. A further

test should be madewhere the stimuli are quantifiedaccordingto generalizing tendency, so that several degreesof inter-list generalizationmay be employed. Relative degree of facilitation would be expectedto follow the above prediction .

54

E. J. Gibson List 1

List 2

List 1

"

........ R a

.;f'"' R a

...... ......

. "

"

Sa

-. Ra

Sc . /

." Ra

Sa """"""

., Ra

Rb "

...". R b .......

.....-

......

"

Sb -

~ Rb

Sd ""' "

." Rb

Sb ....."""

.. Rb

Figure 4.6

IV . Intra-List Interference or Difficulty of LearningHomogeneous Lists If multiplegeneralization occursduring the learningof a list, and if the list is constituted so that the generalizing items have different responses , an

increasing numberof repetitionswill berequiredto reacha givencriterionof learning as the strength of the generalizing tendenciesincreases .5

Supposethat a list in which multiple generalization may occur is so constituted that the generalizing stimulus items have different responses . The generalizingexcitatory tendency, and the right excitatory tendency, in this situation, will conflict (see Fig. 4.1). Two casesrequiring slightly different arguments may result .

Case1: If, in the above situation, the generalizing tendency is stronger than the right excitatory tendencyafter a few repetitions, it is obvious that the generalizing tendency will have to be reducedin strength, sincelearning is definedas giving the right response., But a decreasein generalization must be obtained through differential reinforcement(postulate 5), and the amount of reinforcementrequired to reduce any generalizing tendency to a given strength will increasewith increasingstrength of the generalizing tendency (postulate 6). Therefore, in this case, an increasing number of repetitions will be requiredto reacha criterion suchas completelearning as the strength of the generalizingtendency increases . Case2:

If, in the same situation, the right excitatory tendency is the

stronger tendency after a few repetitions , it is nevertheless weakened in

proportion to the strength of the generalizingtendency (postulate 2). By postulates5 and 6, it is clear that an increasingnumber of repetitions will in this casealso be required to reach a given criterion of learning, as the strength of the generalizingtendency increases . In a list of the sort described, every right excitatory tendency will be interferedwith more or less, accordingto either case1 or case2. Therefore, putting thesetwo casestogether, it may be said that an increasingnumber

VerbalLearning

55

ofrepetitions ofthelistwillberequired asthestrength ofthegeneralizing

tendencies within it increases.

Thisprediction ischecked indetailinanexperiment donebythewriter (7),in whichthelistscompared arequantified according to degree of generalization. Theevidence is conclusively positive. In so faras the generalizing tendencies areactually stronger thantherightexcitatory tendencies andabovethreshold strength, overterrorsofconfusion should

occur during thelearning. These didoccur intheexperiment reported, and witha muchhigher frequency inthehigher-generalization series. V. Retentionof Homogeneous Lists

Difficulty inrecalling orrelearning a listwillbeproportional tothedegree of original generalization, if time has been allowed for spontaneous recovery to occur. In a situationsimilarto the one describedabove,the differentiation

which wasestablished during learning willdecrease overa period oftime, leading tospontaneous recovery ofthegeneralization (postulate 7).The generalizing tendencies willagaininterfere withtherightexcitatory ten-

dencies, inproportion to theirstrength (postulate 2),sothatretention will beimpaired inproportion tothedegree oforiginal generalization. Positiveevidencefor thispredictionmaybe foundalsoin the writers

experiment (7).Retention ofhigh-generalization listswaspoorerthanretentionoflow-generalization listsafteroneday,eventhoughtheformer listsweregivenmorethantwiceasmanylearning repetitions. VI. Learningof Isolated Items

A stimulus-response pairwhichis a member ofa listcontaining other

stimulus items having a strong tendency togeneralize withit willrequire morerepetitions tobelearned thanwould thesamepairasa member ofa listwhose stimulus items havea lowtendency togeneralize withit.

This prediction isaspecial casecovered bytheargument under proposi-

tionIV.It hasbeenstatedseparately, because thesituation asputinthe aboveprediction wasthesubject ofanexperiment byvonRestorff (35).

Shepresented tosubjects series ofpaired associates which contained ahigh

proportion ofhomogeneous itemsanda smallproportion ofitemsofother types.Itemswerealwaysremembered betterwhentheywereisolated

thanwhentheyweremembers ofa homogeneous groupwithin thelist;the listwasgiventhe samenumberof repetitions in the two cases.The prediction wasconfirmed inanexperiment ofthewriters (7)exactly asitis statedin thisproposition.

56

E. J. Gibson List 1

List 2

Ra

Sa

Ra

Sc __ R

Sb

*Rb

Sd Rd _.

Figure 4.7

VII.Inter-ListInterference as a Functionof Similarity

Morerepetitions willbe required to learna secondlist,in proportion

to thestrength ofthetendency foritemsofa firstlisttogeneralize with the itemsof the secondlist.

Suppose thatgeneralization occurs froma firstlistto a second list,as provided forinpostulate 3,andasillustrated inFig.4.7.Astheselistsare constituted, therightresponse to anystimulus willconflict withanygen-

eralized response, sincetheresponses arealldifferent. Now,as list2 is beinglearned, responses fromlistI willtendto occurto stimulus itemsin

list 2 by generalization. According to postulate2, the rightexcitatory tendency willbeweakened inproportion to thestrength ofthegeneraliz-

ingtendencies. Generalization canbedecreased onlythrough differential reinforcement (postulate 5) andtheamountof reinforcement required to

reduce anygeneralizing tendency to a givenstrength willincrease with increasing strength ofthegeneralizing tendency (postulate 6).Therefore, as strength ofgeneralization fromthefirstlistto thesecond listincreases, morerepetitionswillbe requiredto learnlist2.

Internal generalization willalsooccuraslist2 isbeinglearned, butthe prediction asmadeabovewillnotbeaffected, sincethisfactormayremain constantas generalization fromlist 1 to list 2 is increased. Thesituation

describedis the one knownvariouslyas proactiveinhibition,interference,

or negative transfer. Naturally, errorsof reversion to listI shouldbe apparent during thelearning oflist2.Anexperiment donebythewriter (8) demonstrated that list 2 requiredmoretrialsto learnas degreeof generalization betweenthestimulus itemsof thetwolistsincreased. Also,the expectedtype of error occurred. VIII.Interference Measuredby Recall

A secondlistwillbemorepoorlyrecalled aftera periodoftime,inproportion

tothestrength ofthetendency foritems oflistI togeneralize withit.

VerbalLearning

List1

Recall

List2

List2

Ra

Sa

p Ra

Sc

,

-* Rc

7.

Sc

57

of

Ra

R ,Rb

Sb

.Rb

Sd 7

.Rd

Rd

Figure 4.8

Suppose thatgeneralization occursfroma firstlistto a secondlistas in

theabove case, butthatlist2hasbeenlearned tosome criterion (Fig. 4.8). Ithasbeendemonstrated intheabove argument thatlist2 willrequire

moretrialsto learnas generalization fromlist 1 to list 2 increases. As

learning goeson,differentiation between theitemsisestablished. Butifan

interval oftimeisallowed to elapse afterpractice hasbeendiscontinued, spontaneous recovery ofthegeneralization willoccur (postulate 7),sothat

recall should beimpaired inproportion tothestrength ofthegeneraliza-

tion from list 1 to list 2.

Thisprediction amountsto sayingthat interference as a functionof generalization canbemeasured equally aswellby a recalltestoflist2 as bya learning measure. Whitely andBlankenship (43)usedthismeasure to demonstrate interference inrecallofa second listwhenthematerial learned

consisted ofmonosyllabic wordspreceded inthefirstlistbyothersuch

words orbynonsense syllables. Themethod ofcomplete presentation was beenobtained hadgreater inter-list generalization beenpresent (e.g., had employed. Probably an evenhigherdegreeof interference wouldhave

bothlistsbeenmadeupofnonsense syllables). IX.Retroactive Inhibition andSimilarity

Afirstlistwillbemorepoorly recalled as thestrength ofthetendency

for itemsofa secondlisttogeneralize withit increases.

Suppose thata firstlisthasbeenlearned, andhasbeenfollowed bya second listwhosestimulus itemspotentially generalize withlist1.It is

assumed thatlist2hasbeenlearned tosome criterion. Nowsuppose thata recall oflistI isasked for.When thestimulus items oflist1 areagain presented responses fromlist2 willtendtooccur bygeneralization (see

Fig.4.2).According to postulate 2, therightexcitatory tendencies willbe

weakened inproportion to thestrength ofthegeneralizing tendencies.

Therefore, therecall oflist1should suffer inproportion tothestrength of thetendency foritemsoflist2 to generalize withit. 6

58

E. J. Gibson

~ :J,S!I 6u!uJeal u! A:J,ln:>!! !!p J.o aaJ6aa

The situation describedis of coursethe one which resultsin "retroactive inhibition." It is an acceptedfact that retroactive inhibition is a function of similarity of original and interpolated tasks (see Britt, 2, p. 389 ff.). However, degreeof similarity hasusually beenestimatedon an a priori basis. An experiment by the writer (8) has checkedthe prediction that a generalization gradient will correspondwith a gradient of retroactive inhibition. Degreeof Learning X. Interference and Degreeof Learning Difficulty of learninga secondlist will vary with the degreeof learningof the precedinglist, increasingto a peak and then decreasing as degreeof learningof list 1 increases .

Supposethat two lists, the items of which tend to generalizewith one

another , are to be learned in immediate succession. The tendency for items

in the first list to generalizewill increaseto a maximum during the early stages of practice of the list , but from this point on, the tendency to

generalizewill decrease (postulate 4). Now a secondlist will be harder to

learnasthe strengthof the tendencyfor itemsof list 1 to generalize with it increases(proposition VII ), so that difficulty of learning list 2 should

A

B Degreeof learning of list 1

Figure4.9

c

59

Verbal Learning increase tion

with

. But

begins list

, and 1,

list

the

2 , the

A

will

theoretical

of

list

the

list . In

the

due

to

curve

is

erated

to

check

of

this

Retroactive

and

The or

second been

was

of

equal

list

- list

by

the

to

B and of

the

, since

to

the

later

from

the

increasing decrease

B to

,

C . The accel

-

' s ( 7 ) demon

during

the

that on

a positively author

the

-

progress

experiments

in the

stimulus

members

generalizing

of

degree

practice

due

fall

in

decrease

assumes of

B , and

up

amount

of

prediction

to

-

with

2 will

2 as a function

generalization

than

of

field in

. Melton

this

of

learning

above prediction

since

point

of of

should ; and had

McQueen

this

- Irwin

, are

to

be

to be tested of

relearning

learning list

. of

1 should

generalization

2 increases

situation

been

. The

occur there

be

, argu

of

not

drawn

-

( 29 ) definitely

. But

find

time in

(23 ) has

with of the

-

any con -

that

interpolated

learning

earlier

on

indication

increase of

come

, and

. McGeoch is some

degree

degree

will

learned

cannot

did the

then

X .

1 has

retroaction

items

maximum list

and .

1 is then

if degree

and

curve

upheld

at which

of list , and

in

list

theoretical

been

point

and

the

a maximum 2 increases

stimulus

constant

to

well

to list

recall

generalization

have ,

that

in proposition

predicted of

increase

interpolated

in recalling

2 up

how

Learning

generalizing

degree

an accurate

the

1 will

of an

, and

given

on

higher

list

of Interpolated

1 is kept

list

maximum

half

beyond

a

inhibition

as that

might

carried

learning

intra

of Degree

as the

same

first

half

A

set

the

list

difficulty A

from

to

cessation

in

from

is available

for

of

inflection

the

after

been generalize

learning

list

experiment

lists , having

practice

, so that

case , the finned

for

succession

depending

intervals

rise

, rather

, retroactive

point

later

soon

an

of learning

two

decrease

is the

a sharp

prediction

learning

with

then

ment

learning

rise

to

proportional of

generaliza

, differentiation

has

members

, is represented

C , since

inhibition

that

of

its

Fig . 4 .9 . The

the

point

differentiation

increase

by

sort

as degree

2 is varied

increase

early

identical

in immediate

If degree list

, the

as a Function

decrease Suppose

in in

maximum

.

XI . Retroaction

learned

difficulty

is begun

this

used

if

of this

for

list

B to of

-

point

1 passes

being

plotted

show

from

( 18 , 3 ) have the two lists

apply

differentiation

strated a curve of learning . No

of

1 is

curve

drawn

drop

the

list

tendency

is represented

increasing

to

of

generalization

second

generalization

1 up

. Therefore , the difficulty list 1 is carried further .

curve of

will

less

in

list

learning

10

given in

learning

first

of

be

decrease

learning

of

stage

postulate

there

reinforcement as differentiation

of

practice

as the

the

learning degree lists

of

1 and

inflection

2 .

60

E. J. Gibson

XII.Retroaction as a Functionof Degreeof OriginalLearning

Retroactive inhibition fora listI willdecrease asdegree oforiginal learning ofthelistisincreased beyond thepointofmaximum generalization.

Suppose thattwolists,having generalizing stimulus items, areto be

learnedin immediate succession andthatrecalloflist1 is thento be tested;

degree oflearning oflist2 istobekeptconstant while degree oflearning oflist1 isvaried.Now,theinitialincrease ingeneralization aslist1 is given

morepractice willcause nocorresponding increase inretroactive inhibition in thissituation,becausethe effective generalization causedby thisincrease willbe fromlist 1 to list 2 ratherthan fromlist 2 to list 1 (i.e.,wrong

responses willtendincreasingly tooccur inlist2,butlist1willnotbecome moresusceptible togeneralization ofresponses fromlist2,because degree

oflearning inthatlistremains constant). Ontheotherhand,differentiation in list 1 willincreasewithdegreeof learningafterthe peakof generalization has beenreached(postulate4);and if differentiation has been set up

among a number ofstimulus items(inlist1,here), therewillbelesstendency fornewstimulus items (from list2,here)togeneralize withthem, the

decrease in generalization beingproportional to theamountof reinforce-

mentgiven(postulate 10).Since alsoretroactive inhibition isproportional to thestrength ofthetendency formembers oflist2 to generalize with members oflist1 (proposition LX), it follows thatretroactive inhibition will

decrease as degreeof learning of list1 is carried beyondthepeakof generalization.

McGeoch(22)has shownthat degreeof retroactiveinhibitionvaries

inversely withdegreeoflearning ofthefirstlist,whenthelistsaremade upofnonsense syllables andlearned bytheanticipation method. Length of Interval

Manypredictions canbemadewithreference totheeffect ofvarying the lengthoftheinterval between learning andrecall ofa list,orbetween a firstlistanda second.Onlya fewof thepredictions willbe included here

asexamples. 7 These predictions depend principally onthenotion ofspon-

taneousrecoveryof generalization. But sincethe rate of spontaneous

recovery willvarywiththeamount of differential reinforcement which hasbeenapplied, thepredictions mustalways be statedin termsoftwo variableslength oftheinterval andamount ofreinforcement or,roughly, degree of learning.

XIII.Interference asa Function ofLength ofInterval Between Tasks

Theinterval afterwhicha list2 mustbeintroduced inordertoobtainthe maximum interference inlearning it willvarywiththedegree oflearning of

Verbal Learning

61

list1,being zero ifdegree oflearning has been carried only as farasthe peak of generalization or below , and thereafter occurring the later , the higher the degree oflearning oflist1. Suppose thattwolists aretobelearned , and thatthere ispotential generalization oftheitems ofthetwolists . Now , iflearning oflistI is carried only tothe point ofmaximum generalization , orbelow ,maximum interference with list2willoccur immediately ,since there can inthis case be norecovery ofgeneralization inlistI,and since ,according toproposition VII,interference willbeproportional tothe degree ofgeneralization . Ontheother hand , suppose that learning oflist1iscarried beyond the peak ofgeneralization .Differentiation willincrease aspractice iscontinued beyond this point (postulate 4), and differentiation willprotect theitems from generalizing withanew list(postulate 10 ) inproportion tothe amount ofdifferential reinforcement given . Butspontaneous recovery of thegeneralization willeventually occur (postulate 7). Itwilloccur more slowly , themore differential reinforcement has been applied (postulate 8). So , the farther differential reinforcement has been carried , the later the maximum in~erference willoccur . Figure 4.10shows thetheoretical relationship .Maximum interference in learning list2willbeobtained ifitisintroduced atonce ,when the first list has been learned only tothe peak ofgeneralization orbelow ; thecurve of interference would probably falloff , then , astheinterval increases . But if - - - List learned to peak of generalization or below -

Figure4.10

List learned to high degree

62

E. J. Gibson

list 1 hasbeenlearnedto a high degree , little interference would occurat once, but ratherwould be postponeduntil spontaneous recoveryof the generalization occurred ; it might thenfalloff again.Therateof riseandfall represented in the curvesis perfectlyarbitrary, sinceit dependson the rate at whichspontaneous recoveryoccursandthe rateat whichgeneralization tendencies areforgotten, andquantitativevaluesfor thesefactorscanonly be determinedexperimentally . No thoroughtest of thesepredictionsexists, but an experimentof Lepley's (21) indicatesthat, with a high degree of learning,interference will increaseto a maximum , andthen decrease as the intervalbetweenlist 1 andlist 2 increases . Thisresultis consistentwith the secondhalf of the aboveprediction.Underthe conditionsof Lepley's experiment , the peakof interferencecarneat about 3 hours. In a more recentexperimentby Bunchand McCraven(5), no tendencywas found for the deQree '-' of transferto changewith varying length of interval, althoughsyllablelists were learnedto a high degree ; but no testswere madebetweenthe zero intervaland 48 hours, so that thereis really no conflictwith Lepley's results.8 XIV. Retroaction asa Function of Intervalbetween Tasks Theintervalafterwhichmaximum retroactive inhibitionwill occurfor a list 1 will vary with thedegree of learningof list 2, beingzeroif degree of learningof list 2 hasbeencarriedonlyasfar asthepeakof generalization or below , andthereafter occurring thelater, thehigherthedegree of learning of list 2. Supposethat two lists are to be learnedin immediatesuccession , and that thereis potentialgeneralization betweenthe two lists; andthat retention of list 1 is to betestedaftera variableinterval. If thedegreeof learning of list 2 is low (reachingthe peakof generalization or below) maximum retroactiveinhibitionshouldoccurimmediately , by the argumentin propositionXIII above. But if the degreeof learningof list 2 is high, maximum retroactiveinhibitionshouldoccurlater, againfollowing the argumentin propositionXIII. A similarpredictioncanbe madeif the degreeof learningof list 1 is varied, insteadof list 2. No test of the temporalcourseof retroactive inhibitionasa functionof the degreeof learningof the two lists hasbeen made. But McGeoch(24, 25) andMcGeochandMcKinney(26, 27) have investigatedthe courseof inhibitoryeffectswith time, andhavenot found a consistenttendencyfor retroactionto vary with time. Two of their experiments gavesomeindicationof anincrease of retroactionwith lengtheningof the interval, oneshowedno tendencyfor retroactiveinhibitionto vary uniformlyin any direction, anda fourth showeda slighttendencyfor

VerbalLearning 63

retroactive inhibition to decrease withtime.Different materials, learning methods, anddegrees oflearning were employed inthese experiments. The

apparentinconsistency of the resultspointsto the need for a further experiment in whichtemporalcourseof retroactiveinhibitionis studiedin relationto degreeof learning.

On the basisof this proposition, and assuming a constantinterval

between learning listI andrecalling it,itispossible tomake predictions for

thewell-known point ofinterpolation problem, withretroactive inhibitionasa function ofbothinterval anddegreeoflearning (9).Results from

a number ofexperiments onthisproblem aresuperficially conflicting (2, 399If.;38).Highdegrees ofretroactive inhibition havebeenobtained by

oneinvestigator oranother withthesecond listinterpolated atalmost any

pointintheinterval between learning andrecall oflistI.Thediscrepancy maybe due,among otherthings, to thevariable degrees of learning employed indifferent experiments onpointofinterpolation. Theexpectation willvary,depending onwhether learning iscarried toa lowora high degree. MeaningfulversusNonsenseMaterial

Thisgroup ofpredictions isrelated totherestofthehypothesis through thedefinition of meaning whichwehaveaccepted. Meaning hasbeen

definedas onecharacteristic of a verbalor visualitemwhichservesto

differentiate it fromotheritemsin otherwords, generalization is at a

minimum formeaningful material, unless ifoccurs ona secondary basis,as

in thecaseofsynonyms (34).Proposition )(Vis included herebecause the

pattern oftheargument issimilar tothatinthepredictions actually dealing with meaningfulmaterial. XV.Learningof Pre-differentiated Items

Ifdifferentiation hasbeen setupwithin a list,lessgeneralization willoccur

inlearning a newlistwhich includes thesame stimulus items paired with

different responses; andthetrialsrequired tolearnthenewlistwilltendto

bereduced byreduction oftheinternal generalization. Suppose thata list1hasbeenlearned, andisfollowed bya list2 whose

stimulus items arethesame asthoseinlistI, although theresponses are different. Lesstotalintra-list generalization should occur inlearning list2 thaninlistI, because ifdifferentiation hasbeensetupamong a number of stimulus items,it willbe easierto differentiate themagainlater,even

though theyarepaired withdifferent responses (postulate 9).Furthermore, since thenumber oftrials required tolearn a listincreases asthedegree of

E. J. Gibson

64

List 2

List 1

/

Ra

Sa.L ::::~_ .. .. Rx Rb

Sb L. . Ry /

Rc

Sc~:~ :__........ Rz Figure 4.11

generalization

2

within

should

tend

A

clear

because

as

ing

cannot

new

factor

reduced

2

;

.

but

as

only

list

1

to

impossible

in

two

the

list

would

The

first

cases

lists

.

)

XVI

half

of

A

of

in

be

was

verified

in

represented

absent

an

is

of

within

to

use

the

stimulus

in

It

actually

weights

tradi

the

the

-

.

comparing

)

.

be

, "

by

as

which

generalization

same

errors

found

as

be

would

the

2

list

responses

group

inhibition

to

can

errors

learned

quantitative

associative

list

in

(

of

( confusion

such

before

control

knowing

as

-

represented

trials

reduction

2

generaliz

would

learning

responses

prediction

-

A

conflict

of

II

.

list

first

number

and

experiment

of

second

the

writer

'

s

.

.

Ease

of

Learning

Meaningful

trials

nonsense

will

Suppose

two

of

,

meaning

one

a

are

list

syllables

is

to

than

lists

one

nonsense

8

Lists

required

,

that

items

are

be

syllables

stimulus

tion

list

without

learned

original

tendencies

-

effect

generalization

reduction

first

2

,

in

inhibited

their

list

prediction

present

are

list

of

against

the

overt

More

list

inter

in

above

is

the

generalizing

but

,

of

the

,

shows

in

facilitating

pitted

of

successfully

number

conflict

. 11

excite

the

,

represented

be

for

half

tendencies

to

have

The

been

generalization

.

responses

4

represented

whether

case

Figure

stimuli

with

.

name

of

8

the

second

of

have

of

trials

generalizing

are

would

say

factors

tional

the

contend

to

the

(

2

the

.

no

decrease

learning

conflict

these

ones

list

)

reduced

;

,

new

list

conflict

1

for

the

learned

list

list

for

-

the

of

made

inter

Therefore

tendencies

well

in

-

,

number

be

,

in

undertaken

increases

the

intra

tendencies

is

list

reduce

case

a

well

the

to

to

are

,

characteristic

in

learn

a

which

be

in

are

learned

meaningful

or

list

they

a

verbal

visual

items

words

in

those

of

According

item

that

the

to

which

are

. 9

only

while

.

or

stimulus

vary

,

meaningless

of

the

meaningful

which

words

relatively

which

serves

the

other

defini

-

to

Verbal Learning

65

differentiate it fromotheritems . Furthermore , if stimulus itemshaveonce beendifferentiated , it will beeasier to differentiate themagainlater , even thoughtheyarepaired withdifferent responses thanthoselearned when theoriginaldifferentiation wassetup(postulate 9). Therefore , moregeneralization will tendto occurin thelist whosestimulus itemsareless meaningful ; andsinceasthedegree of generalization withina list is increased , moretrialsarerequired to learnthelist (proposition IV), more trialswillberequired to learnthelistwhose stimulus itemsarelessmean ingful . Thefactthatalistofmeaningful wordsiseasier to learnthanonemade upof nonsense syllables isnotorious , andit seems certain thata check of thesituation described above wouldyieldpositive results . XVII. Interference andMeaningfulness More trials will be requiredto learna secondlist whenthe stimulus itemsof bothfirst andsecond list arenonsense syllables , thanwhenthey aremeaningful words .

Suppose thata second list is to belearned followinga firstlist, andthat easeof learninglist 2 is to becompared whenstimulus itemsof bothlists arenonsense syllables , or relativelymeaningless , withthecasewhenstimulusitemsof botharemeaningful words.In theformercase , generalization fromlist 1 to list 2 shouldbe greaterthanin the latter, sincemeaning serves to differentiate theitems . And sincemoretrialswill berequired to learna second list in proportionto thestrengthof thetendency for items of afirstlistto generalize withit, morelearning trialswill berequired when thestimulus itemsarerelativelymeaningless . In a checkof this prediction , it wouldbe necessary to preventsuch factorsas overlearning in the nonsense syllablesituationfrom cutting across themainvariable(relativemeaningfulness ) andconfusing theissue . If learningwerecarriedto a criterion , manymoretrialswouldpresumably begiventhefirstlistwhenstimulus itemsweremeaningless , thusintroduc ing a second variable , andincidentally tendingto evenup thetwo situa tionsas regardsdegreeof differentiation reached in the first list. This difficultymightbeavoidedby delayingpresentation of list 2 for a length of timesufficient for spontaneous recoveryto occur . Also, comparisons shouldin eachcasebemadewith controlgroupswhichhavelearned only list 2, but are otherwiseidenticalwith the experimental situationsdescribed , sincelist2 wouldpresumably beharderto learnwhenstimuliwere nonsense syllables , evenin theabsence of list 1. Degree of relativenegative transferwouldbe calculated separately for the two situations , andthen compared .

E. J. Gibson

66

XVIII. Retroactionand Meaningfulness Retroactive inhibition will be greater for a first list when the stimulus

itemsof both first and secondlist are nonsense syllablesthan when they are meaningfulwords. Supposethat retention of a first list is to be tested after interpolation of a second list in two

cases: one , when

stimulus

items

of both

lists are

nonsensesyllables; and two, when stimulus items of both are meaningful words. More retroactive inhibition would be expectedin the first case, by an argument similar to the one above . Again , a number of variables other than the crucial one will tend to cut

acrossthe experimental situation,andit would be absolutelyessentialin a test of the predictionto provide control groupsin which interpolated learning was omitted , so that actual percentages of retroactive inhibition could be calculated for the two cases. In an experiment by Sisson (37),

retention was comparedfor list 1 when both lists were madeof syllablesof 100 percent associative value with a case in which both contained only

syllables of 0 percent associativevalue. Presumably, a higher associative value is correlated with greater meaningfulnessof the item. Sissonfound slightly poorer retention in the first case; but percentagesof retroactive inhibition could not be calculated, sinceno control groups were provided. More learning presentations were given 0 per cent lists, since a learning criterion was used. Time intervals between tasks were brief . This experi -

ment should be repeatedwith adequatecontrol groups, since it has certainly been generally assumedby investigators that the surestand highest percentages of retroaction are to be found with nonsense material .! 0

PossibleExtensions of theHypothesis It has been possible to present only a sample of the predictions and implications of the hypothesis advanced. Many extensionsmay be made, especiallyif one or two new postulates are added. Among the potential extensionsis a group of predictions involving various phenomenarelating to overlearning. These predictions require a definition of overlearning, as well as the new assumption that the variability of trials needed to learn

individual items will increasewith aggregate generalization. Six or more predictions follow , among them these: increaseof overlearning with greater generalization; later maximal loss as a result of high generalization; various length-of-list phenomena, such as decreasein retroaction with increasinglength of list; decreaseof retroaction as generalizationwithin a first

list increases

.

If a postulate concerning IIdisinhibition" is added, a number of predictions regarding the effect of shockor distraction in learning and retroaction

VerbalLearning

67

situations willfollow. Afurther group ofpredictions ensue ifanassumption regarding theeffects ofcaffeine beincluded. These particular extensions

havebeenmentioned, because theyhavebeenworked outin(9). Evaluation

Ahypothesis relating certain factsofverbal learning tothemore general ingthis,a number ofpropositions dependent onthehypothesis were outlined in somedetail.Sincethepotential number ofdeductions froma hypothesis maybealmost infinite, itdoesnotconstitute a perfect evaluationofthathypothesis tocalculate thepercentage ofverified propositions concepts ofdifferentiation andgeneralization hasbeenpresented. Follow-

which havebeenshown tofollow fromit.Yet,it iscertainly ofinterest to

examine theevidence sofarobtained, sincenegative results, at least,area goodindication thatsomething iswrong.Whereevidence relevant to the proposition wasavailable, it hasbeenreferred to already. To summarize,

conclusive positive evidence exists foreight ofthepropositions presented;

somepositiveevidenceexistsfor fourothers,thoughnot conclusive; no

evidence existsforfour;andfortheothertwo,theexisting evidence is

conflicting andnotentirely relevant. Inother words, none ofthese proposi-

tionshasbeenshown tobefalse. Furthermore, ofthepropositions which

havebeenworked outbutnotincluded here,nonehasbeenshownto be

false.

Ahypothesis maywellbeevaluated withreference to itsscope, also and inthisconnection, thetypesoflearning phenomena which the

hypothesis isnotadaptedto explain, aswellasthosewhichit canhandle well,shouldbe mentioned. Obviously, thehypothesis is framed to deal withthataspect ofverbal learning inwhich development ofdiscrimination

between items isvital; andit islimited to a stimulus-response typeof

analysis. Factsrelating tovarious typesoftransfer-effect, bothintra-list and inter-list, constitute itsprincipal applications. Manyotherfactsof verbal learning arenotpredictable fromthehypothesis asif stands. A number of

thesefacts(serial position effects, reminiscence, etc.)canprobably bestbe subsumed under a hypothesis which develops theimplications oftheinhibitory aspects ofmechanisms postulated asbasicto learning, andthe present hypothesis hasleftunexplored suchpossibilities. 11 Otherfactsof verbal learning notsubsumed bythepresent hypothesis would befactsof recognition-memory, andmemory changes.Herea perceptual analysis, perhaps ofthetypemade bytheGestalt psychologists (19,20,32)might bemoreprofitable thana stimulus-response analysis. It maybethatthe problems ofverbal learning must beapproached from boththese angles,

atthepresent timeatleast.Inspiteoftheselimitations, thewriterfeelsthat

therange ofthepresent hypothesis issatisfactory, since nomore compre-

68

E. J. Gibson

hensivetheorycenteredaroundthoseproblems of learningwhichare broadly classifiable astransferphenomena seems tohavebeenproposed. Thevalueof emphasizing discrimination as a factorin verballearning seemsunquestionable; andthatgeneralization anddifferentiation arecon-

ceptssuitable foritssystematic development seems almost aslikely. Notes

1.Thisarticleis a partof a dissertation presented to theFaculty of YaleUniversity in

partial fulfillment oftherequirements forthePh.D. degree. Thewriter wishes toexpress thanks to Professor Clark Hull for his advice and assistance.

2.Thetermdifferentiation willbe usedin placeof differential inhibition, sincethe term

inhibitionimplies a theoryofdifferentiation whichthepresent hypothesis neednot assume.

3. It willbe notedthat thisdefinitionmakesno referenceto nervousprocessesor to any

physiological explanation. Thisusage isinaccordance withthatfollowed byBass and Hull(1)andHull(16).Thatgeneralization willoccurwithvoluntary verbalresponses in theformofa typical generalization gradient hasbeendemonstrated byGibson (6).

4.Thereis a possibility thata slightincrease in specificity mightbeobtained bypure practice ofa single connection. Thispossibility hasnotbeenincluded inthepresent

system because thegreatoverelaboration necessary would beoutofproportion tothe importance of the possibility.

5.From thispointon,itwillbeassumed thatthegeneralizing stimulus items havedifferent responses, andtheassumption willnotbespecifically mentioned. 6.Thereis a furtherpossibility of explaining a decrement in recallof list1.Aslist2 is learned, itemsfromlistI tendto occur, andmustbedifferentiated fromthelist2 items. Suppose thatthisdifferentiation actually involves inhibition ofthegeneralized responses. ThenthemorelistI tendsto generalize withlist2, themoreitemsfromlistI willbe inhibited. Whenlist 1 is returnedto, not onlywilllist 2 itemstendto occurby

generalization, butalsotherewillbespread ofinhibition fortheresponses oflistIin thiscase,thedesiredresponses. Whether thispossibility shouldbe takenseriously depends partlyonwhether generalization ofresponses fromlist2issufficient toaccount fortheloss.Acomparison ofthelossinthissituation withthatinthesituation described

inproposition VIIIshould berelevant, forinthatsituation therecouldbenoadded inhibition withwhichto reckon. Ifrecallof list2 is notaspoorasrecalloflist1 after thesameinterval, thensomefurtherfactorsuchasinhibition mustbe at workin the present situation.

7. For others,see dissertationon filein YaleUniversityLibrary(9).

8.InBunch andMcCravens experiment, thetotaltransfer effectwaspositive; thiswould not,however,preventrelativedegreesof interference fromshowingup.

9.Anexception tothisprediction should beacaseinwhich thestimulus items aremutually synonymous meaningful words. Thissituation hasactually beenshown, intheSmith College Laboratory, toincrease learning difficulty ascompared withunrelated stimulus words. 10.If thematerial is reallyhighlydifferentiated, beingnotonlymeaningful words,but, furthermore, wordsmaking upprosepassages or poerty,rathersmallpercentages of retroactionare apt to be obtained(26,27).

11.TheLepleyhypothesis (21,14)covers someofthesefactsbyexploitation ofa postulatedinhibition of delay.

VerbalLearning 69

Bibliography

1.Bass, M.J.,&Hull, C.L.Theirradiation ofa tactile conditioned reflex inmanJ.comp. Psycho!.,1934, 17, 4765. 2. Britt, S.H.Retroactive inhibition: a review of theliterature. Psycho!. Bull., 1935,32, 381440. 3.Bruce, R.W.Conditions oftransfer oftraining. J.exper. Psycho!., 1933,16,343361. 4.Bunch, M.E.,&Winston, M.M.Therelationship between thecharacter ofthetransfer andretroactive inhibition. Amer. J.Psycho!., 1936,48,598608.

5.

, &McCraven, V.G.Thetemporal course oftransfer inthelearning ofmemory

material.J. comp.Psycho!., 1938,25, 481496.

6.Gibson, E.J.Sensory generalization withvoluntary reactions. 1.exper. Psycho!., 1939, 24, 237253.

7. 8.

. Intra-list generalization asafactor inverbal learning Manuscript inpreparation. . Retroactive inhibition as a function of degree of generalization. Manuscript in preparation.

9.

. Asystematic application oftheconcepts ofgeneralization anddifferentiation to verballearning. DoctorsThesis, YaleUniversity, 1938.

10.Gulliksen, H.Transfer ofresponse in human subjects. J. exper. Psycho!., 1932,15, 496516.

11.Hovland, C.I.Thegeneralization ofconditioned responses: I.Thesensory generalization ofconditioned responses withvarying frequencies oftone.1.gen.Psycho!., 1937, 17, 125148.

12.

. Thegeneralization ofconditioned responses: II.Thesensory generalization of conditioned responses withvarying intensities oftone.J.genet. Psycho!., 1937,51,

279291. 13.

. Thegeneralization ofconditioned responses. Ill.Extinction, spontaneous recovery, anddisinhibition ofconditioned andofgeneralized responses. 1.exper. Psycho!., 1937, 21, 4762. 14.Hull, C.L.Theconflicting psychologies oflearningawayout.Psychol. Rev., 1935, 42, 491516. 15. 16.

46,

. Mind,mechanism, andadaptive behavior. Psycho!. Rev.,1937, 44,132. . The problem of stimulus equivalence in behavior theory. Psychol. Rev.,1939, 930.

17.James, H.E.0. Theproblem ofinterference. Brit.J.Psycho!., 1931,22,3142.

18.Kline, L.W.Anexperimental studyofassociative inhibition. J.exper. Psychol., 1921, 4, 270299. 19.Koffka, K.Principles ofGestalt psychology. NewYork: Harcourt, Brace andCo.1935. 20.Khler,W.,&vonRestorif, H.Analyse vonVorgngen imSpurenfeld. II.ZurTheorie der Reproduktion. Psycho!. Forsch., 1935,21, 56112.

21.Lepley, W.M.Serial reactions considered asconditioned reactions. Psycho!. Monogr., 1934, No. 205. 22.McGeoch, 1.A.Theinfluence ofdegree oflearning upon retroactive inhibition. Amer. 1. Psycho!.,1929, 41, 252262. 23.

. Theinfluence ofdegreeofinterpolated learning uponretroactive inhibition.

Amer.J. Psycho!.,1932, 44, 695708. 24. 25.

. Studies inretroactive inhibition: I.Thetemporal course oftheinhibitory effects

of interpolated learning. J.gen.Psycho!., 1933,9,2443.

. Studies inretroactive inhibition: II.Relationships between temporal pointof interpolation, length ofinterval, andamount ofretroactive inhibition. J.gen. Psycho!., 1933, 9, 4457.

70

E. J. Gibson

26., 27.

& McKinney, F. Retroactive inhibition in the learning of poetry.Amer.J. Psycho!., 1934, 46, 1 34.

. Thesusceptibility of proseto retroactive inhibition. Amer.1. Psycho!.,

1934, 46, 429437.

28.McGeoch, J.A.,&McGeoch, G.0. Studies inretroactive inhibition: X.Theinfluence of similarity of meaning between listsofpairedassociates. J. exper. Psycho!., 1937,21, 320329.

29.Melton, A.W.,&McQueen-Irwin, J.Inter-serial competition ofresponses duringthe

relearning ofserialverbal material intheretroactive inhibition experiment. Psycho!. Bu!!., 1938, 35, 691692.

30.Muller, G.E.,&Pilzecker, A.Experimentelle BeitrgezurLehrevomGedchtniss. Z. Psycho!., Erg. 1, 1900.

31.Muller, I.ZurAnalyse derRetentionsstrung durchHufung.Psycho!. Forsch., 1937,22, 180210.

32.Ortner,A.Nachweis derRetentionsstrung beimErkennen. Psycho!. Forsch., 1937,22, 5988.

33.Pavlov,I. P. Conditioned reflexes(trans.by C. V.Anrep).OxfordUniv.Press,1927.

34.Razran, G.H.S.A quantitative studyof meaning by a conditioned salivary technique (semanticconditioning).Science,1939, 90, No. 2326, 8990.

35. von Restorif,H. Analysevon Vorgngenim Spurenfeld. I. Uberdie Wirkungvon Bereichsbildungen imSpurenfeld. Psycho!. Forsch., 1933,18,299342. 36.Robinson,E.S.Associationtheorytoday.CenturyPsychologySeries,1932.

37.Sisson, E.D.Retroactive inhibition: theinfluence of degreeof associative valueof originalandinterpolated lists.J. exper. Psycho!., 1938,22,573580. 38.

.

Retroactiveinhibition:the temporalpositionof interpolatedactivity.J. exper.

Psycho!., 1939, 25, 228233.

39.Spence, K.W.Thedifferential response inanimals tostimuli varying within a single dimension. Psycho!.Rev., 1937, 44, 430444.

40. Thorndike,E. L.Humanlearning.CenturyPsychologySeries,1931.

41. A note on assimilation andinterference. Amer.J. Psycho!., 1937,49,671676. 42. Waters,R. H. The law of acquaintance. J. exper.Psycho!., 1939,24, 180191.

43.Whitely, P.L.,&Blankenship, A.B.Theinfluence ofcertain conditions priortolearning uponsubsequentrecall.J. exper.Psycho!., 1936,19,496504.

44.Yum,K.S.Anexperimental testof thelawof assimilation. J. exper. Psycho!., 1931,14, 6882.

Retroactive Inhibition as a Function of Degreeof Generalization between Tasks

EleanorJ. Gibson

Looking backat thisexperiment, thenovelaspectof it seemsto meto be

thepreliminary experiment inwhich a measure ofgeneralization among a setof

items wassecured, independent ofthelearning andrecall called forinthemain experiment onretroactive inhibition. Confusion between items wasthemeasure of

generalization, alsoa novelty. Inchoosing material tobesampled fordegree of generalization among items oftheset,I selected pictorial instead ofverbal material.Similarity wastobejudged, as wellas theactual responses toan item confusing it withanother. Nonsense forms were suggested asappropriate byan

earlier experiment ofmyhusbands U.Gibson 1929; histhesis research, actually), inwhich hestudied memory forforms andfound thatthememory ofa form

changed overtime. I copied some ofhisforms indevising myset. Myfirstexperiment, using thismaterial ascue items forpaired associates in a retroactive inhibition experiment, wasperformed asa group experiment, using

students infivelaboratory sections oftheintroductory course atSmith College. Theresults wereinlinewithpredictions andthusquite satisfactory, butHull distrusted group experiments, andsoitwasrepeated withindividual subjects. The

results ofthefirstexperiment werereplicated, including a finding thatstrict analogy withconditioning would hardly have predicted: thata setofitems, once differentiated, doesnotgeneralize againwhen newresponses mustbelearned to

them,andthatdifferentiation obtains withinthesetas a whole, notas an acquisition ofa particular item.Intralist generalization wasgreatly reduced in

interpolated learning, even when theitems were notthesame ones presented for original learning. What occurs, itseemed tome,wasa kind ofperceptual learning, butnotlearning attributable todistinctiveness ofparticular responses learned to each individual item, asa burgeoning theory ofacquired distinctiveness ofcues

assumed (Miller andDollard 1941). Differentiation, onceachieved, is notlost again,noris it a specific resultofattaching particular responses to individual JournalofExperimental Psychology, 1941,28, 93115.

72

E. J. Gibson

items. A setofconfusable itemsisdifferentiated bymeans ofdiscovering the features thatareshared tovaried extents byitems within theset.Thisidea surfaced yearslaterwhen I wasgroping fora theory ofperceptual learning. Thereweretwomorepapersin thisseries, notreprinted here(onebyoneofmy

graduate students). A hiatus followed dueto World WarIIandcoping witha family along withteaching. Butformedifferentiation asafundamental process in development andlearning wasfirmly established, tobecontinued later .

A hypothesis relating verbal learning to twoconcepts fromtheliterature of conditioninggeneralization and differentiationhasrecentlybeen

presented bythewriter (4)12Thishypothesis hasadvanced thegeneral notionthatdiscrimination is anessential processinverballearning, andthat

thespecial difficulties referred tobysuchtermsasinterference andretroactive inhibition can be understood,at leastin part, as casesof relativelyiow

discriminability of learning material. Whenthisnotionis combined with theprinciple of establishment of differentiation between thepreviously undiscriminated items,many implicationsfollow.These implicationsare

statedasspecifically aspossible intheformofpredictions in thearticle referred to.Figure 5.1below mayclarify themechanism assumed tobeat work.It represents a listof pairedassociates in whichgeneralization is assumed to be takingplaceduringlearning. Generalization is defined as the tendency foraresponse Ralearned to5a tooccur when5b (with which it hasnotbeenpreviously associated) is presented(4,p. 204).In other words, 5a 5b andS possess lowdiscriminability whentheyarepresented

separately. Generalization tendencies (represented bybroken arrows) will

tendto blocktherightresponses (solidarrows) in proportion to the strength of thetendencies. Competitive blocking thusoccurs, rendering learning difficult.

Ina two-listsituationalso,generalization andcompetitive blocking may

occur. Figure 5.2represents therelationships. Whenthesecond listis learned, itsstimulus members maygeneralize withthoseofthefirstlist,and Sa

*Ra / \

s.

/

\_,

Sb f

,,-_\ .

S Figure 5.1

\/\

/%.

Rb

A

Rc

RetroactiveInhibitionas Function

73

theresponses ofthefirstlisttendtooccur, ortoblock therightresponses, during thenewlearning. Likewise, ifList1 isrecalled afterList2,generalized responses from List 2mayinterfere with recall, producing competitiveblocking or overtgeneralization (reversion to thewronglist).The right excitatory tendencies willbeweakened, itisassumed, inproportion tothedegree ofinter-list generalization. This argument leads tothepredictioninthehypothesis (4,p.215)thata firstlistwillbemorepoorly recalled as thestrength of thetendency foritemsof a secondlistto generalize withit increases. Thepresent experiment isdesigned totest thisproposition; thatretroactive inhibition is a function ofdegree of generalization between a primary andaninterpolated list.Theexperi-

mental situation willyielddatarelevant to theprediction ofinterference fromListI to thelearning ofList2 aswellasretroactive inhibition. The

hypothesis predicts thatmorerepetitions willbe required to learna

second list,inproportion tothestrength ofthetendency foritemsofa first

list to generalizewith itemsof a secondlist.

Numerous experiments havedemonstrated that retroactive inhibition

varies withthedegree ofsimilarity between anoriginal andaninterpolated task(1,p.339if.).Thepresent hypothesis isobviously related tothiswork, butwould maintain thatindistinguishability, rather thansimilarity orhomo-

geneity (14)is theimportant variable. Similarity implies a sophisticated phenomenal relationship involving boththe likeness andthe discriminability oftheitems inquestion. Butit istheindistinguishability ofthe items,andthenecessity fordifferentiating themwhichformsthebasisof thepresent hypothesis. Attheoutset oflearning, items usually havevaryingdegrees ofstimulus equivalence, particularly intheverbal learning situation where items arenotpresented sidebysideforperceptual comparison. Itisreasonable tosuppose thatthisequivalence orindistinguishability isessentially similar totheclassical factofgeneralization which appears in

theearlystagesofconditioning.

Sincethe purposeof the presentinvestigation wasto discover the

relationship between generalization andretroactive inhibition, a prelimi-

naryexperiment wasdesigned tostandardize material interms ofdegree of List1

List2

List1

Sa

0 Ra

Se

..R

Sa o-Ra

Sb

ORb

Sd

.Rd

Sb -

. Rd

Figure 5.2

.-Rb

74

E. J. Gibson

generalization. Themethod ofstandardization andtheresulting dataare presented inExperiment I.Thestandardized material isutilized inExperimentII,a retroactive inhibition experiment, to formseriesofprimary and interpolated listsdiffering indegree ofgeneralization. Experiment IIIisa repetition ofExperiment IIusingindividual subjects instead ofthegroups used

in II.

Experiment I.Determination ofDegree ofGeneralization ofStimulus Items Procedure

Fordetermining degreeof generalization of items,a technique previously em-

ployed byYum (15) wasused. Aseries ofstandard forms wasselected, tobeused ascuesinpaired associates learning. Anumber ofvariations fromeachofthese formswasthendrawn. Twenty-four hoursaftera groupofsubjects hadlearned the

listofpaired associates, arecall series wasgiven them, inwhich each cueform was either theoriginal standard, ora variation substituted forthestandard. Themeasure ofgeneralization wastaken astheextent towhich thesubjects responded tothe variation asifitwere thestandard. 3 Yum wasinterested inrelating thefrequency of response tothevariation withthedegree ofperceptual similarity between the variation andthestandard. Perceptual similarity wasdetermined byjudgesratings. Tomake thepresent experiment comparable, a number ofgroups offorms, each group consisting ofa standard andseveral variations, weredrawn oncards and presented totenjudges. Theywereasked toputeachvariation withthestandard whichit resembled most.Whenthiswasfinished, theywereaskedto rankthe variations according to degree ofsimilarity to thestandard. Thirteen groups of formsweretreatedinthisway.Eachfinally consisted ofa standard formandtwo variations fromit.Therewereavailable thenthreeclasses offorms: standard, first

degree similarity, andsecond degree similarity. Included alsointheexperiment wasa fourthclassofformswhichborenoresemblance to anyofthestandards, as determined by the judges.

Relative degrees ofgeneralization forthevariation-forms werenowdetermined.

The13standardformswereusedasthecuemembers of a listofpairedassociates.

Eachstandard waspaired witha monosyllabic word(again inaccordance with

Yums procedure). Theorder ofthepairs wasvaried witheachpresentation, sothat noserialassociations wouldbeformed. Fourtestlistsfordetermining degreeof

generalization were prepared, each composed of13forms, likethelearning list,but

formsofall4 classes wereincluded amongthem.Ina typicaltestlisttherewould

be3 ofthestandard forms, 3 forms highly similar totheirstandards, 3 forms less

similar to theirstandards, and3 formswhichresembled noneof the standards.

(Actually, therehadinanylisttobe4 forms fromoneoftheclasses tomake up 13.)There werenecessarily 4 different testlists,toinclude thetotalof54forms. Novariation everappeared in thesamelistwithitsstandard. Fourgroups of subjects tookpartintheexperiment, varying innumber from19to29.Each group hadtheoriginal learning list,butadifferent testlist.Thegroups werefoursections in an elementary psychology course.

RetroactiveInhibitionas Function

75

Aspecial serial exposure device was used topresent thematerial tothesubjects,

consisting ofaJastrow memory apparatus placed behind aprojector which threw alarge image ofeachform-word pairona translucent screen. Theforms hadbeen printed bymimeograph sothatperfect copies were available forthedifferent lists,

and theresponse words were printed inblack India ink. Exposure time foreach pair

was2 seconds. Five learning trials oftheoriginal listwere given; immediately following each presentation, recall was tested, theforms alone being presented in adifferent order each time. Thesubjects hadtorecord theproper response words

ina prepared booklet within 3 seconds. Thesubjects werethenreleased but returned thenextdayatthesame hourandweregiven oneofthefourtestlists

fordetermining degrees ofgeneralization. This testwas similar inprocedure toone

oftheprevious recall trials. Following thistestseries, anordinary recall testwas given using the originalforms.

Thefollowing instructions were given thesubjects atthebeginning ofthe

experiment:

You willbeshown agroup offorms, each onepaired withaword. Study

these pairs sothatwhen aform isshown alone, youcanwrite theappropriate

word. Youwillbeshown onlyonepairata time. Donottrytolearnthese pairsinanyparticular order, because theorderwillbechanged everytime.

Thepoint is,toassociate aparticular word withtheform withwhich italways Aftereverypresentation oftheforms paired withwords, youwillbe shown theforms bythemselves inordertoseewhether youremember the appropriate word ornot.ifyoudoremember it,write it intheblank proappears.

vided, being sure toputitintheproper space. Donotleave anyblank spaces; ifyoudonotremember theproper word, draw aline through thespace. After wehavefinished oneofthese testtrials, turnthepagesothatyoualways havea blank pagebefore youduring thelearning. Onthesecond day,before thetestseries, these instructions were given: Today Iwant youtolook atsome forms astheyappear andputdown the

wordforeachonewhenever youcan.Youwillbeshown forms bythemselves; tryto writetheappropriate words, justasyoudidinthetestseries yesterday. Youmaynotremember themallthefirsttime, butyouwillbe given another chance later.

Thesubjects could hardly failtorecognize atsome pointintheseries thatthiswasnottheoriginal list,since 3or4oftheforms were totally unfamiliar ones bearing noresemblance tothestandards. This awareness probably hadtheeffect (tojudge from later questioning) ofcausing them towrite responses onlytothose forms which theyfeltsureof,sothatingeneral whenvariations wereresponded toasiftothestandards, theitems were functioning asequivalent stimuli, notasmerely recognizably similar stimuli. Results

Degree ofOriginal Learning Since variations inthedegree oflearning attained areprobably accompanied byvariations inthedegree ofgen-

76

B. J. Gibson

Table

5.1

MeanNumberof ResponseWordsCorrectlyRecalledafter5 Presentations

Group I

Group II

Group III

Group IV

12.5

12.0

11.9

11.9

eralization exhibited, 4 the meannumberof responsewordscorrectlyre-

calledafterthefifthlearning trialforeachgroupofsubjects ispresented in Table5.1.Thedegreeoflearning attained washigh,beingapproximately 12 out of a possible13 correctresponses. Thisimpliesthat manyindividualswereableto recallall 13 responsescorrectly.The differences betweentheaverages of the4 groupsaresatisfactorily small.

Degrees ofGeneralization forIndividual FormsThemainpurpose of the

resultswasto establish fourlistsofformsvarying intheirtendency to call forththe sameresponses as didthe originallist of standardforms.To obtainsuchlists,theresultswerescoredasfollows. First,thenumber of

subjects responding correctly to eachstandard formin thetestlistwas

obtained. Then,thenumberresponding to eachof the variations as if it werethestandard wascalculated. Whentheselatterfiguresareexpressed

aspercentages ofthetotalnumber taking partinthegroup concerned, they

constitutemeasuresof the degreeof generalization with the standard.

Finally, it wasdetermined whether anyformrateddissimilar to thestandardswasresponded to asifit wereoneofthestandards; andwhether anyvariation froma standard wasresponded to asifit werea different standard.

Figure 5.3shows theforms grouped inclasses according to theirtendency to generalize withthestandards. Thestandard forms themselves areClassI. Theformsin ClassII arethosevariations whichproduced

thehighest frequencies ofresponses belonging to theirstandards. Those grouped inClass IIIproduced lower frequencies ofthestandard responses. In eachcase,theformsin ClassII andClassIIIarelistednextto their standards. Thepercentage ofthegroupwhichresponded appropriately is

ineachcasegivenopposite theform.Themeans showa widedifference between Classes I,II,andIIIintermsoftendency to callforththestandard

response. Theforms grouped asClass IVarethosewhich werejudged dissimilar. Fourof theseformswereresponded to by a fewsubjects, but withno consistency as regardsthe responses given.If the meanpercent

response iscalculated regardless ofthenature oftheresponse, thefigure is 2.8percent. Theorderinwhich forms ofthisclassarelisted isnecessaril arbitrary. Noneofthevariations (Classes IIandIII)wasresponded towith anyconsistency asifitwereastandard otherthantheonebeside which it islistedinthechart.Therewerea totalof10wrong responses eachfor

Retroactive Inhibition as Function

77

CLASSIV

() ~. ;J-.-J.j . .

. .

. .

. 84.5

41.1

~ 9Q ~ bb gg % 0 ~ C5~ ~

. ~ I1II Av .

::;J/ / ~~~2:~1.j :;[J ,I,,I 6~-'~ ..D cyu ~ 9.7

Figure 5.3 Stimulus fonns inclasses according tothe percentage ofsubjects giving the stan dard responses .grouped

78

E. J. Gibson

ClassII andClassIllabout 10percentof the totalpossibleresponses in each

case.

Twoof the formsincluded in ClassIIIwerenot responded to at all,

although thesetwoformshadbeenjudgedsimilar to thestandards next which theyarelisted. Besides thesetwocases, therewereonlytwodiscrepancies between similarity asratedbythejudges andgeneralization as determined by quantitative score.Thesediscrepancies causeda change fromClassIII to ClassII in the caseof forms10 and 12. Therewas

therefore a highcorrelation between judgments ofsimilarity andobjective

tendency to generalize. Toshowthiscorrelation, theresults canalsobe scored according toamethod usedbyYum. Hecalculated foreachtestlist themeannumberofformsofeachsimilarity groupresponded to correctly,

ratherthanscoring eachformindividually ashasbeendoneabove. Ifthis

calculation is madein the presentexperiment, the dataareconsistent for thefourtestlistsin showingthehighestmeanresponseto the standard

forms, nexthighest toforms judged offirstdegree similarity, andleasthigh to thosejudgedof seconddegreesimilarity. 5

Experiment II.Degree ofRetroactive Inhibition andofInterference asDependent on Degreeof Generalization Procedure

Sinceit wasproposed to testtherelationship between degreeof inter-list generalization ontheonehand,anddegreeofretroactive inhibition ontheother,the conditions oftheexperiment wereplanned soastoinclude threedifferent degrees

ofgeneralization between primary andinterpolated lists(Conditions II,IIIandIV below). Another condition contained identical stimulus itemsin thetwolists

(Condition I),anda fifthserved asa control forcalculating retroactive inhibition (Condition V).Inallconditions theprimary learning listwasmadeupof the standard (Class I)forms paired withnonsense syllables. Thefourclasses offorms

astheyaregrouped inFig.5.3constituted thestimulus items forthefourinterpolated lists.Theplanof theexperiment wasthenasfollows. Theresponse Learning of

Learning of

InterpolatedList

Retention of

Primary List

Condition

Stimulus Forms

Stimulus Forms

Stimulus Forms

I

ClassI (Standard) ClassII

Standard

II

Standard Standard

III

Standard

Standard

IV

Standard

ClassIII

V

Standard

(Nolearning)

Standard

ClassIV

PrimaryList

Standard Standard

RetroactiveInhibitionas Function 79

members oftheinterpolated listswere nonsense syllables different from those of theprimary list,butidentical inallfourinterpolated lists. Conditions I,II,III,and

IVdiffered, therefore, only astostimulus forms. The syllables forboth primary and interpolated listswere chosen according totheusual rules from Hulls list(7) Only those oflowassociation value were included. Incondition V.thesubjects

were given copies oftheNew Yorker toreadduring theinterpolated interval, with Thedetails ofprocedure were similar tothose ofExperiment I.Since thenature

instructions to selectthebestandtheworstcartoon.

ofthefive conditions precluded using asubject more than once, five large, equivalent groups ofsubjects were secured. The sections ofanelementary laboratory course offered themost available andsuitable subjects. Thesame technique of

exposure wasemployed, again allowing twoseconds perpairforlearning, and threeseconds peritemfortherecall trialwhich followed. Theorderwasvaried withevery learning trialandeverytesttrial. Fivelearning andtesttrials were given. 6 Theprepared booklets inwhich responses were recorded were collected

immediately following theoriginal learning, andnew booklets passed out.Instructions fortheinterpolated learning followed andthistaskwasbegun atonce. The interpolated listwas also presented five times, with tests following each presentation.Assoonastheinterpolated learning wasfinished, books werecollected, others distributed, further instructions given, andrecall andrelearning ofthe original listwasbegun. Inallfive conditions, theinterpolated period between the endofthefifth testoforiginal learning andthefirstrecall testofitlasted fourteen minutes. Seven relearning repetitions ofthefirstlistweregiven. Thenumber was limited to seven, owing to the termination of the class period; it did not allow learningto complete masteryin mostcases.

Theinstructions were almost identical with those given inExperiment I and

need notberepeated. After the5presentations ofthefirstlist,subjects were told: Weshall return tothislistlater on,andgive enough more trials foreveryone to finish learning. When thenewrecord books hadbeen handed out,theywere told: Nowthere will beanew list,ofthesame type, tobelearned. Theprocedure will beexactly similar tothatwehave justfollowed. Please tryagain towork ashard aspossible. Thefinal instructions given before recall ofthefirstlistwerethese: Nowweshall goback tothefirstlistandrepeat ituntil everyone haslearned all ofthepairs. Weshall startwithatesttrialtoseehowmuch youremember. Write every syllable which youcanrecall intheproper space when itsform isshown, and

draw aline intheothers, asusual. Then weshall proceed with alternate learning

and test trialsas we did before. Results

Degree ofOriginal Learning Since thevalidity oftheresults depends on theequivalence ofthe5groups ofsubjects asregards degree oforiginal

learning, themeannumber ofsyllables correctly recalled onthefinaltest

foreach condition ispresented inTable 5.2.Themeans arefairly uniform,

andthevariation could notberesponsible forthedifferences inrecall which

will appearlater.

80

E. J. Gibson

Table

5.2

Degree of Original Learning Condition

Mean No. Learned

No. of Ss

Table

1

II

8.21

III 8.05

7.86

24

22

22

IV 7.92 39

V 7.91 23

5.3

Degree ofLearning andMean Number ofReversion Errors inInterpolated Learning Condition

MeanNo.Recalled atTrial5 Mean Total No. Recalled

Mean No.Reversion Errors

II

I

7.67 22.75

.83

III

9.68

30.91

.45

Iv

10.00

31.82

.41

9.82

33.95

.41

Degree ofLearning ofInterpolated ListsThedegree ofinterpolated learning

achieved inconditions I,II,III,andIVisinteresting ontwoscores. First,the

hypothesis predicts thathigher degrees ofgeneralization between a first

anda secondlistshouldmakethesecondlistharderto learn.Andsecond,

variation inthedegree ofinterpolated learning achieved mayinitselfaffect thedegreeof retention of theprimary list,sincedegreeof retroactive inhibitionhas been shownto be a functionof degreeof interpolated

learning (9,12).Table 5.3shows themeannumber ofsyllables correctly recalledafterthe fifthtrialof the interpolated list in eachof the four conditions, as wellas the meantotalnumberrecalled on all fivetests

ofinterpolated learning. Onlythedifferences between Condition I andthe otherconditions approach statistical reliability, 7 but the totalsshowa definite tendency formoresyllables tobecorrectly recalled asdegree of generalization decreases fromCondition I to Condition IV.Thelearning curvesforeachofthefourconditions arepresented inFig.5.4.Condition I

is lowerthantheothersthroughout thefivetrials.In theearlytrials, Condition IVisnoticeably higherthantheotherthree,butinthelatertrials itscurveis closeto thecurvesforConditions II andIII.Thechieffact shownisthemuchlowerdegreeoflearning achieved inCondition I,which

isinkeeping withthehypothesis under consideration. Itmight alsoaffect thedegree ofretroactive inhibition forthiscondition bylowering it,since

it hasbeenshownthatretroactive inhibition decreases withlowerdegrees ofinterpolated learning (within therangeofpartiallearning).

Whentheresponses givenduring interpolated learning areanalyzed for

reversion errors(casesof overtgeneralization withthepreceding list)it

shouldbe expected, according to thehypothesis, thatmoresucherrors accompany thecondition where learning ismostdifficult. Themeannum-

81

Retroactive Inhibition as Function

a311~J3~ All .J3~~OJS3SNOdS3 ~ ::jO~38l1 'JnN

(g.x- a0- - - - . 0- --

Figure

5 .4

Learning

curves

number

ber

of

of

of

that

more

conditions

according

I

Degree

of

be

ately

is

the

low

Retroactive

the

Table

number

of

groups

show

5

correct

.4

first gives

significant

at

the degree

of

first

of

of

!

( Recall

retroactive

second

the

) ,

the

comes

from

) ,

and

inhibition

. 8

all

the

second

from

All

,

.

list

immedi

recall

a

,

overtly

-

test

relearning

based

four

on

the

experimental in

is

the

first

test

inhibition

test

It

errors

at

recall

the

.

between

of

first

.3

reversion

retention

2

5

between

occur

the

retroactive

and

of

with

,

of

from

Table

difference

errors

degree

trial

percent

terms

coinciding

measures

( Recall

relearning the

I ,

such

:

in

in

greatest

feature

measures

plotted

given

reversion

which

The

is

also

The

striking

with

several

.

and

A

learning

recalls a

.

learning

.

Condition

series

recall

II

interpolated

follows .

both

in

that

Inhibition

from

of trial

are

occur in

frequency

calculated

Degree

each

conditions

errors

learning

.

after

four

Condition

following

which

the

to

and

lists

recalled

such of

Condition

,

interpolated

for

difficulty

however

four correctly

errors

obvious

score

the

responses

such

greater

may

CONDITIONIV CONDITIONIII CONDITIONII CONDITIONI

terms

of

both

,

82

E. J. Gibson

Table5.4 RecallScores , Percentof RetroactiveInhibitionMeasuredby RecallScores , andMean

Number ---------of - Trials toRelearn 9Items Conditions --------- I II III IV V Mean Recall 1 3.04 3.00 4.18 5.79 7.57 - ---- of 9.78 8.36 7.00 6.55 5.92 Mean of Recall 2 Retroactive Inhibition , Recall ! Retroactive Inhibition , Recall 2

Mean ----Trials to'Relearn '

60.6% 41.7%

4.08

59.9% 32.6%

3.73

45.8% 29.6%

3.86

23.5% 14.6%

3.00 2.35

IV Figure 5.5 Degrees of retroactive inhibition plotted as a function of degree of generalization between stimulus forms of original and interpolated list. The Roman numeralsunder the percentages of generalization on the abscissaindicate the experimental condition previously described.

Retroactive Inhibition as Function

83

measuresof recall. Furthermore, the degree of retention varies with the degree of generalizationbetween the original and the interpolated leaming.9 Figure 5.5 shows degreeof retroactive inhibition plotted against the degreeof generalizationas obtained in the standardizationexperiment. As the degree of generalizationincreases , so also does degree of retroactive inhibition, as the hypothesis predicts. When the two curves are averaged, the result is a smooth, negatively acceleratedgradient. It is noteworthy

that Condition I (identical stimulus items) does not

show a significantincreasein retroactive inhibition over Condition II (Class II stimulusitems), particularly in RecallI. A greater differenceis apparentin Recall 2, but even here the critical ratio is only .74. This result may be due

to the lower degree of interpolated learning for Condition I. In other words, had degree of interpolated learning in Condition I been as high as that in the other three series, a higher degree of retroactive inhibition might have been found. Another feature of the curves is the sharp rise from Condition IV to Condition III Condition

from 0 generalization to 10 percent generalization . Yet

IV itself shows 23 .5 percent

retroactive

inhibition

in the first

recall. It may be questioned whether such a degree of retroaction for a condition of supposedly zero generalization is consistent with the theoreti -

cal prediction. Two explanations suggest themselves. The first is that syllable responseswere learned in the interpolated list as well as in the primary list, and that something analogous to generalization may take place among response membersas well as stimulus members. Doubt is cast on this possibility by the McGeochs' study (10), which showed that variation in degree of similarity of responsemembers of interpolated paired associates to response members of a primary list was without

effect in

determining degree of retroactive inhibition. This fact is consistent with Thorndike's finding (13) that similar responsemembersare not frequently substituted

for one another in recall , but that similar stimulus members do

frequently evoke one another's responses . The second possibility is that there was actually generalization between stimulus members of original

and interpolated lists in Condition IV, in spite of the negligible tendency established in the standardization

experiment . It will be remembered

that a

very high degree of original learning was achievedin the standardization experiment , while only 8 out of 13 responses, on the average, were learned

in the presentexperiment. It is suggestedthat generalization, after reaching a maximum, decreasesas learning increases ; and that as generalization decreases , the stimuli

most discriminable

from the standard are the first to

ceaseproducing generalizedresponses . In this case, the forms belonging to ClassIV were ineffective in producing generalizationin the standardization experiment becausethe degree of learning was high and becausethey

84

E. J. Gibson

varied considerablyfrom the standards.But in the present experiment,

theywereeffective to someextentbecausethe degreeof learningwas muchlower.Evidence to supportthisexplanation maybe foundin thefact that somecasesof errorsof reversionto the interpolatedlist actuallydid occurin ConditionIV,thoughnot as manyas in the otherconditions. But

beforethisexplanation canbe accepted, it willbe necessary to studythe curveof generalization in relationto degreeof learningundercomparable conditions.

The tendencyfor a greaterdecrement to accompany higherdegrees

of generalization is furthersupported by the numberof trialsrequired to relearnto a set criterion.Thismeasureis of doubtfulvalue,sincelearning was not carriedto a criterionoriginally.But it shouldbe of some sign-

ificance, sincethe groupswereequatedas to numberlearned,on the average. TableIVgivesthemeannumber of trialsrequired in eachcondition to relearnto a criterionof 9 out of the 13 items. (A higher criterion couldnot be set, sincesomesubjectslearnedno morethan 9 in the seven

relearningtrials.)ConditionI requiredmoretrials,on the average,to

relearn9 items,thandid any of the others.Thereis not a perfectgradient

here,sincethe averagefor Condition IIIis slightlyhigherthanthat for ConditionII.It is interestingto notethat the difference betweenCondition IIIand ConditionIV is againthe greatest,as wasthe casewiththe two other measures

of retention.

Errorsin Recall In the light of the above results,it is interestingto

examinethe meannumberper subjectof errorsof reversionto the inter-

polatedlistwhichoccurred inrecalling thefirstlistinthefourconditions. Thefiguresforthefourconditions are:Condition I, 1.00;Condition II,1.09; ConditionIII,.91;ConditionIV, .36. As was the casewhen retentionwas

measured by firstrecall,thereis littledifference betweenCondition I and

ConditionII.Whatis especiallyinterestingis the largedifference between ConditionIII and ConditionIV, which coincideswith the great difference between these conditions in all three measures of retention.

It is obvious that actual reversionsaccount for only a smallpercentage

of the errorswhichoccurin either interpolatedlearningor recalland

relearning oftheprimary list.Superficially, thisfactmayappearto be out

of linewiththe hypothesis undertest.Butwhenthe natureof the other errorsis studied,their occurrenceseemscompatiblewith the hypothesis.

Omissions arethe mostfrequenttypeof error.Whenthe relearning trials in ConditionI are examined,omissionsaccountfor 67.9 percentof the errors. Theoretically,omissionsmay occur becausethere is no strong

excitatory tendencyto respond,or becausethetendencyor tendencies are

inhibited.The formerexplanationmay accountfor a numberof the omis-

RetroactiveInhibitionas Function

85

sions, since theprimary listwasnotlearned toa veryhighdegree originally.Thesecond explanation isinlinewiththehypothesis, since it would assume thatwhena correct response anda generalized response, or two generalized responses bothtendtooccur atonce,blocking mayresult. Itis

alsopossible thata response, although learned, wasextinguished during theinterpolated learning andisthereby inhibited. Such a possibility re-

quiresan extensionof the hypothesiswhichwillbe considered later.

Theerrorsotherthanomissions maybedivided intothreetypes

reversionsor casesof overt generalization between lists,casesof overt

generalization within thelist,andmisspellings. During therelearning of Condition I,3.2percent oftheerrorswerereversions; 5.6percent were cases ofovertgeneralization within thelist;and23.4percent weremisspel-

lingsofthesyllable responses. Thequestion arisesasto whether theerrors of misspelling areconsistent withthehypothesis. Whentheseerrorsare examined carefully, a veryinteresting discovery emerges. Manyofthem appearto becompromises resulting fromconflicting excitatory tendencies. Forinstance, twoformswithina listhaveas theirresponses ruvandtah, respectively; thesubject, instead ofresponding correctly to eitherofthem, mayrespond withtuvto both.Or,a formintheinterpolated listwhichhas

beenshown previously togeneralize witha formintheprimary list,may haveasitsresponse kiv;theformintheprimary listmaybepaired withhaj.

Butwhenthesubject isaskedtorecalltheprimary list,hemaywriteneither of theseresponses, butinsteadkij.In suchcases,the stimulus formseems

tohavetwoexcitatory tendencies, anassociated oneandageneralized one, whichresultina compromise response ratherthanblocking. Underwhat conditions compromise willoccurinstead of blocking is an intriguing

question. It seemspossible thatthestrengthofbothtendencies maybethe

keyto thisproblemthattwoequalbutweaktendencies mayleadto compromise, strongonesto blockingbutfurtherstudyis required to

solve it.Probably notallthemisspelling errors areofthistype;anattempt wasmadeto estimate thepercentage falling inthisclass, butit proved impossible to classify manyofthecaseswithcertainty.

Theerrorsofgeneralization within thelistwhichoccurred during the learning oftheprimary andtheinterpolated listsshowaninteresting trend. There isamean decrease of59percent insuchcases from primary learning to interpolated learning. Thisdecrease occursconsistently in allthefour

conditions. Thereduction inCondition I is consistent witha prediction regardingthe learningof pre-differentiated itemsIf differentiation has beensetupwithin a list,lessgeneralization willoccur inlearning a newlist which includes thesame stimulus items paired withdifferent responses (4,

p. 222).Thereductionin the otherconditions is hardto understand unless

something liketransfer ofdifferentiation ispostulated.

86

E . J. Gibson

ExperimentIII . A Repetition of ExperimentII Using Individual Subjects Procedure

Sincethe third experimentwasundertaken principallyasa checkon the reliability of the results obtained in Experiment II, conditions were kept as nearly the same as possible. In general, only those changeswere made which were demandedby individual experimentation in place of a group procedure. The two major changes were omission of Condition V , and the use of a learning criterion rather than a set number

of trials . Condition

V (the control

for retroactive

inhibition

in which

the

subjectsread during the interpolated period) was omitted, becausethere could be no doubt of the fact that retroactive inhibition

occurred and we wished only to

verify the gradient observed from Condition I to Condition IV . The subjects always learnedto a criterion in the primary learning, the interpolated learning, and the relearning so as to insure equal degreesof learning in the four conditions. The lists learned by the subjects in Conditions I, II , III , and IV were identical

with those of Experiment II, except that one pair was omitted, making them 12-unit instead of 13-unit lists. The forms were pasted on heavy paper, and the syllables were typed beside them. The learning material was presented on an electrically driven Chicago-type memory drum. Eachpair was again exposedfor 2 seconds. But when the forms were presented alone, in the test trials, they now appearedfor 2 secondseach instead of 3 seconds, since the subjectshad only to speak the right associatesinstead of to write them. Six random orders were preparedfor eachlist, and the order was changedeachtime for both learning and test trials . All the subject 's responses were recorded by the experimenter .

The procedure was as follows. As soon as the subject had received the instructions, which were the sameas those of Experiment II, the first list was presented. When the list had been exposed, the forms alone were presentedand the subject attempted to recall the appropriate responses.This sameprocedurewas continued until the subjecthad recalled8 of the 12 responsescorrectly in a single trial. Then the subjectwas given a four minute rest period while the list was changed. In order to prevent rehearsing , he was given a magazine to read. At the end of this

period, the interpolatedlearningwasbegun. Theprocedurewasexactlythe same as during original learning, and was continued until the subject had learned 8 of the 12 responses.Another rest period ensued, and the subjectread, as before, until the time interval was up. The total time allowed for the interpolated interval was 20 minutes . The subject then tried to recall the responses of the first list , and

relearning followed. Relearningwas carried to a criterion of one perfect recall trial. The experiment required about one hour. Fourteen subjects were secured for each condition . Since it was important to

equatethe groups for number of trials spent in learning the first list, an attempt was made to secure similar distributions for the four conditions in this respect. That is,

an attempt was made to distribute the fast learnersand the slow learnersequally over the four conditions. This could easily be done since it was possible to assign a subject to a condition after shehad completed the originalleaming . The subjects

were all studentsfrom elementarypsychologycourses , and comparableto the subjects in Experiment II .

Retroactive Inhibition as Function

87

Results

Trials toLearn Original List Themean number oftrialsrequired tolearn

theoriginal listto a criterion of8 outof12rightassociates foreachofthe fourconditions areasfollows: Condition I,6.86;Condition II,7.00;ConditionIII,7.07; Condition IV,7.14.Themeans areverysimilar, andtheslight

differences which existareintheopposite direction fromtheexpected differences inretention. Itisinteresting thattwomore trials were required

here,ontheaverage, to learnSresponses, thaninExperiment II,where onlyfivetrialsweregiven.Thedifference maypossibly bedueto the shorterrecalltimepersyllable in thisexperiment, or to theinfluence

of socialfacilitationin the earlier one.

Trials toLearn Interpolated ListThemeannumber oftrialsrequired to learntheinterpolated listtoa criterion ofSoutof12rightresponses is presented foreachofthefourconditions inTable 5.5.Ingeneral, thedata

corroborate thoseofExperiment II.There, fewerresponses werelearned in Condition I thanintheotherthreeconditions, which didnotdiffer reliably

fromoneanother. Here,moretrialsarerequired to learnto thecriterion in ConditionI thanin any othercondition. ButConditionIV alsodiffers widelyfromtheotherthree,requiring fewertrials.Thedifference between Conditions II andIIIis whollyunreliable. 11 Theseresultstendto confirm

theprediction thatlearning ofa second listwillbeharder asthedegree of generalization betweenthefirstandthe secondlistincreases.

Retention of Original List Theretention of the original listin thefour conditions maybemeasured infourdifferent waysby a firstrecallscore, a secondrecallscore,andby twodifferent relearning criteria. InTable5.6

are presentedthe resultsfor Recall1 and2. Thetrendsin the resultsof

Experiment IIareconfirmed beyonda doubt,sincea gradient inrecall from Condition I to Condition IVis againapparent in bothRecall1 and2.12

When thereciprocals ofthese figures areplotted, thecurves areremarkably similar totheonesobtained inExperiment II.Inspiteofequivalence in

degree ofinterpolated learning, Condition I again shows onlyslightly

worseretentionthanConditionII.Thedifference betweenthe two averTable 5.5

Number ofTrials Required toLearn theInterpolated Listtoa Criterion of8 Correct

Responses

Condition MeanTrials to Learn

I

aav.

– 1.07

7.43

II 5.43

III 5.71

IV 3.93

–65

–77

–28

88

E. 1. Gibson

Table

5.6

RetenUon of theOriginal ListasMeasured by Recall1 andRecall 2 Condition 1 II III MeanNo.Retained, Recall 1 2.93 3.00 3.71

IV 6.21

a av.Recall I – .45 – .46 – .64 – .34 MeanNo.Retained, Recall 2 5.00 5.71 7.29 8.36 a av.Recall 2 – .56 – .62 – .67 – .43 11.Thecritical ratios(D/crdlff ) ofthedifferences are:I and11,1.60; I andIII,1.31; I andIV, 3.18;II and III,2.8;II and IV,2.12;IIIand IV, 2.18. Table

5.7

ReteMionof the OriginalListas Measuredby Relearning

Condition MeanTrialsto Relearn8 outof 12

I

aav.,Soutof12 MeanTrialsto Relearn12out of 12

a av.,12outof12

II

3.79

–.80

2.79

–48

III 2.00

IV 1.36

–41

–21

8.29

6.43

5.43

4.79

– 1.30

– 1.10

– .69

– .53

agesisunreliable. 13 Thedifferences between theotherconditions areall

fairlyreliable byonemeasure or theother.Therecallscoresthusbearout ineveryrespecttheresultsof theprevious experiment andtheexpectation of thehypothesis thatretentionshouldbe pooreras degreeof generalization between the two lists increases.

Relearning scoresintermsofthenumberoftrialsrequired to relearnto oneperfect repetition, andthenumber required to relearn S outofthe12 responses maybecalculated. Onlythelatteriscorrectly calledrelearning, since the first list was learnedoriginallyto this criterion.The results

obtained bythesetwomeasures areshowninTable5.7andFig.5.6.Both relearning measures areconsistent withtherecallmeasures inrevealing a gradientin retention fromCondition I to Condition IV.Thedifference betweenConditionI and ConditionII is relativelylargeras measuredby

relearning thanwhenit ismeasured by recall,thoughit is stillnotstatisticallyreliable. 14 The slopesof the curvesbasedon relearning are more

gradual andnearlylinearthanthosebasedon recall, butthetrendof the results could hardly be more consistentlysupported.

Analysis ofErrorsTable5.8showsthefrequency ofoccurrence oferrors

of reversionto the previouslist in interpolatedlearningand relearning. Totals,not averages,are presented,sincethe absolutenumbersare very

small. Thefigures areinconsistent, andshownogeneral trend,exceptthat

the failureof any sucherrorsto occurin ConditionIV is probablysig-

Retroactive InhibitionasFunction 89

o- - .

120UTOF 12 8 OUT OF 12

IV DEGREEOF GENERALIZATIONIN PERCENT

N~V313~ 0.1a3~lnO3~ SlVI~1. ~O~381 /\.jnN

Figure 5.6 Trials torelearn List Iasafunction ofdegree ofgeneralization between the primary and the interpolated lists . The two curves represent number of trials required to relearn to two different criteria .

Table 5 . 8 Frequency Reversion Errors in Interpolated Leaming and Relearning ,T .,.1..1..of . 0 ---~~ '~ & 'bIII Condition I II IV . I L " 1 .V Interpolated Learning 2 4 7 0 Relearning 14 23 -- 4 ~ 0 v-

90

E

.

nificant

J .

.

larger

Gibson

The

most

total

than

interesting

number

in

original original

case

the

.

probably

two

in

McKinney

not

cases

,

for

in

and

question

.

and

spontaneous

that

tain

such

fairly

an

as

of

cases

51

original

of

such

,

criterion

Verbal

Reports

of

that

it

a

of

during

,

as

in

stimulus

between

greater

the

as

have

The

a

cer

list

.

-

-

hypothesis

Experiment

within

condition

on

interpolated

the

learning

original

can

As

in

were

be

A

during

in

Experiment

interpolated

learning

lists

,

the

average

.

during

interpolated

II

members

generalization

and

than

would

.

in

time

) .

each

during

was

lists

the

,

the

.

some

interval

;

too

omission

that

,

short

must

responses

any

case

be

learning

all

the

conditions

learned

to

that

for

decision

to

,

since

and

had

they

could

,

the

same

in

none

was

blocking

case

made

among

trend

a

,

,

of

the

caused

strong

at

by

results

occurred

.

to

tendency

ordinarily

some

respond

responses

sometimes

a

Therefore

the

distributed

the

not

reported

to

of

were

such

had

so

learned

think

responses

they

done

had

not

subjects

two

.

that

who

influenced

was

but

Those

these

that

reported

.

have

there

due

,

,

not

reported

response

was

cues

could

subjects

simultaneously

make

In

subjects

rehearse

form

and

the

periods

to

visual

cues

conditions

all

them

to

of

rest

really

way

four

Almost

Most

during

impossible

without

the

Subjects

the

particular

easily

of

207

was

requires

generalization

and

longer

I

explanation

errors

in

- list

than

original

.

List

in

,

test

a

list

of latter

.

thought

in

15

intra

less

the

p

of

occurred

of

consistently

,

the

generalization

and

frequency

although

( 4

- list

recall the

permitting

a

-

the

equal

recall

of

retro

given

that

thereby

This

to in

about

single

fact

within

inter

of

errors

learning

the

.

a

from

repetitions

just

recall

accomplished

percentage

a

the

, 15

generalization

against

of

and

in

list

are

only

to

learning

assumption

high

classified

total

of

function

make

A

,

related

learning

differentiation

protective

does

experiment

be

interpolated

recovery

sumption

is

' s

) ,

errors

they

a

relearning

transfer

reversion

number

experiment

may

( 11

"

for

in

interpolated in

in

tendency

occur

Ilovert

the

increase

difference

interpolated

learning

,

mere

present

increase

between

original

the

the

the

to

McGeoch

less

from

the

is

list

noticed

than

for

McGeoch

The

intervened

II

reason

results

previous

and

also

list

The

these

the

McKinney

,

interpolated

one

'-'

is

the

.

experiment

to

the

to

learning

inhibition

of

reversions

interpolated

active

aspect

of

,

least

active

to

because

of

the

the

errors

competition

of

.

Discussion

The

ment

that

consistency

II

when

retroactive

of

repeated

inhibition

the

results

and

with

the

individual

is

a

complete

corroboration

subjects

function

of

leaves

the

degree

of

no

room

of

for

generalization

Experi

doubt

-

Retroactive Inhibition as Function

91

between primary andinterpolated list. Butaconsideration ofthedynamics oftheprocess isagain inorder , since thisfactdoes nottellustheactual role ofgeneralization ; nordoesproofof thisrolenecessarily exhaust thepossi bilitiesof explaining decrements in retention . Theprocess assumed to occurby thepresent hypothesis is briefly asfollows . Generalization mayoccurfromlist to list- therewill bea tendency forwrongresponses to occur inlearning anewlistorinrecalling anoldoneafterinterpolation . Activeintrusion of wrongresponses will thusbea factorleading to decrement . Butsincetheseintrusions areinfrequent , importance fallsonthenotionofcompetition ofresponses , which mayobviously leadto omissions or compromises . Generalized andright responses willtendto blockoneanother . Asfarasourowndatago, active generalization andcompetitive blocking seem tobesufficient explanations . A recent articlebyMeltonandIrwin(12) furnishes evidence , however , thatsomeprocess besides theonedescribed aboveoccurs in thetypical retroactive inhibition situation . Thisevidence is in thenatureof a curve showing therelationship between theamount ofretroactive inhibition and thedegree of learning of theinterpolated material . Whena separate curve wasplottedto showthedecrement directlyattributable to intrusions , the curveincreased to a maximum andthendecreased asthedegree of interpolated learning wasincreased . Thisis therelationship predicted forgeneralization by thepresent hypothesis (4, p. 217 ). Competitive blocking , according to thehypothesis , should followthesame curveasovertgeneralization , sincestrength of thegeneralized tendency is thecrucial factor inboth.Asdifferentiation isincreased , blocking aswellasovertgeneraliza tionshould decrease . Buttheresidual retroactive inhibition , afterthecurve yielded by intrusions hadbeensubtracted fromthetotal, didnotfollow thesamerelationship . Somefactoror factorsotherthangeneralization andcompetitive blocking musttherefore havecontributed to thetotal inhibitory effect . Thepresent experiment throws nolightonothercontributing causes of retroactive inhibition . Thewriterwouldlike, however , to examine the possibility ofextinction orunlearning ofthetypefoundin theconditioned response situation asafactor , since suchanhypothesis hasbeenattributed to her(12). Theconditions for extinction arepresent duringinterpolated learning , in sofarasdifferential reinforcement takesplace . If thesubject givesageneralized response , it willnotbereinforced andanother response will beindicated asthecorrect one.In thissituation , therefore , thegeneralized response maybeextinguished through differential inhibition . But thisresponse itselfis notgenerally inhibited , of course . Whena salivary response to a bellis extinguished , thereis notinhibition of salivation in general , butonlyof salivation in thatsituation or similar ones . If aparallel isdrawnin thepresent situation , thegeneralized response hasbeenextin-

92

E. 1. Gibson

guished asfarasthestimulus intheinterpolated listisconcerned; another response hasbeenlearned asthepositive oneforthatstimulus. Buthasthe extinguished response beeninhibited foritsownstimulus ofthefirstlist itspositive stimulus? Generalization ofinhibition is a well-known phenomenon(6,p. 178if.),butto thewritersknowledge it hasnotbeendemonstratedin a discrimination situation.Thereis not evenan analogue,then,to

supporttherequired assumption thattheinhibition ofgeneralized responses

duringinterpolated learning willitselfgeneralize to thepositive stimulus itemsof the firstlist.Supposesuchan assumption is made,however.It is stilla factthatveryfewgeneralized responses actuallyoccurred overtly

during interpolated learning, sothattheopportunities forinhibition of

generalized responses werefew.Itmightbeargued thatnon-reinforcement of anywrongresponse willproduce inhibitory effects, or thatimplicit generalized responses mightbe extinguished. Asfortheeffects of extinguishing anywrongresponse, it seems clearthatsuchextinction couldnot contributeto retroactiveinhibition,becauseunlessthe punishederrors were

dueto generalization, theywouldnotbe at allrelevantto recallof List1. It wouldbe necessaryto assumea conceptof free-floatinginhibitionto

finda wayforsuchnon-reinforcement to affectrecallofList1.Asforthe possibility thatinhibition mayresultfromthenon-reinforcement ofimplicit generalized responses, thefollowing arguments presentthemselves. If a subject thinksofa generalized response, butdoesnotspeakit,oneoftwo

things isprobably happening. Ifitistheonlyresponse hethinks of,hedoes notspeak itbecause healready knows ittobewrong, inwhich casehehas differentiated it. In the writersopinion, the essential condition for retroactiveinhibitionis therebyremoved.If it occursalongwith anotherpos-

sibleresponse,the situationwhichhas beenreferredto as competitive blocking results.In so faras thecompeting generalized responseis nonreinforced, it maybe extinguished in thatsituation. Thepossibilities of extinctionare thereforelimitedto overt generalizedresponsesand compet-

inggeneralized tendencies. Sincein boththesecasesit is necessary to assume generalization ofdifferential inhibition to thepositive stimulus, the theoreticalstructurerequiredis elaborateandnot too plausible. Othertheoretical possibilities forretroactive inhibition mustbe keptin

mindat thepresenttime,especially becauseit hasbeenshownto occurin

certainsituationswhichare difficultto analyzein the presentterminology.

Retention offormsinprimarymemory, especially, maybementioned. Here Gibsonand Raffel(5)have shownretroactiveinhibitoryeffectsfor the

reproducing ofa givenvisualformwhenif hasbeensucceeded byother forms.The retroactiveinhibitoryeffectincreasesprogressivelywith an increasein the numberof differentinterpolatedforms.Subjectsreporteda

blotting-outofthememory imageasnewformsmadetheirappearance. In this situationit seemslikely,therefore,that the principallocusof the

Retroactive Inhibition as Function

93

decrement occurred while theinterpolated impressions werefaking place, andthatinterference atthetimeofrecall played a small part.Norcould

inhibition resulting from differential reinforcement beinvolved. Itispossiblethatwhencontentretention isprimarily involved, asinthissituation (and presumably tosome degree inanylearning situation), aweakening of theoriginal impression during interpolation isanimportant factor. The weakening effect does notnecessarily depend upon theinterruption ofperseveration, ashasoftenbeenassumed; it mighttaketheformofdistortion

or deletionof the originaltrace.

Inconclusion, thewriter feels thatadetermination oftheroleandweight

ofthevarious factors suggested canonlybemadeaftercareful considera-

tionhasledtoprediction oftheeffects which theyshould produce when

theusualdimensions (similarity, degree oflearning, etc.)arevariedinthe

retroactive inhibition situation. Conclusions

1. Whensyllableresponses havebeenlearnedto a listof stimulus

forms, variations fromtheseforms maybeshown togeneralize with theoriginal forms, since thevariations areresponded toina testseries asiftheyweretheoriginals. Gradients ofgeneralization intermsof

frequency of response to thevariations maybe demonstrated, and

suchgradients correspond withgradients ofperceptual similarity. 2. Asthedegree ofgeneralization between corresponding stimulus members ofafirstandasecond listisincreased, when responses inthe

twolistsaredifferent, thereis a tendencyfor the secondlistto be harderto learn.Thesecondlistishardestto learnwhenthestimulus

membersof the two lists are identical.

3. Thedegree ofretroactive inhibition ofa firstlistvaries directly

withthestrengthof thetendency forstimulus members ofaninter-

polated listto generalize withthoseofanoriginal list. 4. Errors of overtgeneralization, or reversion to theprevious list occur during bothinterpolated learning andrelearning oftheprimary list.There issometendency forfrequency ofsucherrors tovarywith the degreeof generalization betweenthetwo lists.

5. Errors ofintra-list generalization occur during thelearning ofboth

theprimary andthe interpolated list,but sucherrorsdecrease considerably fromprimary to interpolated learning. Notes

1.Thisexperiment isoneofa series ofstudies presented totheFaculty oftheGraduate

School ofYale University inpartial fulfillment oftherequirements forthedegree of

94

E. J. Gibson

Doctor ofPhilosophy. Thewriter wishes toacknowledge thevaluable advice andassistanceof ProfessorClarkL.Hullduringthe progressof thiswork.

2.Thisexperiment wasreported at the1938meeting of theAmerican Psychological Association

(3).

3.Thiscriterion is appropriate to ourdefinition ofgeneralizafionthe tendency fora

response Ralearned toSato occur when5b (withwhich it hasnotbeenpreviously associated) is presented (4, p. 204).

4.Hilgard andMarquis (6,p. 182)present evidence to showthatwithconditioned responses thecurve ofgeneralized responses differs withdifferent stages ofpractice.

5.Themeansforthefourlistscombined are:Standard forms,2.47;firstdegreesimilar-

ity,1.12; second degree similarity, .33.Themaximum response would be3,since each testlistcontained formsfromeachofthefourgroups. Thedataarepresented indetail in (2).

6.Fiverepetitions werechosen onthebasis ofa preliminary experiment. Itallowed only partial learning ofthelists(about Soutof13items) which wasdesirable inorder to securea highdegreeof retroactiveinhibition(5).

7.The criticalratio(D/o ff)

of the difference betweenConditions I and IV is 2.84

whendegree of learning is measured bythenumber recalled at trial5, and3.75 whenmeasuredby the total numberrecalled. 8.Theretroactive inhibition scores havebeencalculated bytheusualformula fromthe percent retained, ratherthandirectly fromthenumber recalled. 9.Critical ratios(D/adff) basedonRecall 1are:I andIII,1.69;I andIV,3.83;I andV,5.56 IIandIII,1.56;IIandIV,3.51; IIandV,5.17;IIIandIV,2.11; IIIandV,3.96.These statistical reliabilities areingeneral satisfactory, considering thata gradient isinvolved. Criticalratiosfor Recall2 rangefrom1.18to 3.07whenalternateconditions are compared.

10.Thisproblem hasbeentakenupinaninvestigation tobepublished. 11.Thecritical ratios(D/o ) 1 ofthedifferences are:I andII,1.60; I andIII,1.31,I andIV, 3.18;II and III,.28;II and IV,2.12;Ill and IV,2.18.

12.Critical ratios(D/ad ff) between alternate conditions are:I andIII,.99firstrecall, 2.61 secondrecall;II andIV,5.61firstrecall,3.54secondrecall.

13.Since interpolated learning wascarried toacriterion, andthetotalinterval ofinterpolationheldconstant, lesstimeintervened between cessation ofinterpolated learning and recall oftheoriginal listinCondition I thanintheotherconditions, giving lessoppor-

tunityforspontaneous recovery ofgeneralization in thiscondition. If spontaneous

recovery isanimportant variable, itwould thentendtolower therecall scores ofthe otherthreeconditions ascompared withCondition I, thusbringing themallcloserto ConditionI, wheredifferentiation wasmostrecentlyachieved.

14.Thecritical ratio(D/o ff) is 1.07for8 outof 12,and1.09for12outof 12. References

1.Britt,S.H.,Retroactive inhibition: a reviewof theliterature, Psycho!. Bull.,1935,32, 381440.

2. Gibson, F.J.,A systematic application oftheconcepts ofgeneralization anddifferentiation to verballearning, Dissertation on filein YaleUniversity Library.

3.Gibson, F.J,,Retroactive inhibition asa function ofdegree ofgeneralization, Psycho!. Bull., 1935, 35, 626.

4.Gibson, E.J.,A systematic application oftheconcepts ofgeneralization anddifferentiation to verballearning,Psycho!. Rev.,1940,47, 196229.

Retroactive InhibitionasFunction 95

Retrospectand Prospect:Are TheoriesRecycled?

Some of the concernsthat were motivating psychologists in the thirties are in one way or another alive and well today . If we moved away from them, we have moved back again with some revision . Functionalism, for

example, is one of them. It was discovered, during the fifties, that conditioning does not happen fortuitously and inevitably according to some simple formula for attaching a responseto a stimulus with frequent repetition. Some situations are definitely favored for conditioning to certain

responses dependingon the speciesof animaland its way of life. Garcia , investigating learning of taste aversion in rats, emphasizedthe evolutionary basisof favored kinds of learning for a species(Garcia, McGowan and Green 1972). An animal learnswhat is functional and adaptive for its species.Present-day ecologically oriented views of animal learning abound (Bolles and Beecher 1988, Johnston and Pietrewicz 1985). On

the other

hand , connectionism is also with

us again , in an even

sterner form than Thorndike or Hull could have imagined . Neural nets are

thought to be shapedby repeated exerciseof pathways in the network, without (as I understandit) any recourseto ultimate function or adaptive necessity (Rumelhart and McClelland 1986). This new form of connection -

ism emphasizesmicrostructureand mechanism , and contrastssharply with the functional, evolutionary approachof the ecologicalbiologists. The preoccupationof the thirtiesand fortieswith learningtheory gave way as the IIcognitive

revolution " came in in the late fifties and the sixties and in -

formation processing took over . The boxes in the flowcharts were labelled with terms like attention and memory (indeed, a proliferation of memories ),

but never learning. The cognitive psychologiststo this day have neglected learning, leaving it to the biologists, neuropsychologists, and Skinnerians. Cognitive psychology seemsto derive more from a structuralist legacy than a functional one. As the ecological view strengthens today and spreadsto humans, concern with learning can be expected to strengthen

too I believe . Thistime, perhaps , perceptual learningwill be in theforefront.

II Comparative Research on Learningand Development(1952- 1970)

Introduction to Part II

- ~ In the introductionto part I, I pointedout the significance of a comparative approachfor functionalpsychologyin Americaand the prevalenceof research on animalsaspsychologybecame accepted asa science . It wasthis domain,in fact, that first attractedme to psychologyasan undergraduate at Smith College. It was a severedisappointment to arrive at Yale as a graduatestudentandfind that its primatelaboratory,underthedirectionof Yerkes , wasout of boundsfor women.Thereweretwo youngerprofessors in Yerkes 's laboratory , Henry W. Nissenand CarlyleJacobsen , who were sympatheticwith my second -classstatusand found me minor ways to maintaina concernwith comparative psychology . I watchedandeventuallyassisteda bit in brainsurgeryfocusedon the motor systemin cats (extirpationwas the style of the day) with a Dr. Marshallin the medicalschool. I alsoconductedexperiments on maternal behaviorin micefor a Dr. LeBlond , a postdoctoralresearcher from France who wasvisiting at the medicalschool. The projectinvolvedthe role of prolactinin instigationof maternalbehavior.The ideawas to inject prolactin in maturevirgin femaleanimalsand test their behaviordaily for evidenceof ensuingmaternalbehavior.The testinvolvedpresentingthem with pups of variousages(borrowedfrom other cages ) and observing retrieval to the nest. The older the pups retrievedand the larger the numberretrieved , the strongerthe maternaldrivewaspresumed to be. But the adultmicewerenecessarily exposedmoreandmoreto the presence of infant animalsas the experimentcontinued , so I introduceda control groupof adultfemalesthatreceivedno prolactinbut wastesteddaily in the samemanner . All the animals , controlsand experimentals , exhibitedincreasinglystrongmaternalbehavior.LeBlondpublishedthe data, but without my name,perhapsbecause of a contretemps that mayor maynot have beenmy fault. Themicebelongedto a colonybredfor someotherproject, I think involving cancer , andit wasimportantto returnthe litters to their

102

PartII

proper home cagesafter using them for testing. Somemonths after 1began working in the lab, a litter of mice with brown coats grew up in a cage with

white parents. Although 1 had never cleanedthe cagesand had been very careful (I thought) in restoring litters to their home cages, 1 was accused of negligence. 1 was pretty sure the negligencewas on the part of service people who cleanedthe cages, and who had a high turnover. 1 finished the experiment but received no thanks and retired from the scene.

Although 1taught a coursein comparativepsychology for a year or two after returning to Smith, there were inadequatefacilities for animal research there . 1 had to wait until after World

War II , when we moved to Cornell

in

1949, for my chanceto do researchwith animalsagain. 1 accompaniedmy husband, although not as a faculty member (Cornell had nepotism rules in those days). SinceI had no laboratory to work in, 1 was glad to acceptan invitation

to work

as research associate in the laboratory

dell , at a sort of farm known

as the Behavior

of Howard

Lid -

Farm . The animals available

were farm animals, sheep and goats. Liddell's research was supported lavishly by the RockefellerFoundationand it was supposedto focus on the

so-calledexperimental neurosis . Thatwasnot whatI wasinterestedin, but 1was happy to have the opportunity to work with animalsagain, even if 1 could only do so at someone else's invitation .

Themethodemployedfor producingexperimental neuroses in the sheep and goats was devised by Liddell from the classical conditioning

para-

digm- the animal was given a signal such as a buzzer, followed by shock (inescapable ) to a forepaw. A daily routine of this procedure supposedly resultedin a neurosis, diagnosedby rapid heart rate and irregular breathing. Recordswere accordingly taken daily of heart rate and breathing, and these were indeed disturbed, but the animalswould also struggle to get away when they saw an experimenter coming , which seemed to me a mark of

quite reasonabledislike of a very uncomfortableprocedure. I managedto perform a study of conditioning with avoidable and unavoidable shock (reprinted below), which fit nicely into a controversy current in learning theory at the time and so reinstated my earlier interest in learning . My main interest during the two years I spent at the farm was in a

project I beganin the secondyear on maternalbonding of newborn kids to their mothers. Goats have the convenient habit of producing young in

pairs, sothe twinscouldbe splitup in experimental andcontrolgroupsand given different rearing treatments . 1 attended the births (often on cold February nights ) of eight pairs of twins , destined to be reared with their own mothers , or foster mothers , or a peer group , or alone (the latter two

groups fed artificially from a nipple pail). A series of observations were madefrom birth at frequentintervals to observeevidenceof imprinting and maternal-kid interactionsof various kinds. A couple of interesting observations came out of the research before it came to a distressing (for me) end.

Research on Learning andDevelopment (19521970) 103

I wasparticularly interestedin chemical information as a factorin bonding,andso at somebirthsI removedthe newbornkidfromthe mother

beforeshecouldtouchit andbeforetheappearance oftheafterbirth.The

kidwasthenbathedin a detergent. Ononesuchoccasion, I hadjust

completed thefirstkidsbathwhen itstwinbegan tomake anappearance. What todowiththefreshly bathed oneinahurry? Thefarm manager, who

waswatchingfroma half-door, saidput it on thestand. Thestandwasa

veryhighcamera standwitha pedestal surface abouta footsquare. I said, Butwontit falloff?Heassured methatit would not.I stoodthedamp littleanimal onthestand,andit remained there,upright, looking around theroom, untilIcould carryitofftoitsassigned place. Goats areprepared frombirthtosurvive onaprecipice, andthislesson wasagoodpreparation for the visual cliff research later on. It wasalsoquiteclearfroma numberof observations that the maternal

goat,if deprived of its offspring forevena fewhoursafterbirth,didnot

welcome it andinfactwould buttit rudely awayif it approached and

attemptedto nurse.Lickingthe newbornkid and otherchemical inter-

changes areimportant (asanysheepfarmer couldprobably havetold

me).Thekid,on the otherhand,wouldapproachany adultfemalefor

several daysafterbirth.Imprinting didnotoccurveryearly,butit would

eventuallyto its peergroupor a humancaretaker, if its ownor a foster

motherwas not provided.

Thisresearch ceasedin the springof my secondyearat the Behavior

Farm whenI discovered aftertheEaster weekend thatsomeofmycarefully rearedsubjects hadbeengivenawayasEaster presents by thefarmmanager.It wasnevercompleted andso notwrittenup,although I havea demonstration movieto showforit.Thatdisappointment turnedmeto a moretraditional kindof research thatI describe in partIII.My next periodof comparative research beganfouryearslaterafterRichard Walk

came toCornell. Weembarked together onaseries ofrearing experiments, thistimewithrats.Atleast, noonewanted themforEaster presents. Inthelate1940sa stronginteresthadgrownintheroleofenvironmental

conditions during earlyrearing fordevelopment ofsensory processes and

perception. Hebbsworkwithdark-reared rats(1937) wasa precursor of thisconcern; in 1947,Riesens reportontwochimpanzees rearedindarknessfrombirthto sixmonths waspublished andproduced a realstir.If

impoverishment ofrearing conditions could havedamaging effects onperception, might notenrichedconditions havebeneficial ones? Howmight environmental conditions function to constrain perceptual development? WalkandI,withtwosuperb research assistants, planned andcarried outa series ofrearing experiments withhooded rats,providing quite specifically enrichedenvironments forthemtoviewastheygrewtomaturity, and

after three months testing their perceptual competence withaveryspecifi-

104

Part II

callyrelateddiscrimination task.Moreof thislater,whenI commenton theseexperiments. Theystruckme aftercompletion as the not too productiveresultofjumpingon a bandwagon, but nowin retrospect I think theyprovidea usefulmoral.Development is enormously flexible. Theenvironmental milieu does indeed constrain development, but perception is

not constructedover timefromelementarypieces.Two of the fivepapers

stemming fromthisresearchthefirstandthelastare presented with a backwards glanceandreconsideration of whattheymean.Latein the program ofrearing experiments, WalkandI decided toraisea groupofrats in the darkso as to providea majorcontrastwithour enrichment groups.

Darkrearing animals is a greatdealoftrouble. It is nojoketo cleanthe cagesandfeedandwateranimals in totaldarkness. Wedecided thatwe

shouldmakeall this troublepay offby testingthe dark-rearedrats in more than one situation.What shouldthat be, in additionto the usualdiscrimina-

tionlearning? Walksuggested depthdiscrimination, andhereminded meof

a storyI hadtoldhimaboutdrivingeastfromCalifornia withour two children andworryingthattheyounger,onlytwoyearsold,wouldfall

overa cliffas we picnickedon the edgeof canyons.Thuswasconceived the idea of a clifftest for our dark-rearedrats:wouldthey walkover the

edgeofa cliff, forlackofseeing edgesduring theirearlydevelopment? This

experiment wasthemostfunofanyI haveeverhada handin.Wewent onto perform a trulycomparative experiment, withasmanyanimalspecies as we could collect.

Webeganwithhoodedratsbecause thatwastheanimal wehadbeen rearingin the dark.Wemovedon to albinoratsto compare animals witha similarlifestylebut poorervision,andthenbranchedoutto observe

babychicks, ungulates (lambs andkids),puppies, andkittens. Thebaby chicksand the kids were particularlyinteresting,becausethey were pre-

cocialanimals, capableof locomotion at birthor minutesthereafter. They couldbe observedon the cliffbeforethey had an opportunityto learn,

withoutbeingdeprivedofa normalenvironment whiletheywereattaining

independent locomotion. Boththechicks andthekidsshunned thedeep side of the cliff,but moved onto and freelyaround the shallowside.

Obviously, therearespecies thatavoida drop-off withoutlearning from postnatal experience ofanykind.Forthem,onecanpresumably claim that perceptually controlledavoidanceof a drop-offis innate.

Naturally, wewenton to observehumaninfantson a visualcliff.We

advertised in the localnewspaper for crawlingbabies,providinga tele-

phonenumberforparentswhowishedtheiroffspring to takepartin an experiment. Several colleagues (including myhusband) predicted thatwe wouldgetno answers, thatparentswouldfeartheirbabieswouldreceive shocksor unpleasant tests.Butthetelephone rangfuriously andafterthe experiment wasexplained to them,mostparentsbroughtthebabiesin.We

Research onLearning andDevelopment (19521970) 105

observed onlycrawling infants inthoseexperiments, though onsetand length ofcrawling experience varied. Only three outofatotalofthirty-six infants crawled offonthedeepside,although allparents beseeched their babies tocome overit.While wecould notclaim thatthetendency to avoidthedeepsidewasinnate, sincetheinfants hadenjoyed sixto ten months ofgrowing upinalighted, social environment, itseemed extremely unlikely thattheyhadlearned to avoidthedeepsidethrough fallsor punishment ofanykind. Parents, when questioned, seldom reported opportunitiesfor suchlearning. Iflearning isimplicated inavoiding a cliff inhuman infants, it seems

likely tomethatithappens when aninfant firstattempts locomotion, inthe

courseofexploring a surface (before movingontoit)to discover whatkind

ofsupport it affords. Thevisible andpalpable qualities ofthesurface are sampled andpropulsion oftheentire bodyontoasurface occurs onlywhen thereisa surface perceived asaffording support. Mylaterresearch (dis-

cussedin partVI)is relatedto thisearlierresearch withthecliffwhich

showed, perhaps forthefirsttime, theclose tiebetween perception and action inthehuman infant. Thiswasanexperiment onperceived affordances foraction, although wedidnotconceive ifprimarily thatwayat thetime. Perception doesnt havetobecome meaningful through associative

learning, nordoesif guideactionbybeingassociated viaS-Rbonds. It is worthmentioning thatthroughout allthisresearch wewerenever struck byananimals apparent fearofthedeepsideofthecliff. Allthenon-

human animals, to theextentthattheycouldseeit,simply avoided it.

Humaninfantsoccasionally cried,butthatwasattributable to a frustrated urgeto get to theirmothers,whowerecallingto themacrosstheinvisible

surface. Laterworkby Campos (Campos, Langer, andKrowifz 1970),

supported theobservation thatfearofdrop-offs waslearned, probably after self-initiated locomotion iswell under way. Butbythattime, itcould easily

belearnedfromanxiousparents. DuringtheseyearsI hadnolaboratory ofmyownandhadto collabo-

ratewitha colleague onthefaculty, I wrotegrantproposals andhad

graduate research assistants, butcould notsigntheforms asprincipal investigator or adviser.I hada tinyofficeon the fourthfloorof Morrill Hall, where thegraduate students andtheanimals were housed. I enjoyed thecompany ofboththese groups. Itwasa goodatmosphere forresearch

to flourish, evenwithoutanempireof onesown.

The Role of Shock in Reinforcement EleanorJ. Gibson

Traditionally, a classical conditioned reflex issupposed toconsist oftheoriginal

unconditioned reflex becoming instigated bysome previously neutral stimulus (the

conditioned stimulus)afterbeingpairedmanytimeswiththeunconditioned

stimulus anditsensuing unconditioned response. Myearlier experience with finger withdrawal toshock (paper 1)hadtaught methattheformoftheresponse was notinvariably thesame andthatitwaseven transferable toanother appendage ormuscle group. Watching thebehavior ofanimals persistently shocked ona forelimb wasnevertheless a revelation. Theanimals thatwesubjected toconditioning exhibited a number ofpatterns ofbehavior, sometimes going through a

gamutof responses that appeared almostritualistic. Thebehavior wasnotme-

chanical, buttheanimals could notgetaway.Whatwasgoing onwhen shock

wasdelivered inescapably, overandover?Theconditioned stimulus wasa warning,buthowcould it beuseful andthusleadtoadaptive behavior? Theessential contradiction inconditioning withshockthatcannotbeavoided occurrence oflegflexion totheconditioned stimulus thatisimmediately punishedbyensuing shockhadalready beenthesubject ofconsiderable debate and wasresponsible fora two-factor ortwo-phase theory ofconditioning (Mowrer 1947). Theideawasthattheretakes place, first,classical conditioning ofinternal

emotional responses (autonomic responses such asquickened heart rate,respira-

tion,etc.)andthatmotor responses (e.g., legflexion) areinduced bythisstate.The motor responsebehaviorwould presumably bedefense reactions characteristic

ofthespecies, suchasfreezing orrunning away. Ifthisbehavior were successful

inavoiding theshock, it would recur,become habitual. Ifnot,theanimal should runthrough whatever repertory ofdefense reactions itsnatureandthesituation

permitted. Thesituation I hadavailable wasideal forobserving theresponse repertory andtheanimals behavior as theconditioning procedure wasadmin-

istered andcontinued daily, because theanimals were notrestrained ina frame Journal ofComparative andPhysological Psychology, 1952,45,1830.

108

E. J. Gibson

and couldmoveaboutwithinthefair-sized roomwherethe experiment took place.

Theanimalsdid indeedexhibitan extensive repertory of actions,suchas backwardlocomotion fretreat),rearingand head-lowering, crouching, freezing

(inhibition ofmovement), andotherpostures thatcouldbeunderstood asmanifes-

tations ofdefense oravoidance. I would reword mydescription ofwhattook place

now,usinga terminology taken frommyhusbands theory ofaffordancesthese animalswerelearning whattheconditioned stimulus (andindeed, thewhole situation) afforded. Theirbehavior washighlycomplex andvariable as they responded withdefense andavoidance actions, gradually settling downtogreater economy ofaction astheylearned thatattempts atgrosslocomotion wereineffective.Thebehavior wasa far cryfroma simpleconditioned reflex.

Onemightaskwhether theanimals werelearning to be afraidofthesitua-

tion.Weseemto thisdayto understand ratherlittleaboutfears(maybe about

emotions ingeneral). Thisexperiment certainly convinced methatfearscanbe learned.Later researchwith the visualcliffand with a looming situation

gradually inclined metotheviewthatmost fearsarelearned, although I doubt thatconditioning (asinthepresumed learning oflittleAlbert)istheanswer. Building a theory of learning ona concatenation ofconditioned reflexes is

clearly insupportable ifa so-called conditioned reflex isitself a complex learned actionthat canshiftitsformin an animalseffortto achieve adaptation to

circumstances that it cannotcontrol.A goodtheoryof howperception of affordancesis learnedandguidesbehavioris still badlyneeded . Does electricshockact as a reinforcingagent to strengthena conditioned

response? 12 Thispurelyexperimental question hasfar-reaching implications.Shockmayberegardedaspunishment. Theeffectsofpunishment are of intenseinterestto socialand childpsychologists; but thereare as yet no

commonly accepted, well-founded principles oftheoperation oreffects of punishment. Thewriter became interested inthisproblem inthecourse of training animals inoneoftheroutines usedat theCornell Behavior Farm fordeveloping neurosesin animals. Theroutineinvolves a classical (5)

Pavlovian conditioning technique, withshockto a forelegastheinevitable unconditionedstimulus.As the observer watches the relentless succession

of metronome-beats followedby shock,the questionpresentsitselfmore

andmoreinsistently, Whyshouldtheanimalflexitsforelegto thesound

of the metronome? Thereactionyieldsit nothing,apparently, but another shock.Yetthisroutinehasbeena traditional methodof establishing conditionedresponses. Froma Pavlovianstandpoint, a conditioned response

(legflexion) ispredicted. Theshock regularly produces theflexion; hence, in termsof a contiguity-substitution theory,themetronome shoulddo so aftera numberof pairedstimulations. Butsuchan expectation is wholly

Roleof Shockin Reinforcement 109

contradictory toeffect theory. From thestandpoint ofthetraditional law ofeffect, shock isunpleasant andistherefore anegative reinforcement; it

should tend tosuppress anyaction preceding it.Iftheanimal begins tolift

itslegto themetronome, andtheshock follows, theactionshould be discontinued.

This paradox hasbeen thesubject ofseveral experiments designed evitably with conditioning when theCRavoids theshock. Schlosberg (17,18,8)performed experiments withwhite ratscomparing inevitableshock withavoidance andfound little difference intheefficiency withwhich theprocedures produced a conditioned motor response (e.g., tailorfootwithdrawal). Infact, theCRdeveloped veryslowly to compare theeffectiveness of conditioning whenshockoccursin-

andremained unstable underbothconditions. Butevidence ofemo-

tional conditioning (sharp inspiration andsquealing) wasclearly pre-

sent in both cases.

Twoyearslater,Brogden, Lipman, andCuller(2)foundshock

avoidance much more effective thaninevitable shock forproducing

conditioned running (wheel-Lurning) intheguinea pig.Anavoidance grouplearned veryquickly toa criterion of100percent; butthenonavoidance groupneverroseabove 50percent andshowed, furthermore, anerratic learning curve. In1918, Sheffield (20)suggested that

thepigsactual response toshock might often have been stopping (freezing) rather thanrunning. Insuch anevent, asubstitution theory would notpredict thatrunning should constantly prevail asa CR. Sheffield applied Guthries postremitytheoryto thesituation and attempted totesttheprediction thata CRona given trialshould

repeat theURofthejustprevious trial(running orotherbehavior). Hefound thatwhether theanimal ranorstopped toshock increased ordecreased, respectively, theprobability ofa conditioned runtothe

CSonthefollowing trial.Butconditioned running roseonlyto52 Brogden, Lipman, andCullers study (2)itroseto50percent without suchselection ofcases. Also, conditioned running deteriorated inthe percentwhencasesfollowing othertypesof URwereomitted. In

laterstages oftraining withunavoidable shock. Since otherfactors are

required toexplain these results, contiguity seems unpromising asa singleexplanatory principle. These twotheoretical positions (contiguity-substitution andeffect) do equivalent to drivereduction (6)andisthereby reinforcing. A recent experiment reported byMowrer (16, pp.278if.)renders thisposition less attractive, however. There is, finally, the possibility of a two-phase or two-factor theory(19,3,21,13,15). notexhaust thepossibilities. It canbeassumed thatcessation ofshock is

110

E. J. Gibson

Thewriterproposes to develop theimplications ofa two-phase theory

for the unavoidableshocksituation.(a) It is assumed,first,that an emo-

tionalresponse, involving the typicalinternalemergencyreactions, is

quickly conditioned to theCS.Tosupport suchanassumption, thereis evidencefromthis laboratory, as wellas others,for the occurrence of increasedheartrate,irregularrespiration, and psychogalvanic reactionto

theCS,anticipating theUS(10,pp.510if.).(b)It isassumed, second, that

thisemotional stateinducesaction,in allprobability an escapeor defense reactioncharacteristic of the species.It is likelythat a repertoryof such reactions existsforanygivenspecies, perhaps ina hierarchical relationship. 3

Inanavoidance experiment, theshockplaysa singlerole.It reinforces the conditionedemotional reaction to the signal,which,in turn, instigates one

of theresponses in theanimals escape-defense repertory,as weshall tentatively nameit.If theresponse is successful in avoiding theshock,it willpresumably recur.Itsrecurrence maybe dueeitherto lackof interference bya competing reaction, inasmuch asnopunishment results, orto

positive reinforcement byfearreduction (14).Ontheotherhand,with inevitableshock,the shockplays a dual role. It functionsto promote one

oftherepertory of emergency reactions, but,also,it tendsto suppress or

inhibit thisveryreaction assoonasithasoccurred. Aftera number ofsuch inhibitory effects, a particular reaction maybe eliminated entirely, and anothermemberof the emergency repertorytakesover.Theresultwould

beanappearance oftrialanderror.Ifourassumptions arecorrect, a number ofpredictions canbemadeforthesituation whereshockoccurs inevitably 100percent ofthetime.(a)Thereshould bea rotation orshiftofresponses

through theanimals escape-defense repertory (akindoftrialanderrorof

histotalrepertory); (b)thereshould bea number ofcasesoftotalinhibition ofresponse, aftera motorresponse to CShasfirstappeared; (c)inhibitions

ofthistypeshould terminate ina shorttimebecause theprimarily conditionedemotionalstateis constantly strengthened by the sameshockthat

tendstosuppress themotorresponse; and(d)thereshould begreatfluctuationor variability ofresponse intheunavoidable shockgroup,contrasted

withincreasing uniformity inanavoidance group. A qualitative studyof theresponses astheydevelop inthetwosituations should provide acheck of these predictions. Method

Fourteen younggoats(between 1 and2 months oldat thebeginning ofthe experiment) wereconditioned in a relatively freesituation. Theywerenot re-

strainedin a Pavlovframe,but couldmovenormallyto any part of the experi-

mentalroom,whichwas10ft.squareandbareexceptfortheobservers chair. 4 TheCSwasa darkening oftheroomfor10sec.preceding theUS,a mildshock

Role of Shockin Reinforcement

111

delivered tothekidsright foreleg. Eight oftheanimals were given practice with inevitable shock, following theclassical Pavlovian procedure (group Inev), while

four were allowed toavoid theshock bysustained lifting oftheright foreleg shock was unrelated toacritical avoidance response, making itinevitable, although (group Av).Twoanimals weretrained with25percent random shock. Here, the

itoccurred only25percent ofthetime. Ourprincipal dataconsist ofdetailed, qualitative, written observations and photographs of the development of behavior through manystagesoftraining.

Theshock wasdelivered bywayoftwoelectrode bracelets ofsaline-soaked

cloth, placed oneabove theother ontheright fore-ankle. Metal clips attached the bracelets towires which coiled around theright foreleg andpassed through a buckle ontheanimals back. Thewires coiled upward, forming aspring attachment toa4-ft. long stick suspended from thecenter oftheceiling, allowing completely freelocomotion toanypartoftheroom andmaking thenecessary electrical

connections through a swivel jointsothatnoamount ofturning would leadto tangled wires. Theequipment wasnotuncomfortable, andtheanimal could walk

normally. Intensity oftheshock was regulated byavariable resistance and ranged from 10to20V.; duration was 0.2sec. The shock was adjusted soastobestrong

enough toproduce areliable foreleg flexion, butnotexcited running andvocalizationoftheanimals coat, andsometimes it wasstrong enough toproduce a jumpingmovement (bothforefeetoffthefloor)in theanimal. The overhead suspension served asecond purpose. The 4-ft. stick which, owing tion.Theintensity did,nevertheless, varyslightly withfactors suchasthecondi-

toitsspring attachment totheanimal, always pointstoward it asit moves about

theroom, ismade tomove twoself-synchronous motors, oneforeach ofthetwo dimensions oftheplane ofthefloor. Each ofthese ispaired withasimilar motor ina recording instrument andmoves a pencil toconform withthemovements of thestick. Agraphic recording wasthussecured oftheanimals totallocomotion

about therooma kind ofminiature map ofitsroute and movements throughout the experimental hour. Thesignal (CS) wasa dimming oftherooms illumination. During intervals between trials, theroom was illuminated bya300-w. bulb; atthebeginning ofthe

signal, this light went off, and adim light, barely enough toallow theexperimenter

toperceive theanimals movement, remained on.Theduration ofthissignal was 10sec., followed immediately bytheshock, butnotoverlapping it.Signals and shocks were timed anddelivered byanautomatic clocking device atregular 2-mm. intervals. Theshock fortheavoidance subjects wascontrolled bymeans ofan experimenters key. Theexperimenter watched theanimal closely, raising thekey whenever theanimal raised itsleg,sothatthecircuit wasbroken iftheanimal had itsfootoffthefloor attheendofthesignal. Theanimal wasnotallowed toavoid theshock iftheforeleg wasreturned tothefloor again before thelight came on. Thekeywasalso used foranimals receiving 25percent shock, though inthiscase

theshock wasgiven according toarandom arrangement without regard tothe Atthebeginning oftheexperiment, thekids were brought totheexperimental

animals behavior.

room andthewires attached fortwoadaptation periods without signals orshocks. Theanimals veryquickly became adjusted totheroom andtheharness andceased

112

E. J. Gibson

tostruggle during harnessing. When training began, theyweregiven 20signals perday,theexperimental period lasting 40ruin. Asa ruletheywereruntwoto threetimesa week,thoughthistiming variedslightly withtheseason, fortraining wenton intothesummer. Allanimals received 25 daysof training(500trials). Somewerecarried to 1,000trials,butonlythefirst500willbeanalyzed here.The animalsliveda normallifewiththe restof the herd,in pastureor barn,except during experimental sessions. Results

Varietiesof Reactionto the CS

Ourfirstquestion is,Whatkindsofresponse weremadetotheCS?Dida family or repertory of defense reactions emerge, or wastheremerely foreleg flexion? Theresults definitely indicate a varied repertory. Aclassificationof thedifferent patternsofresponse to theCSthrough25daysof

training hasbeenmade fromprotocols oftheInevgroup. Thefirstdayof training hasbeendisregarded inthedatapresented, because it washardto judge atwhatmoment during thefirst20trials theanimals behavior began to be affected by theCS.Theobserver, of course, alwayshadto makea

judgment astowhether theanimal wasaffected atallbytheCS,butthis wasveryeasyafterthefirstday, for spontaneous activity rapidly decreased. observed follows: A catalogue of the responses

1. Walking or runningbackward, i.e.,retreatingfromthe apparent locus of the shock

2. Walkingor runningforward.

3. TA/heeling (circling)to right or left. 4. Side-stepping.

5. Independent leg-movements without locomotion, e.g.,pawing thefloor, tapping,markingtime,steppingin place.

6. Flexion of eitherforeleg.Flexionvariedas to pattern,but we defined it asliftinga singlelegoffthefloorwithbentknee,whether thelegwasliftedforward or retracted backward (seeFig.6.1). 7. Extension. Extension alsovariedin direction. Wedefinedit as stiff or rigidliftingof the legfromthe shoulder(Fig6.1).

8. Humping withheadlowered to ground. A slightly bowedhead wascharacteristic of manyreactions, but an exaggerated loweringof theheadusuallyaccompanied a humpedback.Thispeculiar postureis

occasionally observed inthepasture ina frightened orangryanimal. 9. Rearing. Rearingis alsocharacteristic of a curious, excited,or

frightened goat.Oneanimal (seeFigure 6.1)reared andwalked back-

wardon its hindlegs.Rearingsustained by leaningwithforefeet againstthewallwasmorefrequently observed.

Role of Shock in Reinforcement

113

Figure 6.1

Fourofthevarieties ofreaction to theCS:a,backward flexion; b,forward extension;

c,rearingwithbackward locomotion; d,humping withloweredhead.

Crouching, lyingdown,andbuttingwereinfrequently observed. Combi-

nations among thenineresponses abovedidoccur, forinstance, wheeling alternately fromrighttoleft,orwalking backward followed byflexion. It isclearthatmanyforms ofaction appear inresponse tothesignal whenthe

animal is not restrained in its movements.

Inhibitionof Response

It waspredicted thatcasesof totalinhibition of motorresponse should sometimes occurduringtrainingin the Inevgroup.Suchinhibition did,in

fact,appearin 24 percentof the trials.Included in thiscountwerecases

where theanimal madeslight headorearmovements butnolegmovement or grossposturaladjustment. Thisabsenceof responsedid not looklike

lackofanyreaction, forevenwhentheanimal stoodcompletely still,itwas frequently notedtolookrigid,tense, oras ifitwerepressing hardonthe

114

E. J. Gibson

floor." Inhibition seldom occurred on more than four successivetrials, and it was generally followed after one or two trials by action again. Frequency of ReactionPatterns There was in the repertory of reactionssome suggestionof a hierarchy of dominance. A comparisonof the four most frequent types of reaction in the Inev group is provided in Figure 6.2. It may be seenthat walking backward is dominant over all other reactions in early training, being the most frequent responseto CS on the second, third, and fifth day of training. On day 2, it was the dominant response for seven of the eight animals. Walking or running forward appearedearly in the courseof training, being second most frequent on day 2, but its curve falls steadily thereafter. Inhibition was actually the most frequent responseon day 4. Not until day 6 did flexion of the shockedleg gain top frequency. Its curve rises as the curves for running fall, but it never becamethe exclusive reaction, occurring only on 52.5 percent of the trials over-all. In general, locomotion was reducedas training went on, both to signals and during intervals between them. Figure 6.3 shows locomotor maps for one animalmadeon the third day and the tenth day of training. By day 10, the animal has cut down both its range and amount of movement. Still later, the animal, unless disturbed, usually stood still at the wall without moving for the entire 40 min. and made single limb movementsto the CS.

120 .

11 0 Z 0

.

/ '

'

.

" " - ,' .

.,. ' _ .. / 1 00

-

.

.-

90

U <{ UJ

80

a:

70

/ "..

. ' . -'

./

i

i

I

./

e,\

/ .\ ,

/ '

'.

,.,-- --

\

- .--' "

.

' \

,e

'\

;'

. . .

"" " ' -

.

FI ""e- .- e

i

' . '. " V

LL

0

60

~

5

z

UJ :) d

40

. ~ I / , ., ~ ;' , ..... - " " ..... \' v / ~ - .... - ... "" ....

30

I I

UJ

a:

20

\

LL

10 0 2

3

4

5

6 7 8 9

101112

131415

16

1718

192021222324

25

DAYSOFTRAINING COMPARATIVEFREQUENCYOF REACTIONSTO SIGNAL Figure6.2 Comparativefrequencyof four typesof reactionto a conditionedsignalwith unavoidable shock.

Role of Shock in Reinforcement

115

On the second day of training, animals in the Inev group made gross locomotor responses to CS on 93 per cent of the trials . But on training day

25, only 30 per cent of the signalsproducedlocomotion; 70 percent of the overt responsesrecorded were single-limb movements. Locomotion was also reduced in the A v group , but here it was related to the experimenter 's criterion for avoidance . Two of these animals were permitted to avoid

the shockonly when the right foreleg was lifted individually. On day 2, 75 percent of their responseswere locomotor. On day 8, 100 percent of their responseswere individual limb movements. These single-limb movements appear to have individuated from locomotion , as has been suggested be-

fore (12, 19). A backward flexion, for instance, is part of a backing movement . In many cases, the observer

wrote , " kid started to back , but lifted

only one leg, and stopped without setting it down." Yet, suppressionof locomotion in the Inev group did not by any meansresult in stereotypy, although the repertory was gradually reduced. Variability with Avoidanceand with InevitableShock It was predicted that frequent shifts from one reaction to another should occur

in group

Inev , whereas

group

A v should

show

a trend

toward

uniformity of reaction to the signal. Suchwas, indeed, the case, as Figure 6.4 demonstrates . Shifts from one reaction pattern to another were very

frequent; both groups averaged13 a day on the secondday of training. By day 25, group Inev still averaged 12 shifts, but group Av now had a mean

of 1. A shift was counted as any changein the reaction to the CS between one presentation of the CS and the next . Repetition of the previous response with

a new

response

added was counted

as a shift . Table

6 .1

presentssampleprotocols, summarizedand codified, for two animalsfrom group Inev. Shifting from one reaction to another is evident in both. -,

JJ~ ~.

i

DAY 3 (TRIALS 40-60)

,-

DAY

10 (TRIALS

180 - 200 )

Figure 6 .3

Locomotor maps showing range and extent of movement of one animal on the third and the tenth day of training.

116

E. J. Gibson "

>-

18

Q

.' 16

a:

W

I

\

I , \

\

1

\

"

~ 14 I 1-

I

12 ' "

\

1

\

r: - -.

\

.

~ :I:

I' 1

'

I

; 8\

' .'

I

,

'.

.

\' - ' ) '

.

25 % Random

lA"

I

~ Reinforcemen

' V'

Inevitable Shock

~ 10 V "'\ .

ffi8 :J 6 ~ 4 ~ 2

\J\

m

~

.

.

\ / .

. \V', .~. I.' .- / \ .1'. ' " .~-- Avoidance .

0 2

4

6

8

10 DAYS

Frequency

of Shift

from

12 OF

14

16

18

20

22

24

TRAINING

One Reaction

Pattern

to Another

Figure 6 .4 Frequency of shifts from one reaction pattern to another under three conditions of shock .

Animal

40 shows

rather

consistent

reduction

in locomotion

and some

stereotypy on day 25, although there is one regressionto walking backward . Animal 7 exhibits no tendency at all to uniformity

and shows a

variety of reactions and frequent shifting on all four days. Avoidance animals, long before day 25, fixated one responseand repeated it consistently . It may be concluded that the data show clearly the extreme variability of reaction under inevitable shock as contrasted with increasing uniformity under avoidance . Change of Reaction with Avoidance There have been several discussions

of the form of conditioned

reaction

shown in shock avoidancewhen the shockwas presentedto the foreleg (3, 9, 22). According to our presentassumptions,the reaction learnedmight be any responseincludedin the animal's escape -defenserepertory. To produce fixation of a given response,the experimenterneedonly wait till the animal makes that response and then consistently omit the shock whenever it occurs

.

As a test of this prediction, we allowed one of our avoidance animals to escape shock by rearing; when the rearing to CS became absolutely regular, the animal was given inevitable shock. After variability of reaction had reasserted

117

Role of Shock in Reinforcement

Table 6.1 Sample Daily Protocols . Animal 40 Animal 7 Day 13 Day 20 Day - 3 Day - 25 Day ~3 Day J13Day - --of20 -- Day - -J25 -I Fl Fl,R Fl H I WB WF WF 2

FI , WB

R

3

WF , WB

WB , FI

4

WF

WB

5

I

WB , FI

FI

H

FI

H

WB

,

FI

PI FI

I

6

I

WB , FI , R

FI

H

7

I

WB , FI (L )

FI

FI

8

FI

WB , R

WB

9

WB

WB

FI

10

WB

WB

WB

11

WB , FI

WB , Wh

12

WB

WB , FI

FI

WB

WF H H LM H I WB WB I

,

FI

FI FI

,

,

FI

FI

FI

WF I FI

FI

FI

WF

I

I

Wh

H

FI

Ex

FI , Ex

H

H

H

LM

WF

FI

Ex

FI

FI

WB

I

WF

FI

I

FI

13

FI , WL

WB , FI

FI

FI

I

WF

14

WB

WB , Wh

FI

FI

I

FI

I WB

I

15

WB , Wh

16

WB , WF

WB

FI

FI

WB , FI

FI

FI

17

WB

WB , FI

FI

FI

H

18

WL

WB , FI

I

H

H

FI

FI

I H

,

FI

WF

,

Wh

WF

,

FI

WL I H ( L )

Ex H Ex I I I FI FI I

Wh

,

WB

Wh

,

WB

Wh

,

WB

,

WF

I

Wh

Wh

I

WF

WF

WF

20 WB ,Wh I : FI FI I I -H - -I .FI Key to abbreviations -flexes IH-head -inhibition FI ( L ) -flexes left leg movements only WB -walks backward Wh -wheel . C \ WF -walks forward Ex -extends leg WL , WR -sidesteps to left or right R -rears LM-independent leg movement (other than flexion or extension ) 19

WB , FI

WB , WF , FI

FI

FI

I

I

E. J. Gibson

118

16 - Avoids shock byrearing

14

I

.

---- Avoids shock byextension of A. forelimb

12

~~ ,

,

! ~,~J~.

i

,

'V

.

I

,

~I

8

,

, .

I

10

.

,

I

.

,

.

,

.

,

I

~ A

I

I

r

'

.

,

6

8

10

12

14

16

18

DAYS

20 OF

22

24

26

28

.

'

I"'

\ I , .~

I

30

\ ,

V

, , ,

4

,

"

~ "

,

2

r "

~ "

I

, I , , I ,

32

34

36

'.

38

TRAINING

Figure6.5 Frequency of shiftsfrom onereactionpatternto anotherfor oneanimalduringtwo phases of avoidancetraining.

A'v'O~3dSJ .:lIHS:lO ~38ViJnN

itself, the animal was again given avoidancetraining, but this time it could escape the shockonly by a sustainedlift of the right foreleg alone. Figure 6.5 portrays the number of shifts of reaction for eachday's training for this single animal. Through day 21, the shock could be avoided by rearing, which the animal beganperforming on its first day of training. Shifts of responseincreasedin the early stages(through day 8) becauserearing was part of a complex pattern involving lying down between trials. When the CS occurred, it rose to a stanceand then reared, placing the forefeet on the wall. As it began to avoid shock, it tended at first to eliminate rearing and to retain the rising from lying position only. After the eighth day, however, the animal gave up lying down, and rearing becamegradually stereotyped. On days 22, 23, and 24, it receivedinevitable shock, with an accompanying rise in the number of shifts, as we should expect from our original assumptions. After day 24, it was allowed to avoid the shock by sustainedlifting of the right foreleg and the number of shifts again diminished until a stereotyped reaction was -achieved ----- - - ---. The - lee was extended forward and waved, usually rhythmically, until the u - rearing light came on. Three days of inevitable shock cut out the reaction completely, and it did not occur again. But a number of other responsesdid, and presumablyit would have been possible to fixate anyone of them.

RandomShock One effect of 100 percent shockis to produce variability of reaction. If the number of shockswere to be decreasedfrom 100 per cent, but in a random fashion so that no one reaction consistently led to avoidance, would variability decrease ? In order to answerthis question, an experimentwas set up with 25 per cent shock, using two animals. This proportion was chosenin order to approximate, at least roughly, the number of shocksreceivedby

Role of Shock in Reinforcement

119

the avoidanceanimals, and it seemed,as an early guess, about right. A new random

distribution

of the five shocks in each group

of 20 trials was used

for every training day. In these animals the reduction of locomotion and the developmentof individual limb movement proceededvery slowly. The animals made a large number of IIconflictful" responses , jerking the head from side to side, circling alternately from right to left, or hitching backward and forward during the signal. But the most interesting result concerns variability. The reducedproportion of shockswas not accompanied by reduction in variability of reaction to the CS, as Figure 6.4 makesclear. Shifting from one reaction to another was very frequent. It is as if trial and error were at a maximum for an insoluble problem, which, indeed, it was. In a sense, the problem is even more baffling than inevitable 100 percent -

shock , for avoidance

occurs without

consistent

.&

relation

to action . Absence

of shock as such, consequently , cannot be the reason for increasing uni -

formity in avoidancetraining. Uniformity is createdby the situation. Consistent avoidance of the shock functions as positive reinforcement . The present

experiment

does not permit

determination

of the nature

of this

reinforcement, but it might be, for instance, either fear reduction or confirmation

of an expectancy .

Discussion

In general, the theoretical predictions made are clearly borne out by the results. But several questions arise concerning additional results which were

not predicted. The theory demandsthat actions following the CS be part of the animal's escape-defense repertory . Some of those observed are

clearly such: locomotion, which begins as running away; rearing and headlowering, which are associatedwith startle or fright in the animal's life history; wheeling, which is a circling type of withdrawal; crouching and butting, which are defensiveand offensivereactions, respectively. But what of the individual limb movements, the segmentalreactions, such as singleleg flexion or extension, or stepping in place? The significanceof these movements can be understood if it is recognized that they are actually

segmentsof the original locomotor response. To quote Liddell: "May it not be that the conditioned limb movements graphically recorded in our experiments are stepping movements , and that these precise reactions represent an experimentally curtailed , or " symbolic " manifestation of run -

ning? . . . The sheep, repeatedly flexing the foreleg in response to the conditioned stimulus, is really trying to run away from a situation about to becomedangerous" (12, pp. 57 ff.). This view is supportedby the late appearanceof single-limb movements in our results. Locomotion clearly dominated the repertory in the early

stagesof training. Schlosberg(17, 18, 8) and Wolf and Kellogg (22) also

120

E. J. Gibson

found that preciseflexion camelate, if at all. The real puzzle to be solved, then, is why locomotion is curtailed . When the animal is restrained in a stock or frame, as were the animals in all previous experiments of this sort ,

the problem is not so evident. But even in the present free situation locomotion was reduced. Liddell, in discussingdevelopment of neurotic manifestations, describedthis curtailment as "self-imposed restraint" (11). But what does this phrasemeanin terms of learning theory? Differentiation and individuation of responseare generally observed in studies of motor learning or development. Perhaps development of motor responsewill tpnd

to QO in the direction

of elimination

of gross

movements

even

when

'-"

the segmentalresponsesare not specifically adaptive- particularly when the gross movements are not themselvesadaptive. This is not generally true of the rodents, as has been pointed out; in their case, freezing or a tense quiescence seems to follow reduction of running or struggling . But in

animalswith , presumably, a somewhatmore differentiatedcortex, locomotion will break up into segmentalresponsesif locomotion is inadequate; if a given segmentalresponsedoes not answer, trial and error persists, yielding either a variety of segmentalresponsesor a combination of thesewith regression to locomotion .

Such, in general, was the picture when the animals received inevitable shock. Even as late as the twenty-fifth day of training, variability persisted in these animals; segmentalresponsesoccurredbut were not consistently repeated. Becauseuniformity did develop in animals permitted to avoid the shock, we concludethat the hypothesis of a dual role for unavoidable shock is supported. If it follows a motor response, this response tends eventually to be suppressedand supplantedby another. The result can be described as an interminable process of trial and error , by which the

problem is never solved. This conclusion is in general agreement with Estes' (4) contention that shockcausessuppressionof precedingresponses , although it does not accomplish complete extinction. In the course of the long trial and error with unavoidableshock observedhere, the animals often regressedbriefly to responsespreviously tried and dropped. The other role of shock- strengthening of a conditioned emotional reaction to the signal- we find consistent with the behavior of both avoidance

and

nonavoidance

animals , for , as we have

tried

to show , the

resulting action is appropriate to a state of internal " emergency ." There is, thus, support for Schlosberg (19) and, later, Mowrer (15) in assuming two

processesof modification. One dependssolely on paired stimulation with CS and US (the conditioned

fear ), whereas a second process follows

the law

of effect (fixation or suppression of the motor reaction to fear). The results tend to support Mowrer 's contention that motor reactions (at least other than very

diffuse

emotional

ones ) are modified

in accordance

with

the

secondprocess. It is impossibleto agreewith Culler that "the CR begins as

Role of Shock in Reinforcement

121

a copyofURandthengrows intosomething different. Inthefirststages, itmaybeindistinguishable fromUR;indeed withdecorticate preparations

it remainsindistinguishable throughout. Normally, however,CRdiffer-

entiates intoa specific preparation fortheoncoming Us,,(3,p. 142).It

appears,rather,thatthe greatestresemblance betweenthe CRandthe UR

comes inthelatestages oftraining whensingle-limb flexion develops. The early reactions to the signal take the form of rapid backing or running and are not copies of the UR. It willbeaskedby thesupporters ofGuthriewhetherthereactions to the

shock itself didnotshowequally widevariation, withthepossibility that

eachreactionto the signalmightvaryin accordance withthereactionto

thelastshock. Thedataasheresummarized donotofferproofto the contrary, butit wastheexperience ofalltheobservers intheexperiment thatnosuchcorrelation existed. Variation ofresponse to thesignal was

enormously greaterthanthatto shock, notonlyinnumber ofshiftsbutin varietyof patterns.In otherwords,anticipation of noxiousstimulation seemsto havemorewaysof displaying itselfthanreactionto thenoxious

stimulation itself.

Summary and Conclusions

I. Inlinewithrecenttwo-factor learning theories it issuggested that

conditioned responses developed withinevitable shockwillexhibit characteristics derivedfromtwo functions of the shock;the shock

reinforces a conditioned emotional statewhichinstigates a motor response ofa defensive character, andit alsosuppresses themotor

response,causingit eventuallyto be supplantedby anotheraction

belonging totheanimals natural escape-defense repertory. Kidswere trained ina situation permitting freelocomotion under threearrangements ofshock: (a)shock inevitable, (b)shock avoidable bya given consistent action, and (c) random shock. 2. At least ten differentreactionsto the CS were observed.The dominantreactionin earlytrainingwas locomotionbackward. All

forms oflocomotion became lessfrequent thansegmental movements,

suchasflexion, in laterstagesof training.

3. Inhibition ofresponse (standing still) occurred on24percent ofthe trialsin the animalsgiveninevitableshock.

4. Withanimals giveninevitable shock, therewerefrequent shifts fromonereaction to another, evenonthetwenty-fifth dayoftrain-

ing;but whenshockcouldbe avoided,therewas a definitetrend

toward uniformityof reaction.

5. Afteran animalhadachieved a uniform response whichavoided the shock,introduction of inevitable shockbroughtrenewedvan-

122

E. 1. Gibson

abilityof reaction. A different avoidance reactioncouldthenbe fixated.

6. Randomdeliveryof shockin 25percentof thetrialsdidnot reduce variabilityas comparedwith 100percentshock.

7. Inevitableshockgave a generalpictureof continuoustrialand error.Therewasno supportforthe Pavlovian viewthatshockactsto reinforcea withdrawalmovement,in the sense of increasingthe

probability ofrecurrence ofthesamemotorreaction; it had,instead, a tendency to suppress a preceding action,withtheresultthatanother took its place. Notes

1.Thisinvestigation wassupported(inpart)by a researchgrantfromtheNationalInstitute of Mental Health, U.S. Public Health Service.

2.Grateful appreciation isexpressed to Dr.Howard Liddell, Dr.A.UlricMoore,Mr.James Block,and Mrs.MiriamSalpeter, colleagues whosharedin the observations to be reported. Dr.Moorewasresponsible fordesign andconstruction ofapparatus.

3.It maybe notedthatJames(7)foundit impossible to condition legflexion in the opossum, although itcould beconditioned toattack ortoplaypossum; Liddell andhis

co-workers (12)foundit impossible to developa preciseflexionin the rabbit,which continued to reactwithstruggling movements evenafterprolonged training; Brogden (1)

notedthattheguinea pigsits tight orfreezes before delivery ofinevitable shock; inthis laboratory a youngramrecently exhibited, instead oflegflexion, pawing andbutting responses to the CS.

4. Fourof theanimals wereaccompanied in theexperimental roomby theirmothers. This feature was irrelevant for the present purpose.

References

1.Brogden, W.1.Theeffectoffrequency ofreinforcement uponthelevelofconditioning. 1. exp. Psycho!.,1939, 24, 419431.

2.Brogden, W.J.,Lipman, E.A.,andCuller, E.Theroleofincentive inconditioning and extinction. Amer.]. Psycho!., 1938, 51, 109117.

3. Culler,E.A. Recentadvancesin someconceptsof conditioning. Psychol.Rev.,1938,45, 134153.

4. Estes,W.K.Anexperimental studyofpunishment. Psycho!. Monogr., 1944,57,No.263.

5. Hilgard, E.R.,andMarquis, D. G. Conditioning andlearning. NewYork:AppletonCentury,

1940.

6. Hull,C. L. Princip!es of behavior.New York:Appleton Century,

1943.

7.James, W.T. Anexperimental studyof thedefense mechanism in theopossum, with emphasis onnatural behavior anditsrelation to modeoflife.J.genet. Psycho!., 1937, 51, 95100.

S.Kappauf, W.E.,andSchlosberg, H.Conditioned responses inthewhiterat:III.Conditioningasa function ofthelengthoftheperiodofdelay.]. genet. Psycho!., 1937,50,2745 9. Kellogg, W. N. Evidence for bothstimulus-substitution andoriginalanticipatory responsesin theconditioning of dogs.].exp.Psycho!., 1938,22,186192.

Role of Shock in Reinforcement

123

10 . Liddell , H. S.The nervous system asawhole : theconditioned reflex . In]. F. Fulton , Physiology ofthenervous system (2ndEd .). New York : Oxford Univ . Press , 1943 . Pp .491 -522 . 11 . Liddell , H. S.Conditioned reflex method and experimental neurosis .Ch . 12in]. McV . Hunt (Ed . ), Personality and the behavior disorders . New York : Ronald Press , 1944 . Pp .389 -412 . 12 . Liddell , H. 5., James , W.T., and Anderson , O. D. The comparative physiology ofthe conditioned motor reflex . Camp .Psycho /.Monogr ., 1934 , II,No.51. 13 .Maier ,N.R Schneirla ,T.C.Mechanisms inconditioning .Psycho /.Rev ., 1942 ,49 , 117 -134 ..F.,and 14 . Miller , N. E. Studies offear asanacquirable drive : I. Fear asmotivation and fear reduction as reinforcement in the learning of new responses .]. expo Psycho /., 1948 , 38 , 89-101 . 15 . Mowrer , O. H-solving . Onthe of.,learning :A reinterpretation of"conditioning " and "problem ."dual Harv .nature educ .Rev 1947 , 17 , 107 -148 . 16 . Mowrer theory and personality dynamics . New York : Ronald Press , 1950 . , o. H. Learning 17 . Schlosberg responses inthewhite rat.]. genet . Psychol ., 1934 , 45 , 303 -335 ., H. Conditioned 18 . Schlosberg , H. Conditioned responses inthewhite rat : II. Conditioned responses based upon shock totheforeleg .].genet .Psychol ., 1936 ,49 , 107 -138 . 19 .Schlosberg ,,H .The relationship success and thelaws ofconditioning . Psycho /. Rev ., 1937 44 ,379 -394 . between 20.Sheffleid ,F,.165 D.Avoidance and thecontiguity principle .].compo physiol .Psychol ., 1948 ,41 -177 . training 21.Skinner , B,.1938 F. The oforganisms : Anexperimental analysis . New York :Appleton Century . behavior 22.Wolf ,1.5.,and Kellogg ,W.N.Changes ingeneral behavior during flexion conditioning and their importance forthelearning process .Amer .].Psycho /., 1940 ,53 , 384 -396 .

TheEffectofProlonged Exposure to Visually PresentedPatternson Learning to Discriminate Them

EleanorJ. Gibson,RichardD. Walk Introduction to Chapters7 and 8

Rearing experiments witheither impoverished orenriched environments were very fashionable inthefifties. Interest began withhistories ofindividuals deprived of normal environmental rearing conditions, themostextreme cases being persons bornblind,whohadlaterhadsightrestored bycataract removal (vonSenden

1932,1960), andpersons thought tobereared inthewild,likethewildboyof Aveyron (Itard 1894, 1962). These cases hadalways stirred theimagination of

philosophers, ofcourse. ButDonald Hebbsbook(Hebb1949), as wellas the

dark-rearing experiments ofLashley, (Lashley andRussell 1934), Hebb(1937)

and Riesen(1947),broughtquestions of whether and howthe environment

exercised a roleindevelopment totheforefront oflaboratory investigation. Hebb thought thatschemasweredeveloped fromquitespecific experiences thataccounted forperceptual organization, withgroups ofcortical neurons exciting and reexciting oneanother. Hebbhadbeeninfluenced bySendens andRiesens pub-

lications, remarking that BothSendens dataandRiesens saidthatthereis no

pattern perception without experience (Hebb 1980,p. 295).(Thisdespite the contraryresultsofhisownexperiments.)

Hebbs book andcontemporary research ontheneural development andorganization ofthevisual system (Hubel andWiesel 1959) inspired a spateofrearing experiments, involving deprivation oflight,orpatterned light(seeE.Gibson 1969,

ch.12,forasummary). Hebbs ownearlier experiments were designed tostudy the

effect oflightdeprivation onvisualdiscrimination inadultrats.Butlateronhe

designed an earlyenvironment program in which younger members ofhis department at McGill participated (Hymovitch 1952,Thompson andHeron 1954, Meizack andScott1957). Thisprogram emphasized positive contributions oftheenvironment. Hebbsaidin hisautobiography, Thisprogram wasless dramatic thansensory-deprivation workbutperhaps moreimportant. It wasa Journal ofComparative andPhysiological Psychology, 1956,49,239242.

126

E. J. Gibson&: R. D. Walk

major inj1uence in persuadingpsychologists that the IQ is not built-in at birth, andso a factor in suchthingsas theHeadStart program" (Hebb 1980, 300). This work, interestingand valuableas it was, seemedto have rather little bearingon Hebb's theoryof visualcorticaldevelopment , with its emphasis on cell assemblies createdin the beginningby eyemovements following linesand angles in quite specificpresentations . As a theoryof perceptuallearning, it was of real concernto me, and I proposedto Richard Walk that we rear infant rats in a controlledrat environmentwith the additionof trianglesand circleson the cage walls. Though thesewere hardly typical of a rat's normal environment , the opportunityto view them daily throughoutearly development might result in appropriatecell assemblies developing , and thus facilitate later learningof a discriminationcuedby them. The circlesand triangleswere cut out of metal, paintedblack, and hungon the wire meshof the cages . When the animalswere testedlater in a discriminationbox with the sameblackfigurespaintedon the doors, they learnedsignificantlyfaster than their litter-matesrearedwithout similar exposure . We might haveleft it there, with a triumphant "Well, well!" But we didn't. We pushedon to further experiments (ninein all), varyingsuchfactorsas timeof exposure(was therea "critical period" early on?), variationsin the patternsat testing, the effectsof differentialreinforcement , and, finally, a comparisonwith dark-rearedrats ( Walk, Gibson, Pick, and Tighe1958; Gibson, Walk, Pick, and Tighe1958; Gibson, Walk, and Tighe1959). Resultsin thesereplications , alas, wereseldomas clear-cut as they had beenin theoriginalexperiment . In fact, rats deprivedof all light during theirfirst ninetydaysof rearinglearnedthe trianglecirclediscriminationas readilyas rats rearedwith an opportunityto view these patternsfor theequivalenttime. We had to conclude that we hadnofirm evidence that "perceptual discriminationof a form is an achievement resultingfrom an integrationprocess of theneuralelements involved" (Gibson, Walk, and Tighe1959). What might accountfor thepositiveresultsof thefirst experiment , and oneor two later ones? Examinationof all the experimentsrevealedone factor that appeared to favor a facilitationof laterdiscrimination learning - that waswhether thepatternshungon the cagewall werecut out, to providedepthat an edge , or werepaintedon a surface , with no edgedepth, like pictures . Our last experiment , consequently , was a test of cutoutsvs. painted patterns . The resultsof this experiment werepositive, in thesensethat thecutoutsprovidedthe only evidence for facilitationby prior experience with thepatterns . But theeffectwas very weak. Locationof an objectas somethingseparateand unitary in the layout is without doubt importantin organizingperceptionandgiving it meaning . But the notion that perceived form of an objectdepends on constructing a schema from repeated exposure to elementary pieceslike linesandanglesthatfall upontheretinaandare assembled in a phasesequence was unsupported . From my presentapproach that perception doesnot start with somethingpictoriallike a retinalimage, but is rathera searchfor informationaboutthingsin theworld- that is not surprising.

ProlongedExposureto VisualPatterns

127

Wegottheresults weshouldhave,andtheymakegoodsense, I think.Passive

exposure totwo-dimensional displays doesnotresultinperceptual learning, even when reinforced, anddeprivation ofsuchexposure hasnoilleffects at allonlater pattern discrimination learning. Exposure ofsimilar patterns isquite aseffective as exposure ofthesameones,whenanykindofeffectoccurs, as it didwithcutouts.

Perceiving anobject segregated fromitssurrounding layoutmayhaveattentional valuethatcarries overtoanother situation, buttheobject neednotbetheidentical one.Weneeded a theoryof perceptual learning, but not theonethat ledus to performtheserearingexperiments.

Recentliteratureon the developmentof discrimination has shownan

increasing trendtowardacceptance ofempiricistic explanations (2,9).That

ability todiscriminate visually presented patterns develops withtheexpe-

rienceandenvironmental reinforcement ofthegrowing animal maybethe case,buttheevidence forthisviewis stillinconclusive. Earlystudies by

LashleyandRussell(11)and by Hebb(8)on the rat favoreda nativistic

interpretation ofthedifferentiation ofvisual qualities, butlatercomparable studieswiththechimpanzee andpigeon(13,14)apparently favored an

empiricistic explanation. Recentexperiments by studentsofHebb(5,6, 10)

haveemployed anenrichment technique, withresults which appear to favor alearning hypothesis. These studies attempted toprovide agenerally rich environment andusedascriteria testsofa rathergeneral type.If

opportunity to viewa variedandpatterned environment is important in thedifferentiation ofvisual qualities, wedonotknowhowgeneral orhow

specificthe relevantexperience mustbe.

Theexperiment tobereported proposed to investigate thedependence

ofvisualformdiscrimination in adultratsona specific variation in visual stimulation duringgrowth. Tothisend,anexperimental groupofanimals

was raisedfrom birth in cageswhichexhibitedon the wallscirclesand trianglesidenticalin formwithoneslaterto be discriminated. Thecontrol

groupwasraisedunderthesamestandard conditions butwithout opportunityto seetheseformsbeforethediscrimination learning began.If the

opportunity to viewspecific formsfavorsdevelopment of the abilityto

differentiate themin a laterdiscrimination learning problem, theexperimental animals shouldlearnfasterandshowa higherproportion of Ss reaching thecriterion thanthecontrolgroup. Method

Rearing

TheSswerealbinoratsrearedfrombirthin identical i-in.wire-mesh cages

measuring 15by 13by 9 in.Thecageswereplacednextto eachotherina small,

128

E. J. Gibson& R. D. Walk

softly

lighted

three

empty

sides

fourth

side

the

, 4

room

wall

metal

of

the

total

the

of

four

, litter first

day

.

The

each

young so

begun

C2

7

white

blank

ft . from

the

the

cage

cardboard

wall

of

mesh

mates

walls

the

room

was

, a

in

two

. on

a

on

on

the

water

were

circles

side

.

the

ceiling

bottle

of

on

one

left

on

the

, two

experimental

four

circles

were

patterns to

were

fastened

The

. These

relationship

used

( experimental

litters

, and

litters

food

and

sides

of

were at

animals

. E2

1 ),

10

)

when

black 3

were

changed

water

. All

the

cages

in

.

in in

during

of

(n

=

(n

pups

age

,

in

separate 90

control

8 )

and

C2

the

approximately

two

experi

. These

( experimental

apart

and

of

were

E2

E1

days

=

weeks

females were

five (n

split

four

and

litter

2 ) . Litters

born

split

and

,

, litter

were

, not

males

the

animals

and

random

( control

weaned

that

when

was

litters

were

divided

top only

a

experimental

3t a

: E1

These

other

the

patterns

follows

between of

by

and

-

.

1 ) , and

born

of

assure

groups

as

interval

the were

were

to

animals

numbered

( control

surrounded mesh

triangles

stimulus

group

A

. At

cages

triangles

experiment

be

was wire

cage

equilateral

occasionally

mental

mesh the

the

two

and

position

the

the

.

,

diameter

cage

from

within

walls

forms

the

ft . from

food

the

. Each

inches

. Visible

, and On

room

, several

and

,

=

at

eight The

2 ),

=

of

2 )

C1

were

the

born

long

within

two

a

days

weeks

.

old

(n

or

will

litter

because

one

cages

days

C1

9 ) were

were

and

,

old

sexes

.

were

experiment

was

.

Apparatus The

apparatus

was

discrimination as

described

side

at

in the

stimulus of

holders

. The

glass

. A

The

stimulus

were for

25

- in

, and four

and

. - wide slid

end into

. bulb

on

the

of

were the 3

25 first

painted

background a

side food

stimulus

holders box

flat

a

by , one

*

, and

flat

white

, and

circle

was

pushing

open the

4i

in

circle

. by

4i

of

section

a

was

. in

door and

one

and

diameter

- in

. square

in

the with

metal

the

side

stimulus by

illumination

circle

Ii

.

by

covered

only

black in

the

,

side - in

and

front

the

- shaped

were

- in

. in

had the

V

constructed

partition

2i

holders

Two

floor

patterns

each furnished

with

( 7 ).

false

. center

floor

stimulus

a

. The - in

the

. The . The

Ii

grooves black

. above

Grice

stimulus

chambers

into a

in

two

the

fitted

in . on

discrimination

and

choice

painted

obtained

by

together

between

white

animal

separate

both

doors was

described

( 1 ) . The

grooves

mounted

holders

one

joined

Lawrence

. Masonite

triangle

each

of were

apparatus -w

painted

equilateral them

10

apparatus

were

modification

Baker

holders

the

a

compartments

.

triangle and

the

doors

center the

in

. There triangle

,

.

TrainingProcedure Pretraining . Animalswereplacedon a 24-hr. feedingcyclefor approximately one weekprior to the start of experimentation . They weregiven threeto four days' trainingin obtaininga smallquantityof wet mashfrom the stimulusholdersby pushingopenthe door in the center.The stimulusholderswerepaintedflat black for this pretraining . Thedooron only onesideof thediscrimination box wasraised at a time. As soonasthe animalobtainedthe food from the food cup, E lowered the door in front of the stimulusholder. The doorbetweenthe two discrimination boxeswasthenopenedfor the next trial, andthe animalsecuredfood by pushing

ProlongedExposureto VisualPatterns

129

itsnoseagainsttheblackstimulus holderat theopposite end.Theanimal ateten timesfromthecupintheholderinthefollowing order:RLLRRLLRRL.

Discrimination training.Duringdiscrimination training bothMasonite doorson

thechoice sideoftheapparatus were raised, exposing thetwostimulus patterns sidebyside. Bothstimulus holders werebaited. Assoonastheanimal pushed

against onestimulus door,theMasonite doorinfrontoftheopposite stimulus was closed. Ifthechoice wascorrect, theanimal wasallowed toeatthewetmashinthe

foodcup. After 60sec., thedoorbetween thetwocompartments wasopened and

theanimal proceeded to theopposite end,wherethenextchoice wasmade.Ifit

wasincorrect, a modified correction procedure wasfollowed. Bothdoorsin front

ofthestimulus holders wereclosed. After60sec.theanimal wasallowed tomake

a choice intheopposite discrimination box.Animals wereallowed upto three

errorspertrial.Following thethirderrorthedoorin frontof thecorrectstimulus

figure remained open, andtheanimal wasallowed toeatfrom it.Thisprocedure

meant thattheanimal ateequally oftenoneachsideoftheapparatus. Tentrials weregiveneachdaywitha maximum ofthreeerrorspertrial. Thepositive stimulus waspresented in thefollowing order:RLRRLLRLLR; LRLLRRLRRL; RRLLRRLRLL; LLRRLLRLRR. Theorderwasrepeated everyfour days.Forhalftheanimals ineachgroupthecircle wasthepositive stimulus, and

for halfthe triangle. Animals wererun untiltheyattaineda criterion of 18

outof20correct responses, withthelasttenconsecutive responses correct (one daysrun),oruntiltheywererunintheexperiment for15days(150trials). After theexperimental session animals wereallowed to eatfoodpellets for1 hr.The hunger drivewasa function ofapproximately 22 hr.deprivation. Each ofthetwo Esranone-halfof theexperimental andone-half of thecontrolanimals.

Results

Thenumber ofdaysofdiscrimination training andtheerrors(initial and repetitive) arepresented in Table7.1forbothgroupsof animals. In the

table areindicated thesexandlitterofeachanimal. Thesecond litters (LE 2 andLC ), 2 it willberemembered, weresplitatbirthandthusprovide a somewhat bettercontrolled population. It is obviousfromthe tablethat thereisa difference between experimental andcontrol groups. Outofthe control group,only1 animal reached thecriterion during15daysof

training. But15of the 18experimental groupanimals did.Bythechi-

squaretest, this difference is significant at betterthan the .001levelof

confidence. Ifwecalculate thechisquare foranimals ofthesplit-litter

groupsonly,usingFishers exacttest(4),thesignificance ofthedifference

isbetween .002and.001.Theerrors, bothinitial andrepetitive, reflect the

same trend.

Afurther check ondifferences inthepopulation studied ispossible by areexactly 50in100thatthereisanydifference between sexgroups.

testingmalesagainstfemales. Whenthiscomparison ismade,thechances

130 Table

E. J. Gibson& R. D. Walk 7 .1

Number J. " - U~' V ' -~ '-' of Experimental &

Days - - J ~ Trained - . - -- -- - and - - - Errors -

for

Group No

.lAnimal ~ ' ~~~' W"

the

Experimental ..

and

Control

Control

Group -

.

No

Days

Initial

Repetitive

Run &, ~ .

Errors - -- - - -

Errors

Animal

12 .

39

44

LCl

Groups

2 J

.

Days

Initial

Repetitive

Run

Errors

Errors

14 .

25

11

LEl

30

d

LEl

31

J

7.

22

34

LCl

~

15

68

52

LEi

32

J

11 .

28

19

LC2

4 ~

15

50

44

LEl

33

J

5.

11

9

LC2

6 ~

15

59

30

LE 1 35

~

15

32

21

LC2

7 ~

15

67

18

LEl

37

J

14 .

40

41

LC2

12

~

15

60

28

LEl

40

J

7.

24

29

LC2

15

~

15

80

80

LEl

41

d

8 -

24

24

LC2

11

0

15

66

72

LEz

44

~

7.

23

28

LC2

5 J

15

68

80

LEz

47

~

10 .

25

16

LC2

13

J

15

74

117

LEz

62

~

15

70

73

LC2

14

J

15

84

92

LEz

63

~

9.

23

16

LE2

64

~

12 .

39

50

LE2

43

J

15

57

51

LE2

45

J

13 .

28

12

LEz

46

J

10 .

25

16

LEz

60

J

9.

30

23

10 .

37

34

20

LEl 616 Mean 10 50 .00 14 .91 63 .73 J.Y.l"-UJ .' ~_.._ - 32 ---.0 --6 30 -.73 . . 56 . I" ~Indicates that animal reached criterion ;the criterion day 'strials are included innumber of days run . Figure 7.1 showsthe learningcurve for experimentaland control groups. Percentageof correct responsesis plotted against days of training. The animalsthat reachedthe criterion are included in the percentageson the assumption that they would continue at their final level of performance. The curves show that the groups begin to diverge by the third or fourth day of training and diverge increasingly thereafter until the tenth day, when a majority of the experimentalgroup had learned. Discussion The results presentedshow conclusively a differencein easeof learning a circle-triangle discrimination between the group reared with these forms exhibited on the cagewalls and the control group. Sincethe control group had the same conditions of training (and pretraining), the same living

Prolonged Exposureto Visual Patterns

131

100

EXPERIMENTAL

GROUP

9

8

7 "

p. . ,

,, "

6

, '

"

A

,

" ,

" " 0

' 0"

,

' d

p' , , ~. o- - - . o- - - - O- " "

"

"

'

, d"

"

...0- - - - 0"

50

,

"

"

p. ,

CONTROL

GROUP

;,; \

,

' t:(

0 1

2

5

6

7

8

9

10

11

12

13

14

15

DAY

Figure 7.1 Learning curves ofcorrect responses per day , forthe experimental and control groups ., inpercentage

S3SNOdS3 ~ l.J3~~OJ::JO39\tl.N3J~3d

conditions , and, in our secondlitters, the sameheredity, thedifference must be attributedto someadvantagearisingfrom the opportunityto look at theforms. Thisadvantage couldbesomethingspecificwhichhappensearly in the animals ' development , analogouswith "imprinting" (12) or with Hebb's postulateddevelopment of reverberating neuralcircuits(9). On the other hand, a learningtheoristwho favors "hypotheses " as a factor in learninga discriminationmight suggestthat seeingthe formson the cage walls favors formationof the correcthypothesis . Sincethe forms were left on the wallsduringthe learningperiod, it is not possibleto conclude that earlyexperience in viewingtheformsis the basisof the effect. Suitable controlsareat presentbeingrun to clarifythis point. Sinceresearch in discrimination learninghascenteredroundthe continuity hypothesisin recentyears,it might beaskedwhetherthepresentresults tend to confirmor deny this hypothesis . The animalsin the experimental groupprofited, in the discrimination task, from an opportunityto view the two formswithout any differentialreinforcement of them. Nondifferential reinforcement in viewing thesecouldhaveoccurred , sincethe animalsate anddrankin their presence . Spence 's 1936article(15) suggeststhat some degreeof positiveexcitatorypotential,irrespective of differentialreinforce ment, would be consistentwith fasterlearningwhendifferentialreinforce mentis introduced . On the otherhand, the valuesselectedfor his analysis

132 are

E. J. Gibson& R. D. Walk

purely

arbitrary

reinforcement Bitterman the

and

course

of

is

a

by

an

with

the

on or

of

test

the

,

.

occurs of

in

differential

further

finding

that

reinforcement

the

. transfer

The from

complications

intro

-

.

the

facilitation

hypothesis

positive

without

described for

the

lack

their

the

reinforcement

problem

despite

nondifferential

stimuli

time

resulting

, by

clearly

of

critical

the

stimuli

of

nondifferential of

differentiation

beclouded

demonstrate

application

effective

statement

perceptual

test is

effect

to

that his

that

retarding

research

specificity

concluded

viewing

optimal

concluded refutes

conclusion

seem

specific

Further

be or

experience this

general

in

duced

cannot

(3 )

sheer

results

.J experience

it

confirms

. But

Dresent

is

so

Elam

reinforcement

there

, either

will visual

of

investigate

whether

experience

,

discrimination

and

there

the

learning

relative

.

Summary

This

experiment

exposure

sought

to

adult

animal

raised

from

Animals

certain learns birth

of

cages

.

and

made

- illuminated

control

errors

with reinforcement

the

them

,

forms

to ,

facilitated

ease

groups by

mounted

control

circle

group

. It

discriminated the

discrimination

was

,

even

an were

cardboard walls

.

of

their

duration

of

days

- triangle

criterion

which

animals

the the

90

the

continued

with

white

on

approximately a

and

of

throughout

reached

be

early

the

Two

had

learned

group

of on

surrounded

birth

were

the

.

also from

groups

than

,

cages

animals

experimental

effect

visually

group triangles

and

fewer

differential

well

the

the

experience

discriminate

and

the

presented

to

When

experimental of

determine

,

experimental

circles

experiment ~

Animals

in

the

black

to

forms

old

,

the both

discrimination significantly

concluded

in learning

that

the

. faster

absence

visual

of

.

References 1. Baker , R. A., & Lawrence , D. H. The differentialeffectsof simultaneous and successive stimulipresentation on transposition . ]. compo physiol . Psychol ., 1951, 44, 378- 382. 2. Beach , F. A., & Jaynes , J. Effectsof earlyexperience uponthebehaviorof animals . Psycho /. Bull., 1954, 5I , 239- 263. 3. Bitterman , M. E., &: Elam, C. B. Discriminationfollowing varying amountsof nondifferentialreinforcement . Amer.J. Psycho I., 1954, 67, 133- 137. 4. Fisher , R. A. Statistical methods for research workers . (11thEd.) New York: Hafner, 1950. 5. Forgays , D. G., & Forgays , JanetW. The nature of the effect of free-environment experience in the rat. J. compo physiol . Psychol ., 1952, 45, 322- 328. 6. Forgus , R. H. The effectof early perceptualleaming on the behavioralorganizationof adultrats. J. compo physiol . Psychol ., 1954, 47, 331- 336. 7. Grice, G. R. Theacquisitionof a visualdiscrimination habitfollowingresponse to a single stimulus . J. expo Psychol ., 1948, 38, 633- 642. 8. Hebb, D. O. The innateorganizationof visualactivity. I. Perceptionof figuresby rats rearedin total darkness . J. genet . Psychol ., 1937, 51, 101- 126.

Prolonged Exposureto Visual Patterns

133

9.Hebb ,D.O.The organization o/behavior .New York :Wiley ,1949 . 10 .Hymovitch , B . The effects of experimental variations on problem solving inthe rat .]. compo physio /.Psycho /., 1952 ,45 ,313 -321 . 11 .Lashley ,K.S.,.& Russell ,J.T./.The mechanism ofvision test ofinnate organization ].genet .Psycho , 1934 ,45 , 136 -144 . .XI.Apreliminary 12 .Lorenz ,K.Der Kumpan inder Umwelt des Vogels .].Om .Lpz .,1935 ,83 , 137 -213 . 13 .Riesen , A . H . The development of visual perception in man and chimpanzees . Science , 1947 ,106 -108 . . 107 14 .Siegel , A..].I.compo Deprivation visual form definition in.the ring dove .I. Discriminatory learning physio /.of Psycho /., 1953 ,46 , 115 -119 15 .Spence ,K..W.The nature ofdiscrimination learning inanimals .Psycho /.Rev ., 1936 ,43 , 427 -449

8 The Effectiveness ofProlonged Exposure to Cutouts vs.Painted Patterns forFacilitation of Oiscrirnina tion Richard D. Walk , Eleanor ].Gibson , Herbert L.Pick ,Ir., Thomas I. Tighe Experiments by the authors (Gibson &Walk ,1956 ;Gibson ,Walk ,Pick ,& Tighe , 1958 ; Gibson , Walk , & Tighe , 1959 ; Walk , Gibson , Pick , & Tighe , 1958 ) have shown that prolonged exposure to visually presented patterns may ,under certain conditions ,facilitate discrimination of these patterns in a learning situation ! An auxiliary finding was that in experiments where the patterns exposed inthe cage were cutouts ,facilitation always occurred , though in other experiments where the same patterns were painted on a plain rectangular background ,there was no facilitation .et This effect has been previously commented on ( Gibson et al . , 1959 ; Walk al . , 1958 ) , but no single experiment has directly compared the effectiveness of cutouts with painted patterns .The present experiment was designed to do so . follows The experiment included four groups of albino rats , treated as : Group Iwas reared with cutout patterns (triangle and circle )on the walls of the living cages ; Group II was reared with the same patterns painted on metal rectangles on the cage walls ; Group III was reared with plain color rectangles ,like Group II,but without the painted pattern ;Group IV was reared with nothing mounted on the mesh walls of the living cages . From the results ofthe previous experiments following relationships among the four groups should hold . ,the 1.The cutout -pattern group (I)of should be superior tothe control group ( I V ) . This is a replication our first experiment ( Gibson & Walk ,1956 ). 2.The painted -rectangle pattern group (II()Ishould not be significantly different from the plain group II ) ( Gibson et al . , 1959 ; Walk etal ., 1958 ). 3. The cutout-pattern group (I) should be superior to the paintedpattern group (II) and the plain-rectangle group (III). This difference Journal of Comparative andPhysiological Psychology , 1959 , 52, 519- 521.

136

R. D. Walk, E. J. Gibson, H. L. Pick, Jr., & T. J. Tighe has not been tested directly, but was hypothesized in two papers (Gibson et al., 1959; Walk et al., 1958). 4. The position of the painted-pattern group (II) and the plainrectangle group (III) in relation to the control group (IV) cannot be predicted from previous research. If only the identical pattern, as a cutout, will produce facilitation, Groups II and III will be indistinguishablefrom Group IV . If the rectanglesfunction as cutout figures and are sufficiently similar to the test figures to yield transfer (Gibson et al., 1958), then Groups II and III should learn the discrimination faster than Group IV .

Method Subjectsand RearingProcedure The5s, 84 albinorats, wererearedin identicalt -in. wire-meshcagespaintedwhite andmeasuring16 by 13by 9 in. Thecageswereplacedin the centerof compartments(27 by 23 by 22 in.) whosefloorsand walls werecoveredwith masonite paintedflat white. Illuminationwas suppliedby fluorescentlight 7 ft. abovethe cages . Twelvepregnantratsfrom RocklandCountyFarmlitteredwithin a spanof ten days. Litterswere split amongthe four groupsso that eachgroup includedoffspringof all 12 mothers . The rectanglesor cutoutswere mountedon the cage wallsof theappropriategroupsbeforethe pups' eyesopenedandremainedon the wallsthroughoutthe experiment . The patternsusedfor GroupI weretwo sheetmetalcircles(3 in. in diameter ) andtwo sheet-metaltriangles(31 in. on eachside). paintedblack. One form wasmountedon eachwall. For GroupII, thesepatterns were paintedon white rectangles , 41 in. by 5 in. Group Ill's cagewalls held four 41-in. by 5-in. rectangles paintedwhite. Discrimination Training Trainingbeganwhenthe animalswere90 daysold. They wereplacedon a 24-hr. feedingscheduleand then habituatedto the apparatusandpretrained . The modified Grice-type discrimination box alreadydescribed(Gibson& Walk, 1956) was used.It containedtwo V-shapedcompartments joinedtogetherso that the animal alternatedfrom oneto the otherwithout handling.Stimulusholderspaintedblack wereusedfor pretraining . Animalslearnedto takefood from holderswith doors openedwide, andthedoorsweregraduallyclosed.Pretrainingcontinueduntil the animalpushedopencloseddoors, alternatingfrom onecompartment to the otherI at leasteight timesin 10 min. Pretrainingrequireda meanof 6.8 days. For discriminationtrainingthe blackstimulusholderswere replacedby white holderson which was paintedeither a black triangle(3 in. on eachside) or a blackcircle(2i in. in diameter ). Whenthe door separatingthe two compartments was raised , an animalfacedthe two stimulusholderssideby side. If the animal chosecorrectlyby pushingthe door of the positivestimulus , it waspermittedto

Pre-exposureand Discrimination

137

eatanda blackdoorwaslowered infrontoftheotherholder, butif theanimal

moved thedoor ofthenegative stimulus (which waslocked within.ofplay), the intheothercompartment. Three errors (oneinitial andtworepetitive) were

doorswerelowered infrontofbothstimuli andtheanimal madea second choice

allowed oneachtrial. When ananimal made a second repetitive error, thedoor

waslowered infrontofthenegative stimulus, andit wasallowed to eatfromthe positive one.Theright-left position ofthepositive stimulus wasrandomized. Ten trialsweregivena day,eachtrialfollowed byreinforcement, untilS attained 18

correct choices in20trials withthelast10correct. Oneanimal inGroup IVthat

hadnotlearned in300trials wasstopped. Tocontrol extraneous cues, eight

stimulus holders wereused.FourofthesewereusedforTrials 1to5,theotherfour onTrials 6 to 10.HalftheSshadthecircle asa positive stimulus, halfhadthe triangle. Each ofthefourEsrananimals fromeachgroup. Results

Themean trialsandmean initial errors foreachofthefourgroups are shown inTable 81.Thetestforreplication is thatbetween Group I (cutouts) andGroup IV(control). Thedifference isinthepredicted direction,buttheVsof1.42fortrials and1.54forerrors areonlysignificant at

the .10level(one-tailed). Actually, themeandifference of 26 trialsand

of12initial errors(15% facilitation fortrialsand20%forerrors) wasas largeasthatfoundinotherexperiments yielding significant differences (Gibson etal.,1958), butthevariance islarger. Thepredicted no-difference

between Group II(painted patterns) andGroup III(white rectangles) was upheld withVsof.39fortrials and.67forerrors (p> .50). Thecomparison

between thecutout group(I)andtherectangle groups (IIandIIIcombined) wasaspredicted withVsof 1.79fortrialsand1.73forerrors(p< .05,

one-tailed). Thequestion whether thestimuli forGroups IiandIIImight of.10fortrials and.40forerrors (p>.50)thatcompared these groups withthecontrols werenotsignificant. Thisexperiment, then,seems touphold thehypothesis from previous yieldsome facilitation canapparently beanswered negatively since theVs

research by theauthorsthatonlythecutoutsyieldfacilitation in the Table 8.1

Mean Trials toCriterion andMean Initial Errors fortheFourGroups Trials

Group

N

I (cutouts)

22

II (paintedpatterns)

21

III(white rectangles) 18 IV(control) 23

InitialErrors

Mean 144.7 175.2

168.4 170.5

SD

Mean

SD

62.9

49.0

24.2

54.9

49.2 56.4

57.0

61.4 61.5

17.6

21.4 28.2

138

R. D. Walk,E.J. Gibson,H. L.Pick,Jr.,& T. J. Tighe

discrimination learning task,butthestatistical levelof thedifferences betweenthecutoutgroupandtheothergroupsisnotverysatisfactory. Discussion

Including theexperiment reported here,wehaveperformed a series ofnine

separate experiments inwhicha totalof 422rodentsweretaughta triangle-circle discrimination, theexperimental variable beingalways sometype of exposure to visually presented patterns. In allexperiments wherecutouts wereusedas exposurestimuli,facilitation was observed(Gibson&

Walk,1956;Gibsonet al.,1958,Experiments I andII;Gibsonet al.,1959,

Experiment I;Walket al.,1958,Experiment I;thisexperiment), butthe

amountof facilitation variedgreatlyfromone experiment to the other.

Painted patterns asexposure stimuli neveryielded facilitation (Gibson et aL,

1959,Experiments IIandIII;Walketal.,1958,Experiment II;thisexperiment). Thereasonsuggested forthefacilitation obtained by thecutoutsis that the cutoutsare attention-getting.Thesesolidpatternoutlinesstandout in contrastto the background as objectsin depth.Whenlaterconfronted

bytwopainted patterns inthediscrimination box,theanimal ispresumably moreaptto differentiate thepatterns fromthetotalsurroundings andto connectthem as cues with the differentialreinforcement.This hypothesis

maybe relatedto the superiority of stereometricoverplanometric stimuliin discrimination taskswithmonkeysand children(Harlow& Warren, 1952;Stevenson& McBee,1958;Weinstein,1941). Summary

An experiment wasdesignedto test the effectiveness of metalcutouts compared withpaintedpatterns as exposure stimuli duringrearing. Four

groupswereused.GroupI hadblackmetaltriangles andcircles mounted onthecagewallsduringrearing. GroupIIhadblackpainted triangles and circleson whiterectangles. GroupIIIhad whiterectanglesin the living

cages.GroupIVhadnothingmounted onthecagewalls.OnlyGroupI

learneda triangle-circle discrimination fasterthantheothergroups,which wereindistinguishable fromeachother.Whilethe experiment tentatively confirmed hypotheses derivedfromprevious experimentation aboutthe

effectivenessof the cutoutsas exposurestimulito facilitatediscrimination

learning, thefirmness of thisconclusion is attenuated by the levelof significanceobtained in this experiment. Notes

1.Thisexperiment wassupported, inpart, byagrath from theNational Science Foundation

Pre-exposureand Discrimination

139

Behaviorof Light:-and Dark-RearedRat:s on a Visual Cliff

Richard D.Walk, Eleanor J. Gibson, Thomas J. Tighe Itwastruly serendipitous thatthetrouble required toreara large group ofratsin

thedark inspired ustoobserve theratsinasecond kind oftest, asortofartificial cliff. Theidea wasthattheratswould walk outonit(glass over a void) ifthey hadfailed todevelop depth discrimination during dark-rearing. T.J.Tighe, atthat

timeourresearch assistant, andI hastily puttogether a simulated cliffwithbits ofwallpaper, glass, androdsandclamps thathappened tobearound thelaboratory. Then we(Walk, Gibson, andTighe, ina party) putthelight-reared rats ona slightly raised center board, chosen asa starting place, andwatched them as,

onebyone, they gotdown ontheshallow side andcrept around, butuniformly avoided thedeepside.Then wetookthedark-reared animals outoftheir seclusion andtried them, onebyone.Their behavior wasindistinguishable from theirlight-reared peers.

Aswewatched them, webegan tobeconcerned thatit might bejustthatside

ofthecontraption thatthey didntlikedrafts orodors, who knows. Wequickly fetched more ofthewallpaper wehadused ontheshallow side, andputitdirectly under theglassofthedeep side.Every ratwasnowgiven a second chance, and theyallcrossed back andforth from sidetoside, darkandlight-reared alike. As wereturned thelastrattoitscage, Tighe said,I wouldnt have believed itifI

hadntseenit. It wasa memorable day.

In themanyexperiments withthevisualcliffthatfollowed thisone,our

first,weconcentrated oncomparative studies withprecocial andlessprecocial animals (Gibson andWalk 1960; Walk andGibson 1961). Itwasseveral years later when weperformed another dark-rearing study, withkittens assubjects. The results wereaninteresting contrast tothestudywithrats.TA/hen thekittens were

brought outofthedarkandplaced onthecliff, they wandered about everywhere,

notfavoring onesideortheother. Butit could notbeargued fromtheirbehavior

thatexperiences with dropoffs wasrequired foravoidance ofthem toappear. They

' AAAS.Science, 1957,126,8081.

142

R. D. Walk,E.J. Gibson,& T. J. Tighe

were putonthecliffdailyfollowing theirintroduction toa lighted environment

After twodaysinthelight, eighty percent ofthekittens avoided thedeep sideand aftersixdaysalldid.Experience should havetaught themthatit wasassafeas the other,if it taughtthemanything.

Darkrearing hasproblems asa method, buttaking theresults withthose of other experiments oneisstruck withthespecies differences, firstofall,inreadiness toactat birthand,equally, withtheamazing degree ofpreparedness toengage in

perceptually guided behavior when anaction system, suchaslocomotion, has matured toreadiness ina normal environment. A light-reared kittenavoids a cliff

when it isready towalk, anddetects theaffordance ofa supporting surface when it is visually specified. Normally maturing visionis essential for theproper outcome, but no learningofspecific S-Rbondsis involved .

Fromthe isth centuryto the present,the empiricistand the nativist theoriesof depthperception havebeenvigorously debated.Oneexperi-

mentaimedat resolving thedisputeis Lashley andRussells (1),in which rats rearedin darknessjumpedto a platformfrom a stand placedat a variabledistancefromit. Theforceof the jumpwasfoundto be gradedin

accordancewith the distanceof the platform.Thisis evidencefor nativism.

But,sincethetestswithgraduated distances werenotgivenuntiltherats thirddayinthelight,andafterpretraining, theconclusion wasnotindubitable.Confirmation by anothertechnique isdesirable andhasbeenprovided in the experimentdescribedin this report (2).

A technique of testingforvisualdepthperception whichinvolves no

pretraining at allthe visualcliffwas developed. It is basedonthe

assumption that,givena choice, an animalwillavoiddescending overa vertical edgeto a surface whichappears to befaraway(3).Theapparatus

(Fig9.1)wasconstructed of twothicknesses of glass(24in.by 32in.), parallel tothefloorandheldbymetalsupports 53in.aboveit.Aboard(4

in.wide,24in.long,and3 in.high)extended acrosstheglass,dividing it intotwoequalfields.On oneside(thenear side),patterned wallpaper wasinsertedbetweenthe twosheetsof glass.Throughthe clearglassof theotherside(thefar side)thesamepatternwasvisibleon thefloorand also on the walls below the glass surface.

Optically speaking, theedgeononesideoftheboarddropped awayfor

a distance of53in.(making thesimulated cliff), whileontheothersidethe

edgedropped away foronly3in.Thus, twovisual fields existed, bothfilled

withpatterned wallpaper, butthepatternofthefar fieldwasoptically much smallerand denser than that of the other and elicitedmore motion

parallax. (More binocular parallax wasalsopossible atoneedgethanatthe

other,but the rat is probablyinsensitive to this cue.)The fieldswere

matchedfor reflectedluminousintensity.The physicalspace,as distin-

Behavior ofLight-andDark-Reared Ratsona VisualCliff 143

Figure 9.1

(left) Apparatus fortheexperimental condition. Thelarger-checked field isthenearside,

optically; theclear glass field isthefar orcliffside. Fig9.2(right) Apparatus forthe

control condition.

guished from theoptical space, wasidentical onbothsides, since aglass surface waspresent ata distance of3 in.Theonlydifference between the twofields, therefore, wasadifference inoptical stimulation. Other possible

cuesforsafedescent (tactual, olfactory, auditory echolocation, aircurrents,

ortemperature differentials) wereequalized bytheglass. In addition to the experimental condition described here,a control

condition wasincluded, inorder tocheck onthepresence ofanyunknown

factors thatwould make forapreference foroneside. Apiece ofwallpaper

wasinserted between theglassonbothsides(Fig9.2);otherwise, the

apparatus wasidentical to thatfortheexperimental condition. Ifcontrols areadequate, animals should shownopreference foreither sideinthiscase. Subjects fortheexperimental condition were19dark-reared, hooded rats,90daysold,and29light-reared littermates. Twenty minutes after

coming intothelight, thedark-reared ratswere placed ontheapparatus. Ananimal wasplaced onthecenter board inabox, toavoid anyhandling

bias.Itwasthenobserved for5 minutes. Results aresummarized inTable 9.1.Thepercentage ofanimals thatdescended onthenearsidewasnot

significantly different forlightanddark-reared rats.Ofthelight-reared

rats,23descended onthenearside,threedescended onthefarside,and

threeremained ontheboard forall5 minutes. Ofthedark-reared rats,14 descended on the near side, three descended on the far side, and two remained on the board.

Buta comparison ofdescent behavior oftheexperimental animals with thecontrols, forwhom thevisual surface wasnearonbothsides, showed

adifference. Thecontrol group, alllight-reared litter mates oftheexperi-

144 Table

R. D. Walk, E. J. Gibson, & 1. J. Tighe 9.1

Comparison ofLightandDark-Reared Animals onVisu alCliff (Experim ental Group) and Comparsion of Bothwitha No-Cliff ControlGroup. Control group

Experimental group

Percentage descending

on near side

Light-reared

Dark-reared

Light-reared

(N = 29)

(N = 19)

(N = 10)

88.5

82.4

50.0

MeanNo.crossings

0.00

0.06

Percentage of time On near

76.0

57.9

On far On board

10.0 14.0

16.9 25.2

1.70 24.1

61.5 14.4

*Thecontrol grouphadnooptically far side.Reference isto thesamephysical sidethat was far

for the experimental

group.

mental group, showed nopreference indescending fromtheboard; five

wentto eachside.Thisgroupdifferssignificantly fromthe experimental group (p < 0.02).

Evenmoreinteresting isa comparison oftheexploratory behavior ofthe

animals.The light-rearedand dark-rearedrats of the experimental group

againbehaved similarly; mostofthemstayedonthesideof thecenter boardthattheyhadfirstchosen. Of the 43 experimental animals that descendedfromthe board,onlyone crossedto the other side.But the

controlanimals explored backandforth,oftencrossing theboardto the

other side severaltimes.The differencein crossingbehaviorbetween

experimental andcontrolgroupsis highlysignificant (p<0.001).The percentage of timespenton thetwosidesconfirms theothermeasurements.Bothexperimental groupsspentmorethantwiceas muchtimeon thesidewiththenearopticalpatternas on the sidewiththe faroptical pattern,whilethecontrolanimals reversed thistrend.

Theseresultssuggesttwoconclusions. First,hoodedrats,90daysof

age,dodiscriminate visual depthordistance. Theyavoid a visual cliffas

compared witha shortvisualdrop-off, andthispreference is eliminated whenthevisualcliffis eliminated. Second, suchdiscrimination seemsto be

independent ofprevious visual experience, sincedark-reared adultanimals behavedliketheirlight-reared littermatesonly 20 minutesafterbeing exposed to the light. Referencesand Notes

1.K.S LashleyandJ. T. Russell, J. Genet.Psychol. 45, 136(1934).

2.Thisresearch wassupported, inpart,by a grantfromtheNational Science Foundation

Behaviorof Light- andDark-RearedRatson a VisualCliff

145

Wewishto thank]. ]. Gibson , for suggestions aboutapparatus andstimulus conditions . 3. Theworkof K. T. Waugh[J. compo Neurol . 20, 549(1910 )] and]. T. Russell [J. Genet . Psycho /. 40, 136(1932 )] makes thisassumption seemplausible . Latency of jumpingor IIdisinclination to jump" apparently increased asdistance increased .

Development of Perception: Discrimination of

DepthCompared withDiscrimination ofGraphic

Symbols

EleanorJ. Gibson Therearing studies andtheresearch withinfants onthevisual cliffrekindled an

interest indevelopment thathadbegun (andbeenfrustrated) withtheworkon

kids attheBehavior Farm. How perception develops became afocal problem for

me,asitstillis.I wrote chapters forseveral books, oneonmethods ofstudying perceptual development (E.Gibson andOlum 1960) andoneonperceptual development fora volume onchild psychology published bytheNational Society fortheStudy ofEducation (E.Gibson 1963). Thelatter wasanopportunity to think about aframework forstudying perceptual development. Theorganization ofthetypical textbook chapter onperception, withsections oncolor vision, form perception, cues fordepth, illusions, etc.,wastotally unsuitable. American psychology offered noframework fordevelopment except learning andconditioning theory, which made noreference atalltoperception. Piaget wasonly beginning to exert aninfluence intheUnited States, andinanycase heconsidered perception primitive, figurative, andmerely preliminary totheconstruction ofintelligence. I came upwithanoutline organized around whattheenvironment offers the

perceiverplaces with surfaces andedges, objects, pictures, andgraphic displays Yearbook thechapter appeared inwasobscure, anda newimpetus forgetting

since these latter figured inmostoftheexisting experimental literature. Butthe

perceptual development intotheforeasa topic forresearch wasneeded. Theimpetus camewitha committee created bytheSocial Science Research

Council in1959, theCommittee onIntellective Processes Research, specifically

focused onthefield ofcognitive development. This committee planned sixresearch conferences, to takeplaceovera five-year period, andthena finalsummer institute. Thetimewasripeforthecommittees efforts. They bore fruitinthe conferences themselves, andfinally ina totally revitalized discipline ofdevelop-

mental psychology. The proceedings ofalltheconferences were later gathered in

Reprinted from 1.C.Wright and1.Kagan (eds.), Basic CogniUve Processes inChildren,

Monographs oftheSociety forResearch inChild Development, 1963, 28,No.2(Serial No.86).

148

E. J. Gibson

onevolume,a veritableclassic(Cognitive Development in Children, published by the Societyfor Researchin Child Developmentin 1970). I was fortunate in being invited to prepare a paper for the second of these

conferences , onecalledJJBasic CognitiveProcesses in Children." My partnership with Richard Walk had broken up in 1961 when he moved away from Cornell, and I shifted mv~ researchin other directions, including researchon older children -

as they were introducedto codedsymbols . This meetinggave me a chanceto considertheway two quitedistinctlinesof research fit into the largerframework of perceptualdevelopment . I openedthe paperwith a comparisonof the kind of informationavailablein thetwo domainsI wasstudying, perception of depthand perceptionof graphicsymbols . The term "information" is not used - that came later- but choosinginformationas the logicalplaceto start is a portentof how a present -day theoryof perception cameabout. Anotherforerunnerof a growing theoreticalview was the lessonsuggested in this paper from the rat-rearing experiments: not only do solid objects, occupying

their own placein thelayout, takeprecedence in a viewer's attentionoverpictures or graphicpresentations , butfeaturesof thoseobjectscanalsobedifferentiated and may transferlater to the perceptualdifferentiationof pictorial copiesor graphic compositions . Herestirs the embryoof a theoryof pictorialperception . It is notable, too, that theorizingaboutgraphicitemsharks backto the theoryof paired associatelearning in my dissertation: that items in a code have to be learned in

two steps , first differentiated from othersin theirsetand only thereafterassociated (if that is what happens) with something called meaning. This paper points ahead to work on reading, which I come to in part V.

It seemsinstructiveto put it here , however , to show how a theoryof perceptual development wasbegun , and how research from ostensiblyquitedifferentdomains was already convergingon it . . . , . . , or . . . . ,

r .. ~ '

-

~ . .

. .;r

-

-

- -

-

-

- o

~

The invitation to speak to this conferenceon my work in the field of perceptualdevelopmentcameat a most welcome moment for me. For the past six years I have worked, with several colleagues, on developmental aspects of two radically different kinds of perception -

the perception of

depthand the perception of outline forms inscribedon a piece of paperthat is, lettersand words. Here was the opportunity to compare the two and,

hopefully, to synthesizethem. Interest in the development of perception (especiallyspaceperception)

goesbackasfar asthe philosophicalbeginningsof psychology . The empiricismof the Britishphilosophers andthe nativismof the Germanshave always formed the core of coursesin the history of psychology. Everyone takes a position in the controversy , usually on the side of empiricism in this

country. Textbooksof child psychologyreflectthis fact; hereis a typical quotation from a well-known one, Goodenough's Developmental Psychology :

149

Discrimination of Depth and Graphic Symbols " Very to

early

in

interpret

and

muscular

tree

looks

qualities

such

other

as

, ,

the

world A

as

urge

"

is

The

term

and

learned

of

. There

which

may

and

interact

We

have

at

with

"

un

of

the

con

of

is

not

into

instance

,

and

that

very

. P . 46

the

term

environ

the is

unlearned

-

enters

. It or

approach

" I

,

,) -

satisfying

from

"

said

learning

learned

and

" innate has

of

which

( 1953

primarily

) .

inference inference

.

to

is

too

easy

primarily

un

processes

into

study

the

processes

-

on

-

develop

impressive

work

laboratory

on

more ,

latter of

there "

the

this

kind

,

may can

);

seem

be

cited

distinctiveness

of and

( for "

of

learning

others

providing

own

scribble

logical

instance

,

that

but

of

the

some

at

features

the

very Haskins

[ Liberman

et

-] ) .

two

cases

they

must in

the

have be stimuli

been

or

described for

can in

the

two

be some situations

attacked detail

by . is

The the

all

these

methods

potential logical

.

But

information starting

-

with

test ,

phoneme

ex

analysis

experimental

roundabout

( the

perceptual

or

( our

procedure

by experi

environment

;

conditions

the

controlled

early

lenses

perceptual

supplemented

by

technique

happened

procedure

" acquired

of

of

characterized

be

and

distorting

is

have

can

species

normal

study

that

control

with

the

country

enrichment

' s

" must

this

animal

the

for program

include

Kohler

these

what

This

in

different

under

Besides of

available testing

and as

practice

periment

methods

methods

such

controlled

My

,

perception

independent into

appealing

psychology

experiment

experiments

available

child

Experimental

deprivation

1957

.

bias

laws ,

the

behavior

" different

developmental

studies .

false

learned

many

The

of

comparative

first

present

. stages

ments

bias

are - called

illusions

for

process

mechanisms more

opposite

so

and

behavior

processes

seem

how

,

abandoned

as

A

asking

' s

.

development early

."

,

of

Hebb of

neural

the

which

enters

on

empiricist

the

the ,

division

control

be

be

,

- effects

objects

perceiving

,

solidity

.

or

problem

even

process

of

from

behavior

learning

togenetic

the

the

dichotomy

should

the

of

the the

) .

that

mechanism

distinct

of

cases

, p . 93

kinds

and

dismissal

find

two

, after

say has

experiments

the

aspects

figural

been

artificial

a

misleading

to

( 1960

an

implying ,

,

of

perceptions

tactual

with

for

learn

of

we

associate

exhibiting

relationship

has

are

factors

a

"

there

" instinct

is

"

that

mental

- ground

to

presence

depth

, we

terms

sensation

enthusiasm

significant and

so

in

visual

learned

statements

the

doing

. . . . When

the

current

find

phenomenon

upright

" acquired

can

" All

criticism

have

are

objects that

continued

pattern

figure phi

recent

and

that

include

organization stancies

, we

we sensations

handling

) . The the

that

visual

only

we

138

confirms

's

. These

from mean

on

, p .

hand

Pastore

learned

the

( 1934 "

aware in

get we

infancy

"

experience the

being

) difference

we

from

rounded

On

without

rounded

that feel

" early

and

( binocular

sensations

trunk

that

al . ,

life

this

place

.

150

E. J. Gibson

Comparisonof Stimuli A standard situation for the study of depth discrimination was devised by Dr. RichardWalk and myself. We called this situation the "visual cliff." The important element of this situation is a drop-off downward, or depth~ at-an-edge. The device consistsessentiallyof a raised center runway with a sheet of strong glass extending outward on either side. Directly under the glass on one side is placed a textured pattern; farther below the glass on the other side, at any desired depth, is the samepattern. The simplest version of the stimulus situation might be conceivedof as a platform with a drop-off to a floor below. Figure 10.1 shows the pattern of light rays projected to the subject's eye from the floor and from the platform on which he stands. If the elementsof the textured pattern are identical above and below, the light rays reaching the eye will differ in density, a finer density characterizing the surface farther below the eye. There is thus potential information in the light itself for the detection of the drop-off. The same situation provides a second kind of differential stimulation if the animal moves. Motion parallax (differential velocity of elementsin the stimulus array) will increaseas the drop increases . There will be a velocity difference, therefore, betweenthe projection of the floor under the animal's feet and that of the sunkensurfacebelow, which will characterize the amount of the drop-off. Finally, the situation provides a third kind of differential stimulation, binocularparallax, if the animal has two eyes with overlapping fields and eye movementsof convergence. It could be said then that this stimulation literally specifiesa drop-off. The proximal stimulus is unique and unequivocal. The information needed is oresent in the stimulation itself. If it is registered, the animal can make ~ an appropriate response. Depending on the kind of organism it is, terrestrial, aquatic, or flying, it may avoid the deep side consistently, or not. A terrestrial animal would be expectedto avoid it; an animal whose way of life includesdiving into water from a height might approachit . If the animal does not behave consistently and differentially, it may mean either that he does not "pick up" the stimulus difference, or elsethat the appropriate responsehas to be learned. But if he does respond differentially and consistently, it meansthat the differenceis discriminatedand that a responseappropriate to his speciescan be made. N ow consider the kind of stimulation presentedby graphic symbols. In the first place, the sourcesof the stimuli are marks on a piece of paper, not three-dimensional natural objects found in all men's environments. They are, in fact, man-madeartifacts. The stimuli do not specify or refer to real objects or situations in space.

152

E. J. Gibson

What a

they

relayed

letters

in

specify

with It

is

able

true

that

speech

drop

printed

- off

. The

is

characteristic

marks

specify

which

would

of

spelling that

symbols

as

unique

relations

in

the

Experiments

graphic

code

experiments

environment

during

plentiful

a

side

same

of

depth

. We

to

first

etc

the

the

.

it

is

in

a

mark be

or

made

up

suggested

must

be

discrimination of

;

Which

often

of

appropriate

could

is

-

that

by

.) . considered

of

the

of

tailored

a a

included

the

unequivocal

graphic specificity

comparative

experimental

some

,

the

choice

rats

,

on

runway

on

,

most

a

,

. and

shallow

. The

side

cues

convenient

very

other

either

the

available

the

the

( both

manipulating

potentially

situation

drop

studies

studies

for

deep

center

rats

side 100

.

variety

to

patterns

depths

Comparative

number

and as

of

shallow ,

as

in

all

drop

glass

was

Figure

side

. A

10

drop

- inch

absence

young

Control of

as

30

,

10

always

significantly

the

shallow

.

experiments side

the

preference

,

. Rats

a the

choices

cent

of

in to

the

monograph

,

First

per

demonstrate

on at

.2

the

shows

experiments . The

,

choice

on

the

deep

days

are were

than

from

85

significant

side

with

was

with

sufficient

equal

described as

often

ranged

was

experiments

texture

descended

more

in

a

discriminating

depths

,

forthcoming as

adults

.

Studies

Experiments

with

size ,

and

and

runway

deep

graduated

kittens

,

,

discrimin

like

letters

code

fact

mastery

cliff )

and

adult

wide

rats

of

development

( 1 )

is ,

Depth

provided

the

center

percent

the

-

.

the

to

one

the

new in

writing

:

visual

cliff

below

the

to

the

runway

Pigmented from

of

growth

diagrammatically

relation

.

small

the

that

system

A ,

of

the

the

morphic

mediated

unequivocal

from

.

speed

(2 )

is ;

have

a

sense

not

writing

( this

symbols

ontogenetic

subjects

one

' s

stages

and

with and

built

,

there

letters

turn

specification

arbitrary

two

Discrimination

phylogenetic

a

of

the

is

better

reforms least

items

on

We

or

at

, in

culture is

well

in have

emanating

given

that

) .

unequivocal stimuli

notice specifies

sounds

meaning

sounds as

perception

Our

a

which

follows

the

the

of

do

proponents It

in

But

eyes

symbols

unique

their for

not

the

graphic

are

But

;

.

the

meanings

information

sounds

code

.

speech to

sounds the

letters

other

of light

( referential

any

speech

sounds The

. Thus

morphemes

from

the .

specify

relationships

ship

in

is

involved

turn

specificity

a

do

sequence

of

baby ,

the

cliffs subject

chicks monkeys

and ,

and

constructed ,

were adult

finally

carried chickens human

on

the

out

on

,

lambs

infants

same

plan

other ,

kids

( Figures

,

but

animals ,

pigs 10

adjusted as

, .3

well

turtles and

, 10

to :

albino

puppies . 4 ) . Many

,

Discrimination of DepthandGraphic Symbols 153

TEXTURED SURFACE

TEXTURED

SURFACE

Figure10.2 The visualcliff modifiedI as a choice situation.

as

Discrimination of Depth and Graphic Symbols

155

Dark -Rearing Experiments

The deprivation experiment is a technique which allows us to control someaspectof the animal's experienceor practiceduring what is, normally,

a developmental stage.Many groupsof ratswererearedin the dark, from birth to 90 days, or to 30 days, and comparedwith their litter -matesreared in the usualcagesin the light . The dark-rearedrats were tested on the cliff shortly after emergencefrom the dark-room, and they behavedas did their S.lN3:JS30 :fa .lN3J ~3d

light-rearedlitter-mates , uniformlychoosingthe shallowside. Rats, therefore, althoughrequiringmaturationtime after birth beforelocomotionis

possible, discriminatedepth without any previous visual experience. Kittens were reared in the dark until 24 days of age, when their normal

controls were walking and had good visual placing responses . When they were first put on the cliff, they presentedthe greatestcontrast to their own controls and to the dark-rearedrats. They crawled on their bellies, fell off the runway, and bumped into walls. Discrimination of drop-offs and even locomotion in the light were impaired. After some days in the light, the

kittenshadcaughtup andbehavednormally. It seemsclearthat development in a lighted environment is required for this species. But learning the differencebetween visually deep and shallow drops through external re100

TO SHALLOW 80

60

4

2

0 1

2

3 DAYS

4 IN

5

6

7

LIGHT

Figure 10.5 Perfonnanceof kittens, reared in darknessfor 24 days, on the visual cliff.

156

E. J. Gibson

inforcement (punishment fromfalling) wasnotthecrucial factor,forthe kittensin thebeginning fellequallyeitherwayto a glasssurface at the samedepthbelowthecenterboardoneitherside.Yetgradually, asFigure 10.5shows,theybeganto avoidonesideandchoosetheother,without reinforcement. Choicesrose to 100per cent by the seventhday. Our conclusionis that discrimination of depth matures,when normal

conditionsof developmentare provided,withoutbenefitof rewardor

punishment or associative learning. Progress mayindeedcontinue after

birth,throughnormalgrowthoforgans.Butearlyfailure ofdiscrimination isnoreasonto adopttheempiricists bias.Differentiation ofperception by developmental stagesis characteristic of phylogenetic differences, andin the case we have just considered, of ontogenetic ones. Enhancement Experiments

Aninteresting question arisesat thispoint.Canweapplyourgeneralizationwithrespectto depthat an edgeto discrimination of three-dimensionalobjects intheenvironment? Ifanobjectcanbeseen around,conditionsforparallax arepresent. Ithas,potentially, theattribute ofdepthand thepossibility ofbeingeasilydifferentiated fromthebackground. We have a littleevidence,fromotherexperiments, that discrimination

of objectshavingdepth-at-an-edge occursrelatively earlyandis responsiblefor sometransferto two-dimensional picturedshapes.

Dr. Walkand I carriedout a numberof experiments of the early

experience type,inwhich wehungcut-out triangles andcircles onthe walls ofliving cages ofhooded ratsfrombirthto90days.At90daysthese

animals, aswellascontrollitter-mates, learned a triangle-circle discrimination.Forthediscriminative learning, thefigures werepainted onflatsurfaces (doorsthrough whichfoodwasreached). Inourfirstexperiment, learning wassignificantly facilitated amongtheexperimental animals. In somefurtherexperiments, we paintedthe wallfigureson a flat

background, ratherthancutting themout.Toourchagrin, thetransfer effectwasno longerobtained.Our variablesin theseexperiments were

presence orabsence ofreinforcement, timeofexposing thepatterns, andso

on. Noneof theseseemedto makeany difference; our originaleffect sometimes appeared andsometimes didnot.Wehaddecided thatwewere

pursuing a will-o-the-wisp untilwenoticed onefactor which divided the results.Transfer occurred whenthecage-hangings werecut-outs; it didnot

occurwhentheywerepainted ontheflatsurface thatsurrounded them. Ourinterpretation of thesefactsis speculative, butrichin hypotheses. Objects havingdepth-at-an-edge areeasilydifferentiated fromtheirsurroundings. Theyaretherefore noticedbytheanimal. Thisnoticeability

weguess, willtransfer to similar pictured objects, thereby helping the

Discrimination of DepthandGraphicSymbols 157

animal to differentiate themfromtheirsurroundings whentheymustbe

isolatedas cuesforresponse.

Whattransfers, exactly, isanother question. Fortherats,notdominantly visualanimals, it mayhavebeenonlydifferentiation offigure from background. Itmight alsobefeatures which serve todistinguish oneobject fromanother, suchascurves opposed to corners, oropenings (indentations)opposedto closure(smoothcontinuity). I shouldliketo returnto this

possibility later,inthediscussion ofgraphic symbols. It shouldbe apparent howdifferent thisviewis fromthe traditional

onethatperception begins withsomething likeatwo-dimensional projectionandprogresses to appreciation ofdepthbyassociational meanings

gainedfromexperiences thataredependent onlocomotion. Theevidence

suggests instead thatdiscrimination ofdepth-at-an-edge isprimitive, both

phylogenetically andontogenetically, andthatdevelopment progresses

toward discrimination offormintwo-dimensional projections. Experiments withGraphicForms

Thejumpfromdiscrimination of depth-at-an-edge to discrimination of graphic symbols is a bigone.Thefactthatonlyhumansubjects are

appropriate forstudyingthiscaseis an indication in itself.Furthermore,

analysis ofthesituation madeit clearthatthestimuli provided bywords andletters donotcontain inthemselves information thatspecifies unequivocally anything abouttheworld. Whattheydospecify canonlybe

foundina codewhichvariesfromonelanguage to another andtherefore

must be learned.

Wehavesaidthatat leasttwostagesof development needto be considered in thisprocess. Thefirststageis the discrimination of the

graphic symbols themselves asunique items. Ourfirstquestion, therefore, was whether there is a developmental process involved in accomplishing the differentiation. Anoldexperiment performed bymyhusband andmyself (Gibson and Gibson, 1955)suggested thattherewas.Wepresented subjects divided

intothreeagegroups (6to 8,8 to II, andadults) withdrawn figures somewhat comparable tothoseofcursive writing. Theywerecalledscrib-

bles.Thetaskwastorecognize a presented figure when itwaspresented laterina series ofsimilar figures. Thestandard to berecognized varied

fromtheothersalongthreedimensions (number of coils,horizontal compression,and right-leftreversal). The subjectlookedat the standardand

thenwentthrough a packofcards onwhich thevariants, some copies of

thestandard, andotherquitedifferent figures appeared (Figure, 10.6). On thefirstrunthrough thepack, theadults werealready veryaccurate intheir recognitions (ameanofonlythreeerrors), butthechildren confused many

158

E. J. Gibson

Figure 10.6 "Scribbles" used in the Gibson and Gibson experiment (1955).

variants with the standard. The youngest group had a mean of more than 13 errors, and someof them did not achieveperfect recognition even after 10 repetitions of the whole procedure. We have recently completed a large-scalecomparisonof the ability to discriminate graphic items in preschool and early grade-school children.2 Our object this time was to study qualitative as well as quantitative differences. A set of letter-like forms, comparableto printed capitals, was constructedby the following method. The letters themselveswere analyzed, to yield a set of rules governing their formation. From theserulesa population of new forms was generated, none of which violated the rules but none of which were actual letters. From among these stimuli 12 were chosen to serve as standards. Twelve variants were constructed from each standard to yield transformationsof the following kinds: three degreesof transformation of line to curve or curve to line; five transformationsof rotation or reversal; two perspective transformations (slant left and tilt back); and two topological changes, a break and a close (Figure 10.7). The master drawings were copied photographically on small cards, and these were covered with plastic so that they could be handled without

Discrimination ofDepthandGraphic Symbols 159

marking. Thetaskgiventhechildren wastomatch thestandard withallits

variants andtoselect andhandtotheexperimenter onlyexact copies ofit. thetoprow(Figure 10.8). Alltransformations ofaparticular standard were Thecardswerepresented in a matrixboard,witha standard centered in

assembled randomly in onerow,accompanied by at leastoneidentical copy.Whena childhadfinished matching fora givenstandard, it was removed andanother inserted initsplace.Demonstration withcorrected

practicewasgivenbeforebeginning, andthenthe childmatchedforall 12

forms.

An errorscore(choosing as same an itemthat did not matchthe

standard) wasobtained foreachchild, andtheerrors classified according

to typeoftransformation. Thesubjects were165children aged4 to 8 years.

Errors decrease rapidly fromages4 to 8.Furthermore, it is veryclear

thatsometransformations areharderto discriminate fromthe standard thanothersandthatimprovement occursat different ratesfordifferent

transformations.

Errorcurves forchanges ofbreakandclosestartlowanddropalmost tozeroby8years. Errorcurves forperspective transformations startvery

highanderrorsarestillnumerous at 8 years.Errorcurvesforrotations and

reversals starthigh,butthecurves dropto nearly zeroby8 years. Error curves forchanges fromlinetocurve startrelatively high(depending on

thenumber ofchanges) andshowa rapiddrop.Thecurvesfortwoand threechanges havedropped tothesamelowpointasthecurves forbreak

and closeby 8 years.

In Figure10.9the datahavebeencombined forthe fourtransformation groups.Ourjustification fordoingso wastwofold:one,the resemblances

ofthecurves within transformation groups, butdifferences between groups

(statistically significant, in fact);andtwo,thehighcorrelations within

transformation groups.

Theexperiment wasreplicated forthe5-year-old groupwithactual letters andthesametransformational variants. Again, thecorrelations with-

intransformation groupswereveryhigh. Interpretationof Error Curves

Theinterpretation oftheerrorcurves forthesetransformation groups leadsusto someinteresting hypotheses aboutthedevelopment ofdis-

crimination of letterforms.Theconceptof distinctivefeaturesis central

to theargument. Thistermisborrowed fromRoman Jakobson (Jakobson

andHalle,1956),whooriginated the conceptof distinctive features of

phonemes. Distinctive features arecharacteristics which aphoneme mayor maynothave; theyareinvariant; andtheyarecritical fordistinguishing

160

E. J. Gibson 0

S

L to C 1

L to C 2

L to C 3

45 R

90 R

0 R -L Rev .

U -D Rev .

. i ..i1 - ..:l ~ l + .-iT

180 R

..ii

Perspective Slant

Trs . LTiltback

Close

Break

~ ..rh . .:1:..

~ [ })) J 4 SZ1<1 ~ [)i t>t ~ l2] c:: I lr )r- )r- V ~ ~ ~ f' - ~

LdldkjlJ

lr V- Ir I~

~ EbJNP1IdLdQ

..cl

: j ::J : J =:J-:::; U t:: ::) ~ I :::) ::j ::j --) .( -{ -(

-{ r

;;"" }. {

T .( .{ {

ld ~ bJ1t::;l ~ a= bJAblbJIdQk

,f

-:J

o\ ~ do <. 9- ~ ~ ~ yJd \ c:l: c:RcA -E ~

~ '-~ ~

m 3- - E - E - E -E - 8 - E

~ X / \ A \ )<) ~ 6 ~ ~ ~ ZX LLLL , C"t ~ ~ 7 [ L [ JL tf \ t1\ AI

\ ~ ' ) AVVf

Figure10.7 Artificial graphicformsandtwelvevariants .

\ A ~ f\

162

E. J. Gibson

one phoneme from another. "Bundles" of such features characterizeany given phoneme. A child, presumably, learnsto hear the distinctive features and can thereafterrecognizephonemicpatterns over a wide range of pitch and intensity variations (sung, whispered, shouted, and so on).3 Taking some license with Jakobson's term, we have assumedthat the solid objects of the world, and also the set of graphemes , possess"distinctive features" - characteristicswhich are invariant and critical for distinguishability within the set. An attempt to analyze all the objects of the world into a classification of distinctive featuresis not our purpose, but someprogresshasbeenmade with letters. Sufficeit to say here that breaksand closes(0 vs. C), transformations from line to curve (0 vs. V), rotations (M vs. W ), and reversals (d vs. b) are all distinctive featureswith respectto which letters may differ. They often occur in some combination as bundles of distinctive features. For instance, A vs. V includes both rotation and closure as critical differences.4 Perspectivetransformations are not distinctive features of letters. Changesof compression(suchaswould result from tilt or slant) often occur in handwriting but are never critical for differentiating letters. "Constancy" of the graphemerequirestoleranceof suchvariance. We assumedfurther (see p. 14) that there will be transfer from discriminating ordinary solid objects of the world to two -dimentional line drawings of the letter-like variety. Positivetransfer should occur when the variable dimensiondistinguishing two line drawings is one which has been critical for discriminating one object from another in the past. No transfer should occur when the variable dimension has not been critical for object identification. Sometimesa variable aspecthasnot only beenirrelevant, but has even been assimilatedto a "concept" allowing free variation along this dimension. For instance, shapeconstancy occurs despite variation in orientation. When transfer is possiblefrom earlier object identification, errors should be few. When it is not, they will be numerousto begin with, but they will decreasewith age if the dimension varied is a critical one (distinctive feature) for differentiating letters. The four transformation types should therefore show the following error trends with age. 1. Errors for differencesof the break and close type should be few initially and show only a small drop (discrimination already being nearly perfect). These changesare critical for object differentiation. TnPTPis no transition from break to close. Pia~et (1956) has shown that suchdifferencesare discriminatedvery early with solid objects. 2. Initial errors for line to curve transformationswill be higher than those for break and close but should drop rapidly. Line to curve differencesare critical for distinguishing rigid objects, but not for live

Discrimination of

Depthand Graphic

Symbols

163

or plastic ones (e.g., the nonrigid transformationsof facial expression). Since they may indicate different states of the same object, perfect transfer cannot be expected. Nevertheless, the error curve should drop rapidly becausesuchdifferencesare critical for letter discrimination .

3. Initial errors for reversalsand rotations should be very high because transformations of this type are not critical for object identifica -

tion. (Rotation is used in fact to study transposability of form;5 it gives information of position of an object, such as leaning or fallen down .) But rotation

is critical

for letter

discrimination

, so the error

curve should drop rapidly after school has begun. 4. Initial errors for perspective transformation should also be very

high. They are not critical for object identification, indicating instead a changein orientation in the third dimension (slant or tilt ). There is little reasonto expecta drop in errors, for slant and tilt are not critical for letter identification

.

The curves obtained do actually follow those expectations . There is,

therefore, support for the view that there is perceptual learning of the distinctive featuresof letters in the stage of development before decoding to phonemesbegins. The kind of learning is not associative; it is instead a processof isolating and focusing on those featuresof letters that are both invariant and critical for rendering eachone unique. Teaching is provided the child in this stage of learning, but it is not of the paired associatestype. It is rather helping the child to "pay attention to" those features that are invariant and distinctive. Learning a name for each letter, on the other hand, is a caseof association, but necessarilya secondary stage.

Grapheme -Phoneme Correspondence We have suggested that the development of discrimination of graphic symbols begins with the differentiation of letters, whose distinctive features must be learned, assisted by transfer from the earlier stage of object

identification. Meanwhile the child has learnedto recognizephonemesand to speakhis language. But the final stage remainsto be accomplished: the decoding of graphemicto phonemicunits. It is not enough to describethis stage as merely associatinggrapheme with phonemepatterns. What are the units to be associated ? In the English language, no single letter hasan invariant phonemic equivalent. Words do, and this fact has led to teaching by the "whole word" method. This method , however , is both

reading of new words.

uneconomical

and insufficient

to generate

the

164

E. J. Gibson

An alternative to thesetwo possibilitiesexists. Dr . CharlesHockett and his collaboratorsat Cornell have shown that rules for predicting pronunciation from spelling can be formulated, if the rules are stated in terms of vowel and consonantIIclusters" and what comesbefore and after. Higher order invariant spelling patterns exist which can be mapped into correspondencewith phonemic patterns. It is the grapheme-phonemeinvariant correspondencesthat the skilled reader has learned. Letter-groups having such correspondencein the languagecome to have a high perceptibility, sincethey form units as stimuli; those lacking suchcorrespondenceare not equally perceptiblebecausethey are not effective units for pronunciation. We have conductedseveralexperimentswith tachistoscopicviewing of pseudowords, somefollowing the rules of correspondence(invariant spelling to sound correlation), somenot.6 Skilled adult readersare consistently more successfulin perceiving correctly the letter groups which are "pronounceable," even though they have no referential meaning. These letter groups have a higher "visibility ." Here we have a final case, at the top of the developmental ladder, of perceptual learning. What the reader has learnedhere (albeit unconsciously) is to perceivethe higher order stimuli as units in reading. Thesestimuli are letter groups (ways of spelling) that exist in the languageas invariants in the sensethat they have a corresponding consistentpronunciation. The readeracquiring skill comesto perceivethese as units by experiencingthe heard and seenpatterns together. He mayor may not be taught them. The letter or spelling units are constituted, formally and objectively, by the rules of correspondencein the language. They are a psychological reality as well, as we have demonstrated. The problem for perceptual learning is to determinehow the unit is constituted. Whether the processof unit formation is an associativeone or one of "discovery" of the invariant relationshipsis yet to be determined. Summary From our comparison of the development of perception of depth and that of graphic symbols, several generalizationscan be drawn. The first, stated below, is a conclusion. The others have the status of promising, partly substantiatedhypotheses. 1. Perceptionof depth at an edge is primitive, both phylogenetically and ontogenetically. Some animals are fully mature in this accomplishment at birth. Animals which have a longer maturation time after birth (e.g., cats, human infants) discriminatedepth at an edge as soon as locomotion is possible. 2. Solid objects, which possessdepth at their edges, are discriminated earlier than two-dimentional pictures or line drawings. If perceptual

Discrimination of Depthand GraphicSymbols 165

learning occursin the earlierphase,it involves a discovery of invariantproperties oftheobjectwhichthestimulation itselfspecifies

and whichare criticalfor distinguishing one objectfromanother. Whatislearnedis isolation frombackground or differentiation rather thananassociative meaning fordepth. 3. Abilityto discriminate thosefeaturesof objectswhicharecritical foridentification maytransfer to outlinedrawings suchasletters,but

somecriticalfeaturesof lettersremainto be discriminated afterfour yearsof age.Thisprocessis againone of differentiation ratherthan

association.

4. Unequivocal referential meaningof lettersmustbe learned;it is notgivenin the stimulation emanating fromthem.At this stage (following differentiation) anassociative process maybeinvolved. 5. Mereassociation of letterwithphoneme,however,is an inade-

quatedescription of theprocessoflearning meaning. Letterclusters

fromhigherorderunitswhich areinvariant forpronunciation and

reading passthrough a learning phaseofintegration sothattheyare actuallyperceivedas wholes.

Thispaperraisesmoreproblems thanit solves. It is myhopethat

stirringup theseproblemswillcreateinterestin a fieldwhichI havefound fascinatingand productive. Notes

1.Harlows discovery thatstereometric (solid) objects, ascontrasted withplanometric (flat) patterns,arediscriminated moreeasilyby youngmonkeysis oneconfirmation of this

hypothesis.

2.AnneDanielson andHarryOssercollaborated inthisexperiment.

3. SeealsoBrown(1958,pp. 202 if.).

4.Notethecomment ofDeutschs infantson(Deutsch [1960,P.1491) whenshowna Why isnt the A crossed?

5.The2-year-old children tested byGellerman (1933) responded toa form asequivalent afterrotation.SeealsoDeutsch(1960,p. 149).

6.Hammond. Collaborators in thisexperiment wereAnneDanielson, HarryOsser,and Marcia References Brown,R. Wordsand things.Free Press, 1958.

Deutsch, J.A.Thestructural basis ofbehavior. Univer. ofChicago Press, 1960. Gellermann, L.W.Formdiscrimination inchimpanzees andtwo-year-old children: I.Form (triangularity) perse.J.genet.Psychol., 1933,42,329.

Gibson, E.J.,&Walk, R.D.Theeffect ofprolonged exposure to visually presented

patterns onlearning to discriminate them.J.comp. physiol. Psychol., 1956,49,239241. Gibson, E.J.,&Walk,R.D.Thevisualcliff.Sci.Amer., 1960,202,29.

166

E. }. Gibson

Gibson , E. J., Walk,R. D., & Tighe , T. J. Enhancement anddeprivation of visualstimulation duringrearingasfactorsin visualdiscrimination learning . J. compo physiol . Psychol ., 1959 , 52, 74- 81. Gibson , J. J., & Gibson , E. J. Perceptualleaming : differentiation or enrichment ?Psychol . Rev ., 1955 , 62, 32- 41. Goodenough , F. Developmental psychology . Appleton -Century , 1934 . Harlow , H. F. & Warren , J. M. Formation andtransfer of discrimination learning sets . J. compo physiol Psycho /., 1952 , 45, 483- 489. Hebb , D. O. Heredityandenvironment in mammalian behavior . Brit. J. Anim.Behav ., 1953 , 1, 43- 47. Jakobson , R., & Halle,M. Fundamentals oflanguage . TheHague : Morton, 1956 . Liberman , A. M., Harris , K. 5., Homnan , S. H., & Griffith,B. C. Thediscrimination of speech sounds withinandacross phoneme boundaries . J. expo Psychol ., 1957 , 54, 358- 368. Pastore , N. Perceiving asinnatelydetermined . J. genet . Psycho I., 1960 , 96, 93- 99. Piaget , J., & lnhelder , B. Thechild 's conception ofspace . London : Humanities Press , 1956 . Walk,R. D., & Gibson , E. J. A comparative andanalyticstudyof depthperception . Psycho I. Monogr ., in press . Walk, R. D., Gibson , E. J., Pick,H. L., Jr., & Tighe1 T. J. Theeffectiveness of prolonged exposure to cutoutsvs. paintedpatternsfor facilitationof discrimination . J. compo ,physiol v . Psychol v ., 19591 52, 519- 521. Discussion

HermanA . Witkin Gibson has approached, in a fresh way, problems that have been with us for a very long time; and the ingenious experiments she has done have provided new insights into theseold problems. Gibson's careful, persistent working through of a variety of particular perceptualfunctions contributes much to our understandingof the role of learning in perceptualdevelopment. In her considerationof perception in children at different ages, and in animal forms of different kinds, Gibson has employed a truly comparative approach. It is gratifying to find such flexibility of researchstrategy in this era of specialization. The comparativeapproachhas many advantages. A maior - one is that observation of a given perceptualfunction in different species , or in organismsof the samespeciesat different stagesof ontogenetic development, is likely to point up differences, which inevitably provide useful leads for further investigation and deeper understanding of the function. An example in Gibson's material is the finding that turtles, the most aquaticform studied, showed the leastpreferencefor the shallow side in the cliff experiment. Another interesting example is the difference observed in this samesituation between kittens and some of the other forms studied. The strategy of considering two quite different perceptual functions, rather than limiting herself to one, has the advantagethat specialfeatures of eachare pointed up through contrast with the other.

Discrimination of DepthandGraphicSymbols 167

It is a reflection onthevalueofGibsons approach andinvestigations thatsomanypromising leadsforfurther research haveemerged andthat herworkhasimplications forthefields ofperceptual andlanguage deve-

lopment,and ultimately for suchpractical issuesas how best to teach

children to read.

Thereareseveralinteresting issuesraisedby Gibsonsmaterial, which mayperhapsbeclarified inthesubsequent discussion or infurtherwork Oneconcerns themechanism underlying performance in thecliffsitua-

tionthe discrimination between theshallow anddeepalternatives and theexpression ofapreference fortheshallow one.Theprocess involved in

eachof theserequires studyin its ownrightfor,as observed in therat,

discrimination maybepossible evenwhena preference isnotexpressed.

Thusfarmoreattention hasbeengivento possible basesofthediscrimina-

tionthanto causes ofthepreference. Although someanimals arecapable atbirthofperception ofdepthatanedge,whatistheprecise natureofthe

processthat occurson firstencounterwiththe cliffsituationthat resultsin

thediscrimination andpreference observed? Whatdifference is therein the

process between suchanimals andothers, likethecat,wherethingsgo differently? Istheachievement oflocomotion important onlyinmaking possible themovement necessary to showthepreference, or do events occur inthedevelopment oflocomotion itself thatcontribute totheprocessesofdiscrimination andpreference formation? Betterknowledge of the

mechanism involvedwouldhelp in the searchfor determinants of its

smooth operation. A search guided bysuchknowledge mightperhaps

revealthat experience playsmoreof a role than is now believed.The

workofSchneirla andhisstudents andofLehrman provide examples of thecomplex andsubtlewaysin whichexperience, evenin activities not

directly involved in thepatternitself,mayaffecttheemergence of an apparently unlearned behavior pattern. Theneedtoidentify underlying

processes andtheirdeterminants existsapartfromthequestion ofwhether experience is a factor or not.

AsGibson described herstudies, I wondered at several pointsaboutthe

fateofindividual subjects inthegroup results shereported. Forexample,

in the studyof graphicformssometransformations wereharderto dis-

criminate fromthestandard thanothers.Thereis a parallel herewiththe findings fromstudies ofGottschaldt figures. Depending onthestructure of

thecomplex figure, it maybeeasyordifficult to findthesimple figure within it.Continuing theparallel, I wonder whether, justassomepeople findtheGottschaldt-figures taskmoredifficult thandootherpeople, there

maybe differences amongchildren in easeof discriminating transformationsfromstandardsin Gibsonsmaterial. I recallheretoo thatthreeof 36

children testedinthecliffsituation seemed to preferthedeepside.Who

weretheseexceptional children? Thoughtheirnumberis small,onewon-

168

E. J. Gibson

ders what made them violate the clear group trend. I hope that in time Gibson may be able to considerissuesof individuality, as we may perhaps call them, raised by questions such as these. It is worth noting here that Wemer has recently commentedon the need, in both researchand theory, for greater attention to the problem of individuality in development. The kind of careful working through of situations in which Gibson has been engagedis a necessaryprelude to the study of differencesamong people in their behavior in these situations; Some of the work in the perceptionpersonality field, for example, reflectsthe consequences of failure to respect sucha sequence . I should like to comment further on the concept of individuality, since it representsa different, though by no meanscontradictory, emphasisfrom that in Gibson's work. In the areaof perceptualdevelopment an interest in individual functioning, implied by the concept of individuality, involves concernwith several issues. One is individual differencesin pace and direction of development in various perceptualareas. Another is individual self-consistencyin mode of perceptual functioning at any given stage of development, as judged from perfonnance across a series of situations. A related issue concerns stability of characteristicsof individual perceptual functioning over time. Still another issueis the relation between the courseof a particular child's perceptual growth and events going on elsewherein his psychological development, in the area of personality, if you will . Finally, there is the issueof the sourcesof individual differencesin perceptualgrowth, which means, in effect, the nature of the forces, constitutional and experiential, which, in interaction, direct a child's perceptual development along the particular courseit follows. I would like to describevery briefly a few of our recent studies,...carried out in collaborationwith Ruth Dyk, HannaFaterson, Donald Goodenough, and StephenKarp (Witkin et al [1962]) that illustrate the possibilities and difficulties of studies directed at the issuementioned. Becauseof our concern with analytical aspectsof perception, someof thesestudiesbear upon Gibson's interest in differentiation of object from background. For some time now we have been conducting cross-sectionaland longitudinal studiesof development, guided by conceptswhich grew out of our earlier investigations of perception-personality relationshipsThe particular studiesI want to mention were basedon the view, very briefly stated, that in early experiencethe body-field matrix is a more or lessfused "perceptual mass." Segregation of self from field and the further crystallization of experiencewithin eachof these "segments" proceedin sucha way during development that progress toward articulation in one area is dependent upon and fostersachievementof articulation in the other. This view implies that a tendency for experienceto be articulated- that is, analyzed and

Discrimination of DepthandGraphicSymbols 169

structuredis apttobemanifested whether theexperience hasitsprimary

sourceinonesownactions, attributes, andfeelings or inobjectsandevents

outside. These notions ledustoexamine childrens mode ofexperiencing,

viewed fromthestandpoint ofarticulateness, ina varietyofcircumstances. Ourearlier workhadbeenconcerned particularly withdifferences among peoplein theextentto whichperception tendsto be analytical. Asone extension of thatworkwe haveexplored the analytical qualityin intellectualfunctioning. Resultsof thesestudies,and similarstudiesin other

laboratories, haveshown thata tendency toexperience inmoreanalytical or lessanalytical fashion characterizes a personsintellectual activityas

muchas his perception. Thereis now considerable evidencethat children

andadultswitha relatively moreanalytical wayof perceiving do sign-

ificantlybetter at intellectual tasksin whichessentialelementsmust be

isolated fromthecontextinwhichtheyarepresented andrecombined into newrelationships. Forexample, Harris 1 hasshownthatpeoplewhose perception tendsto be analytical, or fieldindependent, as wehavecalledit,

findit relatively easy,inDunckers insight problems, to extract a part

froma familiarfunctionalcontextanduseit in anothercontext.As a second

example, Goodenough andKarp(1961) foundthattestsofperceptual field dependence wereloaded onthesamefactorastheblockdesign, picture completion, andobject assembly subtests oftheWechsler Intelligence Scale forChildren, subtests which aresimilar to theperceptual testsinthatthey

require theovercoming ofanembedding context. Theperceptual testsdid

not appearon verbal-comprehension or attention-concentration factors, defined by other WISC subtests.

Otherstudies alsosuggest thatchildren whopassively accept theprevailing organization of thefield,ratherthanexperiencing it analytically, tendto leaveas is stimulus material thatis unorganized andtherefore experience it aspoorlystructured andvague.Thiswasshown, forexample, ina studyoftheextenttowhichchildrens Rorschach percepts reflected an

attemptto imposestructureon the amorphousink-blotstimuli.

Considering analysisandstructuringascomplementary aspects of articulation, it issuggested thatthereareidentifiable differences among

children intheextentto whichtheirexperience tendsto bearticulated, or,

in contrast, global.

Whereasthestudiesmentioned wereconcerned mainlywiththenature of childrensexperience of situationsexternalto themselves, otherstudies

focused moreon the natureof childrens experience of themselves. One approach usedin exploring experience oftheselfwasto studychildrens conceptions oftheirbodies. Thesestudies werecarried outwiththeexpec-

tationthatchildren whoshowa relatively articulated wayofexperiencing

invarious perceptual andintellectual taskswouldtendto havea relatively

articulatedconceptof the body. Sucha relationwas demonstratedin

170

E. J. Gibson

onestudy,forexample, whichevaluated childrens figuredrawings from thestandpoint of sucharticulation criteria as formlevel,integration of

parts, sexdifferentiation, andrepresentation ofrole.Thisrelation hasbeen confirmed in otherstudiesusingmoreexperimental procedures to infera persons view of his body.

Another approach to thenatureofexperience oftheselfwasbasedon theconcept thatprogress towardself-differentiation duringdevelopment entailsthe childsgrowingawareness of needs,feelings, and attributes whichhe recognizes as hisownandtheiridentification as distinctfrom thoseof others.Wehaveusedthetermsense of separateidentity to refer to the outcomeof thistrendin development. A senseof separateidentity

implies anexperience oftheselfassegregated andstructured. Ready and continuing useofexternal frames ofreference fordefinition ofonesneeds, feelings, andattributes wouldbesuggestive ofa limited development 0f sense of separate identity.

A studyofadultsbyRudinandStagner (1958) is illustrative ofmany studies showing thatpersons witha relatively articulated wayofexperiencing tendtohavea relatively developed sense ofseparate identity. Thesubjectwasaskedto imagine himself in foursituations and,using special ratingscales, to describe himself in each.Thescores derived reflected extentof similarity amonga subjectsself-descriptions forthefour situations. As expected,relatively field-independent personsshowedsig-

nificantly lessfluctuation in theirviewsof themselves in the imaginary context,suggesting that theirexperience of themselves wasrelatively stable.

Theevidenceto dateindicates consistency in wayof experiencing the

external field,thebody,andtheself.Degreeofarticulation appears to be a commonqualityrunningthroughmuchof an individualsexperience.

Thesefindings areat leastconsistent withthe concept thatthereis a linkage, during development, among achievement of a developed concept

ofthebody,a segregated, structured self,andanarticulated wayofexperiencing the world.

The evidencethusfarbearsuponone of the problemsof individuality mentioned earlierindividualself-consistency. Someof thefirstresultsof

ourlongitudinal studiesdealwithanotherof theproblems: therelative

stability ofindividual functioning during development. Though wehave onlybeguntoanalyze theresults ofthesestudies, it isalready clearthat, withinthegeneral groupchanges thatoccurwithgrowth, children show markedrelativestabilityintheanalytical aspectofperceiving andinarticulatenessof bodyconcept,evenovera seven-year period. Ourevidenceat onepointsuggesteddifferences amongchildreninpace

ofdevelopment oftheinterrelated cluster ofcharacteristics wewerestudy-

Discrimination ofDepthandGraphic Symbols 171

ing.Webecame involved instillanother oftheproblems ofindividuality whenweundertook to determine someofthepossible sources ofthese differences. Inoneseries ofstudies wehavebeeninvestigating childrens

earlyrelations to theirfamilies, particularly theirmothers. Fromthecon-

ceptual framework weuseinviewing thecontrasting ways ofexperiencing,

itmustbeevident whyweconsider early mother-child interactions aspotentially important inthedevelopment ofthese ways ofexperiencing. The

studies havebeeninprogress forsometime,but,because ofthemethod-

ological complexities involved, theymustbeconsidered firststeps. We havethusfarinvestigated mother-child interrelations through interviews

withmothers and,to a lesserextent,through interviews withchildren and throughchildrens TATstories. Wehavealsodonesomestudiesof the

mother as aperson. Theresults ofthese studies, andaconfirmatory study bySeder, suggest thatmothers ofchildren whose experience offield, body, andselftendstoberelatively articulated haveinteracted withtheirchildren insuch awayastofoster separation from mother aridtoprovide standards bothfordealing withtheenvironment andforhandling impulses. Theprocesses underlying theseverygeneral relations require moreprecise defini-

tion,andtheimportant question ofcauseandeffectremains unanswered.

Wehopesomeprogress maybemadewiththeseproblems through studies ofrelations between characteristics ofchildren ininfancy andthe kinds ofdifferences wehavebeenfinding laterindevelopment. Through thecooperation ofSibylle Escalona andLois Murphy wewere recently able

to see72ofthechildren observed in infancy byEscalona andhercollaborators.Thechildren were6 to 8 yearsoldwhenwerestudied themand

their mothers. Thedata-collection stage oftheinvestigation iscompleted, Ihope thatintime theapproach toperceptual development represented

buttheanalysis ofdatahasnotyetbegun.

byGibsons workandtheapproach represented bythestudies justdescribed willfinda congenial meeting ground.

Note

1. Harris,Frances.Personal communication.

References

Goodenough, D.R..&Karp, S.A.Field dependence andintellectual functioning. Unpublishedstudy, 1961. Rudin, S.A.,&Stagner, R.Figure-ground phenomena intheperception ofphysical and socialstimuli.J. Psychol., 1958,45, 213225. Witkin, H.A.,Dyk, R.B., Faterson, H.F., Goodenough, D.R.,&Karp, S.A.Psychological differentiation: studies ofdevelopment. Wiley,1962.

172

E. J. Gibson

Group Discussion

The discussionof Gibsonspapercenteredaroundher two experimental

situations. Withrespectto the perception of depthat an edge,Gibson made the following additional points:

1. Therewereso fewexceptionsin the preference of humaninfants forthe shallowsideof thevisualcliffthatconsistentindividual differenceswere hard to obtain.The only three infantsin the original

experiment whocouldbecoaxed ontothedeepsideofthecliffwere boys.

2. Theapparentimportance of visualexperience for somespecies (e.g.,dark-vs.light-reared kittens) maynotmeanthatlearning, aswe

usually thinkofit, is required forthedevelopment ofperception of depthat anedge.Hesshasshownthattheresponse ofcertain birds to shadows, a response thatmaturesduringa critical period,doesnot

develop indark-reared animals, simply because shadows arenotavailableduringthecritical period.Maturation afterbirthnecessarily involves all interactions with environmental stimulation.

Levin,Berlyne, andJeffrey raisedquestions aboutthemechanisms of

depthperception inrelation totheenhancement experiments reported by

Gibson. Levinproposed thatstereometric objectsonthewallsofthehome

cages offered tactile aswellasvisual cuesforgenerating distinctiveness of thestimuliandenhancement of subsequent discrimination learning.Gibson

agreed thatthiswaspossible, andconsistent withHarlows findings, but

noted also that the most importantdepth cue in visualcliffperformance

appeared tobemotion parallax. Berlyne proposed thatthecut-out shapes on

thehomecageshadtheuniquepropertyof eliciting attentiveor orienting

responses because ofthemotion parallax taking placeattheiredgeswhenever the animal moved. He proposed that the inherently interesting per-

ceptual eventsofdisappearance andreappearance ofmoredistant objects, whentemporarily maskedby thecut-outs,elicitedorientingresponses to the contoursof the forms,whichin turn enhancedtheirfamiliarity and

distinctivenessas cues. Gibson agreed that attention would accompany

movement, butnotedthatotherdepthcuesarealsorelevantto differentiation of the stimuli.Withrespectto movementof the animal,shenoted muchcliffbehaviorresembling VTEmovementsin all species.Berlyne

pointed outthat,evenifonecould demonstrate thatperception ofdepthat anedgeismoreamatter ofmaturation thanoflearning, thereremains the response

differentiationof animalson the visual cliff.Although they may

perceive which sideisdeeper without muchexperience, learning mustbe

involved in theappropriate responses andpreferences whichtheydisplay.

Kagan suggested thattheresponses might beexplained onthebasisof

a generally nearsighted visualaccommodation onthepartoftheorganism

Discrimination ofDepthandGraphic Symbols 173

anda corresponding level ofadaptation forrateofchange ofvisual pat-

ternsatanedge. When therelative rateofchange ofvisual pattern accompanying headmovements andlocomotion is eithertoofast,tooslow,or

unexpected, anavoidance response maybeelicited. Inotherwords, the organism avoids discrepancies fromtheusual experiences inhisvisual field. Gibson replied thatthelevel-of-adaptation argument would failto

explain thefullydeveloped preferences indark-reared rats,whocouldnot

haveestablished adaptation levels formovement at anedge.Thedark-

reared rats,furthermore, showed thepreference without learning anapSigel proposed thataccessibility andthepotentiality forexploration of theshallow sidemight make it moreattractive, andGibson replied that propriate response.

thefailure ofextinction ofpreference inmanytrialsforratswouldmake

sucha hypothesisunlikely.

Discussion oftheresearch onperception ofgraphic symbols centered on twoproblems. Onewasthedegreeto whichthetransformations selected

byGibson doin factrepresent critical features ofgraphemes. Brown pointed outthat,while adistinctive features analysis ofgraphic symbols is

needed andcanbecarried outinthesame wayastheanalysis ofdistinctive features ofphonemes, someofthefeatures proposed byGibson areeco-

nomical andothers arenot.Analogous transformations forphonemes can

beeithereconomical ornot.Forexample, voicing isaneconomical distinc-

tivefeature because itisalways relevant. Infact,itdivides thepopulation dental, orplosive arenoteconomical features, however. Inmany cases they ofphonemes intotwogroups: voiced andunvoiced. Features suchaslabial,

areirrelevant. Similarly withrespect to graphemes, thetransformations, openvs.closed andstraight vs.curved, areeconomical features. Theyare

distinctions which divide thepopulation intoimportant subgroups. Reversal androtation, however, areoperations linking twographemes, describing nottheirfeatures, buttheirrelation tooneanother. Hence, asfeatures they arelesseconomical andlessdistinctive. Baldwin suggested that,using somearbitrary specification, onecould define allletters asbeingright-

handedor left-handed, thusmaking reversal an absolute ratherthana relative feature; butit wasgenerally conceded thatsomeof thefeatures proposed forgraphemes weremoredistinctive, whileotherswereuseful

transformations ingenerating letterforms, butlacked independent disThesecond problem relating to theperception of graphemes that evoked considerable discussion wasthatofgrapheme-phoneme correspontinctiveness fortheperception ofgraphemes.

dence. Some letterclusters areclearly morereadily perceived thanothers,

buttherewassomequestion as to whether thedifference is dueto the

degree ofinvariant grapheme-phoneme correspondence, asGibson suggested, ortothefluency andcommonality ofcertain grapheme sequences

174

E. J. Gibson

inwritten language. Putanother way,docertain letterclusters havelower tachistoscopic thresholds because theyaremorepronounceable, orbecause theyareletterchainscontaining highersequential probabilities in written language? Wailach reported thatthefinding of MillerandBruner, that thresholdsare lowerwhenthe nonsensewordspresentedcontainhigher orders of statisticalapproximationto English,can be demonstratedin children.Moreover,when childrenare matchedfor their abilityto perceive

randomaggregations ofletters(zeroorder),thentheamountof improvementtheyshowatthefourthorderofapproximation predicts theirreading

andspelling achievement, butnottheirnonverbal achievements. Thatis, onlychildren highin verbalandspelling skillscouldprofitperceptually fromincreasedcorrespondence of sequentialprobabilities to thoseencountered inordinaryEnglishwords.Jeffreysuggestedthatpronunciability

wasbeingusedas anindexof sequential dependency or frequency of occurrence, andBaldwin notedthatthetwovariables couldandshouldbe

separated. Gibson proposed thatnonsense trigrams ofhighandlowfrequency andofhighandlowpronunciability could bedeveloped inorder independently to testtheeffects ofthetwovariables onperception. However,Gibsonpointedout that sequential probability, as such,doesnot

operate toconstitute units,andthereisevidence nowtoshowthathigher order units are formedby invariantrelationshipsof letter patternsto pronunciation.

Ervinstatedthat the use of trigramsignoresthe positionof the letter clustersin the word,a variablethat is criticalfor recognition of whole words.One mustconsidernot onlythe sequential probabilities of letters, but of clusters,phonemes, andsyllablesas well.Gibsonprovidedan exam-

pieofthethree-cluster nonsense wordGLURCK, which ispronounceabl and canbe perceived at relatively low thresholds. Reversing the two

consonantclusters,however,givesCKURGL. Althoughthiswordis com-

posedof threeclusters in theorder,consonant, vowel, consonant, it is virtually unpronounceable, andis perceived onlyat highthresholds. She went on to note that,althoughthe firstpart of the word is mostprominent

andcriticalforrecognition, perceptual differences wereobtainedon the basisof the pronunciability of the last part as well.

Baldwin proposed thattherulesforgenerating grapheme clusters dictate

thatwithinthe clustersthe frequencies of lettersequence matchthosein

English. Inorderto separate frequency frompronunciability, it would be

necessary to considerthe relationsbetweenclustersas well.Levinpro-

posedthatlowcommonality trigrams couldbepresented withcontrolled

frequencies, thusmaking position inthewordirrelevant. Picksuggested the useof deafmutesas controlsfor pronunciability It wasnotedby Kessenthatpronunciability is alsoa problemin response

organization. Ifonecouldpretrain a differentiation ofresponse sounds and

Discrimination of Depth and Graphic Symbols

175

independentlyvary the subject's experiencewith isolated graphemes , then one could predict differential performanceswhen the stimulus is a novel combination of graphemesand the responseis a novel combination of phonemes. Gibson replied that responseorganization is largely a problem

of differentiatingphonemes .

Brown proposed that three different sets of rules need to be considered:

(1) the sequentialand combinatorial rules for phonemes, (2) the sequential and combinatorialrules for graphemes,and (3) the rules of correspondence betweengraphemeand phoneme. The consensusappearedto be that these

threetypesof rulesneedto be studiedboth separatelyanddevelopment ally , in terms of distinctive features, as has been done by Jakobson, Gibson ,

and Hockett .

Sincethe time of Darwinand the acceptance of the doctrineof evolution of species , psychologistshave contemplatedthe phylogeneticdevelopment of behavioras a markof adaptationto an animal's environment . In AmericanScientist , 1970, 58, 98- 107.

178

E. J. Gibson

the late nineteenth century , the comparative psychologist 's eagerness to fit behavior into the evolutionary scheme took some amusing, and by hind sight , naive trends . G. J. Romanes, one of the best of the so-called " anecdotalists ," spoke for them when he said, " I hold that if the doctrine of Organic Evolution is accepted, it carries with it , as a necessary corollary , the doctrine of Mental Evolution , at all events as far as the brute creation is concerned" (1895, p. 8). In two volumes called Mental Evolution in Animals and Mental Evolution in Man he prepared a tree and a chart which served, he thought , to represent the " leading features of psychogenesis throughout the animal kingdom " and also the " principal stages of Mental Evolution in M an." The chart lists, under " products of intellectual development ," a number of faculties which are ranked from lowest to highest - from protoplasmic movements to morality . Then , in a column titled the " psychological scale," there is listed in correlation with the faculties the animal order where each faculty presumably first makes its appearance. " Memory " comes in with the echinoderms , " association by contiguity " with molluscs, " association by similarity " with fish, " recognition of persons" with reptiles and cephalopods, " recognition of pictures , understanding of words , and dreaming " with birds , and "morality " with anthropoid apes and the dog . Along with this, in a third column , is " psychogenesis in man," and we find the order of appearance of the faculties recapitulated ; association by contiguity at 7 weeks, association by similarity at 10 weeks, recognition of persons at 4 months , recognition of pictures and words at 8 months , and so on. Evidence for this order was based almost entirely on anecdote and informal observation . As comparative psychology developed an experi mental method , the work of Romanes and his generation was derided and banished as a shameful page in the history of a new science. Yet the adaptiveness of behavior and evolutionary continuity were never quite forgotten . One no longer looked for faculties but for " laws" of behavior . Hull 's theory of learning , which converted half the psychological world , stressed the biological adaptiveness of the conditioned reflex and the principle of reinforcement which operated by reducing biological drives or need conditions , thereby strengthening behaviors useful to the organism (Hull 1943). The continuity was there too , because it was presumed that one could investigate these mechanisms in the rat and apply the findings to man. Seeking to understand man's behavior by experimenting with the rat has fallen off in fashion, in its turn , but the ethologists have revived the biological tradition begun by Darwin and furthered by Romanes in a new and more sophisticated spirit of naturalism . Behavior that is specific to the species has become of interest and is studied in relation to the ecology of the species, thus revealing its adaptiveness.

Development of Perceptionas an Adaptive Process

179

In this abbreviatedsketch of the influenceof evolutionary conceptson psychology, where does perception come in? Do only executivebehaviors like spinning webs, building nests, running through mazes, or pressing bars have adaptive value? Or is there a phylogenesisof perception and a paralleldevelopmentin the individual? Is there perceptualleamingwhich is adaptive? Or must learning be only on the responseside, as behaviorists

believed ?

Karl Lashleywas one of the first to raisethesequestionsand to point out the role of perception in species-specific behavior and in evolution. In the "Experimental Analysis of Instinctive Behavior" (Lashley 1938) he stressedthe importance of studying the innate components of "sensory organization

," as well as the motor

aspects of behavior . The essential first

step, he said, was "analysisof the propertiesof the stimulus situation which

arereallyeffectivein arousingthe behavior." Understanding of the motor activities "hinges on these perceptualproblems." Much of his work, from the early naturalisticstudiesof terns to later studiesof stimulusequivalence in the rat, was directed at this problem. He never forgot the importanceof evolution for understandingan animal's behavior. In "PersistentProblems in the Evolution of Mind " (1949), he told us that "the limits of capacity of each order of animals are set by the kinds of relations among objects that

it can perceive. The development of the individual is a slow maturation of such capacities" (p. 460). "It is not the fact of learning but what is learned that differentiatesanimalsin the evolutionary scale. The learning of higher animalsinvolves a perception of relations which is beyond the capacity of the lower " (p . 458 ).

The latter statementhe illustrated by comparingthe behavior of a spider monkey and a chimpanzeein a matching problem. The monkey was required to choose a red or a green square, according as a red or a green squarewas given as a model. When the squareswere placedas in row a in Figure 11.1 , the monkey never improved above chance in 1000 trials . But when there was contact between

the model and the test square , as in rows

h, d, e, and g, he quickly achieved errorless choice. He saw the model, Lashley thought, as a pointer or a signal but did not perceive the relation of similarity. The chimpanzee , on the other hand, graspedit quickly. Backgroundof the Theory

To tell you how I seeperceptualdevelopment as a mark of adaptiveness , I must explain first what I think perception is and what perceptuallearning is. Then I will distinguish two modes of perceiving and illustrate with experimentswhat we know about their development in phylogeny and in

the individual

.

180

E. J. Gibson -

. .. . .

a

b

c

d

e

f

g

-

m

.

50%

.

100%

. .

0

. I_IIIIIII [ - 111111. .

50%

100%

I III ~

-~

=

---

Figure 11.1 Arrangements of red and green squarespresentedto a spider monkey and a chimpanzeein a matching task. The percentagesat the right represent the 6nallevel of accuracyattained in each situation (from Lashley 1949; reproduced by permission of the Quarterly Reviewof Biology).

Development of Perceptionas an Adaptive Process

181

Perception is extracting information from stimulation (Gibson 1966). Stimulation emanatesfrom the objectsand surfacesand eventsin the world around us and it carriesinformation about them; though different from them, it specifies them. If we were to consider stimulation only as individual rays of light or vibrations in the air, this specificationwould not be intelligible, becauseinformation about objects and layout of the world around us lies in relations, like edgesbetweenthings; it is not punctate, but structuredover spaceand time. Not only is there information about things in stimulation; there is rich information, far more, potentially, than we utilize. Let me give an example. Some animals, especially bats and dolphins, locate food and find their way around by means of echolocation. The dolphin emits clicking sounds at varying rates from one per second to bursts of 500 or more. Theseclicks are thought to be usedfor food-finding and navigation. To quote a dolphin expert, "The click trains, or sonar, searchthe seascapein front of a dolphin in much the sameway that the cone of light from a miner's headlampshows his way through a mine. In the presenceof reflected light, we see where we look. In the presenceof reflectedsound, or the echoesof their own clicks, dolphins hear where they point their beamof sound. The click-echoesreturned from the environment before the moving dolphin are information-bearing. The echoescontain information about the size, shape, location, movement if any, and texture of the living and non-living things in the water" (McVay 1967, p. 8). It has in fact beendemonstratedthat dolphins candifferentiatein this way objects of different sizesand shapeand even different metallic substances , and can swim an obstaclecoursewithout collision. Three points emergefrom this: one, that potential stimulus information about featuresof the environment is vast; second, that the information accuratelyspecifiesthe layout of the world and the objects in it; and third, that perception is an active process, a search for the relevant information that specifiesthe path an animal needs to travel, the obstaclesto be avoided, the mate or the food to be approached. So perception, functionally speaking, is extracting information about the world from stimulation, a highly adaptive processsince the animal must somehow discover where to go, what to seize, and what to avoid. What kind of world is there to perceive? We can describeit in severalways. I choosea classificationthat refers to properties of the environment. These includepropertiesof the spatiallayout(surfaces , edges, drop-offs); properties of events(motion, occlusion, appearance , disappearance , and reappearance ); and properties of objectsthat make them distinguishable and identifiable. For man at least we can include another class; man-made symbols- coded itemsthat stand for objects and events, suchas speechand writing .

182

E. J. Gibson

Animals perceive the surfaces and objects and events in their surround -

ings by way of stimulation which specifiesthem. But they seldom do this perfectly, and the potential information in stimulation is vastly greater than that which becomeseffective. To understand how potential information becomeseffective we need the conceptsof perceptualdevelopment and perceptuallearning . As the higher-order invariants and structure that uniquely specify objectsand eventsare progressivelyextractedfrom the total stimulus flux. so does perception becomemore differentiated and more specific ~ to those things. This is a processwhich goes on in the evolution of species and also, I think, in the developmentof the individual. How do animals learnto perceive the permanentdistinguishableproperties of the world in the changing flux of stimulation? Not , I believe, by association, but by a process of extracting the invariant information from the variable

flux . I think

several

processes

are involved

, all attentional

ones .

{See Gibson 1969 for a detailed statement of the theory of perceptual learning) One is perceptualabstraction, akin to what JamescalledIIdissociation by varying concomitants" (note the dissociationas opposed to association; something is being pulled out from context, insteadof being added on). Another is filtering of the irrelevant, an attenuation in the perceiving of ,

random , varying , noninfonnative

aspects of stimulation . A third is active ,

exploratory search. The dolphin beaminghis clicking soundsis an example of the latter .

Another example is active touch (Gibson 1962) . When a blindfolded subject is handed an unfamiliar object and asked to learn to identify it so as

to be able to match it visually to one of a larger set of similar objects, what doeshe do? He runs his fingersround its contourssearchingfor distinguishing features, and pressesit with different finger combinationsto detennine its proportions . The stimulation to which he exposes himself is constantly varying and, from the point of view of individual receptors, never the

same. Yet he picks up from this variable flux of tactual-kinaestheticstimulation constant structural properties like curves, edges, and indentations which are translatable into visual properties .

With respectto the searchprocessin perceptuallearning, a very important question is what tenninates the search and thus selects what is learned.

For many years no one questionedthe proposition that external reinforcement (e.g. food or shock) is the selectiveprinciple for learning things like bar-pressing or choosing one ann of a maze rather than another . But is a distinctive

feature selected as relevant

because it wins a reward

or avoids

punishment? Is this the way that higher-order structural relations are detected? Although this might happenin a teachingsituation, I do not think it is the true principle of perceptualleaming. So much of it goes on very early in life and is necessarilyself-regulated. No experimenteris on hand to deliver reinforcement; probably not even a parent could provide it

Development of Perceptionas an Adaptive Process

183

deliberately, sincehe seldom has any way of knowing just what the child .

.

.

IS perceivIng

.

I think the reinforcement is internal -

the reduction of uncertainty. Stimu -

lation is not only full of potential information; there is too much of it . There is a limit to what can be processed, and variable , random , irrelevant stimula -

tion leadsonly to perception of confusion- what someonehas referredto as cognitive clutter as opposedto cognitive order. But distinctive features, invariants, and higher-order structure serve the function of reducing uncertainty, taking order and continuity out of chaosand flux. The searchfor invariants, both low-level contrastive featuresand high-level order, is the task of perception , while detection of them at once reduces uncertainty and

is reinforcing.

: and Individuals PerceptualDevelopmentSpecies

With this brief background of theory, I propose to return to my first question. Is there perceptualdevelopment, in the animal seriesand in the individual, and is it adaptive? Are there trends in what is respondedto, as Lashley suggested7 In order to give some specific answers, I shall compare two modes of perceiving and give evidence, in both cases, of species

differencesand of developmentwithin the life span. The two modes are perception of space and events in space and the perception of objects and permanent items, like written letters , that can be

approachedand examinedclosely. I have chosento contrast thesebecause there is reasonto think that in their phylogenetic development there is a considerabledifference between them. Localizing one-self in the spatial layout or monitoring events going on in the space around one seem to

develop earlier and to be neurologically more primitive than fine-grain identification of objects and outline figures such as letters. This difference

is akin to a distinction

within

visual perception

drawn in

a recent paper by Trevarthen (1968), who speaksof "two mechanismsof vision

." One

of these

he calls

" ambient

vision

." It has to do with

orienta

-

tions of the head, postural adjustments, and locomotion in relationship to spatial configurations of contours, surfaces , events, and objects. The other he calls "focal vision." It is applied to one spaceand a specific kind of object; it servesto examineand identify. "Ambient vision in primates," says T revarthen , " resembles the vision of primitive active vertebrates . . . . At any instant , an extensive portion of the

behavioral spacearound the body is mappedby this ambient visual mode; in primates, somewhat more than a frontal hemisphereis apprehended . With large rotations of the head or whole body, an animal may quickly scanall of the spacecloseto his body and thus obtain a visual impressionof the large features in it. The visual mechanismis strongly stimulated by

184 E. J. Gibson parallax changescausedby translation of the eye, and the receptor mechanism is particularly sensitive to the velocities of displacement of continuities in the light pattern on the retina" (p. 328). "In contrast with this vision of ambient space, focal vision, enormously developed in diurnal primates, is applied to obtain detailed vision." Its scope at any given instant is restricted, but it is extended over time by samplingmovementsof the eyes. An areaof interest may thus be brought to full attention and "analyzed as if carried close in by a zoom lens" (p. 329). Focalvision in primatesappearsto be primarily a function of the cortex. Even in rodents a comparable, though less pronounced, distinction may exist. Schneider(1967) working with hamstersfound that ablation of the visual cortex left the animal with only a minimal ability to tell what he was seeing, but left nearly intact his ability to find whereit was. He was unable to discriminate and identify objects, but could localize them in space. Ablation of the superior colliculus produced the opposite effect; the hamster knew what he was seeing but behaved as if he didn't know where it was. I shall say no more about this neurologicaldistinction, sinceI have made no contribution to it, but it supports the point I intend to make; that discrimination of events in spaceis primitive, both phylogenetically and ontogenetically, while development progressestoward differentiation of form in objects and two -dimensionalprojections. In other words, fine-grain identification of objects or patterns is the later achievement; its development continues over a long time; learning plays a prominent role in it as comparedwith perceiving the spatiallayout and events; and we can expect to find more striking phylogenetic differences. Perception of Space Consider, first, development of the perception of spaceand of events in space. Is there phylogenetic continuity here within the vertebrate phylum? Indeed there is. The similarities between speciesare far greater than their differencesin this respect. We can adduceevidencefor this in three important cases- perception of imminent collision (called "looming"); perception of depth-downward; and perceivedconstancyof the sizesof things. Looming can be defined as acceleratedmagnification of the form of an approaching object. It is an optical event over time. It specifiesa future collision (Schiff, Caviness, and Gibson 1962) . If a vehicle or even a small object such as a baseballis perceivedas coming directly toward him by a human adult, he ducks or dodges out of the way. Is the perception of imminent collision together with its avoidance instinctive? If so, in what species , and how early? Schiff (1965) constructed an artificial looming

Development of Perceptionas an Adaptive Process

185

situation in which nothing actually approachedthe animal observer but there was abstractoptical information for something approaching. In Schiff's experiments, a shadow was projected by a shadow-casting device on a large translucentscreenin front of the animal. The screenwas large enough to fill a wide visual angle. The projected shadow could be made to undergo continuously acceleratedmagnification until it filled the

screenor, on the otherhand, continuouslydecelerating minification . Magnification resulted in a visual impression of an object approaching at a uniform speed. Minification gave a visual impressionof an object receding into the distance. The projected silhouette could be varied in form , so as to

compare, for instance, jagged contours with smooth ones, or silhouettesof meaningfulobjectswith meaninglessones. Subjectsstudied includedfiddler crabs, frogs, chicks, kittens, monkeys, and humans. The crabs respondedto magnification (but not to minification) by running backwards, flinching, or flattening out. Frogs jumped away from the ghostly approachingobject. Chicksrespondedmore often to magnification

thanminificationby running, crouching , andhopping. Kittens(28 daysold) tended to respond to magnification with struggle and head movements, but the kittens were restrained

in holders and well -differentiated

avoidance

behavior did not show up clearly. Rhesusmonkeys (including infants five to eight months of age) were observed in the situation under four condi -

tions (magnification, minification, lightening of the screen, and darkeningof the screen). Both young and adult animals withdrew rapidly in response to

the approachdisplay, leaping to the rear of the cage. Alarm criesfrequently accompaniedretreat in the younger animals. The recedingdisplay brought responseswhich might be described as curiosity, but never retreat. The

lighteninganddarkeningof the screenhadno effect, andthis servedasa control, that is, a changeof mere stimulation as comparedwith changeof magnification. The adaptivenessof the responsesto optical magnification is illustrated by the turtle . Hayes and Saiff (1967) investigated what they termed the

"visualalarmreaction " in the turtle. A looming shadowon a screenwas used, as in Schiff's experiments. The turtles respondedto magnification by withdrawing the head into the shell. What about a human subject? Schiff measuredthe galvanic skin reflex in adult human subjectsin the looming situation. There was decreaseof skin resistancein the majority of subjectsfor magnification but not for minification. Human infants, Burton White found (1969), begin to blink at a rapidly approaching object (with air currents controlled) at about three weeksof age. The reliability of the responseincreasedfor another 10 or 12 weeks. Perhapssensitivity to visual approach of a missile takes this long to mature or be learnedin the human infant; but perhapsanother indicator responsewould show that it is picked up even earlier. Some observers

E. J. Gibson

186

claim that attempted head withdrawal to visual looming occursas early as two weeksin human infants. Does it matter what shapeor object characteristicsthe expandingsilhouette has? Schiff tried objects of different shapesas shadow-casters, but he did not find that silhouettesof theseobjects had a differential influence on avoidance behavior. It was the event of looming as such, not an identifiable object, that controlled the avoidant behavior. The functional usefulnessof this lack of specificity is obvious; quick avoidance of a fast approaching object is often necessaryfor avoidance of collision, while fine-grain identification of the object hardly matters. Now let me give you somephylogenetic comparisonsfor avoidanceof a falling-off place, that is, a drop-off of the ground. Depth downwards is specifiedin the light to the animal's eye. Does this information by itself cause the animal to avoid it? Some years ago, Dr. Richard Walk and I

SUPPORT

LIGHT

DIFFUSING

' \.

.

/ .'

llilllll i1IIIIIIIIi :~

CANOPY

SHALLOW FLOOR

PATTERN

SEEN

THROUGH

GLASS

ON

DEEP

SI DE

GLASS

-

SI DE : BACKED

BY PATTERN - SURFACE

LAMP

FRAMEHANGS TO INCHES ~ ABOVE FLOOR Figure 11.2 OrawinQ of a visual cliff (from Walk and Gibson 1961; copyright by American Psycho...... logical Assoc. and reproducedby permission).

187

Development of Perceptionas an Adaptive Process constructed

an

1960

)

drop

-

.

We

off

other

for

it

a

downward

on

center

board

inches

a

"

more

center

board

tactual

information

.

side

Cliff

because

we

11

Figure

as

-

,

cliff

an

is

inch

rat

or

a

sheet

or

.

of

two

,

chick

.

The

is

from

the

floor

is

extends

the

10

from

board

currents

-

animal

similar

glass

air

all

con

out

A

below

and

eliminate

apparatus

extends

it

Walk

simulated

the

,

and

so

that

echoes

are

.

What

is

density

the

of

visual

to

yield

from

animal

'

s

' s

.

own

,

movement

in

For

-

11

two

the

two

the

11

Cross

section

subject

of

eye

and

American

.

The

best

the

light

might

information

,

produced

-

edge

.

the

by

as

the

Differences

the

cliff

in

edge

texture

SURFACE

. 3

' s

runway

in

parallax

movements

shallow

in

surfaces

binocular

TEXTURED

Figure

difference

perspective

head

to

the

sides

motion

especially

comparison

shows

animals

the

was

. 3

checkered

some

about

suggested

observed

Figure

the

eye

information

experiments

animal

?

texture

the

differential

our

information

optical

projected

by

an

floor

eliminated

a

to

shows

below

A

and

was

attempted

board

.

Gibson

there

. 2

a

two

side

(

because

.

or

other

question

, "

such

inch

floor

the

"

checker

an

the

the

for

off

,

A

,

on

above

-

animals

below

,

"

drop

board

one

. "

visual

the

center

this

cliff

small

on

or

equated

testing

answering

visual

and

for

for

placed

"

,

information

structed

is

apparatus

called

a

from

from

Psychological

visual

a

an

cliff

.

textured

The

diagram

surface

identical

at

surface

Assoc

shows

a

shallow

farther

.

and

below

reproduced

the

depth

pattern

below

( from

by

Walk

permission

of

his

light

station

and

Gibson

) .

projected

point

1961

to

on

;

a

copyright

center

the

188

E. J. Gibson

densitywere eliminatedand a monocularanimalwas usedto makethis observation . Many animalspecieshavebeentestedon the visualcliff: rodents,birds, turtles, cats(includinglions, tigers, and snow leopards ), sheepand goats, dogs, and of courseprimates(Walk and Gibson1961; Routtenbergand Glickman1964) . All thesespecies , saveflying onesor swimmingones, avoidedthe cliff edgeof the apparatus andchosethe safe,shallowedgeon the basisof visual informationalone. Texture must be presenton the ground under the animal, however, for a safesurfaceof supportto be perceived . The animalwill not walk out upona homogeneous , untextured surface - he demands"opticalsupport" aswell asfelt support.This surely hasvaluefor survival. Avoidanceof a drop-off anddependence upon opticalsupportmustbe developmentally primitive. This conclusionis suggestednot only by the continuity of the behaviorwithin the vertebratephylum, but also by ablationexperiments andontogeneticdata. Whenthe striatecortexof the cat is removed(Meyer 1963) patternvision in the senseof identification goes, but a catwill still avoida cliff, if he canmovefreely. Ontogenetically , Walk and I found that cliff avoidancedevelopsvery earlyand, in somespecies , without any opportunityfor learning.Precocial animalssuchas chicksand goatsavoid a cliff a few hoursafter birth, as soonasthey canbe tested. Ratsrearedin the darkavoida cliff assoonas they arebrought out, with no opportunityfor preliminaryvisualexperience.Primatescannotbe testedat birth, but humaninfantsavoida cliff as earlyasthey cancrawl. Monkeys, like humaninfants, arecarriedby their mothersin early infancy. But monkeysplacedon the untexturedglass without opticalsupportat threedaysof age(Rosenblum andCross1963) showedindicationsof emotionaldisturbance (crouching , vocalization , selfclasping , androcking), whereastherewasno disturbance whenthey were placedon the glasswith a texturejust below it. We may concludethat perceptionof a safesurfacein contrastto a drop-off appearsearly in evolutionandearly in life, and that little learningmay be requiredfor its appearance . It is modifiedby biasedcircumstances , suchasprolongeddark rearing, but terrestrialanimalsdo not generallyhave to be taught this usefuladaptation . Now considermy third case : perceivedconstancyof the sizesof things. Informationfor sizeconstancyis givennormallyby motion; motionof the objecttoward or away from a stationaryobserver , or movementof the observertowardthe object. Sinceinformationfor constantsize- the rule relatingsizeand distance - is given in motion, it shouldbelongwith my "primitive" modeof perceiving . Let us seeif it does, if thereis continuity over speciesand early developmentwithout any markeddependence on learningwith externalreinforcement .

189

Development of Perceptionas an Adaptive Process As

for

continuity

strated

in

( 1933

)

( 1968

) ;

and

the

and

locomote

or

were

target

' s

Is

be

of

seize

an

object

the

toward

at

appropriately

take

place

as

from

his

again

.

,

three

the

;

and

30

stimuli

.

the

cube

90

ingenious

were

.

such

a

a

mother

the

it

and

out

a

90

was

and

) .

After

.

placed

Figure

a

then

effective

perception

cube

in

.

one

tests

.

-

head

response

infant

a

( see

presented

the

placed

remarkably

- cm

con

method

of

cube

get

for

to

size

the

generalization

;

,

have

used

the

,

away

were

He

-

him

of

- turning

to

meters

apart

would

at

ap

hand

presenting

white

"

face

his

turning

.

head

turns

away

situation

.

might

motion

' s

evidence

- cm

for

3

the

It

moves

leftward

30

?

on

learning

something

meters

placed

training

,

do

around

with

size

together

found

introduced

3

cube

The

baby

- a - booing

and

to

placed

- cm

original

The

conditioned

stimulus

- turnings

most

.

with

If

,

the

the

might

)

elicit

stimulus

1

perceived

cube

next

imaRe

Does

this

constancy

.

If

11

. 4

counter

often in

at

placed

after

its

the

The

true

These

I

meter

) .

These

- balanced

same

at

3

most

same

meters

responses

- sized

,

' t

training

,

situation size

as

.

the

where

,

Next

it

when

the

it

attained

,

( test

and

-

test

that

;

is

one

stimu

to

often

projected

next

identical

even

evidence

cube

most

the

responded

most

training

,

cube

image

practice

the

it

been

retinal

in

elicit

like

the

hasn

,

objective

distance

to

most

expect ~

constancy

infants

the ,

expected

infant

would

next

the

.

the

one

size

projecting

most

cube

the

response

cube

placed

large

retinal

cube

,

evoke

1 )

be

to

constancy

stimulus

most

the

larger

to

would

appearing

size

one

the

to

one

has

( test

expect

3

situation

the

infant

away

lus

and

training

farther

size

new

" peek

learns

.

the

infant

for

and

things

in

for

But

to

reinforcement

( an

infant

it

;

has

being

thus

up

up

human

response

they

.

head

the

The

The

- cm

and

the

order

.

of

depend

them

.

respond

popping

disappearing

were

the

that

could

changed

constant

to

)

the

) ;

biologist

invariant

moves

in

Freeman

1958

indeed

distance

as

it

( 1966

Kluver

by

( Pastore

opportunities

months

to

eyes

reinforcement

training

and

Bower

two

trained

experimenter

relation

pick

-

?

guaranteed

for

,

was

in

and

sizes

the

to

by

)

How

apparent

position

demon

monkey

thoughtful

as

stimulation

,

conditioning

meter

eyes

are

early

young

infant

away

his

These

the

things

bends

structured

operant

an

.

the

been

duckling

.

extracting

observer

from

time

to

perceiving

she

the

constancy

' s

the

( 1951

in

expanding

observer

the

) ;

if

for

as

very

as

The

of

away

a

with

stancy

the

in

has

Gunter

surprise

or

in

baby

and

hours

no

constancy

) ;

by

size

conditions

or

proaches

is

accurately

involved

the

cat

1968

shrinking

or

since

It

size

( 1915

the

exhibit

things

movement

,

in

( Heller

) .

man

constantly

learning

well

of

1953

than

,

Kohler

) ;

rat

( Herter

species

by

( 1937

weanling

other

them

,

Locke

the

fish

animals

animal

chimpanzee

by

in

in

among

they

carne

last

identical

the

of

all

- sized

.

result

mean

?

Definitely

that

no

not

learning

,

since

is

even

involved

eight

in

weeks

the

development

gives

of

a

lot

of

,

E. J. Gibson

190

CONDITIONED

STIMULUS

TEST

STIMULI

1

TRUE SIZE I::J TRUE

DISTANCE

3

DISTANCE

3

1

3

[:J

1

RETINAL SIZE [:J RETINAL

2

~

CUES

J::J

DIFFERENT

SAME

DIFFERENT

Figure11.4 Schematic representation of cubesof differentsizesplacedat differentdistances in Bower's investigationof sizeconstancyin infants(from "The VisualWorld of Infants," by T. G. R. Bower, December1966. Copyright~ I 1966by ScientificAmerican , Inc. All rightsreserved ).

opportunity have

for

certain

or

could size

)

did

despite

of

not

with

to

methods of

the

a

mere

more

of

, let

cation there

spatial There striking learning

are

turn

to

objects

,

my

in

requires

him

object

be

very

in

their

,

),

therefore

sizes

( Heller

the

light

,

experiences , for

size

, appearing

probably

an

relating

rats into

changes

a of

rule

visual

con

develop

objects

-

when

indicate

perceiving

true

the

Some

. Later

not

reaching

- reared

brought .

could

as

motion

of

necessary

1961

such

. Dark

first

toward

learning

,

perceived

cliff

early

( Piaget

-

rather

than

.

the

, yes

of

of

perceptions

coded

and

, but

change individual

objects to

class use

-

stimuli

early

the to

appearance

fine

- grain

identifi

substitute as

for

there

is

-

them

with

.

the

?

continuity

the

other

species

evolutionary

of

discovery

the

visual

may

used

the

over

perceptions

features

the

so

things

and

continuity

is

for

when

attitude

of

. But

of

any acts

Objects

me

of

space does

analytical

localizing

Perception Now

it

that

paces

constancy on

in , but

many

nearness

size

means motor

opportunity to

moving develop

judgmental ment

an object

it

specific

so

performance

objects

stancy

walking

exhibit

normal

. But

with

provide

projective 1968

experience

association

distance

object

Is

visual

involved

in ,

.

and

The

, representations

differentiate

.

this

instance

also

there

for

human

a

child , and

long must

symbolic

is

a

course

good of

case

for

a

perceptual

learn

the

distinctive

items

that

human

life

Development ofPerception asanAdaptive Process 191

Whataboutthephylogeny of objectidentification? Certainly animals identify someobjects atquiteanearlyage.Theherring gullchick identifies

bya spotofredthebeakofitsparenthovering overit.Thisinformation is referredto by the ethologists as an innate releaser.Releasers seemin

manycasesto be verysimpleunlearned signalsforthedischarging of a fixedunlearned patternofresponses, likethechemical signalthatreleases

attackbehaviorin somespeciesof snake.A meretraceof the chemical in

a boxwillbringtheattack. Sometimes theeffective stimulus patternis

more complex,as is the visualpatternthat constitutesowlness and

releases mobbing behavior in thechaffinch, or thequitecomplex and informative song patterns of many birds, but the role of learning is still minimal in these cases. Dowehavestudiesof learnedobjectidentification in anyanimals but primates? Ofcourse, thestudiesofimprinting inprecocial animals cometo mind.Certainproperties of an objectlikehighbrightness contrastand

motion release inthenewborn animal a following response, andfollowing

the movingobjectserves,presumably, to impress its featureson the mind ofthesubject sothathewilllaterdiscriminate it fromotherobjects andapproach it ratherthanothers.Here,in a mannerinsuring thatthe precocial animal willtaketo hisparentorat leastto hisspecies, isa very

immediate kindoflearning thatseems tocontain therudiments ofperceptuallearning. Itisnota matter ofassociation ofstimulus andresponse; the

response is readyto go at onceand,besides, recognition canbe measured byotherresponses thanfollowing. Thereisnoexternal reinforcement; the mothercanbuttaninfantgoatawayandhewillfollowherjustthesame.

Whatislearned istypical ofperceptual learning: anincreased specificity of

response to visualandauditory stimulation characteristic ofthereleasing object.Towhatextentthereis increased differentiation wereallydont

know, forearlyimprinting isquickly followed byopportunities forlearning to discriminate feature-contrasts thatinsuremoreprecise differentiation. Wecanstudythislatterprocess mosteasilyintheyounghuman animal,

soI shalltracesomeofthestepsinhislearning to differentiate complex objects inhisenvironment. Doeshebegin, liketheprecocial animal, with

innateattentionto high-contrastvisualstimulationand to motion?Some

peoplethinksoandliketo compare theturning oftheeyesorheadtoward

a voiceor a shinymovingthingto imprinting. One of the firstand most prominentobjectsin an infantsworldis the faceof hiscaretaker.Studiesof

development ofrecognition of a humanfacetellusmuch(Gibson 1969, Ch.16). Atfirst,it appears to bemotionof thehead(likea nod)thatis compelling, butveryshortly theeyesemerge asa prominent featurethe

dominant feature fora discriminative response. Theyarebright, andthey

move.Aftera time,the facialcontour,the contoursof browand nose

aroundtheeyes,andlaterthemouth(especially inmotion) differentiate as

192 E. J. Gibson critical features. But not until nearly four months must these features be present in an invariant and " face-like " relation for the recognition response to occur . At four months , a " realistic " face is smiled at more . Not until six months are individual faces differentiated , and not until much later still

are facial expressions. We know little as yet about how this learning goes

on, but it is perceptuallearning; there is increasingdifferentiation of more and more specificstimulus information. Motor responsesplay little or no role , nor does reinforcement

.

"Learning-to-leam" about objects was demonstratedin a long-term ex-

oerimentwith infants6 to 12 monthsold by Ling (1941). Shepresented infants with a pair of solid wooden objects, differently shaped. Both were colored bright yellow and were of a graspablesize. They were presented to the infant on a board within reachingdistance. One was fastenedtightly

J.

to the board . The other was removable

and was furthermore

sweet to taste ,

having been dipped in a saccharinesolution. The infant learned over a period of days to reachat once for the shapethat was sweetened.Then five seriesof problems were presentedto him. The first serieshad four problems: circle vs. cross, circle vs. triangle, circle vs. square, and circle vs. oval. After successivemastery of these, the child progressedto a seriesin which one of the forms was rotated ; then a series in which sizes were transformed ; a fourth , in which

the number

of " wrong " blocks

was increased ; and a

fifth , in which the positive and negative shapes were reversed. There was

evidence of more and more rapid learning as the series continued, as well as transfer of discriminationswith rotation and changeof size. What were the babieslearning that transferred? Distinctive featuresof the shapes they were comparing, to be sure, but something more general too, Ling thought. They learned search strategies of systematic observation and comparison , IIattention to form differences, rather than improvement in

form discriminationperse." Comparedwith control babiesof the sameage, they made a more immediate and minute examination of the stimulus patterns and inhibited extraneous bodily movement . Now I want to finish with the top -level achievement of fine-grain iden-

tification- the identification of written symbols. Only man does this and only a well -grown , well -tutored man at that . Monkeys can indeed learn to discriminate a pair of fairly small line drawings from one another , but they are much

children

slower

at this

task

and

make

many

of four or five (Hicks and Hunton

more

errors

1964 ; Hunton

than

human

and Hicks 1965 ).

Both phylogenetically and ontogenetically, this is the peak of perceptual achievement . Here we find that education

is most essential . How

is one

letter discriminated from another ? I think by learning the distinctive features

for

the

set of letters . There

is evidence

to show

that

this

is the way

letters really are discriminated , that there is a set of distinctive features, not

idiosyncratic to an individual perceiveror to a given graphic characterbut

Development ofPerception asanAdaptive Process 193

characterizing thesetandpermitting eachlettertobedistinguished byits

unique pattern of features within the set.

Thesetofletters,inthiscase,mustbedifferentiated asa setfromother outline drawings orsimilar things.Linda Lavine, at Cornell, foundthatthis

isdonequite early. Children from three tofivewere shown asystematically

chosensample of graphic items:handwritten Roman letters,numerals,

lower-case letters, cursive handwriting, artificial letters, scribbles, andsimple linedrawings ofobjectsaflower, a stick man,a childs attempt atdraw-

inga houseora face,andsoon.Thechildren wereaskedto tell mewhich

ofthesearewriting. Mostchildren ofthreeandfourcould separate the

drawings ofobjects fromtherest,butnotthescribbling. Mostchildren of fivecould differentiate thenumbers andletters, whatever typeorcase, from bothscribbling andpictures although theycould seldom identify individual

numbers orletters. Habits ofobserving differences between small objects,

likethose shown byLings babies, probably carry overfrom object percep-

tiontotheperception oflinedrawings, andthesetofgraphic symbols is somehow differentiated from other marks on paper before individual items can be identified. Todiscover whatfeatures areactually notedin comparing lettersand

identifying them, weperformed a number ofdiscrimination experiments

usingbothchildren andadults(Gibson, Schapiro, andYonas1968). These

experiments, which simply required thesubject todecide whether a pairof

letterswasthe sameor different, allowed us to construct a confusion matrixfora setofletters.Thetimetakento judgesameor different and

theerrors wereentered inmatrices to showwhich pairsweremostoften confused andwhich least.Thenthematrices wereanalyzed to findwhat

proximities or clustersunderliethe structureof the matrix.Thistellsus

whatfeatures arebeingusedbytheobserver whenhemustdecide whether

a given pair is the same or different.

Themethod of analysis wasa hierarchical cluster analysis (Johnson 1967). It looksprogressively (insteps)forthemostcompact andisolable

clusters, thenforthenextmostcompact, andsoontillonewinds upwith loose clusters andfinally thewhole set.Theresults canbeturned upside

downanddiagrammed ina treestructure. Figure 11.5shows ontheleftthe treeresulting froman analysis of 48adultsubjects latency datafor9

letters. Thefirstsplitseparates thesharpletters withdiagonality from all

theothers. Ontheleftbranch, theroundletters, CandG,nextsplitoff fromtheothers. Atthenextbranch, thesquare right-angular letters Fand

Fsplitofffromlettersdifferentiated fromthembycurvature. Theerrordata fortheseletterswiththesamesubjects revealanidentical structure. On the rightof Figure11.5is thehierarchical structure for60 sevenyear-old children, withthesameletters.It is similar to theadultsbutnot quitethesame. Thefirstsplitisa simple curve-straight one.Onthesecond

E. j. Gibson

194

Figure

11

Tree the

.5

structure left

branch

,

are

yielded

was

by

obtained

the

now

round off

to

quential Gestalt

- like

of

and

adaptive

kind

with

the

,

tied of

this

with

the

,

while

stage

and

R .

This

is

doing

The

square

very

of

highest

on

children

.

letters

neat

,

and

it se

progressed

to

structure

, but

the

processing

structure

- old

straightforward

orders

speculative

, achieving

- year

.

have

higher is

. The

seven

P

are

adults

up

of

judgments

right

from

. This

economy

- different

the

diagonality

picking

relations

development

greatest

off

at

features

processing

redundancy

on

with

children

processing

same

; that

split

those

that

making

subjects

are

from

me

in

adult

letters

split

suggests

confusions

with

it

would

level

of

a given

be

a

-

more by highly

differentiation

.

Conclusion

Is

the

the

development

means

of

such by

of

development

of

detecting

the

dangers simple

sets

as

of

.

objects

process

,

based

too

But

on fine

is

where

an .

in

and

for

.

It

of

only

as

much

so

around

first and of

of

high

objects

vividness

multidimensional

is

complex

and

through

as the

avoiding

Discrimination

scheme

achieved

is

insured

characteristics

evolutionary

is

have

getting

differentiation the

?

to

missiles

physical

- grain

high

,

process seems

needed

pitfalls

single

adaptation

adaptive

Nature

information

obstacles

signs

primitive

perception locomotion

in

development

education

,

a

.

References

Bower

, T . G . R . 1966

Freeman

,

R . B .

1968

monocular

cats

Gibson

, E . J ., and

Gibson

,

E . J .,

obtained Contract

F . with No

. The .

visual

. Amer

a . OE6

,

of

infants

. Scient

determinants

. J . Psycho

R . D . Walk Schapiro

world

Perspective

. 1960 and

latency - 10 - 156

A

! . 81

: 67

. The

, Cornell

-

. .

Pp

cliff

1968 .

. Amer

visual

. 215

size

: 80

-

92

- constancy

.

in

binocular

and

73 .

" visual

. Yonas

measure

of

76

University

."

Scient

.

Confusion

-

96

in and

. Amer

. 202

matrices Final the

report

: 64 for

,

U . S . Office

-

71 .

graphic

Project of

patterns No

Education

.

5 - 1213 .

Gibson,J. J. 1962. Observations on activetouch. Psychol . Rev. 69: 477- 91. Gibson,J. J. 1966. Thesenses considered asperceptual systems . Boston: Houghton-Mifflin.

,

Development ofPerception asanAdaptive Process 195 Gunter, R.1951. Visual sizeconstancy inthecat.Brit. J.Psycho!. 42:28893. Hayes, W. N., and E. I. Saiff. 1967. Visual alarm reactions in turtles. Anim.Behav. 15: 10206.

Heller, D.P.1968. Absence ofsizeconstancy invisually deprived rats.J.Comp. Physiol. Psycho!.65: 33639. Herter, K.1953.DieFischdressuere unc!ihresinnes physiologischen Grundlagen. Berlin: Akademie-Verlag. Hicks, L.H.,andV.D.Hunton. 1964. Therelative dominance offormandorientation

indiscrimination learning bymonkeys andchildren. Psychon. Sci.1:41112. Hull, C.L.1943. Principles ofbehavior. NewYork: Appleton.

Hunton, V.D.,andL.H.Hicks. 1965. Discrimination offigural orientation bymonkeys and children.Percept.Mot. Skills21: 5559. Johnson, S.C.1967. Hierarchical clustering schemes. Psychometrika 32:24154. Klver,H.1933. Behavior mechanisms inmonkeys. Chicago: Univ. Chicago Press.

Khler, W.1915. Untersuchungen amSchimpansen undamHaushuhn. Abh. preuss. Akad. Wiss.(phys.-math.) No. 3, 170. Lashley, K.S.1938. Experimental analysis ofinstinctive behavior. Psycho!. Rev. 45:44571. Lashley, K.S.1949. Persistent problems intheevolution ofmind. Quart. Rev.Biol. 24: 2842.

Ling, B.C.1941. Form discrimination asa learning cueininfants. Comp. Psycho!. Monogr. 17, Whole No.86. McVay, S.1967. Howhears thedolphin? Princeton Alumni Weekly, October, pp.69.

Meyer, P.M.1963. Analysis ofvisual behavior incatswithextensive neocortical ablations. J. Comp.Physiol.Psycho!.56: 397401.

Pastore, N.1958. Formperception andsizeconstancy in theduckling. J. Psycho!. 45: 25961. Piaget,J.1961.LesmØcanismes perceptifs. Paris:Presses Universitaires deFrance.

Romanes, G.J.1893. Mental evolution inman. NewYork: D.Appleton. Romanes, G.J.1895. Mental evolution inanimals. NewYork: D.Appleton. Rosenblum, L.A.,andH.A.Cross. 1963. Performance ofneonatal monkeys onthevisual cliffsituation.Amer.J. Psycho!. 76:31820.

Routtenberg, A.,andS.E.Glickman. 1964. Visual cliff behavior inundomesticated rodents, landandaquatic turtles, andcats(panthera). J.comp. physio!. Psycho!. 58:14346.

Schiff, W.1965. Theperception ofimpending collision: A studyofvisually directed avoidantbehavior.Psycho!. Monogr.79,WholeNo.604. Schiff, W.,J.A.Caviness, andJ.J.Gibson. 1962. Persistent fearresponses inrhesus monkeys to theopticalstimulus oflooming.Science 136:98283.

Schneider, C. E.1967.Contrasting visuomotor functions of tectum andcortexin the GoldenHamster.Psycho!. Forsch.31: 5262.

Trevarthen, C.B.1968. Twomechanisms ofvision inprimates. Psycho!. Forsch. 31:299337.

Walk, R.D.,andE.J.Gibson. 1961. Acomparative andanalytical study ofvisual depth perception.Psycho!.Monogr.75, No. 15. White, B.L.Inpress, 1969. Child development research: Anedifice without a foundation. Merrill-Palmer Quarterly.

Retrospect andProspect : Comparative Psychology andAnimalCognition

. ~. . Althoughcomparative psychology asenvisaged by the earlyfunctional psychologists received a certainimpetusfromtheethologists , oneof the effectswasthesevering of thetie between studyof animalbehaviorand concern with learning theory.Theethologists , unliketheanimalpsycholo gistsin theearlierdaysof thecentury , werenot onlyuninterested in how a mazewaslearned , they werechieflyinterested in so-calledspecies specific behaviorandhadstrongbiasestowardnativism , inventingconceptssuchas"innatereleasers ." Theold fieldof comparative psychology allbutdisappeared fora while.Theemphasis ontheanimalin its niche , and onecological andstructural constraints , seems to havebeena healthyone, however , andecologically inclinedpsychologists areagaininterested in learning , nowwithaneyetowardwhatananimalhasto learnto survivein itsnaturalenvironment . I thinkmyowncomparative research fitswellwith thisemphasis , especially theresearch with thevisualcliffsinceit hasto do with perceptually guidedactionthat contributes to an animal 's survival . Theconcern withlearning is shiftingto exploratory activityandits contributionto perceptually guidedbehavior in activities suchassafelocomotion , foraging , andinteractions with otheranimals , especially maternal -infant interactions . In contrastto thistrend,therehasemerged in recentyearsaninterest in so-called"animalcognition ." Thisconcernhasits forerunners too, in Hobhouse andRomanes andlater MargaretWashburn 's AnimalMind (1908 ). These ancestors areseldom referred to, however . In a recentreview of animalcognition(Gallistel1989 ), animalcognitionis described as a "computational -representational approach ." Gallistelis firmlyopposed to anyformof empiricism , quotingLeibnitzto theeffectthat"thereisnothing in themindthatwasnot firstin thesenses , excepttheminditself." This "cognitiverevivalin animalbehavior " (Dewsbury 1989 ) stemspartlyfrom thevery fashionable studiesof language learningin primates otherthan

198

E. J. Gibson

man , but also covers processes related to ideas of space , time , and number

(Gallistel 1989), intentionality , and self-recognition . Romanes' chart of the

phylogenetic evolution of mind (Romanes1989), fits in nicely- many of his terms , in fact , have been borrowed .

The old dichotomies do not

go away !

nativism vs. empiricism and mind vs. behavior -

III Perception:Pyschophysicsto Transformations (1954- 1959)

Introduction to Part III

. ~~ - - - AlthoughI hadperformed noresearch onproblems of adulthuman percep tion in my ownright, I wasprettycloselyin touchwith thefieldbecause of my husband 's work. DuringWorldWarII, especially , I did somethinking aboutperception . I wasa campfollowerfor fouryears , devotingmy timeto moving , runningahousehold undertryingconditions in Texasand Southern California , andbeinga mother . In California , wheremy husband spenttwo andahalfyearsat theSanta AnaArmyAir Basedoingresearch for theArmy Air Force , I hadopportunities to discuss problems like the bestwayto trainpeoplein aircraftrecognition (a niceproblemin percep tuallearning ) andhow distance mightbe perceived overlong stretches , andtargetslocatedwhilebothobserver andtargetweremoving . These werenot theclassic laboratory problems , andtherewaslivelyargument in James Gibson 'sunitonthebestwayto conceive of themandhowto work on them.Hisunitincluded someyoungofficers , mostimportantly Robert Gagnewhohadbeena studentof hisbrieflyasanundergraduate at Yale, andsomenoncommissioned officers whowereinducted partwaythrough theirgraduate trainingin psychology . We hadgatherings on weekends fairlyoften(bakedbeans andbeerfor supper ) thatwerenotunlikegraduate studentgatherings in normaltimes . Thosegatherings werea heaven -sent opportunity for meto listento andtakea turnin somepsychological shop talk, aftermy rathertiresomeweekof copingwith wartimedomestic problems andintellectual aridity.Wehada lot of discussions of perceptual learningthatinterested meparticularly andpaidoff a fewyearslater. In part II, I relatedthe disappointing endingof my rearingresearch with kids at the CornellBehaviorFarm . Luckily , a promisingresearch opportunity cameupsoonafterward . Twowartimefriends , ArthurMelton andRobertGagne , hadonceagainleft academia andgoneto workfor the militaryat lacklandAir ForceBasein SanAntonio, Texas . Meltonwas Technical Directorof theHumanResources Research CenterandGagne

202

Part III

was Directorof Researchat the Perceptualand Motor SkillsLaboratory.

Thisgrouphadfundsto support external research in addition to their

internalresearch program. Wewereinvitedto applyforresearch funds,

andI wasgladto doso.Myhusbands research wasbeingsupported by theOfficeofNavalResearch (andcontinued to beformanyyears),butthe

Perceptual andMotorSkillsLaboratory actually wantedresearch onperceptuallearning,andI wantedto do it.

GagnØ suggested thatI beginby puttingtogetherallthe available

research on perceptual learning to date,sinceit hadneverbeencollected. It was clearfroma few olderstudiesand fromthe WorldWar II projects

(seeJ.1.Gibson1947)thatperception waseducable, butit hadseldom (if ever)beenthoughtof as an areafor research.Psychophysical experiments sometimesfoundevidenceof it (it was a nuisanceto them)and industrial

gradersandtastersoccasionally referred to it. ExceptforWilliam James (James 1891)it wasnotmentioned intextbooks. It wasa farcryfromall

thelearning research andtheoryofthethirties andforties, inviting a fresh approach.

I spenta yearormoredelving through obscure sources andproduced a lengthy report, with206references, forthePerceptual andMotorSkills Laboratory. I alsoprepared anarticle forthePsychological Bulletin (E.Gibson 1953)andmanaged to improve theorganization ofthematerial greatlyin the courseof revisions.I havealwaysbeengratefulto the editor(Wayne Dennis, at that time).

The actualresearchto be performedwas contractedfor. The topicwas

to be theeffectof trainingonjudgments of distance, andthelocationwas to be a realoutdoorone,not a tabletop or a corridor,asinmosttraditional studies.SinceI had no laboratoryof my own,workingout-of-doorswas a

pieceofgoodluck.Fortunately, theCornell campus wasverylargeand wasnotyetcrammed withbuildings. Therewasa good-sized athletic field

nearthe stadium,witha clear550 yard stretchof grass.We somehow

persuaded theadministration to letususethisareaexclusively (literally keepitclearofeverything butourownequipment) throughout thesummer vacation. Wewerealsoableto commandeer thebigquandrangle of thearts

campus. Weusedtheseareasforthreesummers, doingopen-airpsychophysics, as we calledit. Twoexcellent graduateresearch assistants, RichardBergmanand JeanPurdy,helped.Bergman had startedout in

engineering, andknewhowto dosurveying, whichwasuseful. Purdy, as an undergraduate at Mt.HolyokeCollege, hadhadrigoroustrainingin

psychophysical methods, alsouseful. Thethreeexperiments thatfollow weremodeledon traditionalpsychophysical researchmethodsbut were

performed ina real-world setting thatmadethemveryunusual at the

time.It wasalsounusualto introduce trainingdeliberately withina psychophysical paradigm as themajorvariable oftheexperiment.

Perception: Psychophysics toTransformation (19541959) 203

Thebiggest bonus ofdoing research fortheAirForce wasthesubjects.

When wewere ready togo,abusarrived every morning from Sampson

AirForce Base, aboutthirtymiles awayonSeneca Lake, fullofnewrecruits (Sampson wasaninduction center) anda sergeant to steerthemaround. Having submitted to several arduous daysofphysical andmental tests,

these young menwere happy tospend a dayontheCornell campus. We

weresadwhenSampson wasclosed downa couple ofyearslater.

Thesethreeexperiments, inmanywaysconforming strictly to traditionalpsychophysics, arein strongcontrast to thetwothatfollow, althoughthelatterwereperformed onlya fewyearsafterward. Thereisa

dramatic change inthewayofviewing theperception ofdepth anddistributed byevents asopposed to staticfeatures ofstimulation. James tance,a newemphasis on the roleof motionandthe information con-

Gibson hadbegunthinking aboutmotion perspective andhowevents enterintoperception ofthelayout during thewaryearsintheArmy Air

Force, asheconsidered thekindofinformation available toa pilotora formulating anewtheory about howweperceive where andhowfaraway things areandtheinformation provided byinvariants dependent onchangingpointsofview.Perspective transformations became partofouressentialvocabulary forunderstanding depth perception. I should saythatJames Gibson wasmytutorasregards perception, butalthough weoften argued gunnerin a moving airplane. Bythe1950she waswellon thewayto

aboutexperimentshow to dothemandwhattheymeantwe never really disagreed, aswedidsometimes aboutlearning. Theexperiment on information in perspective transformations as in-

formation forshape andformutilized psychophysical procedures, tobe sure,butthequestions posedhadto dowithinvariance overtransformation,a novelty intheexperimental literature atthattime. Thepaper on motion parallax inquires intotheroleofchange asinformation forperceived separation indepth. Itled,eventually, toquiterevolutionary ideas

aboutthe importance of occlusion anddisocciusion as information for

onethinginfrontoforbehind another. Thisexperiment alsoprovided an

opportunity fortestinga classic theoryofperceptual learning, welcome

because Ihadbegun searching forways ofputting thetwofields, percepInvolvement inthese experiments gavemea grasp ofpsychophysical methods, some insight intotheimportant questions toaskabout perception,andtraining in howto setupa perception experimentnot to mention gradual understanding ofanewwayofviewing perception. This background paved thewayforlater research onperceptual development in infants and children. tion and learning,together.

TheEffectof Trainingon AbsoluteEstimation of Distance over the Ground

Eleanor J. Gibson, Richard Bergman

There hadbeen several studies oftraining judgments ofdistance during World WarILsometimes simply when aiming ata target through empty airandsome-

times ona firingrange, buttherewasgoodreason tothinkthatthese studies had

little generalizability andthatthesubjects hadsimply made specific associations between local cues,likefamiliar size,anda verbal response. This suspicion wasparticularly convincing because ofanexperiment performed at thevery beginning ofourproject byJoann Smith (later Kinney) andmyself. Itoccurred to

usthattraining thatcould beconducted withphotographic material would be

extremely convenient andtimesaving ina practical situation. Afineseries of photographs existed, made by].J.Gibsons unitattheSanta AnaArmy AirBase, ofvariable stretches ofdistance over ground, witha setofdistant stakes of different heights astargets forcomparison witha standard stake close by.Size

constancy judgments overdifferent stretches ofdistance werethought tobede-

pendent onaccurate estimations ofthedistance from theobserver tothetarget (sometimes expressed as allowing forthedistance). The experiment weconducted presented anumber ofphotographs tothesubject formatching judgments, gavecorrection forerrors toonegroup ofsubjects butnot toagroup without training, andthenlooked fortransfer toanother setofphotographs. Notransfer atallwasfound. Evidently, ifthecorrected group hadlearned anything, it wasnotbetter differentiation ofdistance. There wasevidence of deliberate association ofdetails inaphotograph with particular verbal judgments. Anytraining thatresulted intrueperceptual learning, therefore, would havetobe conducted soastoprevent rotememorizing ofverbal responses toparticular targets andavoid idiosyncratically marked paths fromthesubject toa target. It should besetupsothatneither stimulus layouts norresponses were repeated. Ourfinal plan gave us108 stretches ofdistance, alldifferent numbers ofyards. Wecontrived thisbyusing multiple station points forthesubjects judgments as Journal ofExperimental Psychology, 1954,48,473482.

206

E. J. Gibson &: R. Bergman

well as multiple targets. Thefield was carefullysurveyedand laid out, so that everystretchof ground usedfor a judgmentwas different. The subjectsmade egocentricjudgments in yards (from themselvesto a target) and were corrected during training . Learning unmistakably occurred, as the data in the paper show. The interesting question was (and still is) what exactly did the subjectslearn? As

the experimentwas conducted , they could not have memorizedspecificpaired associates . Constanterrorsin theuseof a yard scalewerecorrected ratherquickly during training. Was that all they learned , a correctionof a conceptual scaleof somekind? It was disappointing to think so.

Thereis informationin a gradientof densityoverthegroundfrom a subject 's feet to a target's base . We thought that judgmentsmight have improvedin precision(as witnessed by reductionin intrasubjectvariability) dueto morefinely differentiatedperceptionof thesegradients . I now wonderif therewas another source of information that subjects learned to detect and use- target height in relation to eyeheight. Targetswere of three heightsand anyone of the threewould cut the horizon in the same ratio to the subject's eye height at any distancefrom him. Relating this ratio to a scale would give a kind of invariant that could be used in increasing precision of any given judgment. This extremely subtle and

certainlyunremarked -on potentialinformationdid not occurto usat the time, but if the invariant were, unaware to the subject, detectedand used, it would be a true

caseof perceptuallearning. It is more than likely that we usethis information unwittingly in our daily business of gettingaroundin theworld. ~ , c. v " " ~f ~O " ~

. . . .., ..... .

.. . ..rr . v

- - -- - - - - - -

-J

~

.....

Does the judgment of distancedependon learning? This questionhasbeen

of theoreticalinterestto psychologists for manyyears,but its practicalimportanceis equally evident.t .2 Even with advancesin instrumental control of ranging and aiming , the necessity of human judgments of distance in many military operations has not been eliminated . Previous studies have shown that training can improve absolute estimations of the distance of a target from the observer (4, 5, 6, 7, 8, 9). The conditions under which

greatestimprovement will occur and the relative specificity of the resulting skill are questions in need of further investigation . The training methods and situations used, to date, may have re-

sulted in rather specificcue-responseassociations . In two studies(4, 8) Ss were given training in making absolute estimations, in yards, of the distance to aerial targets . Since distance cues for an object seen

through the air are nonexistent, except for those in the object itself, judgment was probably based primarily on the apparent size of a familiar target (airplane). The Ss probably associateda certain perceived size of the airplanewith the number of yards called out by E during training trials. Suchtraining might have limited transfer value to unfamiliar targets or new field conditions .

Estimationof Distance

207

Ontheotherhand, when thetarget islocated ontheground,

thesurface stretching between Sandtarget would provide perspectiveandtexture gradients incorrespondence withvarying distance stretches. 3 Training under these conditions might yield more generalizable improvement, since cues independent ofanyparticular target would beavailable. Onestudy withtargets ontheground (5)has beenperformed, butsince thetargets werefamiliar existing landmarks, such as telephone poles, it is conceivable that improvement againwaslinkedto thespecific targets. ThatSsarelikely tohitupon rather specific cues during practice andachieve improvement byassociating them withthose responses

which arereinforced wassuggested byanexperiment byGibson and Smith (1).Itwasproposed tomeasure transfer oftraining ofdistance estimations tosizeconstancy judgments (which would presumably be analternative measure ofdistance judgment). Theexperimental group wasgiven training inestimating distance (inyards) topictured targets (stakes) inphotographs. Sizejudgments werethenobtained forthe stakes, following thematching procedure used inaprevious study (2).

Acontrol group made thesizejudgments without anypre-training. Nosignificant transfer wasfound bymeasures ofconstant oraverage

error, though group variability wasreduced bythetraining. TheSs showed a strong tendency to formhighly specific associations betweenparticular identifying cluesinthephotographs anda verbal

response ofsomanyyards. Thisaccomplishment, which thecorrec-

tionprocedure ofthetraining encouraged, wasapparently irrelevant forthejudgment ofthesizeofanobject atagiven distance.

These considerations suggested thatthecrucial question tobeexplored iswhether itispossible totrain Ssinjudging distances insuch a way thatgeneralizable improvement isobtained. Ifthere arechanges from training totestinthetarget objects, stimulus values, ortypeofjudgment required, cananytransfer ofimprovement beexpected? The experiment to bedescribed wasplanned todetermine whether training canimprove the absolute estimation ofdistance overtheground when adifferent stretch of distance from Stothetarget istobejudged onevery trial. Ifnoopportunity isgiven tomemorize thedistance inyards toanytarget, willtraining by corrective reinforcement stillresultinimprovement ofestimation ofother distance stretches? Inother words, thedistance stretches presented as stimuli forjudgment during thetraining series were notrepeated inthetest

series, sothatit couldbedetermined whether anyskillhadbeenachieved which would transfer tothejudgment ofdifferent distance stretches.

Theintention wastolimit thepotential cues tothose provided bythe ground surface, insofar aspossible, rather than topermit thetarget itself

208

E. J. Gibson & R. Bergman

toprovide thecues forjudgment. Consequently, thetarget could notbe unique foreach judgment, andthekind ofterrain chosen forthejudging ground wasmost important. Weused alevel athletic field, 132X488yd., witha surface ofmowngrass.It wasfreeofallobjects exceptthetargets, whichweremetalrectangles mounted onstakes. Thefieldwasbounded at

thelowerend,whereSswerestationed, by a terraceanda groupof

buildings. Theylooked toward theupperend,bounded bytreesanda gymnasium. The21targets werearrayed permanently onthefield, but none stood in line with another.

AsSlooked towarda targetto makehisjudgment, heviewed a stretch

ofgrassofa givendistance bounded bya targetwithnoparticular contextualorassociative value.Therewasnointerposition, noshadows, and nolinear perspective. Sizeperspective, inthesenseofgradients oftexture density ontheground surface, waspresent, andmight havebeenenhanced bybinocular disparity andmotion parallax. Some oftheothertargets were visible in theperiphery andmightalsohaveprovided gradients of size perspective. Thetargets varied inupanddown location inthefield, since a horizon wasclearly visible. Thequestion is,couldSimprove hisskillin

making absolute estimates byresponding tovariables oftexture density andsizeprovided bythegound andtarget array, plusvertical location of thetarget,withoutmemorizing specific cue-yard associations? Method

Arrangement oftargetsTheprincipal problem inplanning theexperiment wasto

arrange thetargets ontheavailable ground soastoprovide different distance stretches foreverytraining trial,andforthetestseries. Sixobserver station points werelocated inthelowerrighthandcorner ofthefield(indicated onthemap,Fig. 12.1, as1,2,3,4,5,and6).The21targets (AtoUonFig.12.1) werearrayed over

thefield insuchawaythatalarge number ofdifferent distance stretches would be

provided when theywerecombined variously withthesixstation points. Onehundred andeightstretches werechosen from thetotalpossible number sothatno target everstood ina linebetween S andtheonedesignated forjudgmen andsothattherewasafairly evendistribution overthetotalrange. Truedistances tothetargets which Swasasked tojudge averaged approximately thesame atall station points. Station points andtarget lociwere staked outbya surveyor and were accurate to a few inches.

Targets It would havebeendesirable tousea single target, which could be

moved tothevarious targetloci.Thiswasimpossible, duetothelargesizeofthe

field,thenecessity ofaccurate location, andthefactthattwoSsweresometimes

runatonce. Thetargets hadtobepermanent andstationary, andtheyhadtodiffer

sothattheycould beindicated toS.Labels bynumber orletter were notpossible sinceevenverylargeonescouldnotberead400yardsaway.Furthermore, some

11111111111111111111

Figure12.1 Map of field, showingstationpoints(1 through6) andtargetlocations(A throughV).

210

F. 1. Gibson & R. Bergman

colorsweredifficult to identifyat a distance. A 3-plate,3-colorsystemof target

markingwasthereforechosen.One to threemetalplates,18 x 24 in.,were attachedto a woodenstake.Theseplateswerepaintedwhite,yellow,or black.A

platewasalways fastened atthebottom ofthestake, closetotheground, sincea raised targettendstogrounditselfandmay,byappearing higher inthefield, be increased inapparent distance. Allthetargetswerereadily identifiable fromthe station points.

Planofexperiment TwogroupsofSs,anexperimental groupanda controlgroup,

weregivena pretestat Station Point1 (18judgments of distance, in yards,to targets ranging from52to395yd.)anda posttest withthesamedistance stretches. Allthesejudgments wereuncorrected. Theexperimental groupreceivedtraining betweenthe two tests,consisting of 90 differing judgments of the distanceof targetsrangingfrom39to 435yd.(targetsviewedfromStationPoints2,3,4, 5, andO).AfterS hadmadehisestimate,it wascorrectedby E.Thecontrolgroup received notraining, butspenttheinterval between teststakinga paperandpencil test of mechanical aptitude.

Sincethepretestandposttestutilized thesame18depthstretches andtargets,

tworandomordersweremade.HalfoftheSsjudgedin orderA forthepretestand

B for theposttest;andhalfthe reversearrangement. Thecontrolgroupand experimental groupweredividedin the sameway.Theexperimental groupwas

also dividedinto two subgroupswhichdifferedin the order of stationpoints

duringthe trainingseries.Thisalternative sequence of stationpointswasnot

thoughtnecessary as a control, butwasadoptedbecause it permitted twoSs to be runat the sametimewithoutoneS andhis Eeverobstructing the viewof

theotherS.Thegroupof18judgments madeat a stationduringthetraining will be referred to as a trainingtrial.Sincefivestationpointswereemployed during training,there were fivetrainingtrials.

ProcedureThe S was firstaskedto read a Snellenchart,and then given his instructions. Theseincluded practicein identifying picturedtargetsusingthesame

3-plate, 3-color systemasthoseonthefield.Hewastoldthathewasto estimate howfarawayfromhima targetwas,inyards,butnodemonstration ofa yardwas given. Hewastoldthathewould judgethesametargetmorethanonce,butnever fromthe samestationpoint.The Ssin GroupF, who were givencorrection,were

furthertold that therewas no pointin memorizing any one figurebecauseno distance stretch would be exactly repeated.

Records werekeptof allof Ss estimates, in yards,aswellashisage,civilian

occupation, andpotentially relevant experience (suchastrackandfootball). The timeofdayandtheweather werealsorecorded. Twochecklistsforweather were

utilized, a descriptive one(sunny, sunnyhazy,cloudyscattered, cloudybroken, cloudy overcast) anda visibility rating. Subjects werenotrunwhenitrained.

Subjects A groupof 30 SsdrawnfromtheCornell population wasrunfirst. Thesepeople variedinage,sex,andacademic background. Asecond groupof92 Ss wasdrawnfromthe groupof airmenin basictrainingat SampsonAir Force

Base.Theywerehighlymotivated andobeyedinstructions to theletter.

Estimation of Distance

211

Results

Analysis byTarget Distance TheSs'estimates of distance to the18targets onthepretest andposttest weretransformed to four-place logarithms , since previous studies (4, 5, 6) haveshown thedesirability of suchaconversion . Frequency distributions of estimates in yardswereskewed , sincetheyardscaleusedby 5 was "open -ended " at thetop. Thelogarithmic transformation yielded reason ably normal distributions and also permitted use of geometric means of the estimates . Whentheestimates areplottedasafunction of truerange , therelation shipisseen to belinear . Thegeometric mean estimates plottedin Fig. 12.2 aretakenfromthepretest andposttest of GroupE, airmen . Theplotsfor GroupC, airmen , andfor theCornell groupareverysimilar . Themeans forthepretest showaslighttendency to underestimation , whilethemeans fortheposttest aregenerally higher . These plots(andthosefor theother groups aswell) leveloff at thelasttarget . TheSsarenotdiscriminating between thelasttwotargets - perhaps asa resultof "end -anchoring "thatis, S'sestimate oftheendoftheseries mayinfluence atargetortargets closeto it. It is clearfromtheotherestimates , however , thatapparent distance ofanobjectincreases asitsrealdistance increases , inapredictable relationship . 400 f-0 350 UJ 2 - 300 fUJ U ) Z . 250 UJU ) 20 20 -U >-, cr :UJ f- Z U . UJ
8 x x 8 x

x8

x

/

88 8,,(X x

XPRETEST 8PQSTTEST 50 45AIRMEN 0 ~ t"'" 50 100 150200 250 0 TRUE DISTANCE , YDS . Figure 12.2 Function relating estimated distance totrue distance (data from Group E,airmen ).

212 Table

E. J. Gibson & R. Bergman 12.1

Percentage ofErrorofSs Estimates forTargets atVarying Distances (Airmen Target

Distance

ControlGroup Pretest

Posttest 0

52

7

68

14

17

107

15

126

Ii

139

19

154 166 188

11

+3

+9 2

+12

18

1

+ 8

11 19

9

+12

14 + 3

1 +11

11

14

+

4

10

13

+

4

14

+

9

9

18

Experimental Group Pretest Posttest

219

10

8

232

10

5

252

7

274

15

8

+

3 2

3 9

5

+

287

+ 5

304

18

+ 3 12

328

+1

+1

342

10

369

12

9

7

395

13

13

13

9

5

+12

17 6

8

1

0

9 +2

12

2 6

9

*Theerrorofthegeometric mean oftheestimates (inlogscores) fromthetruerangehas beenusedtoobtain thepercentage oferror; theantilogarithm oftheerror(theratioofthe

geometric mean estimate ofrange tothetruerange) x 100yields thepercentage oferror. Thestatistical methodfollowed is essentially thatdescribed in reference 6.

Errorin relation to distanceThepercentage errorat eachtargetfor the controlandexperimental groups,pre-andposttest,isgiveninTable12.1. Theseerrorsare calculatedfromSs geometricmeanestimates.Signswere

kept,sotheseareconstant errors. Theerrors arepredominantly ofunderestimation, withtheexception oftheposttestforGroupE,whichfollowed

thetraining series. Allbutonetargetinthisgroupsposttest showsa shift inerrorawayfromminusandtoward plus.Nosuchtrendappears forthe control group, which hadnointerpolated correction. Anotable feature of

theseresultsis the absenceof anytendencyforthe directionor magnitude

ofthepercentage errorto varywiththedistance. Similar calculations for theCornell sample showed thesametrendsa shiftawayfromunderestimationas a resultof training,andindependence of distanceandmagnitude of percentage error.

Variability inrelation todistance Table12.2givespercentage ofSDsof

estimatesof the varioustargets.The most notablefeatureof theseSDs

Estimation of Distance

213

Table 12.2

Percentage ofSDsofSsEstimates forTargets atVarying Distances (Airmen) Target Control Group Experimetnal Group Distance Pretest Posttest Pretest Posttest 52 68 107 126 139 154 166 188 219 232 252 274 287

69 77 77 85 74 85 75 72 76 85 86 89 88

55 58 60 62 66 60 73 65 76 71 80 72 69

58 72 112 81 97 124 85 64 67 135 139 86 135

29 29 25 27 25 20 21 20 20 15 21 19 20

304

90

76

36

22

328 342 369 395

85 77 87 82

69 80 71 76

99 85 111 112

17 17 22 19

TheSD,inlogarithms, isconverted intoa ratiobyfinding itsantilogrithm; this,times

thegeometric mean, gives a score oneSDabove themean; and,divided intoit,onebelow.

Forinstance, theantilog oftheSDforthecontrol group at52yd.is1.69. Thegeometric Values table have been converted topercentages bysubtracting I andmultiplying by 100.inthe mean is49;thescore oneSDabove is83yd.,whereas thescore oneSDbelow is29yd.

istheirsize; thevariability oftheSs,before training, wasenormous. They saidthathehadnever seena yard-stick. Theeffect ofthetraining isto reduce theSD,asthefigures fortheposttesf oftheexperimental group clearly show. Variability drops slightly forthecontrol group, probably asaresult ofsomewhat greater consistency ofindividual judgments, aswill were,actually, verydifferent; oneairman, an 18-year-old PuertoRican,

beshown later. Thepretest figures andtheposttest forthecontrol group

shownoconsistent tendency forpercentage ofSDtovarywithdistance. Butthereissometendency fortheSDto decrease asdistance increases,

aftertraining. Thistendency isconfirmed bythedataforGroup Eofthe Cornell sample. Since it ismostevident forthelastfewtargets, it may againbea function ofend-anchoring; i.e.,thefarthest pointonthescale becomes a well-defined reference point andcauses Ssjudgments nearthe

endofthefieldto centerclosely around anestimate associated withthe

end of the distance scale.

214

E. J. Gibson& R. Bergman

Analysisby Subject Accuracy Since the principal object of the experiment was to determine whether training brought about improvement, a comparisonof individual error and variability before and after training is of primary interest . Of the

45 Ssin Group E, 41 improved; that is, the meanconstant error is reduced, whatever its direction. Of the 47 Ssin Group C, 26 improved. The percentage of Ss improving in Group E is 95.6; in Group C, 55.3. The figure for Group C is not significantly different from chance. The two percentagesare significantly different from eachother (p < .0001). If the mean net gain is determined by subtracting the amount of decrement from the amount of

improvement and dividing by N, the resultant figure is nearly 16 times as great for Group E as for Group C (mean net gain = .1625, in logs, for Group E and .0128 for Group C). It is interesting to compare the types of change, also, for the two groups . In Group E, 43 of the 45 Ss shift away from the previous constant error ; for instance

, an underestimator

may

shift

to overestimation

, but he is

very unlikely to enhance his error of underestimation . In the control group ,

however, the original constant error was actually enhancedin 17 out of 47 cases.If the frequencyof shift away from the previous CE is compared in the two groups by the chi-squaretest the differenceis significant at the .001 level .

Variable error A question of major theoretical interest was whether or not training caused Ss to become more consistent in their judgments , whatever the tendency to over - or underestimate . Does the variability of the estimates of the individual 5s decrease with corrected practice; or , for that matter , with uncorrected practice? To determine variability for individuals a formula4 which corrected for man and target mean was

employed since distancestretcheswere not repeated. It is clear that variability was reducedin the experimentalgroup; the meanpercentageof 5D was 34.6 before training and 18.3 after it . There was a small reduction for

the control group also; the mean dropped from 29.4% on the pretest to 25.5% on the postiest. The significanceof thesedifferenceswas tested by Wilcoxon's (10) non-parametric test for paired replicates. The difference between pretest and posttest for Group E is significant at the .001 level; that for Group C at the .04 level (using a one-tailed test). The training which Group E received was therefore very effective in reducing individual variability . The practice which Group C obtained on the pretest alone,

without any specialtrainingor correction , seemsalsoto haveproduceda small reduction in variability of judgment . Since constant errors did not shift toward any common norm in this group , it seems likely that this

Estimation of Distance

215

reduction in individual variability, eventhoughsmall, accounted forthe similar smallreduction ingroupvariability.

Analysis ofvariance Inorder tocheck some ofthese trends, ananalysis ofvariance wasperformed onthedatayielded bytheCornell sample, Group E.Table 12.3presents theresults ofthese analyses forpretest and posttest. Theestimates havebeenconverted intologscores, asintheother sample. Ourprincipal interest, again, wasina comparison ofthepretest andposttest, between which training wasgiven. Thegroup variability (vari-

anceduetomen) wasreduced, asintheairman population. Andtheindividualvariability (variance duetomanx target) 5 wasreduced, likewise, from .0157 to .0056.Thisdifference is significant at betterthanthe.01level,

confirming thefinding thatwithin-S consistency isincreased bytraining. Theanalysis of variance permits, also,an examination of theeffect oforder ofpresentation oftarget. Order assuch doesnotproduce significant variance, andonthepretest, order-target interaction isbarely

significant atthe.05level. Butaftertraining, thevariance duetothissource isclearly apparent, perhaps because othersources ofvariance which masked it before havebeenreduced. Itmaybethat5,injudging, always decided

firstwhether thedistance being judged atpresent waslonger orshorter thanthejustprevious one,andperhaps evenwhatfraction ormultiple it was.Thesizeofthedistance stretch orstretches preceding a given judgment, therefore, would haveaneffect onanygiven judgment; infact,a different effect depending on the relative farness or nearness of the particulartargetcurrently to bejudged. Courseof Learning

Having established theeffectiveness oftraining forabsolute judgments ofdistance, thequestion ofthecourse ofthelearning arises. Whatkindof Table 12.3

Analysis ofVariance ofDistance Estimates (inLogs) Made byCornell Ss,Group F Pretest

Source Order

df 1

Mean SquareF .0306

Between Ssinsamegroup

18 1.4778

Targets Order x Target

17 .0216 17 .0260

S x Target

306

Total

359

= .05 <.01.

Posttest

.0157

Mean SquareF

.0207 .1186

1.1455

.0983

1.3758 .0204 1.65O1.0143 .0056

4.0000 2.5536

21O

E. J. Gibson & R. Bergman

40

s

--0---S

MEDIAN

10

0

PRETEST

1

2 TRAINING

3

4

I-

I

5

POSTTEST

TRIALS

Figure 12.3

Course oflearning asmeasured bycrudemedian error(datafromGroupE,airmen).

learning curve isexhibited bythesechanges? Since thecalculation ofgeometricmeanswastoolaborious to be carriedoutforallthetrainingtrials,

medians wereemployed forstudying thetrend.Thesemedians werecalculatedwithoutregardto sign,sinceplusandminuserrorsmightotherwise balanceeachother out and give the appearanceof no error.Figure12.3

shows a plotofthemedian errorexpressed asa percentage oftruedistance forthetargetsof thepretestcombined, foreachof thetraining trials consecutively arranged, andfortheposttest. Atraining trial, inthiscase, meansall18judgments madeat a givenstationpoint.Thefirstandthird quartiles areincluded intheplotalso.Themedian errorbefore training is

high(33%), butonthefirsttraining trialit isreduced bymorethanhalf

(to 14%). No furtherreduction occurs, exceptthatthequartiles tendto approach themedian moreclosely inlatertraining trials. Themostinterest-

ingaspect ofthesecurves isthepercentage oferrorfortheposttest. Since no furthercorrection wasgivenon the posttestand sincenoneof the

judgments duplicated anymadeduringtraining, it mighthavebeensup-

posed thattheerrorwould riseagain. Nosuchtendency occurred, how-

ever.Theeffectsoftrainingpersisted undiminished, inspiteofnonidentity ofthestimuli inthetestserieswithanyofthetraining ones.Ouroriginal

question isthusanswered positively: improvement inabsolute judgment is possible, whendistances arejudged overground, eventhoughidentical

judgments arenotrepeatedandno opportunity formemorization occurs.

Learning trends forover-andunder-estimators Whathappensto persistent constanterrorsof individuals, whentheirjudgmentsare corrected? To

answer thisquestion theairmen, GroupE,weredivided intoover-and underestimators on the basisof theirgeometricmeanerrorson the pretest

Estimation of Distance

50

.

40

2I 7

OVERESTIMATORS

- - 0- - UN DE REST

! MATORS

C( 0 C(

30

C(

~ 20

9?

z

W U

1

q \

C( W

/ 1-

a..

:

<{

bA b

I. / I

Z - 10 ,

() w

<> - 20

~

n

-3

rro .

R

a

R

:

~ "., \\ 1 \ ,

/ I

" ,, / \ I \ I \, : ~

\~.- 'tS ' \tJ $S ~ ~

\ 9 \ I

' I

~

6

18

1

- 40

1

PRETEST

ORDER

1st

TRIAL

TRAINING OF

18

1

POSTTEST

18

JUDGMENTS

Figure12.4 Changesin constanterrorasa functionof training.

(thesewere constanterrors). For eachsubgroupthe medianerror was calculated asa percentage of truedistance , retainingthe signs. Theseerrors areplottedin Fig. 12.4 for eachjudgmentof the pretest,eachjudgmentof the first trainingsequence , andthe posttestjudgments . Thejudgmentsare consecutivelyarrangedin the order in which 5 was askedfor them. The line drawnthroughthe centerrepresents zero constanterror. During the pretestthe overestimators showa consistentpluserrorfor everyjudgment. Theunderestimators , on the otherhand, stayconsistentlyminusbelowthe zeroline. But on the secondjudgmentof the first trainingsequence (after onecorrection ), the curve for the underestimators rises, and that for the overestimators drops. Thereare, of course , fluctuations , but by the ninth trainingjudgmentthefluctuationsbeginto paralleloneanotherfor the two groups. Thecurves , hereafter , evenin the posttest,areremarkablysimilar. Thetrainingdoes, therefore , causea changein constanterror. This change occursvery rapidly. Discussion The principalquestionof the experimenthasbeenansweredin the affirmative: absolutejudgmentof distancecanbe improvedby training, even when no opportunityis given for memorizingthe distanceto particular targets.Underthe conditionsof the experimentthereis transferof training to thejudgmentof "new" stretchesof distance . Thesearenewin the sense

218

E. J. Gibson & R. Bergman

that the distancefrom 5 to the target is not the sameas that given as a stimulus for judgment in any training trial. (The targe~ marking ~he end of the stretch has appearedbefore, during training, but is irrelevan~ as a cue, since it marked different lengths of ground during training.) The ground surfaceitself provided a basis for consistently varying cues (texture perspective, retinal disparity, motion parallax, vertical placement) and these dimensionsof variation were present for every judgment. What 5 learned was to evaluate points along the relevant stimulus dimensions in accordance with a yard scale so that he was able to identify a given distance stretch by naming a quantity of yards. The hypothesis will be advancedthat what is learned here can best be describedas a conceptual scaleof yards in a psychophysicalrelationshipwith stimulation provided primarily by a receding stretch of ground. This scale might be characterizedby having: 1. a unit (a yard, or more likely someconvenient and rounded multiple of yards); 2. boundaries or anchors at both ends (the near anchor, zero, is given by 5' s own position, and the far anchor is acquiredduring the training; perhaps it is defined for 5 as the longest distance ever mentioned by the experimenterduring training); 3. differentiatedareaswithin the scale, which are fractions of the total and multiples of the basicunit (the 5 fractionateswithin his scale, we assume , when making an estimate in yards for a target at some intermediatepoint along the scale). Does this hypothesis of a conceptualand perceptual scaleexplain the measurablechangesbrought about by learning? It certainly explains the correction of constanterrors, for either a far anchor or the yard unit would provide a basisfor shifting from any bias causedby an inaccuratescaleunit. But what of the reduction in variable error found in comparing individual SO' s on the pre- and posttest? It would appear that 5 becamemore discriminating in his assignmentof categorieswithin the scale. This might occur becausehe acquired a consistent unit and end anchor during the training to use in identifying or labeling the distance stretchespresented for judgment. His conceptof yard might have fluctuatedbefore the training sessionspresentedhim with an external anchor. However, it is also conceivable that he learnednot only to identify consistently (by naming yard quantities), but also improved in the orderingof points along the stimulus dimension, whatever the categoriesused. This implies an improvement in discrimination- a better differentiatedperceptualscale. Whether the ordering of stimuli, as well as categorizing, was affected by the training is a problem for further investigation.

Estimation of Distance

219

Summary

Previous experiments haveshown thatabsolute estimations ofdistance, inyards, canbeimproved, butthereis reason to believe thatthesetraining situations resulted inimprovement which wasspecific tothetargets andotherconditions of

thetraining. Thepresent experiment wasdesigned toseewhether training would result inimprovement whenthetargets themselves provided nocuesandwhen memorization ofspecific cuesandyardnumbers wasnotpossible. Targets wereviewed overa mowed grasssurface fromsixstation points.

Twenty-one targetsweresoplaced that108judgments, allofdifferent distances,

could bemade byrotating station pointandtarget combinations. Theexperimental group judged thedistance, inyards, to18targets (pretest), thenmade90 corrected judgments (training series), andfinally repeated thepretest. Thecontrol

groupperformed thetwotestseries without intervening practice.

Resultssuggestedthe followingconclusions: I. Improvementin absolutejudgmentof distanceoccurredas a resultof

training, eventhoughnoneof thedistances presented forjudgment were

repeated.

2. Trainingtendedto correctconstanterrorsof both over- and underestimationin the experimental group.Learningcurvesshowedthat thisshift occurredvery quickly.

3. Variable error wasreduced from pre-toposttest inboththeexperimental and the control group, but the reduction was greater fortheexperimental group. 4. Thefunction relatingtruedistance to estimated distance wasshownto be

linear.

5. Thedevelopment ofa conceptual scale ofdistance ina psychophysical relationship with stimulation provided by a receding stretch ofground was hypothesized. Notes

1.Thisresearch wassupported inpartbytheUnited StatesAirForce underContract No.

33(038)-22373 monitored bythePerceptual andMotorSkills Research Laboratory, Human resources Research Center, Lackland AirForce Base. Permission isgranted for reproduction, translation, publication, use,anddisposal inwhole andinpartbyorforthe United States Government. Amoredetailed report willbeavailable inaResearch Bulletin

of theAFPersonnel andTrainingResearch Centerseries.

2.Mrs.NoelNaisbitt, Mr.AlfredSteinschneider, andMrs.NaomiWolinassisted with thestatistical analysis atvarious stagesoftheproject. MissJeanGraham drewthecharts.

Thanks aredueDr.T.A.Ryan foradvice onthechoice ofstatistical procedures. 3.SeeGibson (3)foradiscussion oftheroleofsurface inspace perception. = 4.ForeachS.variability wasdetermined bycomputing (X XM X+XT) 2 inwhich X= Ssestimate ona giventarget, XM= themeanofallSsestimates, XT= themean

forthegiventarget,andX thegrandmean.

5.This istheerror term. Itisassumed thatthisterm reflects degree ofindividual variability since variance duetoindividual differences andto targets hasbeenremoved. Experi-

220

E. J. Gibson& R. Bergman

mental error should remain constant from pretest to posttest, so that a differencebetween the error terms in the two tests wou1dindicate a changein ~he fac~ors producing within -S varIance

.

References 1. Gibson, E. J., Smith, J. A . The effect of training in distanceestimation on the judgment of size - at -a-distance . USAF , Hum . Resour . Res. Cent ., Res. Bull ., 1952 , No . 52 - 39 .

2. Gibson, J. J. (Ed.) Motion picture testing and research. Washington: U.S. Government Printing Office , 1947 . (AAF , Aviat . PsychoI. Program Res. Rep. No . 7.)

3. Gibson, J. J. Perception of the visual u'orld. Boston: Houghton Mifflin , 1950. 4. Horowitz , M . W ., Kappauf, W . E. Aerial target rangeestimation . (OSRD, 1945; Publ. Bd., No . 15812.) Washington: U.S. Dep. Commerce, 1946. 5. Princeton Branch, Fire Control Division , Frankford Arsenal. Analysis of range estimation data . Branch n - Princeton direct

Memorandum

Branch , Fire and differential

No . 20 , 1943 .

Control . Branch

Division

, Frankford

Memorandum

Arsenal . Visual

range

estimation

,

No . 34 , 1943 .

7. Rogers, M . H., Sprol, S. J., Vitereles, M . S., Voss, H. A ., & Wickens, D . D . Evaluationof methodsof training in estimatinga fixed openingrange. (OSRD, 1945; Publ. Bd., No . 4021.) Washington; U.S. Dep. Commerce, 1946. 8. Viteles, M . S., Gorsuch, J. H., Bayroff, A . C., Rogers, M . D ., & Wickens, D . D . Learning rangeestimationon thefiring line. (O5RD, 1945; Publ. Bd., No . 4023.) Washington: U.S. Dep . Commerce , 1946 .

9. Viteles, M . S., Gorsuch, J. H., Bayroff, A . C., Rogers, M .D . & Wickens, D . D . History and .final reportof projectN-I05. (OSRD, 1945; Publ. Bd., No . 4018.) Washington: U.S. Dep. Commerce

, 1946

.

10. Wilcoxon , F. Somerapid approximatestatistical procedures . Stamford, Conn.: American Cyanamid Co., 1949.

13

The Effect of Prior Training with a Scaleof

Distanceon Absolute and RelativeJudgmentsof Distance

over Ground

EleanorJ. Gibson, RichardBergman,JeanPurdy - --

-

-

The previous experimentleft us still in doubt as to whether precision in differentiating stretchesof distancecould really be improved. The subjectshad achieveda better correspondence betweena yard scaleof distanceand stimulation provided by

ground stretchesin one location , without duplicatingparticular stretches . But might the subjectsonly begaining adroitnessin using the scalewith a particular

kind of target and place? It seemeddesirableto test for generalizabilityby conductingtraining in a differentsettingand with differenttargetsthan the test situationemployed . It alsoseemed desirableto makea moredirecttestof theeffect of training on relativejudgments , not just on specificidentifications . The result was two rathergrandioseexperiments that not only kepta teamof experimenters busy but also providedentertainment at timesfor summerschoolspectators as they watcheda bicyclecruisethe quadrangleand stop at the blast of a police

whistle .

The quadrangle was used in theseexperimentsfor training, and the testswere

conducted againon theathleticfieldsix blocksaway. Subjectswereliterallytaught a scale, as decribed in the paper that follows. The method required subjects to

divide up the total space , which was markedas training proceeded with large postersgiving distancesin yards. Absolute judgments in yards were again col-

lectedduring testingin the athleticfield. Thegroup given training did indeed improve relative to a control group, so there was some generalization of the

training. Informationprovidedby gradientsalongthegroundmusthaveplayeda role, sincejudgments of distant landmarks not located at the end of uncluttered ground stretcheswere no better than thoseof a control group. The secondexperimentrequireda radically different kind of judgment in the test

series , following thesamekind of trainingas thefirst. Thesubjectlookedat a fixed targetat oneof threedistances alonga radiusat an angleto oneside. He then watched a variable comparisontarget that moved along a different radius on the

otherside. Theangleof separationwas sufficientlygreat (120) that the subjects Journalof ExperimentalPsychology , 1955, 50, 97- 105.

222 had

E. J. Gibson , R. Bergman , & J. Purdy to

look

" equal , " greater

back

and

' farther precision

group

ground

of

being

the

merely

, and

ideal

of

a " mental

present

the face

of

when

were

duplicated

by

by be

a ground In

the

ment

of

the

to

distance

The on

reduced

with

that practice

on

perceptions

of

described

, to

( cf . 3 , p . 481 ) on

one

effect

transfer

of

was

precision

a different

this

new

concerned . If the ) than

relative

with

distance

control

,

test

with

a

achieved stimula

-

conceptual provided

ground

surface

experimental judging

.

group ground

. The

first

and experi

-

absolute

estima

-

distances

. Would

the

with targets

by yards

or

and

. The

at varying

a control

group

, to

?

the

effect

experimental

the

S had

yards

on

comparison

with

ground

field

training

targets

, by field

distance

of

on

sur -

training

of

particular

with

in

stimulation

the

narrow

training

that

target of

5 of

unfamiliar

a new

the

any

, each

distances the

of

those

too

distance

of

is not

a ground

for

scale to

from

.

that

stretch

gradients

quite

far

concerned

over

hypothesized

receded with

distance

experiment

judgments DL ' s ( greater

cluded

it

with

judge

training

second

as

be

, in yards

judgments

relative

a conceptual

not

determine

preliminary

absolute

of

to

a particular

between

to

" a scale required

sought

tion

but

experiments

was

of

, it was

ground

research

presented

prevented

, are

learning

is much

of

small

scales , a physical

to discover

viewed

distance

correspondence

surface

" taught

then

association

was

the

in

of

vision

two

estimation

control

textured

back , was

( 3 ) demonstrated

absolute

. The

of

. That

of

distance

show

gradients good

gradually

a program of

stretches

correspondence

would

was

the

value

provided

scale

of

of

were

not

rather

. Perceptual

assumed

experiment

improved

. Since

yard

part

with

, but

did

involved

, looking

between

as I was

yards

group

in

learning

tacitly

judgments

. A previous

none

particular

tion

on

method

even

work

control

old

situation

correspondence

of. learning -

are

training

a correction

a better

kind

years

this

in

training

, which

information

perceptual the

made

given

the

judgments is

. But

studying

not

was

) than

these

. There

show

increasing

this

out - of - doors

were that

men , eighteen

DLs

experiments

effect

at

" one , as this

a way .." to conceive

The

young

one for

a matter

one and

good change

to it , as the

judgments

. " The group

differentiallimens

quite

distance

surface

sensitive

. The

" nearer

( lower

, in fact , was

magnitudes

forth

, " or

of

group group

become

, it

better

this were could

training to

show

be

con -

differentiated

.

ExperimentI Method Training procedure The training was carried out in a large grassy quadrangle approximately 350 yd. long. Trees and buildings were visible at the sides, and

Judgmentsof Distanceover Ground

223

several walks cutacross theterrain, butShadanuninterrupted viewofitsentire length. A300-yd. flatstretch waschosen forthescale, anda largemarker labeled 300wassetup.Theboundaries orendsofthescalewerethusprovided bySsown

stationpointandthe300-yd. marker. Theonlyraisedobjectalongthisstretch was

afirehydrant which stoodexactly 10yd.infrontofSsstation pointandwasused

to markoffa unitforhisscale.Theunitandthe endwerethusdefinedforS at

beginningof training.

Thelearning oftheintervals within thisscalewasaccomplished byhaving S divide the300-yd. stretch intoprogressively smaller fractions. Amoving marker

(abicycle equipped withwhitemetalflapsonthefrontaxle)wasstartedat oneend

ofthe300-yd. stretch andkeptmoving untilSthought it hadreached theap-

propriate division point.TheE,whostoodbeside 5,thensignaled thebicycle rider tostopbyblowing a policemans whistle. Theriderdismounted atthispointand

walked tothecorrect division point, where hesetupa marker, painted white and labeled withthenumber ofyardsdistant fromS.Thisprocedure enabled Stosee

the direction andapproximate extentof his error.Theriderthenremounted

thebicycle andbegan theprocedure ofthenextfractionation judgment.

Thetotalstretchwasfirstbisected, andthenfurthersubdivided intohalvesor

thirdsuntilmarkers hadbeenplacedat 25-yd.intervals alongthescale.(An

exceptionwasmadeat 275yd. sincethe intervalwasverydifficultto detectat this

distance.) Thuspractice consisted of10corrected fractionation judgments, ending witha scale marked at25,50,75,100,125,150,175,200,225,250,and300yd.

Forany givenS the bicyclealwaysapproached, or alwayswithdrew,for all the judgments,but the directionof movementwasvariedfor differentSs.Theentire procedurerequiredabout 20 mm.

Testing procedure TheSswhotookpartinthetraining justdescribed constituted

theexperimental group. Attheendoftraining, eachSwasconveyed byauto-

mobileto a largeathleticfield(488x 132yd.).Herehe wasaskedto make

absolute judgments inyards ofthedistance oftargets spaced from52to395yd. awayfromhisstation point.Theprocedure forobtaining thesejudgments, the

targets,andtheirlayouton thefieldwasexactlyas described by Gibsonand

Bergman (3).Eighteenjudgments, eachof a differentdistance,weremade.No

anchor pointwasdefined forSonthisfield. Hewasnottoldhowlongitwas,or

thedistance toanypointonit.Thefield resembled thetraining areainhaving a

flatmowed grasssurface, buttherewasnoresemblance inotherscenery orinthe targets.Thetargetswereconstructed of metalplatesmountedon stakes.The

plates werepainted white, yellow, orblack sothateachtarget wasunique and could beidentified. Notarget intercepted another asSviewed them. Thejudgments asked forwereidentical withthepre-andposttest oftheearlier experiment (3),andweremadefromthesamestationpoint. Inadditionto these18judgments, S wasaskedto estimatethedistanceto four

landmarks visible fromhisstation point. Thelandmarks werea tallchimney of theUniversity heating plant, a towerofanother University building, a lowathletic building, andahouse onalowhilladjacent tothecampus. Trees andotherobjects

intervened beyondthelimitsofthefield,so therewasnota continuous stretchof

ground between S and these structures.

224

E. J. Gibson, R. Bergman, &: J. Purdy

Procedure

for

group

,

group

,

except

all

the

that

control

group

there

airmen

in

was

basic

was

no

identical

with

preliminary

training

that

training

from

.

Sampson

Air

for

There

the

experimental

were

Force

Base

35

Ss

in

each

.

Results

Analysis

by

converted

subjects

three

experimental

that

estimations

of

of

metric

test

the

sign

for

statistical

place

the

,

the

two

training

with

of

at

distance

9

.

.

distance

scale

unfamiliar

were

by

of

This

effects

an

nonpara

-

that

that

expectation

in

unfamiliar

,

indicating

the

improvement

on

a

ranked

exceeded

,

verifies

the

obtaining

' s

error

confidence

an

targets

s

for

then

Wilcoxon

'

finding

were

scores

by

errors

group

level

.

estimates

compared

These

control

. 0002

over

to

The

the

error

compared

)

the

carried

the

the

groups

(

group

S

all

individual

were

each

replicates

effectively

,

The

group

for

and

experimental

training

.

control

estimation

paired

treatment

logarithms

and

error

regardless

the

-

group

constant

of

Before

into

absolute

field

.

Analysisby targetdistance Table 13.1 gives the geometric mean estimate in yards and the percentageSO for each of the 18 target distances. The mean estimatesof the experimental group are more accuratethan those of the control group for 15 of the 18 target distances. By a sign test, this difference is significant at better than the .01 level, indicating again that training influencedfavorably the yard estimatesmadeon the new field. The SO also is decreasedmarkedly as a result of training. The percentage SUs are very large for the control group, and show no tendency to vary with distance. Not only are they reduced for the experimental group, they also show a tendency to grow smalleras distanceincreases . The function relating true distance to estimated distance is plotted in Fig. 13.1. The relationship is linear, as previous studieshave shown (6, 3); perceiveddistanceincreasesas real distancedoes. But it will be noted that the control group tended to underestimate ; 14 of the 18 targets yielded negative CE's. The experimentalgroup, on the other hand, underestimated for 9 of the 18 targets and overestimatedfor the other 9. The five most distant targets are underestimated.They are all beyond 300 yd., the point at which the practicescaleended. An analysis of variance (Table 13.2) was performed on the errors of estimation for each group. The component variancesare all lower for the experimental group. Between-S variance was especially reduced by the pretraining (p < .001 when the variances are compared by an F test). The residualvariancesare interpreted as an estimateof individual variable error (7). When no judgment is repeatedby the individual, the residual variance ("error" term) is the best measureof intraindividual variability,

Judgmentsof Distanceover Ground

225

Table 13.1

Geometric Mean Estimates inYards andPercentage SD* forTargets atVarying Distances True

GroupE(N= 35) Estimated

Distance

Distance

ofTarget 52 68 107 126 139 154 166 188 219 232 252 274 287 304 328 242 369 395

(Yards)

63.6 71.4 107.7 115.4 148.7 181.4 163.0 188.6 190.7 238.5 255.0 267.2 315.4 280.0 311.3 309.2 360.9 333.8

GroupC(N= 35 ) Estimated

SD

(%) 52 53 42 45 46 42 43 31 38 31 27 29 23 30 36 25 22 24

Distance

(Yards)

47.3 60.5 86.8 102.2 105.6 124.6 131.0 151.2 171.9 234.8 218.5 240.3 348.4 274.2 356.5 288.3 400.0 353.5

SD

(%) 113 87 92 82 77 100 82 101 75 139 90 105 125 113 114 101 102 105

The SD,inlogs,isconverted intoa ratiobyfinding itsantilogrithm; this,times the

geometric mean,givesa score1 SDabovethemean;anddivided intoit, 1 below. Values

inthetable have been converted topercentages bysubtracting I andmultiplying by100. sincecomponentvariancesdue to target and individualdifferences have

beenremoved. Furthermore, anyvariance caused byexperimental error

shouldbeequalforthetwogroups.If,therefore, thecontrolexceeded the experimental groupin the residualvariance, it canbe inferredthat intraindividual variability wasreduced bytraining. AnFtestreveals a difference in the expecteddirectionat a littlebetter than the .05 level of

confidence.

Estimation ofdistance of landmarksForthe distanceestimatesso far con-

sidered, therewasanuninterrupted grasssurface extending fromSsstation pointtothebaseofthetarget. Whether thepreliminary training effects any

improvement whenanuninterrupted flatsurface isnotpresent canbefound

intheestimates ofthelandmarks. TheCESwerenotconsistently smaller fortheexperimental group(theyweregreater forthisgroupintwoofthe

226

E. J. Gibson , R. Bergman , & J. Purdy 450

400

0

(/)

0 >-

350

w u z

300

0

0

~ ~ 250 w

(/ )

~ () ~ 0 a: ~

w ~

w
0 200 150

-

w ~ 100 ~ w

0 CONTROL I

50

.

0

50

100

150

TRUE

200

250

DISTANCE

EXPERIMENTAL

300

, YDS

350

400

450

.

Figure 13.1

Geometricmeanestimateddistanceasa functionof true distancefor the experimental and control

group .

Table 13 .2 Analysis - ofVariance ofErrors ofDistance

Estimation

(in Log- Units )

Mean Square

Group Group Source df E C F: C/E~ p -\J\JU .i """ FrJ Targets 17 .0625 .1091 1.74 > .05 Men 34 .1558 1.4029 9.00 < .001 Residual 578 .0107 .0169 1.579 < .05 Total 629 .l v"uJ . ~- . . . . .. . These comparisons by theF testarepresented asa matterof interest , although the variances forthetwogroups areclearly unequal . Theprincipal conclusion issupported by nonparametric tests .

Judgmentsof Distanceover Ground

227

fourcases). Variance wasgreater forthecontrol groupinthreeofthefour cases, butnotsignificantly. Themeanerror, regardless ofsign,wassmaller fortheexperimental groupinallfourcases, significantly so(p< .02)in

three. Theevidence fortransfer oftraining tothese judgments istherefore

equivocal. Theabsence ofacontinuous ground stretch isprobably responsible forthelackofa clear difference between thetwogroups.

Comparison withtraining bycorrection Intheprevious experiment (3)in

which Ssweretrained bya correction procedure, training andtestjudg-

ments weremadeonthesame field. Inthisexperiment Sshadpreliminary

training bya scaling method, ona different field. Improvement inthetwo experiments canbecompared bycalculating thepercentage transfer ineach case. 2 Forthe previous experiment, therewas79%transfer basedon

constant errors; forthepresent one,62%. It seems, then,thattraining with a scale ofdistance isnearly aseffective ascorrection ofjudgments onthe

very same ground where estimation is tested. ExperimentII Method

TheSswereagaindivided intotwogroups, experimental andcontrol, andthe

experimental group wasgiven preliminary training identical withthatgiven the experimental group inExp. I.Both thecontrol andexperimental groups were then required tomake relative judgments forpairs ofdistances overa ground surface. Asingle distance wasthought ofasanimaginary lineextending radially away fromS.Thequestion was,howaccurately could S compare a givenstretch of

groundalongoneradiuswithanotheralonga different radius.

TheSswere taken tothesame athletic field used inExp. Iforrelative judgments. Theground wasmowed grass, andthefield wasempty except forthetwotargets. These wereattheendsoftwodistance stretches making anangle of120withSs

stationpoint,so thathehadto tumhisheadto compare thedistance to eachof

them. Thewideangle wasusedsothatSwould actually compare thedistance

stretches, notmerely theupanddown location oftwotargets located onthesame radius. Thetargets wereofdifferent shapes. Thestandard targetwasa white

triangle (altitude 94cm.andbase90cm.). Thevariable tagetwasa rectangle from Sbyaman concealed behind it.Itransmoothly onbicycle wheels andalways madecontactwiththe ground. Thepaths ofthetwotargets formed a V.withSatthejunction point andthe (91cm.wideand1S3cm.high). Thevariable target wasmoved toward oraway

standard andvariable located at somepointalongthearms.Therewerethree distances forthestandard target, 50,100,and200yd.(Themaximum distance of

thepathalong each armoftheVtothecomers ofthefield was275yd.)Thefield theselines. Theyweredriven flush withtheground, sothattheywereinvisible to

hadbeen surveyed byanengineer andstakes were laidat1-yd. intervals along

228

E. J. Gibson,R. Bergman,& J. Purdy

Ss.Thinwirewasstretchedalongeachlineso thatthe targetmancouldfollowa straightcourse.Thewirealsowasinvisibleto S.

The methodof judgmentwas a combination of a methodof limitsand a

constantmethod.The variabletarget startingat a distanceeithermuchgreater thanor muchlessthanthat of the standard,madea run whichtookit up to the

equal pointandthena considerable distance beyond it.During therun,it stopped 10timesatassigned positions toenable Stomakeajudgment offarther,equal,

or nearer. The total path traversedon one run by the variablestimuluswas

12yd.forthe50-yd. standard, 24forthe100-yd. standard, and48forthe200-yd. standard. Fora typicalrunwherethe standardtargetwas50 yd.,the variable targetwasstartedat 57yd.andstoppedforjudgments at 55,53,52,51,50,49, 48,47,and45.A typicalrunforthe200-yd. standard beganat 228yd.withstops at 220,212,208,204,200,196,192,188,and 180.Approachandwithdrawal runs werealternated,and the startingpointwas alwaysvaried.For eachstandard distancethe variablemade three approachand three withdrawalruns, so that S

made60 judgments foreachstandarddistance. Thetargetmankeptin constant touchwithEby radio, 3 so thatjudgmentsandstopswereco-ordinated. Forhalfthe Ss,the standardstimuluswason the left,andfor halfon the right. Sincethe illumination of the targetschangedfrommorningto afternoon,these

right-left positions wererotated soastobeequally divided between morning and

afternoon. Halfof theSsbegantheirjudgments withthe 50-yd.standarddistance andhalfwiththe200-yd.standarddistance. All60judgments fora givenstandard werecompleted beforegoingon to the next standard.One practicerun (10

judgments) wasgivento makesureSunderstood thetask.A restperiodofabout 10mm.wasgivenaftercompleting judgments foreachstandard. Thejudgments for all three standardsrequiredabout 2 hr.

FourSstookpartin theexperiment at onetime.Eachmanrecorded hisown judgments. TwoEswerealways present andwatched constantly to insurethatS wasrecording intherightblank.TheSsfacedstraight aheadaftereachjudgment, whilethe variabletargetwasmoved.TheE saidJudge whenthe targetwasin

place. TheSwasinformed whena newrunbegan, andwastoldthatforeachrun the directionof movementof the variabletargetwouldbe the same.He wasalso

toldthatthestartingpointwouldvaryforeachrun,andwasaskedto be sureto

compare thestretch ofground between himself andthetwotargets before making his judgment.

The four Ss at each session includedtwo membersof the experimentalgroup

andtwoofthecontrolgroup.Theystoodbehindoneanother,andpositions from front to rear were rotated for the two groups so that any advantageof station

pointwouldbe equalized. Allfourmencouldseethetargetseasily, however. Therewere32 Ss in eachgroup,allairmenin basictrainingat SampsonAir Force Base.

Results

Theeffect oftraining ondifferential sensitivity todistanceTwomeasures of sensitivity wereobtained,grouppsychometric functionsand individual

229

Judgmentsof Distance over Ground 100

U) ill U)

Z 0

7

a. . U) ill

cr:

LL

5

STANDARD

200

CONTROL

GROUP

0 I-Z

DESCENDING

ill U CI:

..., ~ " E "

25

ill a. .

'

I

,

,

,

x,

...

... "' ...

'

,

.... x ;

~

,

. . . .

" "

"

. .

. . . . . . .

SERIES

...

; ~ ;

.

~ ,

,~ "

,

,

YDS

. . .

" "

'

. . .

"

' X- -

- -

-

0 2

DISTANCE

Figure

13

Group

a

function

approaching

function

DL

.

the

To

of

distance

tions

,

tive

yd

errors

(

.

The

The

variable

DL

' s

runs

for

(

10

by

respective

p

.

SD

the

' s

experimental

at

all

A

t

require

distance

of

according

.

and

plotted

as

and

lower

while

The

all

200

-

yd

the

point

. ,

condi

of

-

subjec

standard

distance

positive

constant

ascending

re

variable

at

of

gave

the

-

,

series

each

the

p

gave

differential

' s

those

case

105

)

is

was

' s

concluded

sensitivity

obtained

the

points

responses

,

PSE

None

.

the

' s

The

the

Wood

and

-

their

limens

for

group

this

the

t

that

to

the

significant

since

of

,

control

difference

,

of

on

by

mean

for

made

.

be

those

and

described

DL

variances

therefore

S

s

groups

no

.

'

method

than

in

,

to

DL

experimental

but

equal

similar

mean

lower

( 2

must

all

very

the

,

,

of

the

group

the

Individual

to

presents

of

not

yd

is

of

.

from

slightly

assumption

,

were

.

observations

did

200

percentage

typical

responses

obtained

distances

the

from

)

method

control

are

It

at

farther

.

were

3

.

and

control

series

individuals

,

Table

confidence

,

function

fairly

outward

standard

)

)

both

paired

the

of

of

.

for

standard

by

level

)

set

equal

a

the

is

descending

this

group

three

test

or

standard

420

plot

displaced

of

( PSE

each

,

standard

,

,

for

This

than

' s

the

nearer

as

function

nearer

obtained

with

plotted

distribution

was

farther

equality

six

the

,

DL

,

of

was

.

variable

the

subjective

shows

case

set

(

averaging

worth

236

.

functions

farther

approach

ones

by

and

expected

this

group

responses

psychometric

,

. 2

the

in

204

, YDS

.

approaching

showing

negative

scale

13

control

of

group

equal

variable

equality

at

. 05

the

,

Figure

the

the

percentage distance

nearer

.

with

of

The

variable

obtain

sponses

not

.

of

' s

VARIABLE

. 2

psychometric

variable

OF

' s

does

reaches

the

training

distance

.

test

with

.

Limens

a

230

E. J. Gibson, R. Bergman , &: J. Purdy

obtained from the group psychometric functions likewise showed no significant changeas a result of the training. The effect of illumination of the standardand the effect of order (beginning with the lOO-yd. standard vs. beginning with the 50-yd. standard) on the 0 L were calculatedseparatelyfor experimentaland control groups. For illumination, no significantdifferencein DL was found for either group. That is, it madeno differencein the limen whether the standardtarget was illuminated by direct sunlight or not. Order had no significant effect on the DL for the experimental group, but the control group had significantly lower DL's for two of the three standards(50 yd. and 100 yd.) when 5 began his judgments with the standardat 50 yd., rather than at 200. Constanterror as a function of training and luminanceof standard The PSE did not differ significantly betweenexperimentaland control groups. However, there was a small constant error for both groups when ascending and descendingtrials were averaged(seeTable 13.3). The CE, small as it is, tends to be positive- that is, the variable looked the samedistanceaway as the standardwhen it was slightly farther. It might be askedwhether bright sunlight shining directly on the standard target would make it appear nearer than when it is not directly illuminated. Manv- Ss commentedon the apparentdifferencein brightness dependingon the sun's location with respectto the targets. The PSE's were therefore compared for the three standard distanceswhen the standard target was in a position to be directly illuminated and when it was not (when it "looked brighter" or "looked shaded"). None of these differed significantly from the standarddistance. In other words, no CE was producedby direct or indirect illumination of the standardtarget.

Table ._ 13 3 -Mean -Distan Indivi DL ' s ( , i Yard ) . n and PSE ' s ( i Ya ) n at Th St D Diffe Lime Po of Su E Cont Exp Co .-200 A ~ _ Mean SD Mea SD Me SO M S 50 yd 1 . 3 . 5 1 . 2 1 . 5 9 50 . 6 4 1 . 4 2 5 . 2 1 . 7 0 100 2 7 1 3 6 0 8 2 10 3 3 0 9 1 4 8 3 5 2 4sign 4 572 1 6the 20 9 7dis 6,2 7as5-positi ~ J .two -CEo These value diffe from sta sh

OL asa functionof distance As Table 13.3 shows, the DL in yards increases with distance. The SO likewise increases , and is notably large. There was considerablevariability in DL's for different individuals. The increment

Judgmentsof Distanceover Ground

231

required toproduce achange injudgment canbeexpressed asapercentage of the standard (AD x

100

Forthethreestandard distances, themeanDLisconstant at about2.5%. Overtherangeofdistances used, Webers Lawappears tohold. Effect oftarget separation onSD Thejudgments ofdistance obtained inthis

study required thatScompare twostretches ofground receding radially from him, thetworadii separated byanangle of120.Bycomparing the present results withthoseofTeichner, Kobrick, andWehrkamp (8),the effect ofincreasing angular separation ofthestandard andvariable target canberoughly determined, since theirtargets wereseparated byonly3 mm. ofarc.They used asestimates oflinear threshold theSDofSssettings around theCE.Comparable measures werecomputed forourdata.Foreach Sforeach run,weobtained thefirstvariable distance judged equaltothe

standard distance. TheSDofthisdistribution is givenbelowforeach

standard distance (infeet,ratherthanyards,to conform withtheother

study):

Distance

SD

l5Oft. 300 ft.

10.83 ft. 21.48 ft.

600 ft.

40.80 ft.

TheseSDsareallhigherthanthoseobtained byTeichner et al.forthe

same distances. Here theyareabout 7%ofthestandard distance, compared toabout -% 2 intheother study. Thus judgment ofcomparative distances overground giveshigher threshold values thandoesjudgment ofthe coincidence oralignment oftwotargets at a distance. Butreasonably

sensitive judgments canclearly bemadeforcomparative distances oftwo

ground stretches. Discussion

Theresults ofExp. I showed clearly thatpreliminary training witha scale

ofdistance improves absolute estimation ofthedistance to anunfamiliar targetinanewlocation. Theground surface itselfprovides a stimulus basis

forSsjudgment inbothtraining andtestfields. TheSobserves along an imaginary linestretching from hisfeettothetarget. Certain optical propertiesofanysuchlineonanylevelterrainwouldremain constant in the stimulus array: farther stretches (relative tonearer) arecharacterized by decreased sizeandincreased density oftexture particles, byincreased

232

E. J. Gibson , R. Bergman , & J. Purdy

uncrossed disparity, and by increaseduncrossedparallactic motion (4). Vertical position of a target in the field of view also moves upward with increasing distance. The other classical"cues" for distance are probably '--'

irrelevant for the present situation . Familiar size and interposition

have

been eliminatedin the experiment, and the kinestheticcuesare probably of little value for the distancesemployed.

Thesevariables , we believe, are in a psychophysical correspondence with impressions of distancealongtheground. The conceptualscaleof yards would have to be related to the concomitant gradients of stimulation by meansof learning. The relationship, in fact, was vastly improved by the training provided in Exp. I. The S can bring with him a scalerelationship which will have a beneficial

effect on estimations

in a new location ; even

the absence of any objectively confirmed reference point on the new field seemed to be no handicap. And if the yard scale was related, in the training

process, to stimulation provided by a continuousstretch of ground, it is not surprising that transfer should be partial or absent when the intervening surfaceis interrupted by objects or hills or gulleys, aswas the casewith the " landmark " targets . The present experiment does not permit us to evaluate the three compo nents included in the scale training . The " unit " of 10 yd ., for instance, may

have played little role. That the fractionation into 25-yd . intervals had an effect is revealed by a comparison of the categories used for estimation by

control and experimentalgroups. Both tended to use responsecategories which are multiples of 25, but the experimentalgroup used many more564 to Group C's 459. The difference is significant (p < .01). The far boundary of the training scale(300 yd.) appearedto exert an effect on the judgments of distancesgreater than this value. Further experimentwith the componenetsgiven singly might throw more light on what 5 learns. ExperimentII was designedto discoverwhether the training receivedby members of the experimental group increasedtheir sensitivity to small differencesin distance magnitudes. Since the DL' s for the experimental group were not significantly lower than the control group's, it must be concluded

that differentiation

within

the distance

dimension

was not in -

creased.It should not be concludedthat the relative judgment of distance cannot be improved, however; the amount of practice may have been too little, for 5s' performancewas already quite good and was perhapsat an asymptote. Also, the judgment madein the training situation was not very similar to that made in the test of relative judgment. The training actually was concernedwith judgment of grosserdistancemagnitudesthan was the test .

The results of Exp. II do, however, strengthen our assumptionthat a

psychophysical scaleexistsin which stimulationderivingfrom a ground surfaceis correlatedwith perceptualjudgments of greateror lesserdistance.

Judgments of Distance over Ground

233

Highly typical psychophysicalfunctions were found (see Fig. 13.2). Since the targets as suchprovided little basisfor a judgment of their distanceand a judgment of alignment was impossible, it follows that for any object in contact with the ground the perception of distance of the object may

be determinedby the impressionof distanceof the background surfaceat the point of contact . For an object not in contact with the ground , the perception of distance

must dependlargely on stimulation deriving from the object itself yielding impressionsof edges, depth, etc. The classicallist of cues, as well as factors of knowledge and inference, would be relevant. An experimentby Bourdon 1; and 10, p. 670) obtained DL's for judgments of relative nearnessand famess of two luminous circles, placed at right angles to one another at distancesaveraging around 20 meters. The surroundings were dark. The DL's for this judgment were about 22% of the standarddistance, compared to 2! % in our study (DL defined as one-half the interval of uncertainty). The absence of stimulation from a ground surface in Bourdon 's experiment

is probably the principal reasonfor the differencein results. Convergence, retinal disparity, and motion parallax were all operative to some extent in Bourdon 's situation , since when he modified it by restricting the conditions

to monocular vision and a motionless head, he obtained completely equivocal judgments. For a pair of objects in contact with a surface, but located so that the stretch of distance to one coincides or nearly coincides with that to the other , the perception of distance as such will be accompanied or even

supplantedby a different type of tridimensionalperception, the impression of in front of or behind. For this judgment, there is a different basis in stimulation . The object behind will be optically above the one in front and

vice versa. We cannot agree with Teichner, Kobrick, and Wehrkamp (8) as to the prevailing importance of this type of stimulation for impressions of distance. When the two target objects are not optically adjacent, the observer must discriminate the respective stretches of distance in order to

decidewhich object is farther or nearer. This is a frequent type of judgment in daily life, and it is probably associatedwith locomotion of all kinds. Summary The purpose of the two experiments described was to determine the effect of

training on (a) absolute judgments of distance, in yards, and (b) relative (nearerfarther) judgments of distanceto variable targets. Training with a scaleof distance was given the experimentalgroup in both experiments. A 300-yd. stretch along a grasssurfacewas bounded for 5, and a la -yd. unit designatedfor his scale. Then 5 made corrected fractionation judgments until the ground was divided into 25-yd. intervals. Following training, Ss of Exp. I were taken to a different field where they

234

&: J. Purdy

madeabsoluteestimationsof the distance, in yards, to unfamiliar targets at varying distances(52 to 395 yd. away). A control group made the absolute estimations without previous training. In Exp. II, Ss were taken to a large field and asked to make relativejudgments of the distancefrom S to two targets separatedby 120 . The standardwas set at 50, 100, or 200 yd. from S. The method of judgment was a combination of a method of limits and a constant method. A control group made the samejudgments without pretraining. Resultsof Exp. I indicated that Sswho receivedthe pretraining were superior to control Ss in both constant and variable error. Absolute estimation was improved even though Ss were not tested in the same field where they were trained, the targets were unfamiliar, and the distancesvaried. It was proposed that S learneda scale relating responses , in yards, to gradients of stimulation deriving from the ground surface. Pretraining with a scaleof distancedid not, however, lower DL's for distancein Exp. II. Sensitivity of the psychophysicalrelationship is apparently not increased by this training, although the yard-responsesbecome more accurately tied to it . The DL for distancewas about 2! % of the standarddistanceand was constant for the three standard distancesemployed. The fact that consistent psychophysical functions of the expected kind were found with wide angular separation of the standard and comparison target strengthens the hypothesis that gradients of stimulation made availableby the ground surfaceitself are in correspondencewith impressionsof distanceof varying magnitudes. Notes 1. This research wassupportedin part by the UnitedStatesAir ForceunderContractNo. 33(038) 22373monitoredby the Perceptualand Motor SkillsLaboratory , HumanResources Research Center,LacklandAir ForceBase . Permission is grantedfor reproduction , translation , publication , use, anddisposalin wholeandin partby or for the UnitedStates Government . Theresearch wasassisted by Dr. LeighMinturn, who actedasexperimenter for the scaletraining, andDorothy Serrie , who rodethe bicycle. 2. Percentage transferwascalulatedby the formula

Group C-Group . Group C-Ex100

A transfer comparison wasmaderatherthandirectcomparison of thetwo experimental groups because of thedesirability ofcomparing eachGroupEwithitsowncontrolgroup 3. Portable radioequipment waskindlyloanedby the CornellDepartment of Military Training . References

1. Bourdon, B. La perceptionvisuellede l'espace . Paris: SchleicherFreres, 1902. 2. Dixon , W . J., & Massey, F. J. Introductionto statisticalanalysis. New York: McGraw-Hill , 3.

4.

1951. Gibson, E. J. & Bergman, R. The effect of training on absolute estimation of distance over the ground. ] . expoPsycho I., 1943, 48, 473- 482. Gisbon, J. J. Theperceptionof the visual world. Boston: Houghton Mifflin , 1950.

Judgmentsof Distance over Ground

235

5.Holway ,A.H .influencing ,Jameson ,the D .A .,Zigler ,M .range J.,Hurvich ,in L.free M .,sWarren ,A .B .,&vision Cook ,. .Cambridge E B . Factors magnitude of e rrors pace and telescop Division of Research ,Graduate School ofBusiness Administratio ,Harvard Univ ., , 1945 . 6.Horowitz ,15812 M .W .),Washington &Kappauf ,W ..E ..Aerial target range estimation .(D.S .R .D .,1945 ;Publ . Bd . , No . . : U S Dep . Commerce , 1946 . 7.Princeton Branch ,Fire Control Division ,Frankford Arsenal .Analysis of range estima data . Branch Memorandum No . 20 ,1943 . 8.Teichner ,on W .H .,discrimination Kobrick ,].L..,Quartermaster &Wehrkamp ,Research R .F.Effects of terrain and observ distance depth and Developme Comma , Environmental Protection Division , Report No . 228 , Natick , Mass . , 1954 . 9.Wilcoxon ,R .Some rapid approximate statistical procedures .Stamford ,Conn .:Americ Cyanimid Co . , 1949 . 10 .Woodworth ,R .S .&perimentaJ psychology .New York :Holt ,1938 .

14

DistanceJudgmentby the Method of Fractionation JeanPurdy, Eleanor] . Gibson

As this project progressed , we becamemore and more interestedin how accurate

peoplereallyare in judgingdistances to betraversedand how theydo it. Theyare seldomaskedto estimateverballydistancestretchesin yards or in any absolute scale . But they maketheir way around(in cars, for example ) with no problem. Cars at any distanceare perceivedas constant in size. Is size constancyperceived

peculiarlyfor objects , perhapsderivedfrom knowledge offamiliar size?] .] . Gibson had suggested(1950) that constancywas not a characteristicof object perception but rather was a characteristic of perception of the total layout within which objectswere situated. It should be discoverablein perceptionof stretchesof ground surface, not only objects located on thesesurfaces. It seemsnow that a scalefor such layout constancyshould also be relative in someway to the perceiver. It may begiven in the perceiver's eyeheightin relation to the horizon, which is a constant peculiar to every individual and provides automatically a neat unit of measure-

ment. We did not think of that in 1955, but the desirabilityof studying the consistency (or nonconsistency ) of people 's scalingof the layout and objectscontainedin it had become evident.Still in a psychophysical tradition, we turnedto the method of fractionation. If there were consistencyof perceivedscaleof layout, it should show up with this method. The subjectsmadefractionation judgments as they had in the previous experiment, but this time we were not so much concernedwith learning as with the consistencyand unity of the dimensions of their perceivedworld.

Theresultswerevery illuminatingfor our question . Thesubjectswere, on the whole, surprisingly accurate. More important, errors were not related to visual angle subtendedon the retina; the subjectsdid not make "retinal matches" but showedgood constancyfor stretchesover ground. As Gibson said (1950, p. 181), "Scale, not size is actually what remains constant in perception." But what gave

thescale ?Not correction , certainly, as it waspresented in this experiment . A large group of subjects that receivedcorrection was not superior in fractionation to a group of naive subjects. We had no answer at that time. What is the information

238

J. Purdy & E. J. Gibson

for the remarkableconsistency of perceiveddimensionsof the spatial layout? If learningis involved, what kind of learning? Certainlynot simplereinforcement of responses .

Two questions beggedfor an answer. What information is available for

perceivingconsistent , accuratedimensions of thegroundand its furnishingsin our environment ? What kind of a theory of perceptuallearningcan we look to if simple correctionor reinforcementis not the answer? -

-

r

~

The literature of space perception abounds in both experiments and theo -

retical discussionsof the perceivedsize and distanceof objects.! The continuum of distanceitself, however, has seldom been specificallystudied as

a psychological dimension . That distancemustbe perceivedaccurately for objects to remain constant in their dimensions is a generally accepted proposition . Errors of constancy , indeed, have been attributed to errors in the estimation of distance (3; 5, p. 151). It is time to inquire , therefore , what

kind of psychological scalefor a dimension of continuing distanceexists. Can an observer tell when far and near stretchesof distanceare equal, or when one is half or twice the other? Gibson predicted that constancy would hold for the distancesbetween objects (2, p. 165) on the hypothesis that constancy is a property not merely of perceived objects but of perceived tridimensional

space itself . Little

evidence is available

on this ques -

tion . Gilinsky (3, p. 473 ff .) had two Os bisect distances from 8 to 200 ft .

long, startingfrom o . Shepredictedthat physicallyequalunitsof distance would

appear shorter as their distance from 0 increased . This result appar -

ently was found in her data, but the small number of 0 ' s used make further

experimentsdesirable, especiallysince the results of recent size constancy experiments indicate that constancy or even " over -constancy " (4) holds for

objects at very long distances. It would appearreasonablein the light of such a result, that " constancy ," or better , objective accuracy, should char-

acterizethe judgments of all the physical dimensionsof a surfacelike the ground. In the presentexperimentthe nature of a psychologicalscalefor distance was investigated by a method of fractionation. The 0 was askedto bisect or trisect distancesalong an imaginary line stretching over the ground from 0 to an indicated

marker . A further group of Os was available whose

judgments were corrected as they were made. These fractionation judgments had servedas a training procedurein another experiment (1). Method

The experiment took place in a quadrangle350 yd. long. The ground was mowed grass, intersectedby three walks. Trees and buildings lined the sides, but the view

DistanceJudgmentby Method of Fractionation 239 Table 14.1

Planof Fractionation Judgments

Serial

Bicycle Approaching

Bicycle Withdrawing

No.of

Stretches

Stretches

Stretches

Stretches

Judgment Bisected (yd.) Trisected (yd.) Bisected (yd.) Trisected (yd.) 1 2 3

300(300to 0) 150(150to 0) 150(300to 150)

300(0to 300) 150(150to 300) 150(0to 150)

4

75(300to225)

5

75(225to 150)

50(25to 75)

75(150to 75)

50(100to 150)

75(75to 0)

50(175to 225)

6

50(200to 150)

7 8

75(75to 150)

50 (125 to 75)

9 10

75(0to 75)

75 (150 to 225)

50 (50 to 0)

75 (225 to 300)

totheendofthequadrangle wasunobstructed. A300-yd. stretch waschosen for

0s scaling operations. The0 stood atoneendofthequadrangle, accompanied byE.Abicycle andridermoved upordown thequadrangle, asneeded, asan indicator ofthedivision point which 0 wasasked tofind. Thebicycle rider hada prearranged schedule ofmovements tofollow andcould besignalled byawhistle

when 0 wanted himtostop. Thearea hadbeen surveyed andhadmarks at25-yd.

intervals to guidetherider,butthesemarkswereinvisible to 0.

Each0 made10fractionation judgments. Sixofthesewerebisections andfour

weretrisections. A !/bt

istheactofstopping thebicycle at halfthedistance

covered bythestretch; atrisection istheactofstopping thebicycle atone-third thedistance covered, thejudgment inthelattercasebeingthefirstthirdorthe nearest thirdofthestretch covered. Table 14.1shows theactual stretches divided, andtheorder inwhich thefractionations were made. Theplanwassuchthatthe truedivision points forthe10judgments made upa series at25-yd. intervals from 25through 250yd.asindicated intheleft-hand column ofTable 2.TheOswere

divided intotwoequal subgroups depending onwhether thebicycle rodeaway from 0 ortoward him. Theorder ofjudging, andwhether 0 halved ortrisected, varied between thetwosubgroups soastokeep within bounds themileage required of the bicycle rider. The first three bisections, however, were comparable forwithdrawal andapproach trials. Before 0 wasasked tomake agiven judgment, markers were setuptoindicate thetwoends ofthestretch thathewastodivide. Themarkers were plain white rectangles, unlabeled. Insome trials, 0s ownstation point wastheorigin ofthe distance stretch indicated, sothatonlyonemarker wasrequired. The0 wasbrought tohisstation point, given hisinstructions, andasked toturn hisback while themarkers were setup.When these were inplace andthebicycle atthestarting point ofthestretch tobedivided, Easked 0 toturnandjudge when themoving target hadreached thedivision point ofthespecified fraction. TheE

240

J. Purdy & E. J. Gibson

signalled

the

one

rider

further

when

interval

of

the

a

was

permitted

markers

priate

labeled

,

as

of

53

Airmen

the

the

40

same

of

with

as

,

the

half

with

judgments

his

distant

error

were

.

and

Air

,

the

rider

were

completed

Sampson

setting

during

.

extent

yards

first

way

bicycle

There

the

approach

at

his

other

withdrawing

and

bicycle

training

.

made

were

the

basic

O

bicycle

of

with

the

by

to

direction

number

satisfied

faced

corrected

judgments

watched

in

not

measured

the

were

the

was

then

invisible

with

the

as

whom

were

see

a

He

was

was

who

with

points

If

error

half

as

to

division

group

of

. "

.

that

and

group

Stop

The

point ~

approaching

a

yard

.

confered

separate

that

"

permitted

judgments

fixed

bicycle

The

said

was

between

aid

a

adjustment

set

and

at

were

it

Base

the

appro

67

watched

Force

except

permanent

up

There

14

,

,

as

-

in

this

withdraw

.

All

.

Results

The

data

Table

obtained

14

for

. 2

each

.

Group

of

error

from

too

10

large

;

a

0

the

segment

,

interval

the

error

14

.

Accuracy

1

fractionation

-

14

given

of

. 1

or

of

-

very

largest

CE

to

error

is

is

3

be

%

7

bisection

.

'

equal

given

positive

him

if

. )

be

was

that

the

angle

small

can

are

A

so

that

visual

too

s

.

stretch

noted

,

The

the

subject

the

nearer

length

found

of

by

It

size

the

third

the

consulting

"

three

target

or

third

2

,

distance

)

. 2

is

the

a

.

stretches

are

directly

related

error

. "

judgments

of

50

-

yd

the

.

,

the

two

means

t

of

yd

is

a

20

On

'

but

The

,

the

average

fractionated

the

the

300

-

yd

direction

subgroups

significantly

are

distance

for

by

s

even

for

be

affected

-

average

total

. 2

uncorrected

differ

CE

,

the

the

to

be

0

tendency

.

to

21

act

When

corre

)

fractionated

to

one

.

the

.

There

stretch

to

the

results

group

bisection

appear

When

Nine

these

fraction

fractionated

length

for

.

not

for

the

uncorrected

fractionated

the

3

of

first

total

of

,

.

were

Only

6

the

point

who

above

stretch

Table

Os

.

go

distance

half

( see

sign

as

the

a

of

judgments

CE

of

division

feature

the

mean

relation

actual

striking

of

of

the

group

most

distance

direction

of

The

the

to

main

many

the

the

to

from

and

.

the

distance

increasing

movement

in

of

large

ranges

point

for

zero

%

concerns

way

measured

total

when

,

The

. 2

regard

.

-

the

the

-

does

from

( without

increases

SO

in

nearer

the

question

the

for

one

with

of

large

-

data

20

or

only

. 1

fractionated

compared

half

different

CE

of

first

of

error

of

well

significantly

the

and

segment

be

much

presented

divided

divided

may

be

yards

the

points

errors

of

out

one

-

their

amount

indicates

the

that

It

terminal

these

fractionation

of

third

presents

small

sponds

so

in

stretches

a

(

would

The

way

knowledge

the

its

)

are

.

half

Figure

CE

the

extents

,

and

of

judged

.

optical

geometry

fractionated

Table

small

trisections

of

that

too

into

optical

stretch

indicates

was

stretch

(

points

divided

him

by

and

errors

division

negative

nearer

divided

bisections

constant

true

that

segment

is

mean

the

means

the

are

at

150

.

DistanceJudgmentby Method of Fractionation

0 '1 . H 0 'I1 . 0 H ~ 0 H '\~ "Q + ~ ~ r:-. 'r~ ":-. I 0 H C (')1 .-tF )V 0 q ~ ' " " 0 'H Q~ e 0 + .cQ :~- J ~ . ' " 'c:-0 ~~ ..:,u ('t3 -.c-0 :t)=I0 0 tF )U ~ '~ "

t.... HHt* *..... \Q *~ *~ *~ H a rf') ar) to -.: ~ ~ ~ I + + + + + HH I I

' 0

t....

to. . . . .

~

C( )

~

~

~

lr )

\

~

Q

~

~

\

~

~

Q

C( )

* LI ' )

I - t

I

+

0 \

*

~

LI ' )

I - t

~

r:-.. :

Or

~

~

t--:

~

N

rf

rf

)

+

t-...

rl

~ )

rf

'

~ )

rf

H

H

*

I - t

~

I

)

H

LI ' )

+

I - t

HLr

I - t

ON

~

~ ~ 0\

\

~

:-.. :

~

I

I

+

0

Lr )

~ )

*

~

Q

rf

~

~

~

\

N

)

Q

H

* 1 ~ --q ~ 0'--: 0 H H I + + +

~ ~ I

* ~ ~ ~ 0 ~ Lr) I I I

Lr ) rf") '0

~

cx: >

t....

0\

N

to .. ~ (t:) to ... 0 t't') ~ \Q ~ ~ ~ ~ ~ ~ H ~ ~ r~-.;j ..0\ Lr) f") "' 0 ~ 0' 0' t-t' + + I +

"a a "' 0 ~ t-t' ~ I I +

0\

N

CXJ

t.....

H

\ Q

lr )

f' t' )

0 lr )

lr ) t.....

0 ONlr HHH

lr )

0

l/) t'-... ~

0 ON NN

lr )

t-t t-t' I

)

241

242

J. Purdy & E. J. Gibson

+ 20

.

+ 10 .

0

0 .

.

0-

Q

0

-

.

.

0

.

.

0

0

.

0 0

- 10

UNCORRECTED

0

GROUP

. TARGET

APPROACHING

0 TARGET

WITHDRAWING

S FROM

S

- 20 0

25

50

75

100

TRUE

125

150

175

200

225

250

1 / 2 OR 1 / 3 , YDS .

Figure 14.1 Plotof thegroupmean constant errorof thejudged fraction (in yd. fromthetruepoint ) relative to thetruepointofdivision , arranged in acontinuum ofdistance .

'SOA'.lN3W90nr

/ l ~O l / l ::10 3J N~ 3W

and225yd. Themeansfor approachandwithdrawalalsodifferon fractionationsrepresentedby points at 250 yd. and 25 yd. However, these lasttwo differences arenot necessarily dueto thedirectionof movementof thetargetalone.Otherdifferences confounded with directionof movement for all judgmentsafter the third were: (a) the judgmentswere madein differentorders,(b) Oswerenot alwaysmakingthe samekindsof fractiona tion of the samedistancestretch(seeTable 14.1). Thus the difference betweenapproachandwithdrawalCE's for fractionations at 25 and250yd. couldhavebeendueto any of thesevariationsin procedure . In general , the error appearsto be positiveon approachtrialsbut muchlesspositiveor negativeon withdrawaltrials. At the Iso-yd. divisionpoint, for instance , 88% of the approachgroupbut only 53% of the withdrawalgroupmade positiveerrors. No differencein the variability of fractionationwas producedby the directionof targetmovementandthe kind of accompanying fractionation exceptat 200yd. wherethe SDfor the approachconditionwashigher. A comparisonof the accuracyof bisectionsandtrisectionscanbe made by comparingthe meanCE's of halving 50-yd. stretcheswith the mean CE's of taking one-third of 75-yd. stretches . By sucha comparison , the sizeof the CE is no largerfor division into thirds than for divisioninto halves , thoughgroupvariabilityis largerfor the one-third judgments . Variability was also influencedby the distancefrom 0 of the stretch to be fractionated . Trisectionsof 75yd. canbe comparedfor four stretches

Distance Judgment by Methodof Fractionation 243 which begin at 0, 75, 150, and 225 yds. from O. As the distance of the stretch from 0 increases , the SD also increases .

Indicationsof distance"constancy " It might be supposedthat perceptual judgments of distancemagnitudesare basedwholly or in part on the visual angles subtendedat the retina by the stretchesof distance. If two physically equalstretches, one nearand one far, were comparedon this basis, the resulting relative subjective magnitudes would not be equal; the farther stretch would subtend a smaller visual angle and would appear less. In order to makeit appearequal, the further stretch would have to be actually greater (a negative error in our experiment ). Thus if 0 made fractionation

judgmentswholly or partly on the basisof suchinformation, a negative CE would be the result. Gilinsky (3) may have assumedsomething like this in stating that " Perceived distances are foreshortened . The perceived distance

d increaseswith the true distanceD but at a reducedand diminishing rate" (p. 462). The presentdata, however, do not substantiatesucha hypothesis. The errors in general are positive -

that is, a made the nearer segment too

large in comparisonwith the farther. And the high degree of accuracyin general indicates that distance judgment exhibits a kind of constancy analo-

UNCORRECTED

8

GROUP

8 TARGET

APPROACHING

cn 150

0 TARGET

WITHDRAWING

0 >-

x

S

X/

FROM

S / /

N~

/

" "

~

/

0

/

w

/

(, 9

/

0

~

/

w

U Z

/0

MEAN

/

/

8 / 0

75 /

r-

/

C/ )

/

0 Z
/ / / /

w

2

/

25

./

0

50

150

300

LENGTH OFTOTALSTRETCH BISECTED , YDS. Figure14.2 Therelationship betweenthe pointjudgedhalfwaybetweena anda marker , andthe distance of themarker in yd. Thelineat 450onthegraphrepresents accurate fractionation .

244

J. Purdy& E. J. Gibson

gous to sizejudgments. The term lengthor depthconstancy , as distinguished from width or height constancy , might be applied to O's judgments. The fact is that a distance stretch beginning at his own feet can be matched fairly accuratelywith another stretch beginning at a distancefrom his feet along the sameaxis. In Fig. 14.2 the mean distancejudged one-half has been plotted for the three distance stretcheswhich began at O's feet: 50, 150, and 300 yd. Judgmentsfor approachand withdrawal trials are presentedseparately, as well as the means for the two combined . The 450 line describes correct half

judgments . Thoughpointsfor the approachconditiondivergefrom accuracy, the mean values lie extremely close to that line. There is a slight indication that as the distance to be halved increases, as subjective halfway point may move farther away from him so that the nearer segment is

physically larger than the farther segment. Figure 14.3 shows the error that resulted from the fractionation

of

distance stretches as the distance of those stretches from 0 increased . The

points connected by solid lines represent bisections of 50-yd . stretches. +5

v> 0

-

-

-

o0

- : ; -- ~

/

>-,

""

Z 0

0

~ ,

- - -~~~

/,

,

0

.

.,

-

--

-

-

O ..:::::~ ~ ""' ~

- -

.

- - ' - - ---:-: . -

-

- -

. . .,. " .. . ,

~

. ..

, .o ~ ,....." . ... . ....,_.,..._."........,....,....- o

...'

- -- ---

5 -

Z

I

- -~

" '"

-

~

-

-

-

- - -

-

- - =: - - 0 -

-

-

-

-

-

-

-

...-

0 -

,/

~ U

/ /

......-

- 10

/

a:

/

LL

I

/

LL

0 w

/

- 15

/

I

u

/

Z

,

~

I

~

UNCORRECTED GROUP

I

- 20

I

. TARGET

APPROACHING

0 TARGET

WITHDRAWING

S

I

I

FROM

S

I

I

-

- 25 0

25

50

I 75 TRUE

-

' 100 POINT

1_ 125 OF

1/ 2

...,... 1/ 3

JUDGMENT

_150

DIVISION

175 , YDS

200

225

250

.

Figure14.3 ThegroupmeanCEof (a) halvinga 50-yd. stretch(solidline) and(b) dividinga 75-yd. stretchintoone-thirdandtwo-thirds(dottedline) asa functionof thetruedivisionpointin yd. Thedashed curverepresents theexpected errorif 0 wereto dividethestretches sothat theresulting fractionsubtended appropriate fractions of visualangle .

Distance Judgment by Method of Fractionation

245

Theerrorforeachbisection is plottedat thetruehalf-way pointofthe stretchbisected. Forinstance, thepointat 125yd.represents a bisection

of a stretchextending from100yd.to 150yd.Thedottedlinesconnect errorvaluesthat resultedfromdividinga 75-yd.stretchinto intervals subjectivelyequalto one-thirdand two-thirds.

Thehorizontal linedrawn atzeroerrorindicates theexpected judgments

if constancy for distancestretcheswereperfect.Thedashedcurve,on

theotherhand,describes thejudgments expected froma hypothetical 0

(6 ft. tall)whofractionated the stretchesby makingthe two fractions subtend(a)equalvisualanglesin thecaseofhalving50yd.,or (Ii)visual

anglesequalto one-third andtwo-thirds ofthetotalanglesubtended by 75yd.Theobtained judgments arecloseto thelineofobjective accuracy anddo not followthe hypothetical visualanglejudgment curve.The judgments at thefarstretches showsometendency to benegative. Since theerrorissmall andsincethetwohypothetical curves areclosetogether at thefarstretches, it isnotpossible to conclude thattheplotted points follow eitheroneortheothercurveat 175,200,and250yd.Thepoints plottedat 25,50,100,and125yd.followthe constancy line.

Thisdemonstration suggeststhe interdependence of all spatialcon-

stancies. Allthe dimensions of a surface, bothfrontaldimensions (termed size)andthethirddimension (termed depthordistance) showconstancy of

magnitude. Gibsons moregeneral statement maythusbeapplicablethat Scale,notsize,isactually whatremains constant inperception (2,p.181). Theeffectof correction onfractionationSo far,onlythe resultsof uncor-

rectedOshavebeenconsidered. Anotherlargegroupof Oswhoseerrors werecorrected as thejudgments proceeded canbe compared withthem. Table14.2givesmeanCEsandSDs for bothgroups.Thecorrected grouphad12of20CEssignificantly greaterthanzero,whileonly9 ofthe uncorrected groupsweresignificantly greater.A t testcomparing the magnitude oftheCEsofthecorrected anduncorrected groupsshowsthat

fiveoftheCEsaresignificantly larger (p< .05)inthecorrected group(at 125and200yd.withbicycle withdrawing, andat 25,200,and250yd.

withbicycle approaching). ThemeanCEisinnocasesignificantly larger in theuncorrected group.Correction, therefore, doesnotappearto increase theaccuracy of fractionation, whichis,indeed, verygoodalready. The fractionation withcorrection did,however, serveasaneffective training

procedurefor laterabsoluteestimatesof distance(1). Summary

Thepresent studyisanapplication ofa psychological scaling operation (fractionation)toperceived distance. A300-yd. stretch ofgrasscovered flatground wasused

246

J

for

as

.

Purdy

&

judgments

judgment

.

by

Six

so

(

as

.

.

angle

3

no

sub

.

positive

the

related

as

the

bicycle

4

.

of

approaches

withdraws

lie

( up

Perceived

and

every

.

ground

25

and

300

yd

to

distance

follow

on

. )

yd

.

into

of

of

from

ap

-

.

a

the

halves

distance

hypothetical

curve

magnitude

of

the

visual

.

motion

,

to

magnitudes

stretches

direction

bicycle

distance

based

distance

the

would

his

point

.

judgments

being

different

indicated

division

intervals

magnitudes

the

0

:

.

their

and

correct

different

point

of

physical

for

to

of

either

the

target

less

,

the

positive

CE

or

tends

to

negative

be

when

.

Fractionation

was

with

from

by

is

.

,

the

points

yd

accuracy

tendency

tended

Error

-

stretch

with

conclusions

good

result

,

division

stretches

well

is

this

considered

made

25D

following

down

he

true

the

very

would

or

what

the

divide

correspond

There

of

through

the

with

which

at

were

one

)

can

to

up

rider

trisections

suggest

pear

moved

that

thirds

2

bicycle

point

Observers

or

Gibson

four

chosen

results

1

.

the

station

The

J

A

and

stretches

.

stopping

bisections

zero

E

of

variability

distances

was

reduced

not

improved

by

correcting

0

' s

errors

,

nor

.

Note

1

.

This

No

research

.

33

( 038

Resources

tion

States

was

) -

supported

22373

Research

,

translation

Government

in

monitored

Center

,

publication

part

by

,

Lackland

,

use

by

the

Air

,

and

the

United

Perceptual

Force

disposal

States

and

Base

in

.

whole

Air

Motor

Force

Permission

in

part

Contract

Laboratory

is

and

under

Skills

,

granted

by

for

or

for

Human

reproduc

the

-

United

.

References 1.Gibson , E. J. Bergman , R., &: Purdy , J. Theeffect ofpriortraining withascale of distance onabsolute and relative estimation ofdistance over ground .J. expo Psycho /., 1955 ,50,97-105 . 2.Gibson ,J.J.Perception ofthe visual world .Boston :Houghton Mifflin , 1950 . 3.Gilinsky , A. S. Perceived size anddistance invisual space . Psychol . Rev ., 1951 , 58, 460 -482 . 4.Gilinsky ,A.S.Perception ofsize ofobjects atvarious distances . USAF Personnel Trainin Res . Cent .Bull ., 1954 ,No.TR -54 -92. 5.Smith , W.M. Amethodological study ofsize -distance perception ,J. Psycho /., 1953 , 35 , 143 -153 .

15

Continuous

Perspective

Perception

of

James

J

.

Constancy

in

of

There

move

don

' t

rigid

of

had

"

the

effect

had

the

transformation

-

sisting

forward

of

for

closely

the

to

the

Wallach

rather

to

than

The

looking

shadow

innovative

these

It

was

displays

that

the

static

such

we

objects

and

pathways

here

and

as

-

approach

we

things

them

these

.

Johansson

in

,

;

constancies

' s

if

?

they

What

in

was

information

for

absolutely

shape

and

stimulus

to

for

the

screen

did

was

perfectly

Journalof Experimental Psychology , 1957, 54, 120- 138.

was

more

later

empha

and

and

stimulation

of

-

form

.

memory

over

role

-

brought

,

time

.

perspective

.

in

the

depth

events

rigidity

the

rigidity

was

per

here

experience

investigate

there

under

he

of

displays

convincing

information

of

of

to

so

is

although

past

,

invariance

-

perception

by

)

depth

underlying

rigidity

,

on

kinetic

1953

of

the

Wertheimer

monograph

exposing

and

transformations

designed

for

used

(

in

.

the

theory

perception

effect

)

and

' Connell

time

for

motion

1957

' s

for

on

of

(

,

the

that

depth

information

O

,

at

role

Gibson

Johansson

change

work

the

.

1950

But

location

principles

J

,

and

.

of

shape

.

in

Wallach

kinetic

of

J

1923

geometry

the

;

theory

perception

as

here

setup

as

appeared

effectiveness

,

explain

caster

.

the

projective

reprinted

as

The

by

' s

for

experiment

transformations

resemble

time

of

sought

,

a

fate

in

Wertheimer

importance

just

constancy

substance

understand

by

"

motion

things

first

related

sized

is

not

perceived

grow

we

to

common

discovered

of

that

properties

led

Perception

role

,

discussed

that

things

of

expounded

of

been

with

that

Event

recently

concern

do

was

suggested

of

or

How

surface

was

constancy

away

.

ground

it

layout

and

substance

law

in

the

and

shapes

move

these

and

suggested

and

,

?

like

shapes

Configurations

a

we

rigidity

them

things

surrounding

,

as

questions

the

Gibson

perceived

environment

for

was

entire

of

shrink

information

It

the

maintain

perception

of

.

paper

constancy

the

to

they

the

of

also

in

appear

J

previous

scale

is

around

is

the

the

.

Eleanor

dimensions

with

and

Motion

,

perceived

characterizes

are

Rigid

Gibson

connection

there

Transformations

most

not

the

experiment

skeptical

need

flat

to

,

was

witness

but

have

the

also

watching

depth

transformations

itself

or

248

J. J. Gibson& E. J. Gibson

were

inevitably

made

seen

movies

Francisco

of

,

address

where

to

Oxford

,

in

could

The

) ,

all

be

children

in

rotation

,

each

of

time

the

years

ago

of

,

' oeuvre

( off

discoveries

went

on

nonchange

at

he

the

she

never

rectangular

.

I

swinging

)

his

film

,

it

wire

the

to

a

if

.

transparent

paths

in

seminar

experiment

wire

of

movement

ourselves

surface

to

wondering

of

film

on

around

the

and

the

then

sudents

( made

San

presidential

travelled

show

the

We

to

and

demonstrating

made

a

The

.

along

,

still

surface

We

)

.

hear

separated

.

screen

also

that

is

,

experiment

it

rigidity

,

and

a

" blobs

"

fared

referred

at

a

as

theory

and

a

our

practised

also

two

path

of

An

own

-duality - - -- - -

the

stimuli

"

,

tive

, "

a

in

is

,

it

in

underwent

,

J . J .

Gibson

all

detected

need

and

irregular

use

,

of

1979

,

pp

in

when

.

in

182

ff

later

arriving

in

objects

.) ,

act

be

was

great

,

an

object

,

observer

-

shape

in

phenomenon

of

the

The

swinging

of

shadow

This

An

is

familiar

movement

of

of

a

six

combination

does

of

any

,

of

ways

exhaust

it

six

-

in

. !

ways

dark

motion

the

,

world

another

,

case

it

the

( 6

in

that

gave

on

a

plane

kind

was

light

of

a

three

mechanical

of

this

of

always

background

of

and

of

yields

other

perception

not

suggested

sequence

transpositions

possibilities

could

7 )

pattern

continuous

a

,

a

any

stimulus

a

,

which

world

shapes

three

transfonnations

but

The

underwent

the

motion

the

motion

space

not

grouping

,

continuous

in

plane

optical

motions

geometrical

motion

a

of

of

screen

in

one

.

,

.

Constancy

the

was

and

varieties

kind

picture

transformations

the

rigid

in

that

motion

moving

which

a

motion

texture

on

.

locomotion

perception

one

of

geometrical

forms

concept

objects

( see

of

the

that

impression

"

a

cast

although

shadow

object

are

of

that

change

,

the

.

memory

the

during

movement

for

hypothesis

an

and

about

survey

as

no

and

of

straightfonvard

the

space

object

types

casting

rotation

more

.." .

exploratory

serve

,

the

" ideal

perception

only

of

through

Furthermore

,

is

and

change

detects

or

which

path

moving

more

observer

that

through

its

saw

the

more

both

or

statically

screen

arc

geometrical

not

his

~ ~ ~~ - -

of

duality

visual

about

rperceptual - ' - - r

a

as

that

us

perceive

two

or

and

impress

to

perceive

the

the

never

well

information

but

.

subject

perceptual

of

obtains

'of"' J

as

to

of

event

the

to

behind

form

particular

since

even

to

system

completely

angle

perception

,

was

visual

demonstrates

constant

simultaneous

involved

that

the

and

object

the

the

,

presented

the

detects

both

experiment

of

time

The

of

this

ability

sees

shape

observer

of

the

same

.

or

of

was

simultaneously

If

year

to

arcs

d

gate

them

with

year

layers

a

.

One

If

sabbatical

more

different

moving

like

took

and

Association

that

four

chef

rather

,

connection

embarrassing

of

our

in

colloquia

or

in

was

a

little

film

,

reason

them

spend

( thirty

once

,

a

depth

that

Psychological

of

is

the

at

helped

to

number

me

in

for

American

about

a

visible

,

showed

it

moving

beautifully

as

to

display

moving

the

although

really

final

of

were

,

object

husband

3

fact

,

that

net

my

we

occasionally

an

experiment

Division

where

Europe

as

this

.

perspec

rigid

-

surface

rotations

movement

geometrical

.

kind

Perception of RigidMotion

249

(but only a few exampleswere presented) it arousedperceptions of nonrigid or elastic surfacemotions, of the kind exemplified in the movements of organisms. Of the six rigid phenomenalmotions, three (rotation around the line of sight, transposition up or down, and transposition right or left) are inducedby a stimuluswhich common sensewould call motion; one (transposition along the line of sight) by a stimulus which common sensewould call expansion or contraction; and only the other two (rotation around a horizontal or a vertical axis) by a stimulus which common sensewould call a transformation . Optics, however, demandsgeometrical terms. All six projectedmotions are different parametersof continuousperspectivetransformation, and they are mathematicallyakin. Common sensetells us that the first three optical motions shouldgive the perceptionsthey do (a motion yields a motion) and that the last three should not (how can a changeof size or shapeyield a motion?). The assumptionis that a visual experience hasto resemblevisually the optical stimulusthat producedit. But a better assumptionis that experiencesneedonly correlatewith their stimuli, not replicatethem, and the presenthypothesis says of - that any ~ continuousseQuence - - -. 1 ------- - perspectivetransformationsis the correlateof perceptuallyrigid motion. There was in the film some evidence to suggest that this hypothesis must be qualifiedif the perspectivetransformationsare those obtained with parallel projection instead of polar projection, that is, with the special case of transformations when the focus of projection is at infinity . The two apparentrotations around a horizontal or a vertical axis then seemed to becomesomewhat ambiguousas to rigidity or elasticity, and apparent reversals of direction of rotation appeared. The apparent approach or recessionalso fails of necessityin this casebecausethe changeof sizeof the stimulus disappearswith parallel projection. If the above observationsare verified, the hypothesis should specify perspective transformations with polar projection. Previous experimental work on the kinetic depth effect (14, 15) or on other appearancesof depth in moving fields (3, 12) does not supply evidencefor or againstthe amendedhypothesissincein generalthe changesof shapestudied in them were not polar projective. Theseexveri~ ments, moreover, are mainly concernedwith what canbe calledthe appearanceof internaldepthof an object, whereaswhat we are here talking about is the appearanceof slant depthof the face of an object. The distinction is made clear in the film. The apparentmotionin depthpreviously studied by Smith (13), however, is relevant to our hypothesis. One or two of Wallach's many experiments on the kinetic depth effect (14, p. 212 ff.) are relevant indirectly if the changes of such impoverished stimuli as line segmentsor anglesare restatedin terms of perspectivetransformations. This hypothesis is comparableto, but different from, the principle involved in Wertheimer's "law of common fate" (16) in several respects.

250

J. J. Gibson & E. J. Gibson

Both refer to some kind of motion in or of a grouping of spots or forms, but Wertheimer's law predicts the organization of a figure in the visual field, whereasthis predicts the quality of rigidity of a surfaceor surface-like experiencein space. Wertheimer's law seemsto imply that the various parts of the complex are united by sharing a common motion (such as moving in the samedirection with the samevelocity) but this hypothesis assertsthat any perspectivetransformationis a single motion mathematically, including the size and slant transformationswhere, analytically considered, every part moves in relation to every other. Wertheimer's law leadsto experiments, on "configurationsof motions" (9, 12) in which each part of the complex undergoescomponentsof translation or rotation but the part is not itself transformed; a geometricaltransformation, however, is somethingthat permeatesevery part as well as the whole of a texture, and the apparatusused in the presentexperiment satisfiesthis condition. It might be noted that the problem of how we discriminatethe rigidity of rotating solid objectsand of approachingor recedingsolid objects in the environment is closely connectedwith the traditional problems of shape constancyand size constancy. Langdon has recently shown that the shape constancy of an object is considerably increasedunder conditions highly unfavorable for it when the object is made to rotate (10). Likewise the question of why we see a rigid environment when we move among the solid surfacesaround us is closely connectedwith the traditional problem of spaceperception (8). Aim of the experiment The experiment to be reported sought to answer four questions. First, doesthe appropriateparameterof continuousperspective transformationswith polar projection always give the perception of the changing slant of a constant shape? Second, are the judgments of amount of change of slant away from the picture-plane in good psychophysical correspondencewith the "extent" or "length" of the transformation sequence ? Also, how variable are thesejudgments? Third, is the outcome dependent on or independent of the kind of shape or texture on which the transformation is imposed? Fourth, how accurate, if at all, is the judgment of slant away from the picture-plane when only the static end product of the transformation sequenceis presented to 0 but not the motion leading to it?

Method Apparatus and Stimuli The optical geometry of the apparatusused is shown in Fig. 15.1. The device can be termed a "shadow transformer." Essentially, it presentsto an eye an optic array

Perception of Rigid Motion

251

TRANSLUCENT SCREEN WITHSHADOW CHANGE CHANGE

OF

OF

OF

" V I RTUAL

SLANT OF CASTER

SLANT " OBJECT

)

SHADOW

~

/-

PO SO

/ / /

-

?- -

-

-

. . ~~~- - - - -

VIEW

OF APPARATUS

FROM

ABOVE

PRODUCING

A SLANT

TRANSFORMATION

Figure15.1 The shadowtransformer .

of limited scope within the boundaries of which either static patterns or continuous

perspectivetransformationscan occur. In this optic array, unlike those of everyday vision, the differential light intensities and their structure are under E's control; the pattern is the same for either eye, and the need for differential convergence and

accommodationis eliminated. All the "cuesfor depth," in short, tend to determine a fIat plane except those of form and motion, which are thus isolated for study. The source of this converging array is a window in a translucent screen.

This optical stimulus is artificially produced by the diverging ray sheaffrom a point sourceof light, into which shadowsare introduced by opacities of one sort or another attachedto a transparentplane mount. Rotations or translationsof the mount (on bearings or tracks outside the ray sheaf) yield corresponding transformation sequences of the shadow . This experiment utilized rotation on a vertical axis. The stimuli were the mirror reversals of these moving shadows , visible on the other side of the translucent screen. If an apparent rotation of a " virtual object " is

induced by such a stimulus it should always be opposite to the rotation of the shadow caster, without ambiguity . ...,

-

The seated0 , in a dimly illuminated room facing a large white surface, saw a luminous

square window

36 cm . on a side at a distance

of 1.80 cm ., made

of

translucent plastic 1 in. thick. The light source was fixed at the same distance behind the window as the eye was in front . It was a 300 -w . Sylvania point source

carbonarc lamp, but any lamp with a single filament of small diameter(up to 1 mm. or more) will serve the purpose. The window was visibly flat. Binocular vision was permitted 0 after preliminary work failed to show any differencebetween the use of one or two eyes. The mounts were transparentrectangularsheetsof i -in. plastic,

252

J. J. Gibson& E. J. Gibson

of such size (30 x 100 cm.) that when they were centered and rotated on a

turntable placed midway between the point sourceand the window they could be turned 700 from the parallel plane without the edges being projected within the window . The turntable could be rotated back and forth through an arc of variable length by an adjustableeccentriclinkage, geared to a motor with a variable speed drive. A speedwhich gave 2-sec. cycles of semirotation was chosen, after exploration indicated that an optimum might be in this neighborhood, although the rate was not critical for the experiment . The quantitative variable of this experiment ,

then, was the "length" of the transformation sequence , as expressedin degreesof angular excursion of the turntable. We shall return to this point later. Five degrees of semirotation were presented: 15 , 30 , 45 , 60 , and 70 . Each cycle began with and returned to the parallel plane. Theforms transformed The variety of forms, patterns, and textures that can be projected with this device has been suggestedelsewhere(7). Four were used in the experiment: an amoeboid group of amoeboid dark shapesor spots (the irregular texture), a solid amoeboidcontour form (the irregular form), a squaregroup of dark squares(the regular texture), and a solid square(the regular form). Eachwas cut out of gummed paper and attached to the central area of a transparentmount so that its shadow was projected to the center of the squaretranslucentwindow . With the mount parallel, the regular shadows extended 20 cm. each way in the 36 cm. squarewindow , and the irregular shadows about the same. It may be noted that the "regular" stimuli are constituted of rectilinear contours and alignmentsand the "irregular" stimuli of randomly curved contours and alignments. There are also differencesin symmetry, and perhaps other geometrical properties. The "forms" are bounded by a single closed contour and the "textures" by many closed contours; the total contour length is much greater in the latter stimuli. A texture might be described as a "form of forms," as distinguished from a form as such. These textures were, however, very 'Icoarse"; there were 36 squaresin the "platoon" and 36 "amoebas" in the /,colony." Thevariableprotractor For recording judgments of changeof slant, 0 had before him a sort of protractor with its baselineparallel to the plane of the screen. It bore an adjustablepointer which could be moved to indicate an angle of semirotation. The top side was blank but the bottom side carried another pointer and a scale which could be read accurately by E after each trial . Instructions

and Procedure

The experimental group Before receiving any formal instructions for the experiment , each a was seated and told : " You see in front of you a screen with

a window

in it which will be illuminated during the experiment. I will first show you a moving pattern of dark lines filling the window . If what you seeis a movement of some kind of object , describe it ."

A network (woven wire fencing of a common type) was then placed on the turntable and turned through various excursions. Although the question was intended to suggest neither a deformation in the plane nor a rigid rotation out of

Perception of Rigid Motion

253

the plane, alIOs reported seeing the latter, and spontaneouslyreported different amounts of rotation. The suggestions in the following instructions were hence consideredpermissible: "During the experiment proper, a dark form or pattern will appearin the middle of the window . It will seem to rotate back and forth on a vertical axis- to turn away from the plane of the screenand return. Your task is to judge how far it rotates, or the maximum angle it makeswith the screen. Use the circular model in front of you to make this -judgment." ~ One of the four patterns was then presented at one of the five degrees of transformation for 20 sec., which permitted 10 cycles of stimulation. The 0 had no difficulty in making his judgment during that interval. Twenty such trials (five for each pattern) in an order counterbalancedfor the group were made, and then another 20 trials in reverse order to determine whether a practice effect would appear. The 0 was not told his errors. There were 20 Os in the group. Thecontrolgroup A separategroup of 30 Os was treated as similarly as possible except that the four stimulus patterns were motionless. Only the end product of each transformation sequencewas presented, and only one degree of transformation was used- that with the mount at 60 . For the preliminary exposure, 0 was shown a motionless pattern filling the window, half the group seeingthe network of lines and the other half a less objective cloud-like pattern (this making no difference in the outcome) and he was asked if he saw an object of some sort. Then 0 was told that he would seea form or pattern in the middle of the window . It might be parallel or slantedaway from the screen. If he saw it slantedaway from the plane he was askedto judge the angle it made using the model in front of him. Four trials 1were given (one for each pattern) in an order counterbalancedfor the group. Results

Theexperimental group The first question is whether alIOs saw the changing slant of a rigid shape. As stated above, all did at the outset. During the 40 trials which followed, eachof considerableduration, many spontaneous descriptionswere offered, and 8 of the 20 Os observedat some stagethat the display could be seenas a compressionof a two-dimensionalpattern. They were all psychologists. Twelve did not so report, and stated at the end that they had never observedit . The two-dimensionalimpressionsdid not persist long enough to prevent the requestedjudgments of changing slant. There was no difference in this respect between the regular or irregular forms or textures. The secondquestion is whether the judgments of changeof slant are a function of the amount of changeof form. The "length " of the transformation sequenceis expressedas the inverse angular excursion of the shadow caster, and this variable is plotted on the horizontal axis of Fig. 15.2. The

254

(S33 'd930 ) S.lN3V'J90nr ;:jONOI .lVI /\ 300 'd~ ON~ .lS

J. J. Gibson & E. J. Gibson

CHANGEOF SLANT IN OPTICALSTIMULUS (DEGREES )

Figure 15.2 Judgmentsof change of slant as a function of the length of the transformation sequence.

9 .0

.~'.. '. :' 0 - ' : '0

8,

:

:

7

.::.'/ '

:.~ . ' .

! ' ::.

0

: .

: "

""0

" 0. , ....'"".; ' -- --- - -

.

" V

:: /

:

"""' "

" 0.

/ ,, "" . .

.

.." .'0.

\

\

0

\

,

"

:' j

\

"

".

5.

,

\

\ \\

'

\ ,

,

,

\\

\

1

-- ~

:',,/

I

40

.

\

!::./ ' ,1

.

,'. :'

\

\

,

.: , .' / !

6

- - -e

,

0

e./

D - - 0- - REGULARFORM

0

DD

/

00

3,

-

- 0 -

- REGULAR

PATTERN

Q - - -e--- IRREGULARFORM

0\\ .'."'X"'.'- I RREGULAR PATTERN 0

0

10 CHANGE

20

30 OF

SLANT

40

50

I N OPTICAL

60 STI

70 MULUS

Figure 15.3 Variability of judgments as a function of the length of the transformation sequence.

Perception

judgments

are

of

40

plotted

.

underestimate

,

also

for

can

is

the

the

be

Figure

and

slant

"

They

. 3

in

"

,

The

the

but

. 1

the

) ,

the

four

kinds

,

is

obvious

of

.

discover

4

of

and

approach

8

90

.

,

An

the

at

analysis

of

whether

either

errors

variance

multiolicitv

J

.

or

regularity

of

Multiplicity

at

the

or

only

the

weak

,

"

Table

5

"

%

patterns

significantly

"

level

. 2

The

is

form

. "

related

not

.

is

in

of

the

might

strengthened

of

Variance

of

Errors

of

~

Judgments

,

of

Change

of

excursion

turntable

( A

or

of

4

irregular

texture

Practice

( P

x

direction

by

the

( R

( T

)

significant

interac

-

F

1437

1

)

. 07

320

)

72

5

. 12

.

. 56

.

. 93

109

. 47

. 53

R

X

5

19

57

. 69

X

5

19

52

. 63

23

. 07

A

X

R

4

60

. 54

4

. 77

.

A

x

T

4

67

. 09

2

. 66

--

T

x

A

A

x

T

x

:

tests

R

x

R

Note

F

1

5

x

T

19

R

x

76

5

x

76

5

(

those

11

,

.

p

=

. 05

.

.

p

=

. 01

.

p

.

interactions

330

and

which

338

)

are

included

13

. 15

12

. 69

25

19

Only

299

appeared

to

in

.

. 90

453

1

76

5

13

. 68

1

19

5

X

is

theory

Slant

T

P

but

Gestalt

of

( 5 )

A

.

significant

MS

)

or

Form

is

from

'"

judgment

..,

df

Subjects

regularity

made

perhaps

-

Source

Regular

of

expected

be

,

errors

. 1

Analysis

Degree

to

effect

is

that

It

The

tendency

predictions

good

15

is

texturedness

considering

about

.

.

change

between

used

to

group

.

differences

pattern

necessary

the

about

to

edge

.

the

of

begins

an

and

transformation

within

between

slant

,

irregular

a

function

range

becomes

any

a

striking

transformed

are

as

to

more

that

judgments

maximum

Even

gets

mean

tendency

and

what

judgments

255

a

small

regular

suggests

errors

show

however

This

the

surface

.

the

of

the

a

angles

independently

variable

as

30

Motion

representing

for

for

.

Rigid

point

except

and

functions

of

The

not

with

15

s

each

,

15

variable

'

again

does

obtained

SD

,

linear

stimuli

,

phenomenal

graph

the

the

how

.

fall

a

)

such

stimulus

limit

Table

as

is

shows

the

rise

which

5

multiple

to

axis

is

of

question

15

to

similarity

single

related

vertical

function

4

responded

A

the

The

( about

however

(

on

reproductions

of

the

Table

be

of

.

some

. 24

. 13

interest

or

were

used

for

the

.

256

J. J. Gibson& E. J. Gibson

tion between angle and regularity which seemsto reflect the tendency, barely noticeable in the graphs, for the irregular forms ~o depar~ slightly more from linearity at ~he larger angles. All ~he forms in ~his experiment were apparently good enough to carry the transforma~ion, and i~ was ~his which mainly determined the judgments. This answersthe third question. The form of the changeseems to be what is important , not the form itself .

A conception of the various forms that optical changemay take is, however unfamiliar, probably necessaryfor an understandingof the perceptual process .

It may be noted from Table 15.1 that no significant practice effect appearedbetween the first and secondblocks of 20 trials. They have been pooled in Fig. 15.2 and 15.3. The two halves of the data independently warrant the same conclusions, when the curve of Fig. 15.2 is plotted separately for them . The control group

The outcome of the control experiment was radically

different inasmuchas the judgments of slant dependedon the regularity of the form or texture presented. The irregular stimuli, in fact, generally appearedin the planeof the screen(85% of 60 judgments) while the regular stimuli generally appearedat a slant from the screen (97% of 60 judgments). Even for the regular stimuli, however, the mean degree of slant perceived was only 240 (SO about 120) whereasfor the moving regular stimuli the mean had been 610 (SO about 60 ). This is gross underestimation

for the motionless and great accuracy for the moving stimuli . The under-

estimation of slant is consistent with previous researchon static optical forms and optical textures under similar conditions. A trapezoidalform can sometimes arouse an impression of slant, but an exact linkage between the

apparent shapeand the apparent slant (a "psychological invariant") is not obtained (2). A static optical texture with a compressionof texture on one meridian relative to the other induces a perception of surface slant, but

even when the texture is regular the slant is underestimated , and when the texture is lessregular the slant is more underestimated(4, p. 380). The irregular form and the irregular texture displayedin this experiment were evidently not of such a kind as to appear slanted when altered by a

slant transformation, since they generally still looked frontal. The Os, of course, had never seen them beforetransformation .3 A truer statement of the matter is that the family of perspective transformations of the amoeboid stimuli has no unique member with immediately identifiable properties . The

family of perspective transformations of the quadrilateral stimulidoeshave such a member - the square. At the outset there is one in this family which the other members can be transformations of, but none in the former family .

Hence, rectilinearcontour and alignment- a rectilinearstructure- provide

Perceptionof RigidMotion

257

a primarybasisforpresumptions of slantwhichnonrectilinear structure

does not.

Oneof thewritershasarguedthatthereexistsa betterbasisthan

contour forpresumptions orevenperceptions ofsurface slant, namely,

internal texture anditsdensity variables (4) 4 Itmight beexpected, there-

fore,thatthetexture ofamoeboid spotswouldinduce slantmoreoften thantheamoeboid form,andthetexture ofsquares moreslantthanthe square. Neither expectation wasfulfilled. Theexplanation maybethatthe

textures used(36elements, oronlyabout6 eachway)weretoocoarse to

make the density variablesdeterminate.

Evidently a configuration withonlythefeeblest stimulating power for

depthperception, ornoneat all,cannevertheless carrya transformation sequence whichyieldsaccurate depthperception. Thisanswers thefourth

question.

Individual resultsTime didnotpermit more thantwojudgments foreach condition per0 intheexperiment justreported. Itwasthought desirable to runa moreexhaustive seriesin orderto determine the extentof the variable erroroftheslantjudgments forsingle individuals. Accordingly, two0s wererunforfivedayseach,sothat10judgments percondition were available. 5 Table 15.2 shows themeans andSDsofthese judgments. Themeans arestrikingly likethoseforthegroup, especially for01,who underestimates thesmaller anglesandis moreaccurate or overestimates onthelarger. Thesecond 0 tendstounderestimate throughout. Allthe curves, when plotted, areclose tolinearity, andtheparticular pattern makes little difference. TheSDsareofthesame order asforthegroup. Discussion

Kinetic depth effect andmemory Wallach believes thatthekinetic depth ofpastexperience onpresent momentary experience (15,p. 364).He argues thatsince anysingle member ofthesequence looks flatinisolation, thepresent member hasdepth onlybecause thememory traces ofpast members enterintothepresent perception. Heassumes thatonlythe present member ofthesequence canbea stimulus fortheeye.Thisisa perfectly logical extension oftheclassical theory which strictly separates effect ofwhatwecalla transformation sequence mustbeduetosomeeffect

traces andstimuliasdeterminers ofperception (orbehavior). Butdoes itnotreduce thetheory toanabsurdity? Doesa stimulus lastfora second,

a millisecond, or a microsecond? Andwhataboutthedoctrine thata

stimulus isalways achange? Isitnottheoretically preferable tosuppose

thata transformation isa stimulus initsownright,justasa nontransforma-

I

!9 o( 3 ~ uosq !9 ofo(

uos

Perceptionof Rigid Motion

259

tionisa stimulus? Or,stillbetter, thatsequence, aswellaspattern, isa

variable ofstimulation? Isitnotbetter totaketheriskthattraces might vanish from psychological theory thantheriskthatthestimulus might vanish? Perhaps theaddiction ofGestalt theorists totheconcept oftraces iswhathasprevented someofthemfromstudying temporal forms as effectively as theyhavespatialforms. Wallach hasconvincingly demonstrated (andthepresent observations

confirm it)thatamotionless nonsense pattern ofacertain sortona shadow

screen which attheoutset usually looks flatwillusually lookdeepafter0

hasseen ifintransformation. This result may well beconsidered agenuine effect ofmemory, oratleast ofrecognition. Butitisfarfrom proving that

memory tracesarenecessary fortheperception ofrigidmotion. Ifmight provenot an effectof memoryon depthperception but an effectof

memory onthedepthinterpretation ofanambiguous staticpicture. It is possible thattheroleoflearning inspace perception isquitea different

affair, consisting notoftheenriching ofbare stimuli butthediscriminating

of rich stimuli.

The anchoring ofa transformation sequence andtheidentifying oftheobjectA motionless pattern ofsufficient irregularity appears intheplaneofthe

screenevenwhenits shadow-making patternis slantedto the screen.

What, then,would happen tothejudgment ofchange ofslantiftheshadow

casterwererotatedbetween,say,300and600insteadof between00and

60?Would 0 thenseethepattern asdeparting fromandreturning tothe frontal parallel plane? Istheneatcorrespondence between length oftransformation sequence andchange ofslantshown inFig.15.2destroyed in thesecircumstances? These questions leadintofundamental problems of psychophysical scaling andshape constancy; complete answers cannot be given, butenough evidence hasbeenobtained tobeworth reporting here. If the30 to 600cycleis presented to a naive0 witha rectilinear or

regular pattern, judgments aremade withsome accuracy. Theapparent surface doesnotbegin andendintheparallel plane. Ifthecycle ispresented withanirregular pattern, however, thenaive 0 reports a large

change ofslantwhich seems difficult toestimate, andtheobject does seem

to returnto theparallel plane.Thecrucial experience seemsto behisfirst

viewofthemotionless pattern. Heseesanobject buthisfirstimpression of

it isanobject thinner thantheshadow-casting pattern. Allobservations sofar suggestthatif he is thusledto identify at theoutsetan objectof the

wrongshape, thejudgments ofchange ofslantwillpersistently beoutof scale. Hisscale ofslantwillbestretched anddisplaced, asifwere, untilhe

is giventheopportunity ofanchoring if at 00and90 which,at thesame time,enables himto identify therigidshapeoftheobject.

260

J. J. Gibson & E. J. Gibson

Thesizeanddistanceof the virtualobjectareneverdetermined by the

optical stimulus in ourexperiment. Similarly, theslantandshapeofthe

virtualobjectmaybe misdetermined at the outsetby showingit, even

briefly, asa continuous nontransformationa staticfrontal pattern. This

establishes a falseshapeconstancy for the phenomenal object.Whenit rotates,anomalies of spaceperception willoccurof the sortdemonstrated

strikingly by Ameswiththerotatingtrapezoidal window(I).Ifcareis takenbyE,however, to avoidtheprocedure above,theevidence indicates thatan intermediate transformation cyclecanbe correctlyjudgedfromthe

outset,whenever thelengthofthetransformation sequence is sufficient. If a naive0 is firstshownan anchoredtransformation sequencefrom0 to

90, orevenifheisfirstshownthemotionless patternat 45 butistoldnot

to assumethat the objectis necessarily in the planeof the picture,thena 30 to 60 cycleis judgedapproximately as such.Thetentativeconclusion is that a motion consistingof a perspectivetransformationsequence

candetermineboth a definiterigidshapeand a definitechangeof slant

in perception, for a whollyunfamiliar object,withoutthe needof any presumption whatever abouttheprobable shapeof theobjectbasedon memory (2).

Itsoundsveryparadoxical toassertthata change offormofthestimulus canyielda constant formwithchange ofslantinthepercept. Theparadox probably arisesbecause twodifferent meanings of thewordform are employed (5),thefirstbeinggeometrical andthesecond substantial. Inany event,theassertion isinaccurate sincetheevidence nowsuggests thatthe formofthechange offormofthestimulus iswhatdetermines theperception of rigid motion.

Iftheimpression ofsurface rigidity invisual perception canbeaccounted for,theconstancy of shapeofobjects is explained at thesametime.The faceof a uniquesolidobjectis ordinarily givennot as a formbut as a unique family oftransformations to theeye.Thedifference between one solidobjectandanother, incontrast, isnotgivenasa relation ofperspective transformation, noristhedifference between anearlieranda laterstateof a physically changing object.Theperceptual problem oftherecognizing or identifying ofunchanging objects bytheirshapehasto beapproached in the light of these facts. Summary

Continuous perspective transformations ofvaryinglengthwerepresented in2-sec.

cycles toeach0 onthevisibly flatsurface ofa translucent screen. Judgments of the amountof changeof slantof the apparently rigidobjectwerein good correspondence withthelength ofthetransformation sequence, without dependingonthekindofpattern which carried thetransformation. Thepatterns differed

Perceptionof RigidMotion

261

withrespect toregularity vs.irregularity andform vs.texture. Regularity may

havehada small effect onthevariability ofjudgments buttexturedness didnot.

As a control,the samepatternswerepresented motionless at the endof a

transformation sequence. Ingeneral theirregular patterns appeared tobeinthe frontal plane butaltered inshape; theregular pattern appeared atsome degree of slant, butthejudgments were notaccurate. Evidently impressions ofchanging slantareprecise whereas corresponding impressions ofunchanging slantareambiguous orweak. Rectilinear contours andalignments seem toprovide some basis forimpressions ofunchanging slant. Asequence ofperspective transformations, on

theotherhand, seems toyieldanimpression ofchanging slantwhether ornotsuch

regularityis present.

Forirregular unfamiliar patterns there wasevidence tosuggest thattheperceivperceiving ofchange ofslantoftheobject. Misidentification oftheshape atthe ingof therigidshapeof thevirtualobjectis intimately connected withthe

outsetwasaccompanied byanomalies intheperception ofslant.

Theeyeappears tobeverysensitive toacontinuous perspective transformation

intheoptic array. Psychophysical experiments arepossible iftheparameters ofthis stimulus are isolated and controlled. Notes

1.Thisexperiment wasreported bythefirstauthor aspartofanaddress entitled Stimulation

andPerception delivered asretiring President oftheDivision ofExperimental Psychology,

APA,in September, 1955.Theworkwassupported in partby theOffice of Naval

Research under Contract NONR 401(14) withCornell University. Anearly form ofthe apparatus tobedescribed wasconstructed, andpreliminary experiments were performed byH.R.Cort.Thewriters arealsoobligated toDr.0. W.Smith forideas andassistance. 2.isThe 5% level is probably not acceptable here, since some inhomogeneity ofvariance evident in the data. 3.An0 whohadbecome familiar withthefrontal aspect oftheirregular formortexture (suchaseitherE)couldseeanother aspect asslanted. Soalsocouldan0 whohad

persistently observed these patterns undergoing continuous transformations. Presumably these Oshadlearned toidentify thesurface, i.e., torecognize apreviously unfamiliar object in nonfrontal aspects. This, we believe, is Wallachs memory effect in the perception of tridimensionalforms(15). 4.Inthestudy referred to,thecorrelate ofslant wassaidloosely tobea gradient of texture density. Thecorrelate might aswell beexpressed asanunequal density along two meridians, ortheratioofthese densities, oraspecial sortofcompression oftexture. These areallcomprehended inthegeometrical notion ofaperspective transformation.

5. Mr. John Hay kindlyobtainedthese data.

References

I.Ames, A.,Jr.Visual perception andtherotating trapezoidal window. Psychol. Monogr., 1951, 65, No. 7 (Whole No. 324). 2.Beck, J.,&Gibson, J.The relation ofapparent toapparent slant intheperception of objects.]. exp.J. Psycho!., 1955,50, 125133.shape

262

J. J. Gibson &: E. J. Gibson

3. Fisichelli , V. R. Effect of rotational axisanddimensional variations onthereversals of apparent movement inLissajous figures . Amer .]. Psychol ., 1946 , 59, 669- 675. 4. Gibson , J. J. Theperception ofvisual surfaces . Amer . ]. Psycho I., 1950 , 63, 367-384. 5. Gibson , J. J. Whatisaform ?Psycho I. Rev ., 1951 , 58, 403-412. 6. Gibson , J. J. Optical motions andtransformations asstimuli forvisual perception . (Motion picture film). State College , Pa.: Psychol . Cinema Register , 1955 . 7. Gibson , J. J. Optical motions andtransformations asstimuli forvisual perception . Psychol . Rev ., inpress . 8. Gibson , J. J., alum,P., & Rosenblatt , F. Parallax andperspective during aircraft landings . Amer . ]. Psychol ., 1955 , 68, 372- 385. 9. Johannson , G. Configurations inevent perception . Uppsala : Almqvist andWiksell , 1950 . 10.Langdon , J. Theperception ofachanging shape . Quart .J. expo Psychol ., 1951 ,3, 157 - 165 . 11.McNemar , Q. Psychological statistics . (2nded.) NewYork:Wiley , 1955 . 12.Metzger , W. Tiefenerscheinungen inoptischen Bewegungsfeldem . Psycho I. Forsch ., 1935 , 20, 195- 260. 13. Smi~h, W. A. Sensitivity to apparent movement in depthasa functionof "propertyof movement ." J. expo Psycho /., 1951 , 42, 143- 152. 14 Wallach . H., & O'Connell , D. N. Thekineticdeptheffect . J. expo Psycho /., 1953 , 45, 205- 217. 15. Wallach , H., O'Connell , D. N., & Neisser , U. Thememory effectof visualperception of three -dimensional form.J. expo Psycho /., 1953 , 45, 360- 368. 16. Wertheimer , M. Untersuchungen zurLehrevanderGestalt . II. Psycho /. Forsch ., 1923 , 4, 301- 350.

16

Motion

Parallax

as

a

Determinant

of

Perceived

Depth

Eleanor H

J .

.

_ 4

The

Gibson

-

paper

for

layout of us

to

specify from

{ if

]

.

.

Gibson

,

the

of

the

problem

O

.

W

.

Smith

,

or

. Helmholtz

projected

the

in

it

motion of

the

occurs were

and

.

previous

, the

some

of the

Note

the

The

indeed

to

be

questions

deeper the

-

.

experiment

the

exactly

informa

does

conditions

really

how

world

as another

movement

was

up

real

of

to

pattern

of

a

met

,

to

be

problem

of

information

following

eye

in

( p . 43

.

It

seems

that than asserted the

field

corresponding

transformation

of is

not

information

of

can paragraph

elements

register

pick

their the

will

a be

, and

immediately about

up

is

the

Gestalt

the

mutual

directions

to

.

changes

of

displacements difference directly in this

in

or

between related

space

.

may

be

evident objects

is

field the

or

absolute

that

a

As

responds

objects pattern

.

positions

eye

to

in

them to

than

rather have

the

to

eye

suppose

elements

the

can

, rather

,

between

eye the

better

two

specifying the

to

velocities

an

what

easier

sensitive

for

angular

pattern

might

distance

primarily

their

is

involve

is

,

change

velocities

is

this

solution

and

,

,

velocities

eye

in

from

of front

differential

exist

pick

role in

made that

which

how

elements

separation

velocities

to

can

positions

accordingly

tive

things

emphasized

of

angular

film }

awaited eye

understanding

have

separation

method

that

with of

difference

extent

the of

trouble

,

the

convinced

the

surface

:

difficulty

And

one

correspondences

information

theorists

be

of

for in

to . The

metric

arrangements introduction

The

investigation time

phenomenon

any But

further this

needed

surfaces

whether

kind

a ,

this we

separate

explored what

was

features

demonstration convinced

and

James

.. ..

following

tion

serve

,

Flock

what an

array

the to

Such

the rela

what the of

-

the best light

) .

Thisstatement soundsratherlike Marr's muchlaterpronouncements about computation (Marr 1982 ). Butthequestion herewaswhether differentialvelociJournalof Experimental Psychology , 1959, 58, 40- 51.

264

E. J. Gibson , J. J. Gibson , O. W. Smith , & H. Flock

ties carry information for separation in depthand whethertheycanspecify absolute , aswellasrelative , amounts of separation . Theshadow caster setupwasusedagain , thistimewith translucent surfaces in front of thesubject . Theactualsurfaces wereplexiglass andtheywererandomly textured bysprinklingthemwith a cleaning powder . Whentheyweremotionless , theappearance wasof onesurface . As soonasany motionwasintroduced , the surfaces sprangapart, butextentof separation wasambiguous . Notethesugges tion in thediscussion that the information for separation might bedisruption of adjacent orderbetween the two setsof texture . Thesuggestion foreshadows J. J. Gibson 's laterworkingoutofaccretion anddeletion at anedge asinformation for specification of separation . Whatwasemphasized in bothcases is change in opticalstructure , a discovery that hasbeenmadeby othersseveraltimessince then(Ullman1979 ). Thesecond experiment presented a gradientof velocitydifferences ratherthan just twoseparated surfaces , attempting to create a continuous flowfieldof velocity change , on thehypothesis that themetricambiguityof a difference in just two surfaces couldbeovercome by a smoothflow of changing velocityratios. There wasgoodindicationof a correspondence between differential velocityratiosand perceived slantin this case . Theseresultsare interesting in the light of recent research findingson thesamequestion (Anderson1989 ) that undersomecircumstances positiveresults(metriccorrespondence withoutsmoothcontinuityof change ) areobtained . It ispossible thatcontinuityof change becomes afactoronly whenno informationis availablefor an opaque surface . I founda comment writtenin themarginoff . J. Gibson 's reprintof thispaper , "Thetransparencies shouldhavebeen emphasized ." Thesetupwastoomuchlikeemptyspace , perhaps . I conducted mostof thisresearch duringa coupleof summermonthswhenI wastheonlyseniorinvestigator in thelaboratory , my husband andothershaving departed for Brussels wherean international congress wasgoingon. Duringa recentmove , I foundtwo letterswrittento my husband givingdetailsaboutthe earlyprogress of theexperiments : I wasverydiscouraged abouttheexperiment yesterday . 1ranHowardFlock andhecouldn 't discriminate anythingexcept thesmallest separation (1/2 in.) from all therest. WhenI plottedhisjudgments theywerescattered anywhereat all. Then1 friedhim wifh rating(1, 2, 3, 4, 5). Still - worse ~. But todayI havebeengiving him training , demonstrating with an arbitrary scaleandcorrecting him. After25corrected judgments hewasnearlyperfect (didnotconfuse 10different settings ). Buthehaddeliberately attended to the ratiosof thetwomotionspeeds of thepatterns andattached thescaleto the ratios.I amgoingto keeprunninghimandseeif thissticks , andwhether it evergetsconverted to a depthexperience . If it did, it wouldbea realcaseof learning todiscriminate a complex stimulusvariable andgettinga phenome nological change alongwith it (July24, 1957 ).

MotionParallaxin Perceived Depth

265

Laterreportsweremorecheerful:

Theexperiment isgoing fine, really. I have3kinds ofsubjects. Thesummer school people arenttooeager, soIonly askthem toDescribe what yousee. They allsaytwolayers orlevels.Then Iaskhowfarapart?One said 5ft today, andonesaid4 in.Thisiswiththelargest separation. Then I show thesmallest. Some cantseetheseparationthen 1increase ittillthey do.Thesecond class ofSsareasked tomake 20judgments, random order (takes about40mm.). Thethirdclass ofSsareFlock andHay,whoare

practising dailywithcorrection, for10trialsandthennextdaywithout (July 30, 1957).

Thelastmentioned subjects, senior graduate students Howard Flock andJohn Hay,were incorporated intoexperiment IIIofthepaper, a study oftheeffect of giving varied amounts ofinformation about theactual setup tothesubjects. If some sortofHelmholtzian inference principle were toexplain howvelocity differences functioned asa cue todepth, wemight have found aneffect ofconversion ofbidimensional impressions intoexperiences ofdepthreferred tosometimes as

percept-percept coupling (Epstein 1982). Butthisdidnothappen. Thetwomost

practiced subjects made inferences about depth, tobesure, onthebasis ofperceived velocity ratios. Butthey never reported aperceptual change inthedepth perceived. Thisexperiment discouraged anytendency I hadtogoalong witha Helmholtzian ora Brunswikian theory ofperceptual learning. However, itdidnotdiscourage my interest inperceptual learning, andmypursuit ofa viable, adequate theory of perceptual learning took offinnewdirections, ascontinued inpartIV . Motionparallaxis the opticalchangeof the visualfieldof an observer

which results froma change ofhisviewing position. 1 Itisoftendefined as

thesetofapparent motions ofstationary objects which ariseduring locomotion. Psychologists assert thatitisacue forperceiving thedepth oftheobjects, buttheoptical factofmotion parallax must bedistinguished from itscapacity toinduce perceptions. Ithasnotbeenexperimentally demonstrated thatmotions inthefieldofviewwillactually yieldcorresponding judgments ofdepth. Thisisapurely psychological problem. The optics of motion parallax, on the other hand, is a problem for geometry and ecology. Recently, thesuggestion hasbeenmade thata continuous gradient of motions in thefieldofviewwillinduce theperception ofslant-depth

(J.J.Gibson, Olum, &Rosenblatt, 1955) inasmuch astheperception of depthisintimately connected withtheperception ofsurfaces (J.J.Gibson, 1950). Thisstatement alsoneedsexperimental test.Thepurpose ofthe present studyistoinvestigate whatkinds ofmotion inthelightentering

266

E. J. Gibson , J. J. Gibson , O. W. Smith , &: H. Flock

an eye do in fact consistently arousecertain judgments of depth, and what do not .

The experiments must be carried out with artificial motions in a field of view

rather than those obtained

in a natural

environment

if we wish

to study the effect of motion parallaxin isolation from other cuesor stimuli for depth. The variables of size, density, linear perspective, differential blur, and binocular parallax should be eliminated or so reduced as to be ineffective in the array of light entering as eye. A method of achieving this result has been devised, and a suitablecontrol employed.

The experimentalmethod shouldalso precludeactualmovementor locomotion of o . If the cue of motion parallax is so defined as to require activehead movement or locomotion, proprioceptive and vestibular stimulation is also present. This definition is unjustified , since passive locomotion

in trains and airplanesshould be admitted as circumstanceswhen motion parallax occurs. Certain patterns of motion in the field of view of 0 do induceimpressionsof being moved through spaceif we acceptas evidence the illusions of locomotion obtained in viewing a panoramic motion pic-

ture, or in a training device for simulating aerial flight . Perception of absolutedistanceand of relativedepth The apparent displacements of the sensations of objects are said to be cues for perceptions of

their depth. What kind of depth? The question arises whether their distancesfrom the perceivercan be judged or whether they will only appear to be separatedin the third dimension. Helmholtz, in his description of motion parallax, assertedboth hypotheses. On one page he describedthe appearanceof objects IIgliding past us" as we walk through the countryside, and asserted that IIevidently under these circumstances the apparent

angular velocities of objects in the field of view will be inversely proportional to their real distancesaway; and consequentlysafe conclusionscan be drawn as to the real distance of the body from its apparent angular velocity " (Helmholtz , 1925, p. 295). On the next page he described the

appearanceof an indistinguishabletangle of foliage and branchesin a thick woods as a man stands motionless , but noted that " the moment he begins

to move forward everything disentanglesitself and immediatelyhe gets an apperception of the material contents of the woods and their relations to

eachother in space, just as if he were looking at a good stereoscopicview of it " (Helmholtz

, 1925 p . 296 ).

In the first quotation Helmholtz says that angular velocity is a cue for the perception of absolute distance. In the second, he suggests that a

differencein angular velocity is a cue for the perception of separationin depth, or of relative distanceonly. Thesetwo hypothesesare by no means the same, and they should be consideredseparatelyand tested separately.

We will be concerned hereprimarilywith the second .

MotionParallaxin PerceivedDepth

267

Two-velocity motion parallax andflow-velocity motion parallax Although

motion parallax hasbeensaidto applyto thewholearrayofobjects inan

environment anda largearrayofapparent motions inthefield, theexperi-

mentsperformed haveinthepastbeenconfined to twoobjects andtwo

velocities in a restricted field of view.

Bourdon (1902) reported experiments inwhich 0 looked withoneeye

at a pairofluminous spotsina darkcorridor. Thesources wereat different distances butthespotswereofthesameangular size.Whentheheadwas

fixed withabiting-board, 0 couldnotjudge atallaccurately which light

wasthenearer,butwiththeslightestmovement of theheadfromsideto sideit waseasyto judgetherelative depthofthetwo.Buttheabsolute

distanceof neitherlightwasdetectable.

Tschermak-Seysenegg (1939) improved onthisarrangement withwhat hecalled a parallactoscope, byanalogy withthestereoscope. Hedefined motion parallax as arising frommovement eitherofa groupof visible

objects ontheonehand, ortheposition of0s eyeontheother, emphasiz-

ingtherelativity of the situation. But,he studiedonlythe detection of

depthoftwoobjects withvoluntary headmovement. Hisapparatus wasa binocular depthperception. It permitted 0 to moveoneeyefromsideto side,witha sliding headrest, soasto obtain successive impressions equivalent tothesimultaneous impressions obtained withbotheyesopen. The modification ofthefamiliar two-pins setupusedto obtainthethreshold for

averageerrorof equatingthedistanceoftheverticalwireswassmallunder

these conditions, although notassmall astheerrorwithbotheyesopen. When onlyoneeyewithafixed headwasused, theerror wasverylarge. Graham, Baker, Hecht,andLloyd(1948; seealsoGraham, 1951)obtained thethreshold forseparation indepth oftwoneedles pointing toward

oneanother, as seenon a uniform fieldthrougha window. Theneedles

moved fromsideto sideona common carriage. Theyappeared to be

aligned at thecenteroftheirmotioncycleandoffsetat theextremes ofthe

cycle unless theadjustable needle hadbeensetintothesame frontal plane asthefixed needle. Graham thuseliminated forthefirsttimeinthistype ofexperiment theadditional sensory information produced byvoluntary head movement. Moreexactly, whatGrahamobtained wasthejustnoticeable difference

between twoangular velocities ina field ofview, under probably optimal conditions. Thethreshold wasextremely lowabout30sec.ofarcper second oftime. Itisnotable thatthereports ofwhatOsperceived, how-

ever,werenot unanimous. Somesawthe separation in depthas such; othersperceived eitherthe difference in velocityof the twoneedlesor

noticed thechange ofalignment oroffset oftheneedles. Although the

latterimpressions maybecuesfortheformer, theexperiment wasnotcon-

cerned withtheeffectiveness ofsuch cuesforproducing depth impressions.

268

F.J. Gibson,J.J. Gibson,0. W.Smith,& H.Flock

Somewhat: later,a caseof motionparallaxdifferent fromthetwo-velocity casewasdefinedmathematically by Gibson,Olum,and Rosenblatt(1955).

Thisisthearrayofangular velocities oftheoptical elements projected from a surface toa moving stationpoint.Thereisa flowofvelocities ratherthan

a pairofvelocities in suchanarray,andthephenomenon wasnamed motionperspective to distinguish it frommotion parallax asit hadbeen studied up to that time.

Forthestudyoftheperceptions induced byflow-velocity ina fieldof

view,including gradients ofvelocity, skewmotions, andtransformations, a different sortofapparatus isrequired fromthatpreviously employed. The two-velocity experimenters useda pairofrealobjectsat realdistances to produce theoptical motions. Morefreedom isachieved byusinga projectionscreenor someotheropticalmeansto producethem.Theexperiments to be describedusedshadowson a translucent: screen.Accommodation is thereby controlled.

Severalexploratory experiments havebeenpublished on flowingmotion.Theyareof varioustypes,andtheyhavebeenproduced in various

ways. 1.J.Gibson andCarel (1952) attempted toinduce theperception of a receding surface ina darkroom witha bankofluminous pointswhich

carried a gradient ofvelocities. Thisstimulus failedtoarousetheperception of a surface,however,and the depthjudgmentswereambiguous. 0. W. Smithand P. C. Smith(1957)investigatedthe perceptionof convexityor

curvatureof a texturedsurfacewithvariouscombinations of depthcues,

including theflow-velocity typeof motionparallax. Although motionin

the fieldcontributedto the judgmentof convexity,in no casedidmotion causea surfaceotherwise judgedas flatto be judgedas curved.Hochberg and0. W. Smith(1955)studiedthe perception of depthinducedby the

centrifugal flowof luminous patternelements in thedark,theexpansion phenomenon. J.J.Gibson andE.J.Gibson (1957) investigated theperceptionoftherigidrotationofanapparent surface elicited by thecontinuous perspective transformation ofregular andirregular patterns orforms. Theseexperiments differed in thestructure of theopticarrayusedto carry themotionin question, andtheyalsodiffered in thedegreeto whichperceptions ofspacewerearoused. Theyledto thechoice ofthe kindof randomtextureemployedin the presentexperiments, whichis

intendedto yieldthe experienceof a planesurface. What is now neededis an experimentalcomparisonof the judgments obtainedwith twovelocitiesin a fieldof viewand thoseobtainedwith

manyvelocities ina fieldofview.Although noclearlinecanbedrawn

betweenthem,the two-velocity type of motionparallaxappliesto the

problem ofperceiving a groupofobjects inotherwise empty space, while theflow-velocity typeof motionparallax applies to theperceiving of a background surface suchasa wall(orsubstratum). Thesearenotthesame

MotionParallax in Perceived Depth 269 problemfor perceptioneven though it may be difficult to distinguish sharplybetweentheir respectivekindsof stimulation . Opticalgeometryof motionparallax The environmentalsituationwhich leadsto an array of differentmotionsin a visualfield shouldbe defined more carefully. This is the optical geometryof motion parallax , as distinguishedfrom the visualappearance of motion parallax . Graham(1951, pp. 878ff.) hasgiventhe geometryof certainspecialcasesof this situation. J. J. Gibsonet al. (1955) haveanalysedthe caseof an extendedsurfacesuch asthe ground. What will be discussed hereis the caseof an environment of discreteobjects. Whenlight raysfrom permanent objectsof anenvironmentconvergeto a point, theyconstitutewhatmaybecalledanopticarray, andtheelements of this arrayconstitutea pattern. An eye or a cameraat the stationpoint can registerthis pattern of luminouselements . If the point moves, the patternis alteredin a way whichdependson both the displacement of the point andthelayoutof theobjects.How theeyerespondsto this alteration of patternis our problem. Thefirst questionis how to specifymathematically thechangeof pattern in a way thatis relevantfor vision. By choosinga coordinatesystemfor the array, one can specify the absoluteposition of eachelementand the displacement of eachelementper unit of time, that is, its absoluteangular velocity. It would then be true in a certainsense , as Helmholtz(1925, p. 295) said, that "the apparentangularvelocitiesof objectsin the field of view will be inverselyproportionalto their realdistances away." Butmore exactly, it wouldbetrueonly if thelinearvelocityof the stationpoint were constant(J. J. Gibsonet al., 1955). A given angularvelocity is a cuefor distance , or permitsa "safeconclusion " aboutdistance , only if the speed anddirectionof one's locomotionis known. Thetroublewith positionsandangularvelocitiesof elements in a fieldis the difficulty of understandinghow an eye can registerthem. As the Gestalttheoristshaveemphasized , what the eye seemsto pick up is the mutualseparationof elements , their pattern, ratherthan their positionor directions . And, accordingly , it is easierto supposethat the eye responds to changesof separation , or changeof pattern, rather than to absolute displacements or velocities . Helmholtzmight better haveassertedthat a differencebetweenthe angularvelocitiesof two elementsin the field will be directlyrelatedto the differencein distancebetweenthe corresponding objectsin space . Suchrelativevelocitiesinvolve a transformationof pattern, and this may be what the eye is primarily sensitiveto. It is not immediatelyevidentwhat the bestmethodis for specifyingthe information aboutobjectsin an arrayof light projectedto an eye.

270

E.J. Gibson,J.J. Gibson,0. W.Smith,& H.Flock

Butonefactshouldbe clear.Onlyif thereis an eyeat the pointof

projection andonlyif it is sensitive to themotions in theopticarray, relativeor absolute,doesa psychological questionarise.Willthe possessor

oftheeyeseemerely thechange ofpattern ofthearray? Orwillhesee moving objects inthefieldofview? Orwillheseestationary objects at differentdistances? In orderto showthatmotionparallaxis effective for

theperception ofdepthitmustbedemonstrated experimentally thatdifferentialmotionsinanarrayoflightto aneyewillyielddifferential judgments

ofdepth. Andthearrayshould besuchthatwhenthemotion iseliminated thejudgments ofdepthwillcease, foronlythenwillmotion parallax have been isolated from other cues for depth.

Experiment I: MotionParallaxwithTwoVelocities Problem and Method

Thetwo-velocity experiments wererepeated with(a)twospotsina fieldto carry themotions, and(b)twosuperimposed textures filling thefieldto carrythem.In bothcasesthevelocity difference wastakento be theessential cueforpossible judgments of depth,nottheabsolute velocities. In thisexperiment, reportswere

obtained fora largevelocity difference, a small velocity difference, andnovelocity difference, that is,a motionless field.Thelastwasa control.

Apparatus andstimuliThelightentering 0s eyecamefromthetranslucent screen

ofapointsource shadow-projector (J.J.Gibson, 1951; J.J.Gibson &E.J.Gibson, 1957).Hesawonlya luminous rectangular fieldin whichdarkcircles or textures

could bemadetoappear andtomove. These wereactually theshadows ofopaque substances attachedto a transparent mountbehindthe screen.Thiswasa large

sheetofglassorplastic whose edgeswerenevervisible. Differential translatory velocity oftheshadows wasproduced withtwomounts, onebehind theother, whichcouldbe madeto moveparallel to thescreenon a common carriage. The

arrayoflightto theeyewassimply thereverse ofthearrayprojected to the screen, sincetheeyeandthepointsourceweresymmetrically locatedequidistant fromthescreen (Fig.16.1). Thewindow was32.2x 36 cm.at a distance of 126cm.fromtheeye,subtending anarray14.5highand16.20 wide.Thewindow

was viewedthroughan apertureby a seated0.

Thecarriage which borethetwomounts rolled silently ontracks andcould be

pulled from sidetosidethrough anexcursion of45cm.Itwasoperated byhand

toproduce amotion cycle inabout8sec.Asmall shutter closetothepointsource enabled Etoeliminate theshadow between trials,leaving thescreenilluminated by diffuse light.

Thetwoadjacent spotsin thefieldwereproduced by attaching smallpaper

circlesto eachmount,at differentelevationsso that theirshadowsdid not pass

through oneanother astheymoved across thefield. Thefaster spotwasabove the

slower spot.Thediameter ofbothwas5.2,onepapercircle beingcompensated in size to match the shadow of the other.

MotionParallax in Perceived Depth 271 VARIABLE

STANDARD

MOUNT

MOUNT

APERTURE

EYE

POINT

TRANSLUCENT

WINDOW

Figure 16.1

Theshadow projector viewed from above. Ina unitoftime, theshadow ofa spotatthe center ofthestandard mount sweeps through a certain angle andthatofa corresponding spotonthevariable mountsweeps through a lesserangle,asshown. Thetwomounts roll onthesame carriage. Iftheyareclose together, there isnodifference inangular velocity,

butasthevariable mountis positioned farther fromthepointsource andcloser to the screen, theangular velocity ofitsshadow decreases. Withthisapparatus, itcandecrease to aboutonehalfoftheangular velocity ofthestandard. Dytrigonometry, theratioofthe lesser(V)to thegreater(S)angular velocity is equalto theinverse of theratioof the distances oftheirrespective mounts from thepoint source. Inthediagram above, itisabout 0.7.

Thesuperimposed random textures were produced byatechnique ofsprinkling talcum powder overthesurfaces ofthetwotransparent mounts. Thisyields an

opticaltexturewithindefinite contours andindefinite elements. Whenthetwo weresuperimposed butmotionless, theyconstituted a singletexture withnocue

forsuperposition, andgavetheappearance ofa single surface, something likethat of a cloud. This apparent surface filled the whole window and appeared atan indefinite distance from 0. Asnoted, thetwoangular velocities assuchwerenotuniform, decreasing to

zeroat eitherendofa motion cycle, andchanging direction alternately. Minor variations invelocity alsooccurred asa consequence ofmoving thecarriage by

hand. Theindependent variable ofthisexperiment wasthedifference invelocity between thetwoshadows. It wasexpressed as theratioof theslower (the variable) velocity tothatofthefaster(thestandard) velocity, orV/S.

Procedure Each 0 wasseated attheapparatus, asked toapply hispreferred eyeto

theaperture, andinstructed simply to describewhathesawinthewindow.He wasfirstpresented witha motionless fieldforas longas he neededto makea

report, which wasrecorded. Hewasthenpresented withcontinuous cycles of motion atthemaximum velocity difference (V/S= .51)untilhisreportwascompleted. Finally hewasgiventheminimum velocity-difference (V/S= .97).TheE

made nocomment atanytime,since wholly spontaneous reports weredesired.

272

E.1.Gibson,J. J. Gibson,0. W.Smith,& H. Flock

Theorder ofpresentation wasintended tominimize theeffect ofsuggestion onthe perceiving of depth.

Agroup of26Oswentthrough thisprocedure withthespotfield andanother groupof46withthetextured field. Formal judgments andanswers toquestions wereobtained afterwards fromsomeOs,whichwillbedescribed whenrelevant. Theywererequested inthetermsusedspontaneously bythe0. Results

ThewordsusedbytheOsto describe whattheysawvariedwidely, and

the effortto identifythingswasreminiscent of descriptions of cognitive inference (Vernon, 1957).Butthe reportscouldlaterbe classified easily

withrespect to depthordistance. Themotionless textured fieldwasunanimously reported tobea single surface without anydifference indepth. Themotionless spots,however, werereported at different distances by4

of the26 Os.Thespots,therefore, didnotwhollysatisfytherequirement

thatimpressions ofdepth beabsent intheabsence ofmotion, although the combined textures did.

A largevelocity difference (.51)forthetextured fieldalways gavea perception oftwosurfaces separated indepth, asevidenced bythereports of all46 Os.Forthe spotfield,thereportswerenot unanimous, but22 out of 26 Os diddescribea difference in depthof the two objects.

Thesmallvelocitydifference (.97)wasevidently closeto thethreshold.

Noneof the Osreported twoseparated surfaces forthetextured field, andonly7 outof26reported different distances forthespots. The directionof the differencein depth reportedwas not unanimous

foreitherthespotsor thetextures. Insofar as two-velocity parallax is a

reliable andeffective indicator ofrelativedepth,thefastervelocityshould

correspond to thenearer object orsurface. But7 outof26Ossawthe slower spotasthenearer object instead ofthereverse, and10outof46Os sawtheslowertextureas thenearersurface. Somedegreeof ambiguity as

to thedepth-difference is alsoindicated by thefactthat7 Osreported spontaneous reversals ofthefront-back relationship between thetwosurfaces at one time or another.

Theamountofseparation indepthbetweenthetwoobjectsor thetwo surfaces wasestimated on formalrequestby a sampleof theseOs,after

themainprocedure. Theestimates werehighlyvariable. Forthe spots

theyranged fromzeroto 5 ft.Forthetextures theyranged from2 in.

to 3 ft. SomeOswereunwilling to judge,saying,It dependson what it is, or It couldbeinfinity.Evidently theimpression ofhowfarapart theseentitieswerewasindefinite, as alsowastheimpression of howfar away they were.

Figure 16.2represents anidealtheoretical possibility ofwhattheOs mighthaveseeninthisexperiment, butit cannot beasserted thatthis,

Motion Parallaxin PerceivedDepth -

-

-

-

-

-

0

-

-

-

-

-

-

-

-

273

-

-

-

-

-

1_-

1_- - -

- -- - - - , - - - - - - - - - - - -

is whattheydidsee . Thereportsindicated thattheyperceived two things of somekind in somekind of spacebehindthe screen , but neitherthe direction in the third dimension was definite . nor the amountof theirseparation

Discussion Thesignificant result of thistwo-velocity experiment isnotsomuch the effect ofmotion ondepth perception asitseffect inseparating onesurface intotwo.Withthetextures , alIOssawasingle frontal surface when there wasnodifferential motion , andalIOssawtwofrontal surfaces when there wasasufficient degree ofdifferential motion . Thisseparation isnotwhatis ordinarily meant by depth , since it wasnotalways clearwhichsurface appeared infrontandwhich behind . Thephenomenon is similar to theIIdisentanglement " of foliage and branches whichHelmholtz notedwhenhebegan to movein thethick woods . Butthisis notthesame ashisIIapperception of depth ." The separation isprobably related toWertheimer 's(1923 ) demonstration thata group ofspots interspersed among others willbeunified bywhathecalled

274

E.1.Gibson,J.1.Gibson,0. W.Smith,& H. Flock

their common fate if they movedtogether.Other conditionsfor the

seeingofonethinginfrontofanother, fortransparency andsuperposition,

havebeendiscussedby Koffka(1935).Althoughthe phenomenonmaynot

seemrelevantforsomekindsofspaceperception it iscertainly relevantfor object perception.

Howcanthisresultbe explained? Insteadof appealing to a processof

organization, or a lawof commonfate, onemightlookforits basis

in thegeometry of the opticalstimulus. Geometry distinguishes between

(a)perspective transformation offorms, (b)topological transformations of forms,and(c)disruptions. Thesecorrespond roughlywiththedistinctions in

physics between rigidmotions ofbodies, elastic motions ofbodies, andthe motions ofbreaking, tearing, orsplitting. Inthisexperiment, themotionof one set of textural elementsrelativeto the other was a disruption,geomet-

ricallyspeaking. Whensufficiently differentvelocities wereimposedon them,the adjacent orderof elements in the textureswas destroyed.More

exactly, therewasa permutation ofthisorder.It wasa particular sortof permutation, to be sure,foreachof two setsof elements retainedan adjacent order,butthedisruption oforderasbetween thesesetsbrokethe original continuity. Andthisproduced theperception ofdifferent surfaces withseparation between. Thedetection bytheeyeofcontinuity orsolidity as compared withdiscontinuity, disruption, or separation, is probably a fundamentalkindof perception.The continuityof a singlesurfacein two

dimensions maybegivenby a staticopticaltexture.Butthecontinuity of a solidobjectinthreedimensions probably depends onthekindofoptical motionpresented to theeye.Perhaps it wasthislackof solidcontinuity or rigidconnectedness betweenthenearerandfarthersurface in ourexperimentwhichprevented theidealpossibility represented inFig.16.2 from having been realized.

Theearlierinvestigators ofmotionparallax werewillingto assumethat

aneyewassensitive to thestimulus ofmotion, buttheydidnotseemto realizethat differentialmotionnecessarilyentailsa changeofpattern.In our

experiment thereweremotions oftheelements relative tothewindow but there were alsomotionsof one set of elementsrelativeto the other.For

example, whenbothsetsofelements weremovingto theleft,relative to

the window,and one moved fasterthan the other. the slowerwas moving

to theright,relativeto the faster.Spontaneous reportsof thisappearance

weregivenby severalOs.Whyshouldnot a differential velocity be perceived justas directlyas the twocomponent velocities? Whentwo moving elements arefarapartinthefield,onemightsuppose theslower and the fastervelocitymighthaveto be comparedin orderto detecta difference betweenthem.Butwhenthe elementsare adjacentin the field the difference is givenby the changeof pattern.Permutation of orderis

onetypeofchange ofpattern. Inorderto studythesensitivity oftheeye

275

Motion Parallaxin PerceivedDepth

toform ,to change of form ,and forms ofchange ofform ,ataxonom of these variables isdesirable . tothe Experiment

II

Problem

If

and

a

flow

two

: Motion

- velocity

field field

was

modified

velocities

of

bottom

of

the

was

If

this

or

pattern

in

producing

motionless was

,

the

so

a

Flow

texture then

,

as

is , a

of

Velocities

an

of research

angle

size

and

of

a

which

, the a

texture

to

fast

field

by

source 450

( Fig

yield

elements

slanted

surface

.3 ) .

By

even

the

principle away

texture

the , were

in

been

when

).

not

'" '" POINT \.

SOURCE

'" '"

-

TRANSLUCENT WINDOW

~

CARRIAGE

TRACKS

WITH

a

of from

a

grid

was But

sharp

~

\.

on

effective

MOUNT

EYE

.

texture

1955

-

the

mount

slant

spaced

Gibson

hor

motionless

. This

have

a

to

paint

.

(J. J. the

be

screen

would

demonstrated

the top

spattering

apparent

will

already

the

the

16

,

which

also

regularly

spacing

composed

in

and

an

depth

apparatus

from

could

produced

of

should

has

, the

point

eye

composed of

eye slow

were the

the

the

before

screen

at

gradients

- transitions

to

of

comparison

from

. As

the

at

impression

previous

present

gradient

perceptions

this

varied

screen

been

consistent make

between

the

had

to

to

on

gradient

the the

brightness

as

interposed

the

arouse

order

elements

shadows

,

,

not

In

the

toward

reversal screen .

s07

, that

sheet

slanted

mirror the

field

The

transparent

does

do

izontal

Apparatus

with

Method

- velocity

described

Parallax

as ,

it the

276

F.J.Gibson,J.J.Gibson,0. W.Smith,&H.Flock

elementshad indefinitesizesand shapes,and the staticgradients,if present,were

not effective in producing a perception of slant.Thistexturewasnevertheless sufficient to producethe perceptionof a continuoussurface.

Asbefore,0 lookedthroughanaperture whichprevented hisseeingeitherthe

edgesofthewindow oranyotherpartoftheapparatus. Thefieldwas82 wide by52 high.Theeyeandthepointsource wereeach75cm.fromthewindow. ProcedureAllOs werenaive.Eachwastoldthatwhenhe appliedhiseye to the

aperture hewould seea grayfieldofview, andwasasked toreportwhathesaw. Theyweredivided intotwogroups, oneofwhich waspresented withthemoving texturewithouteverseeingit static(GroupI),and the otherwiththe moving textureafterhavingseenit static(GroupII).Theirspontaneous descriptions were recordedandlaterclassified. Questionswereaskedonlyafterthesereports,andin the same terminology used by 0. Results

Nineteenout of 21 Os(GroupI)reporteda rigidmovingplanesurfaceof somekindslantingawayfromthemat thetopof thefield.Theinclination of the surfacewasestimatedwithoutdifficulty, and theyjudgedit to be receding upward. Ofthe2 remaining Os,onesreportwasuninterpretable, and the others was of a surfaceperpendicular to his lineof sight. With the statictexture,25 Os out of 28 (GroupII)reportedsomething whichcouldbe classedas surface-like, but whichwas in no caseslanted

backward intospace.Oftheremaining 3 Os,2 sawa surface whoselower

partwasperpendicular butwhoseupperpartwasslanted back,andone sawthe surfaceslantingbackintospace.Whenthese28 Os werelater

presented withthemoving field, 26sawitunambiguously tobea receding rigidsurfaceat a fixedinclination. Of the remaining two,one sawit perpendicular to thelineofsight;theothersreportwasuninterpretable. Estimatesof the slant of the apparentmovingsurfacewere obtained fromeach0 in termsof degreesof inclination backwardfromthe frontal

plane. ForGroup I,the19judgments varied from20 to60 withamedian of 40. ForGroupII,the 26judgments variedfrom12 to 55 witha medianat 37 . Thetheoretical valuebasedon thegradientof velocities alonewouldbe 45. The mediansshow a constanterror of underestima-

tion.Thesurfaceneverappearedto be at 90, thatis,it neverlookedlikea groundon which0 mightbe standing.

Estimates wererequested of howfarawaytheapparent movingsurface

seemedto be from0, but thesejudgments,in contrastto those of slant,

couldnot as easilybe made.Forthe totalof 49 Os,theyvariedfrom 3 in. to 5 miles.Someof these Os reportedthat it was possibleto see

themselves movingwithrespectto the surfaceinsteadof the surface movingwith respectto them.

Motion

Parallax

in

Perceived

Depth

277

Discussion

These

results

velocities receding of

a

in

depth

respects put

it

skew

of

in

a certain

rigid

surface

another

that way

pattern were first

the

kind

of

were

for

, ( a ) the

ever

, was

nected

recession for

the

Experiment III Under Various

In

and

Exp

.

I,

has

visual

been

velocities

would

be

the

there about

the

to

limits

about

length

the

greatest

ments

of

a

experience the

first

absent

in .

, but

distance

in

reluctance

II depth

of

in

the

"

two

and

, as

more so

kinds

third

a

this they

of

than

far spatial

a level

. This

ground

- Pairs

-

( b ) the , how

may

was

and

percep and

situation

distance

one

described

dimension . Neither

of

this

separation

of

Exp one asked

of

of

in

of

be

-

con

induced

Judgments

of , in

occur ,

or

-

.

of

Depth

between

, would

be

. This

three

degrees

a

not

alter

the

the

Helmholtz

,

interpret

to

perceive

suggestion with

that

instructions

perceptions

tested

in

a

this

not

learned

,

was

of did

, the

objects to

prediction

of of

between

spirit

theory

the

expected

the

because

this

perception

explanation

continuity

had

. On

the

. An

however

distance distance

depth

induce

absence

not

depth

Exp

and

. III , using

. I. of to

of

reproduce

between

the

an

nearer

described what

was for

both

and , was

see , "

( V IS that

of the

=

or

adjustable

, initially

you

difference

suggestion

on

, the V / S ratio with it .

apparatus separation

the

did

in

not

consistency

explanation

previously " Describe

did

great

terms

of

velocity

.97 ) . The

amount

velocity

difference

element

, as

, the =

Os

Velocity

perception

separation

request

( V IS

in

an

were

AlIOs

stimulus

of of

introduced

was of

situations

of

any

possible

given

the

ence

with

of velocity difference could be correlated

Procedure to

optical

spots

, and

The degree judgments

gradient property

was

with

dimension

Between

judgments

were

depth

1 m . in

tions

of

and as

third

perception

as a difference

always

textures

of

judgments

in

as

consistent

Naive

the

no

expected

velocity

the

lead

the

motions

was

in absolute

suggested

that

two

separation

distance

. Another

differential

like

elements

made

kind

naive

of

: Correspondence Instructions

difference

in

result two

the

was motion

experimental

by

that

with

. Judgments

were

the

Method

a

difference

fact

differential textural

one

two

judgments

of

with

Problem

than

. The

sufficient

is

combined

present

and

continuous . This

---

phenomenon

phenomenon

is

a slant

experiment

of was

more

parallax

sufficient

tion

. It second

variable . -

evidently

with of

of

elements

highly exoeriment

motion

direction . The

continuity

. is

parallax judgments

I permutation

of

1 There

motion

consistent

except

the

experiment were in

that

induce

continuous

most To

indicate

does

the

farther

the a

textures

and

surface

varied

to

. I . Each

the

scale

apparent

spontaneous

response

.5 I ) , and

Exp

information

unmarked

systematically

made in

verbal but

.

that

observa

the

-

motionless

least

velocity

then

made

spots

so

. The

20

differ

-

judg

-

variable

278

E. J. Gibson, J. J. Gibson, O. W . Smith, & H. Flock

transparentmount was so set as to produce 10 velocity ratios of .51, .54, .57, .61, .65, .70, .76, .83, .90, and .97. Eachratio was presentedtwice, in random order, in the texture seriesand the spot series. Half the Os began with one seriesand half with th_e other. An 0 was assignedto one of three groups, and then instructed as follows :

Group I (Leastinformation):

The 0 was shown the sliding scalebeside his chair

and was told that it could be used to indicate

the distance between

the nearer and

farther of the two (surfaces, spots). If 0 had used other terms instead of "nearer"

and "farther," these terms were employed. He was then told that he would be shown a number of different settings of the apparatus, and that each time he would

be asked to make a judgment (degree of separation, distance between, etc.). No othPT information was given. The 0 was encouragedto report as he went along ~----- -------but no commentswere made on his performance. There were 16 as. Group II (Maximum and minimum): After the adjustablescalehad been demon'-'

-

strated , these Os were given another demonstration of the greatest and the least velocity -difference and were told which one was the maximum and which the

minimum. The procedure thereafter was the same as for Group I. There were 17 as .

Group III (Most information):

After the preliminaries described, these as were

told : "These are shadows of two (surfaces, objects ) which are actually at different distances from you . One is farther away from you than the other . This is what they look like at their maximum separation , which is 18 in . (The 0 was shown 18 in . on the adjustable scale.) This is what they look like at their minimum separation ,

which is 1 in. (The 0 was shown ! in. on the scale.) Each time I will ask you to estimate how far apart the (surfaces , objects) making the shadows are." The procedurewas thereafter the sameas for the other groups. There were 20 Os. When each 0 had finished his judgments he was questionedby E as to how he had made them unless he had already made this clear. The questions were : How

did you make your judgments 7 Did you go by appearanceof depth, or did you try to use some other

cue ? If some other

cue , what

was it ? Did you ever see the

front and back (surfaces , spots) changeplacesor fluctuate? Did the two (surfaces, spots) ever appearto be connected, like parts of a rigid object?

Results For each 0 a rank order correlatiort was run between his 20 judgments of separationand the correspondingvelocity-ratios. Table 16.1 summarizes the median for the three kinds of instruction and for the two ---- --------- coefficients ~ kinds of apparent objects, surfacesand spots. They range from .51 to .84. Eighty-seven of the 106 correlations were significant at the 5% level (r , > .44). , The data thus demonstratesome correlation between amount of differential velocity and degree of depth judged. The spread of the individualcorrelationswas very great, however, ranging from - .52 to + .99. Group III, which was given the most information, had higher correlations (medians of .83 and .84) than the other two groups. If individual correlations are considered, 19 out of 20 were significant for both the

MotionParallaxin PerceivedDepth

279

Table 16.1

Medians andRanges ofCorrelations Between Velocity Ratios andJudgm entsofDepth Measure

GroupI

GroupH

GroupIII

Stimuli (N= 16)

(N= 17)

(N= 20)

Median r

Surfaces + .72

+ .67

+ .83

Spots

+51

+84

Range

Surfaces .23to + .95 + .34to + .85

.18to + .96

Spots

.45to + .99

+57

.52to + .94

Number rs significantSurfaces 12/16 at <.05

Spots

14/16

.14to + .85

15/17

19/20

8/17

19/20

surfaces andthespots.Thecorrelations of GroupsI andII do notdiffer

fromone another.The demonstrating of the maximum and minimum

stimulus presentations attheendsofthescale didnot,therefore, improve

theordering oftheestimates. Buttelling 0 thathewasseeing shadows of

twothingsseparated by a givenamountofspacedidso.

Whentheinstructions weregiven, theyseemed tomake0 beginsearchingat once,andquitedeliberately, forcueswhichhecouldputintosome

order. Onlya minority oftheOsreported thattheyhaddepended on

anyappearance of depth in making theirestimates. Considering allOs,

70%reportedthat theyhadusedthe relativemotion, or the difference

inspeed,orhow faronepasses theotherasthebasisforjudgment.

Evidently, mostof themsawmotionsof somekindof entitiesbutdidnot clearlyseetheamountofspaceseparating them.

Theappearance ofa connection between thetwosurfaces or spots

likepartsofa rigidobjectwasalmost neverreported inanswer to the question. Theappearance of exchanging placeor fluctuating in depth

was reported by 17 out of 53 Os.

Because themorespecific instructions raisedthecorrelations, twoOs (both

psychologists andfamiliar withtheapparatus andtheproblem) wererunon consecutive dayswithreinforcement, to seehowhighthecorrelations mightgo

andhowstable theymight become. Thespots were notusedinthisexperiment.

Thefirst20trialsoneachdaywererunwithout comment. Butonthenext10trials Owascorrected aftereachjudgment. TheEdidsobymarking offonthescalethe actualdistancewhichseparated the standardandvariablemountsbehindthe translucent screen. Thisprocedure of 20uncorrected and10corrected trialswas repeatedfor9 dayswithone0 and7 dayswiththeother. Thecorrelations improved rapidly fromcoefficients of.20and.30to onesof.90 or more.Theintroduction of newandunfamiliar textureshadlittleeffecton the

correlation. Neither didtherequiring ofone0 tousea verbal rating scale ofI to 10insteadof the adjustableslidingscale.

280

E. J. Gibson , J. J. Gibson , O. W. Smith , & H. Flock

Theseas definitelytook a problem-solvingapproachto the taskandchecked their methodsof judging againstthe correctionsgiven by E. One estimatedthe ratio of the two velocitiesand tried to expressthis numericallyeachtime. The othersaid, aftera few days, that the taskhadbecomesimilarto memorizingparied associates ; he wastrying to link up ordered"cues" with particularscalepositions. The training did not have the effect of producingor enhancingan immediate appearance of depth.

Discuss ion After verbalsuggestion , information , or trainingconcerningseparationin depth, a correlationwaspresentbetweenthe degreeof velocity-difference and the degreeof separationjudged. It was raisedby informationand correctedtraining. But the reportsindicatedthat the as generallysaw motionsrather than depths, and that the appearance of depth was not inducedby informationor training. They wereled to look for andusecues for the requiredjudgmentsof depth, but not to reportthat they perceived it. They couldinterpreta velocity-differenceasa depth-differenceif given instructions , and they couldimprovethe consistencyof their estimatesif givenreinforcement . Theycouldlearnto perceivein onesenseof the term, but they did not learnto seea differentialvelocityasa differencein depth. This resultdoesnot supportthe theory of "unconscious inference " or point to any processfor the conversionof bidimensional impressions into perceptions . It shouldbe remembered , of course , that in theseexperiments motionhasbeenisolatedfrom otherdeterminants of perception . Thisdoes not occurin everydaylife. Motion usuallycomesin conjunctionwith size, shape , density, anddisparity. But it might be supposed that the fact of this conjunctionoveryearsof pastexperience wouldhavegivenan associative cuevalueto the motionsin our experiment . In that casethe velocity-pairs shouldhaveproducedspontaneous judgmentsat once. Sincethey did not, this particulartheoryof cuelearningis not supportedby our results. Summary

Thecommon assertion thatmotionparallax isa cuefor depthperception is vague . Theopticsof differential velocities of the elements in a fieldof viewwereexamined andtwocases weredistinguished : thatof twovelocities in thefieldandthat of agradient of velocities in thefield.Thetwo-velocitycaseyieldedconsistent perceptions of theseparation of onesurface intotwo. Theflow-gradient case(motion perspective ) yieldedconsistent perceptions of slant , or rateof recession in depth. In neithercasewerethereconsistent judgments of distance fromO. Stillanother case , thatof two differentvelocities of two differentspotsin anotherwise empty field, didnot yieldconsistent space perceptions of anykind. EvidentlyHelmholtz waswrongaboutan"immediate apperception " of thedistances of bodiesor the depthbetween themsolelyfromanimpression of thevelocities of spotsin thefield

MotionParallax in Perceived Depth 281

ofview. Theonlyconsistent orimmediate impressions obtained werethoseof separation indepth (inresponse toapermutation ofadjacent order oftexture), and recession indepth(inresponse to a transformation ofthetexture). These werespontaneous perceptions inthesensethatnootherinformation was given0 thanthatcarried intheoptical stimulation. When hewasgivenverbal information aboutdepthinthetwo-velocity situation hewasableto correlate a judgment ofdepthwithanimpression ofmotion. Butthecorrelations werenot perfect, anda minority ofOssaw thedepth. Onemight conclude fromthese

facts thatthere aretwokinds ofoptical stimulation forexperiences ofspace: akind which requires additional information toyield consistent judgments, andanother

kindwhich doesnotrequire itwhichiscompelling orcoercive. Thefactsalso

suggest thatthere aretwokinds ofexperience ofspace: emptydepth, asexem-

plified byonesurface infrontofanother, andfilleddepth, asexemplified bytheslant orrecession ofa surface. Thedepth ofa surface isperceived more consistently thanisthedepthofthespace between surfaces, inoursituation.

Asregards theperception ofabsolute distances, thisprobably depends onthe

perception of a terrainor groundsurface, the conditions forwhichwerenot

reproduced inthepresent experiments. Fortheinvestigation ofthisproblem, a

verylargefieldof viewis required. Note

1.Thisworkwassupported bytheOffice ofNaval Research under Contract NONR 401

(14)withCornell University. Reproduction inwhole orinpartispermitted forany

purpose of the U. S. Government.

References

Bourdon, B.Laperception visuelle delespace. Paris: Schleicher, 1906.

Gibson, J.1.Theperception ofvisual surfaces. Amer. J.Psycho!., 1950, 63,367384.

Gibson, Rev., J.J. 1957, Optical 64, 288295. motions and transformations asstimuli forvisual perception. Psycho!. Gibson, J.J.,&Carel, W.Does motion perspective independently produce theimpression ofa receding surface?]. exp.Psycho!., 1952,44,1618. Gibson, 1.J.,&Gibson, E.J.Continuous perspective transformations andtheperception of rigidmotion.J. exp.Psycho!., 1957,54, 129138.

Gibson, Amer. J.J., 1.Olum, Psycho!., P., 1955, &Rosenblatt, 68, 372385. F.Parallax and perspective during aircraft landings. Graham, H.Visual perception. InS.S.Stevens (Ed.), Handbook ofexperimental psychology. NewC. York: Wiley, 1951, 868920. Graham, C.H.,Baker, K.E.,Hecht, M.,&Lloyd, V.V.Factors influencing thresholds for monocular movement parallax. J.exp.Psycho!., 1948,38,205223. Helmholtz, H.Physiological optics. Vol. III.J.P.Southall, Ed., Optical Soc. America, 1925.

Hochberg, J.,&Smith, 0. W.Landing strip markings andtheexpansion pattern: I.

Program, preliminary analysis, andapparatus. Percpt. mot.Skills, 1955, 2,8192. Koffka, K.Principles ofGestalt psychology. NewYork: Harcourt, Brace, 1935.

Smith, 0. W.,&Smith, P.C.Interaction oftheeffects ofcues involved injudgments of curvature.Amer.J. Psycho!., 1957,70, 361375.

Retrospect andProspect: Psychophysics to

Computation

These experiments, performed during the1950s, paved thewayforfuture developments inresearch andtheory onperception. Theyunderlie James Gibsons laterworks, ofcourseThe Senses Considered asPerceptual Systems (1966) andTheEcological Approach toVisual Perception (1979). Buttheyalso, especially thelasttwopapers, pointtoward Marrscomputational theory of perception (Marr1982). Marrtried,as didGibson, to analyze the

information available fordetecting layout intheworld. Hisapproach is morelikeGibsons earlier theorizing thanthelater, since Mannever gave upthefixed retinal image asthefundamental datum forknowing aboutthe world. Whether hewould eventually havediscovered theopticarrayand thenecessity oftreating information asa flowtobeexplored, asGibson did, well never know. Psychophysics, asa method andevenasinstigator ofproblems for research, hasneverdiedoutandistodayholding swayovera bigfieldof

research, mislabeled by someas perception. Themostrecenteditionof

Stevens Handbook ofExperimental Psychology (1988) demonstrates thepervasiveinfluence ofpsychophysics asbotha method anda wayofthinking aboutpsychology upto thepresent moment. A recent meeting ofthe Psychonomic Society devoted manyhoursofitsprogram to psycho-

physical research. Isthisbad?Inonesense, notat all,because theresearch

islikely tobetechnically elegant. Butitcanbestultifying too.Veryoften, psychophysical research seems tobesetbytheparadigm, rather thanbya

truepsychological question. Thebackingof the Establishment doesnot makeupforfailure to asktherightquestions. Thesituation is a littlelike

theroleoftheEstablishment inartinEngland inthenineteenth andearly twentieth centuries. What washungintheannual exhibitions oftheRoyal Academy wasdetermined byestablished convention andtaste, notbynew ways of viewing the world. Inonesense, psychophysics ismorethana method. ToBoring, when

hewroteThePhysical Dimensions ofConsciousness (1933), it wasa kindof

284

E. J. Gibson

statementof the task of psychology. One lays out the physical dimension, an intensive one, for example, or frequency, like pitch, or distancefrom one point to another measuredin some physical scale, and then looks for the psychological scalethat corresponds. The question of correspondencebetween two domainsis as old as scienceitself, absolutely basic. Theoretical disagreementsor the possibility of a futile searchfor correspondencesthat matter are likely to come in the choice of domains and scales. Are the physicists' measurementsthat are so often used appropriately scaledfor relevanceto human perception? We can measuredistancesin the universe in light years, or wave lengths in millimicrons, but these scaleswill not be appropriate for enlightening us about human perception in a human-sized environment.

IV PerceptualLearning(1955- 1969)

Introduction to Part IV

The two previous sections included several experimentsdesigned to enhanceperceptuallearning by introducing correction or reinforcement, none of which worked very well . The field experiments worked in a way , since constant errors were all reduced and there was some evidence of an increase in precision . But the shift in constant error could have been

simply judgmental, a deliberate correction of a scale of verbal responses rather than a change in what was perceived. In one of the rat rearing experiments(Walk, Gibson, Pick, and Tighe 1958), we introduced during rearing reinforcement of one of the forms to be discriminated in later

learning. The rats obtained food by pushing their noses through a small door in a food holder that displayed one of the forms to be discriminated, whereas animals in a control group obtained their food through an unadorneddoor. There was a small advantagein later discriminationlearning for the reinforced

animals , but it turned

out that it made no difference

whether or not the form reinforced during discrimination learning was the sameone that had been displayed during rearing. In a secondexperiment, reinforcement (differential or otherwise ) had no effect whatever and so had to be ruled out as a facilitator . In a different context , the two -laver denth .. Ir a

- -

-

-

__

experimentwith prolonged training of humansubjects(reprinted in part III) resultedin no phenomenalchangein the depth experiencedby the subjects. Despite these failures, there was ample reasonto think that perceptual learning occurs. I reported an experiment in which it did with children and adults at the meeting of the American Psychological Association in 1950 (Gibson and Gibson 1950). The trouble was that the old learning theorieswere inadequateto handle it . We neededa new approach. Neither rewarding responsesdifferentially nor providing correlatedinformation as a basisfor making inferencesaccountedfor perceptuallearning. Perceptual learning becamethe important problem for me by the 1950s, and I gradually expanded it to include perceptual development. Infancy is the time

288

Part IV

whenperceptual learning shouldbemostprominent, sincelanguage isnot yet availableand the responserepertoryis at a minimum.

Fora periodof aboutfifteenyears,I performed experiments in the general areaofperceptual learning anddevelopment andgradually evolved a theory,culminating withmybook,Principles ofPerceptual Learning and Development, published in 1969.JamesGibsonandI spenttheyearof 195960 as fellowsat the Institutefor AdvancedStudyin Princeton.My

planwastospend thatyearwriting abookaboutperceptual learning. I did makea start,but it wasnot longbeforeI feltthat I wasnot readyfor a

bookyetandneededmoretimeforthinking. I putawaythefewchapters I hadcompleted andworkedinsteadon chapters I hadbeeninvitedto contributeto otherbooks,good preparationfor the one I intendedto write

on myown.I foundI hada lotofworkto do if I wantedto include (or

rather create)the wholefieldof perceptuallearning,criticizealternative theories,and propose my own theory.

During theyearsbeforethebookfinally appeared, manythingshappened

to furthermy educationandbroadenmy knowledge baseacquaintance withsomeprominent European thinkers (notably PiagetandseveralRussian

psychologists), a trip to Genevaand to Russia,a memorable summer conferencehosted by Held and Hem at MIT on the effectsof visual adaptationto prismsand what they meant,and a brief attractionto informationprocessing,among other things.Most importantwas a year at the Institute for Advanced Study in the Behavioral Sciences in 196465

whenI got backto thebookseriously oncemore,threwawaywhatI had writtenearlier,plotted it all out, and began to write again.I madegreat progressthat year,but the bookturnedout to be a muchbiggerproject than I had first envisionedtwenty

chapters in alland

covered history

of the subject,animalsand children,as wellas my own theoreticalviews.It has four introductory and historical chapters, four chapters on my own

theory,four factualchapterson differentresearchareas,two chapterson animals,five on humanperceptualdevelopmentin children,and a final theoretical chapterthat focusseson generaltrendsin perceptualdevelopment and some classicdevelopmentalissues like priority of perceiving wholes or parts.

Sincethe book cant be reprintedhere,I includesomemilestonesalong the way and excerptsfromthe books finalchapter.But firstI presenta

brief statement of the central ideas that I espoused.The basic concept is the

emphasison differentiation as opposedto association or any formof

add-on.Theproblemof perceptionwastakento be howwe keepin touch with the world around us. How we can know the world has been the

age-oldproblemfor theoriesof perception; the problemfora theoryof perceptual learning mustbe howweimproveourknowledge of theworld and make better use of the information about it. A differentiationtheory

PerceptualLearning(19551969)

289

assumes, necessarily, that the information aboutthe worldis abundant,

ratherthanminimal andimpoverished. Perception istheprocess ofobtainingthisinformation. Perceptual learning goesonbecause inthebeginning

onlygrossimprecise information isobtained. Development proceeds via

differentiation ofinformation thatspecifies thingsandeventsintheworld,

by discovery of invariants as changes in stimulation areproduced by

movementsof thingsand of the observer,and by encounterswithnovel and broader environments.

Thetheorydivides intothreeparts:whatis learned, theprocesses

involved,andfactorsthatselectwhatislearned.Whatislearnedconcerned

me morethananythingelse,becauseperceptual learningan effectof learning onperception, notresponse learning ormemory ofverbalmaterial had beenso neglected by modernpsychology. Whatwaslearnedwas

takentobedifferentiation ofthedistinctive features ofthings, invariants of events,andhigherorderstructure ofboth.I wasmuchimpressed by the

notionofdistinctive features, whichI hadreada gooddealaboutwhenwe wereat thePrinceton Institute. NoamChomsky wasthereat thesametime,

andI picked upa bitaboutlinguistics, hitherto unknown to me,through him.I usedwhatwaslearned, especially theemphasis ondifferentiation, as thebasisforseveral principles thatmighthaveapplications. Theprocesses I thoughtwereinvolved inachieving differentiation were

abstraction, filtering,and peripheralmechanisms of attentionusedto tune

theperceiver foroptimal search (important because I thought ofthepro-

cessas a search,andan activeone).Thispartof the theoryhasbornethe

heaviest bruntofcriticism bypeoplewhoareimpressed withmechanism ratherthanfunction. I thinktheyreallywantedassociation or inference,

which I hadruledout.I foundmorepromising thenotionofstripping

awaythenoiseandunwanted garbage soasto sculpta cleanoutlineofthe

information thatspecified thepersisting structure ofthingsandeventsin theworld. Thisprocess couldinvolve bothextraction andfiltering (inhibition).I amgladto seethat thisideahasfoundsomefavorlatelyin neuroembryology (e.g., Changeux, Heidmann, andPatte1984). Myhus-

bandthoughtof andreferredto perceptual learningas the educationof attention,a good phrase, but I thought it toogeneral to carrymevery far. It wasimportantto tie motivationin withthe theory,sincein the last

anaysisbothperception andlearning areselective processes. Whatdoes

motivateperceptionand perceptuallearning?Not metabolicneeds,like Hullsolddrivestimulus. Perception is a searchforinformation, an intrinsic

motivethatis exhibited especially strongly in younganimals, including

humaninfants.It underlies perceptual learning as well.Butwhatselects

whatis learnedif not reductionof a drivestimulus? I maintained thatit is the discovery of economical structuresuchas the distinctive feature,the

290

PartIV

invariant that underlies all the transformations, the higher order structure that is superordinateover all embeddedstructures. I calledthis principle the reductionof uncertainty . It is a kind of economy principle, by no means unfamiliar in the history of science (Hatfield and Epstein 1985).

In a nutshell, I thought of my theory (and referred to it) as a specificity theory of perceptuallearning, meaning the kind of achievementreflected by the perceptualchangesas they come to corresponduniquely to what is going on in the world. Paul Weiss consideredspecificity a major question for science, defining it as IIa sort of resonance between two systems attuned

to eachother by correspondingproperties" (Weiss 1970, 162). In the case of perceptual learning, three systemsare implicated: the world of events and objects in an environmental layout , the information in ambient arrays

of energy that specifiesthe happeningsand layout of the environment, and the changing perceptions of the observer. Achievement of specificity of perception to the information specifying the world is the function and the result of perceptuallearning. The description of these correspondenceswas not far advanced, and their relations were not so clear to me when this book was written as they became later . The papers that follow

reveal stages on the way to a more

sophisticateddefinition of the true questions. This questcontinued through publication of my husband's last two books (J. Gibson 1966 and 1979) and still requires our attention in pursuing an adequatetheory of perceptual learning . It was clear from the start, however , that fortuitous association

is an unlikely conceptualmechanismfor a theory focussedon the description and matching of correspondences .

17

Perceptual

Learning

James

J

This

is

a

making

a

.

of

how

the

,

I

was

of

learning

.

pinch

course

The

to

subjects

a

of

,

until

he

ment

proceeded

the

or

she

,

.

which

The

to

run

subject

errorless

.

run

.

to

was

shown

No

correction

at

( as

an

a

)

matter

to

,

of

show

accurately

entire

them

,

set

was

as

variables

was

,

the

do

midyear

dimensional

idea

identify

-

to

identification

used

in

The

My

identifica

( except

We

-

"

really

.

on

defected

.

built

subjects

was

call

my

about

responsibilities

scribbles

the

little

how

member

had

systematic

requiring

an

for

subjects

as

since

fine

Richard

Solomon

on

September

"

order

achieved

ready

flawlessly

the

forms

them

the

in

questions

to

the

written

classified

Gagne

teaching

me

I

learning

those

come

in

had

very

discussed

faculty

for

of

often

no

a

said

to

had

I

over

given

just

one

and

as

over

the

experi

-

.

,

was

with

had

long

,

differentiated

Robert

we

was

.

be

back

we

when

nonsense

called

so

first

learning

kind

get

with

,

experiment

I

what

what

me

.

to

research

on

left

?

before

to

forms

II

course

standard

groups

spring

That

Cornell

random

his

the

years

material

or

War

an

chosen

was

age

We

the

time

kind

but

ten

in

but

long

of

learning

drawn

.

in

scribble

to

,

accomplished

a

,

than

nonsense

for

designing

set

drawn

1955

More

differentiation

time

systematic

the

in

.

World

move

for

kinds

published

because

more

about

)

three

I

a

hitter

set

of

wanted

during

of

of

had

aircraft

we

set

interested

experiment

so

a

that

Enrichment

Gibson

of

at

was

and

husband

and

of

all

importance

differentiation

entailed

,

.

attached

learning

consisted

or

J

stories

the

verbal

Differentiation

Eleanor

papers

some

about

as

research

the

three

has

dissertation

two

,

series

research

a

Gibson

and

tion

:

well

I

.

I

had

got

first

experiment

(

access

Gibson

had

and

ages

to

Ithaca

1950

Psychological Review, 1955, 62, 32- 41.

time

)

.

The

and

I

no

for

experiment

adult

children

' s

( one

conference

data

to

!

present

lay

fallow

of

during

subjects

summer

learning

me

children

experiment

( my

my

a

included

the

groups

of

for

me

in

scout

two

gave

,

performed

of

when

thus

1950

children

two

finished

Gibson

of

The

shock

come

and

was

spring

.

to

a

run

the

adults

rather

who

the

in

as

the

it

for

)

of

)

But

groups

performed

others

at

A

a

.

them

while

did

. P

. A

.

until

,

in

292

J. J. Gibson & E. J. Gibson

my husbandwas invitedto give a paperon military training at a panelmeeting on Training Devices in Washington. We got out the experimentand gave more thought, then, to what the results meant for learning theory and training . That

paperseemsnot to bein existence any longer- perhapsit wasjust delivered from notes. But I remember the discussions at the meeting and the interest aroused in the

question of whether we can learn something by perceiving alone, or only by performing something. This was the era of the so-called "new look" in perception (Blake and Ramsay 1951), so views on perception were shifting away from

psychophysics perse. But the "newlook" wasnot at all thenewview that we had in mind. The prefaceto Blakeand Ramsay's book (1951) says, for example : "Perceptual activity suppliesthe materialsfrom which the individual constructs his own personallymeaningfulenvironment " (p. iii). Bruner, one of the contributorsto the book, presented a theoryof perceptionas a constructiveprocess , with hypotheses , sets, and checkof hypothesesas the working concepts . Stimula-

tion was thoughtto be impoverished and unreliable , leavingthe door openfor autisticfactorsto determinewhat is perceived . Our view was radically opposedto thesenotions. An opportunity to say so to a psychologicalaudiencecameup rohenJamesGibsonwas invited, as Chairperson of SectionI, to organize up a program for the Christmas meeting of The American Association for Advancement of Sciencein Berkeley, 1954. We

werespendingthefall term at the Universityof Californiaat Berkeley , onleave from Cornell, so we hadplannedto attend. LeoPostman , our goodfriend on the faculty at Berkeley , agreedto preparean opposingview, to ensurea lively debate . Postman had coauthored some papers with Bruner in the Iinew look " tradition , but

he was more interestedin a traditional learning view, as his paper that follows

makesclear. He spokeeloquentlyfor associationand learningtheory, and for perceptual learningasa matterof behaviorchange , presentingthe bestarguments that couldhave(or still have) beenmadeagainstour proposals . We repliedwith arguments about what is learned (not connectionswith responses , but changesin

what is responded to). It wasa lively andfriendly argument , thekind I would like to seemoreoftenin psychology . The views in our paperemphasize differentiationof available information, rather than an add-on of anything, whether it be response , ideas, affective states,

etc., as thebasisof perceptual learning. Thiswasthemajor hypothesis of my book on perceptuallearningand I am still committedto it- more than everas the ecological approachto perception hasprogressed in sophistication in recentyears. -

-

-

-

L ;

>

I

I

.

.

.

-

The term "perceptuallearning" meansdifferentthingsto differentpsychologists.! To someit implies that human perception is, in large part, learned - that we learn to seedepth, for instance, or form, or meaningful objects. In that case the theoretical issue involved is how much of perception is

learned, and the correspondingcontroversy is that of nativism or empiri-

PerceptualLearning: Differentiation or Enrichment?

293

cism. To others the term implies that human learning is in whole or part a matter of perception- that learning dependson comprehension , expectation, or insight, and that the learning processis to be found in a central processof cognition rather than in a motor processof performance. In this second case, the theoretical issue involved is whether or not one has to study a man's perceptions before one can understand his behavior, and the controversy is one of long standing which began with old-fashioned behaviorism. Thesetwo sets of implications are by no meansthe same, and the two problems should be separated. The problem of the role of learning in perception has to do with perception and the effect of past experienceor practice on it . The problem of the role of perception in learning has to do with behavior and the question of whether we can learn to do something by perceiving, or whether we can only learn by doing it. The questions, then, are these: (a) In what sensedo we learn to perceive? (b) In what sense can we learn by perceiving? Both questionsare important for the practical problems of education and training, but this paper will be concernedwith the former.

In WhatSense Do WeLearnto Perceive ? This questionhasroots in philosophy and was debatedlong before experimental psychology came of age. Does all knowledge (information is the contemporary term) come through the senseorgans or is some knowledge contributed by the mind itself? Inasmuchas sensorypsychology has been unable to explain how as much information about the world as Wp manifestly do obtain is transmitted by the receptors, some theory is required for this unexplained surplus. There has been a variety of such theories ever since the days of John Locke. An early notion was that the surplus is contributed by the rational faculty (rationalism). Another was that it comesfrom innate ideas(nativism). In modem times there have been few adherentsto thesepositions. The most popular theory over the years hasbeenthat this supplementto the sensationsis the result of learning and that it comes from past experience. A contemporary formula for this explanation is that the brain stores information- possibly in the form of tracesor memory images, but conceivably as attitudes, or mental sets, or general ideas, or concepts. This approach has been called empiricism. It preservesthe dictum that all knowledge comesfrom experienceby assuming that past experiencesomehow gets mixedwith present experience. It assumes , in other words, that experienceaccumulates , that tracesof the past somehow exist in our perception of the present. One of its high-water marks was Helmholtz's theory of unconsciousinference, which supposes

294

J. J. Gibson & E. J. Gibson

that we learn to see depth by interpreting the clues furnished by the depthlesssensationsof color. Another was Titchener's context theory of meaning , which asserts that we learn to perceive objects when a core of sensations acquires by association a context of memory images.

Over a generation ago this whole line of thought was challengedby what seemed to be a different explanation for the discrepancy between the sensory input and the finished percept -

the theory of sensory organiza -

tion. The gestalt theorists made destructive criticisms of the notion of acquiredlinkages among sensory elements and their traces. Instead they assertedthat the linkages were intrinsic, or that they arose spontaneously , taking visual forms as their best example. Perceptionand knowledge, they said , were or came to be structured .

The theory of sensory organization or cognitive structure, although it generateda quantity of experimentationalong new lines, has not after 30 years overthrown the theory of association. In this country the old line of empiricist thinking has begun to recover from the critical attack, and there are signs of a revival . Brunswik (2, pp . 23 ff .) has followed from the start the line laid down by Helmholtz . Ames and Cantril and their followers

have announcedwhat might be calleda neoempiricistrevelation (3, 11, 14). Other psychologists are striving for a theoretical synthesis which will include the lessonsof gestalt theory but retain the notion that perception is learned. Tolman, Bartlett, and Woodworth began the trend. Leepertook a hand in it at an early date (15). The effort to reconcile the principle of sensory organization with the principle of determination by past experiencehasrecently been strenuouslypursuedby Bruner(1) and by Postman (16). Hilgard seemsto acceptboth a processof organization governed by relational structure and a processof associationgoverned by the classical laws (10). Hebb has recently made a systematic full -scale attempt to com-

bine the best of gestalt theory and of learning theory at the physiological level (9). What all these theorists seem to us to be saying is that the organization process and the learning process are not inconsistent after all, that both explanations ~

are valid in their way , and that there is no value in

continuing the old argument over whether learning is really organization or organization is really learning . The experiments on this issue (beginning with the Gottschaldt experiment ) were inconclusive , and the controversy itself was inconclusive . Hence, they argue, the best solution is to agree with both

sides .

It seemsto us that all extant theoriesof the perceptualprocess, including those based on association, those based on organization , and those based

on a mixture of the two (including attitudes, habits, assumptions , hypotheses, expectation, images, contexts, or inferences) have at least this feature in common: they take for granted a discrepancybetween the sen-

PerceptualLearning: Differentiation or Enrichment?

295

sory input and the finished percept and they aim to explain the difference. They assumethat somehow we get more information about the environment than can be transmitted through the receptor system. In other words, they accept the distinction between sensationand perception. The development of perception must then necessarilybe one of supplementingor interpreting or organizing. Let us consider the possibility of rejecting this assumptionaltogether. Let us assumetentatively that the stimulus input contains within it everything that the percept has. What if the flux of stimulation at receptorsdoes yield all the information anyone needsabout the environment? Perhapsall knowledge comes through the sensesin an even simpler way than John Lockewas able to conceive- by way of variations, shadings, and subtleties of energy which are properly to be called stimuli. TheEnrichmentTheoryversusthe SpecificityTheory The entertainingof this hypothesisfacesus with two theoriesof perceptual learning which are clear rather than vague alternatives. It cuts acrossthe schoolsand theories, and presentsus with an issue. Is -perception a creative ~ processor is it a discriminative process? Is learning a matter of enriching previously meagresensationsor is it a matter of differentiating previously vague impressions? On the first alternative we might learn to perceive in this sense: that perceptschangeover time by acquiring progressivelymore memory images, and that a context of memoriesaccruesby associationto a sensorycore. The theorist can substituteattitudes or inferencesor assumptions for imagesin the above Titchenerianproposition, but perhapsall this does is to make the theory less neat while making the terminology more fashionable. In any case perception is progressively in decreasingcorrespondence with stimulation. The latter point is notable. Perceptuallearning, thus conceived, necessarilyconsists of experiencebecoming more imaginary, more assumptive, or more inferential. The dependenceof perception on learning seemsto be contradictory to the principle of the dependence of perception on stimulation. On the secondalternative we learn to perceive in this sense: that percepts change over time by progressive elaboration of qualities, features, and dimensionsof variation; that perceptualexperienceeven at the outset consists of a world, not of sensation, and that the world gets more and more properties as the objects in it get more distinctive; finally, that the phenomenalproperties and the phenomenalobjects correspondto physical properties and physical objects in the environment wheneverlearning is successful . In this theory perception gets richer in differential responses , not in images. It is progressively in greatercorrespondencewith stimulation,

296

J. J. Gibson& E. J. Gibson

not in less. Instead of becoming more imaginary it becomes more dis-

criminating. Perceptuallearning, then, consistsof responding to variables of physical stimulation not previously responded to. The notable point about this theory is that learning is always supposedto be a matter of improvement -

of getting in closer touch with the environment . It con-

sequentlydoesnot accountfor hallucinationor delusionsor, in fact, for any kind of maladjustment .

The latter kind of theory is certainly worth exploring. It is not novel, of

course , to suggestthat perceptualdevelopmentis a matterof differentia tion. As phenomenal descriptionthis wasassertedby someof the gestalt psychologists, notably Koffka and Lewin. (Just how differentiation was related to organization , however , was not clear.) What is novel is to

suggest that perceptual development is always a matter of the correspondencebetween stimulation and perception- that it is strictly governed by the relationships of the perceiver to his environment . The rule would be that , as the number of distinct percepts a man can have increases,

so also the number of different physical objects to which they are specific increases . An example may clarify this rule. One man, let us say, can identify sherry, champagne , white wine, and red wine. He hasfour percepts in responseto the total possible range of stimulation. Another man can identify a dozen types of sherry , each with many varieties , and numerous blends, and so on for the others . He has four thousand percepts in response

to the range of stimulation. The crucial question to ask about this example of differentiated perception is its relation to stimulation .

Stimulusis a slippery term in psychology. Properly speakingstimulation is always energy at receptors, that is, proximal stimulation. An individual is surrounded by an array of energy and immersed in a flow of it . This sea of stimulation consists of variation and invariants , patterns and transforma tions , some of which

we know

how to isolate and control

and others of

which we do not. An experimenterchoosesor constructsa sampleof this

energywhenheperformsa psychological experiment . Butit is easyfor him to forget this fact and to assumethat a glass of wine is a stimulus when actually it is a complex of radiant and chemical energies which is the stimulus. When the psychologist refers to stimuli as cues, or clues, or carriers of information he is skipping lightly over the problem of how stimuli come to function as cues. Energies do not have cue properties

unless and until the differencesin energy have correspondingly different effectsin perception. The total range of physical stimulation is very rich in complex variablesand theseare theoretically capableof becoming cuesand constituting information . This is just where learning comes in .

All responsesto stimulation, including perceptual responses , manifest some degree of specificity, and, inversely, some degree of nonspecificity.

PerceptualLearning: Differentiation or Enrichment?

297

The gentlemanwho is discriminating about his wine shows a high specificity of perception, whereas the crude fellow who is not shows a low specificity. A whole classof chemicallydifferent fluids is equivalent for the latter individual; he can't tell the differencebetween claret, burgundy, and chianti; his perceptions are relatively undifferentiated. What has the first man learned that the secondman has not? Associations? Memories? Atti tudes? Inferences ? Has he learned to have perceptions instead of merely sensations ? Perhaps,but a simpler statementmight be made. The statement is that he has learnedto taste and smell more of the qualities of wine, that is, he discriminatesmore of the variablesof chemicalstimulation. If he is a genuineconnoisseurand not a fake, one combination of suchvariablescan evoke a specificresponseof naming or identifying and another combination can evoke a different specific response. He can consistently apply nouns to the different fluids of a classand he can apply adjectivesto the differencesbetween the fluids. The classicaltheory of perceptuallearning, with its emphasison subjective determinationof perception in contrast to stimulusdetermination, gets its plausibility from experiments on errors in form perception, from the study of illusions and systematicdistortions, and from the fact of individual differencesin and social influenceson perception. The learning processis assumedto have occurredin the past life of the experimentalsubject; it is seldomcontrolled by the experimenter. Theseare not learning experiments insofar as they do not control practice or take measuresbefore and after training. True perceptual learning experiments are limited to those concernedwith discrimination. One source of evidence about discriminative learning comes from the study of the cuesfor verbal learning. The analysisof these cuesmade by one of the authorsin terms of stimulusgeneralizationand differentiation (4) suggeststhe present line of thought. It has also led to a seriesof experiments concernedwith what we call identifyingresponses . Motor reactions. verbal reactions, or percepts, we assume , are identifying responsesif they are in specificcorrespondencewith a set of objectsor events. Code learning (13), aircraft recognition (7), and learning to namethe facesof one's friends are all examples of an increasingly specific correspondencebetween the items of stimulation presentedand the items of responserecorded. As a given responsegains univocality, the percept is reported to gain in the feeling of familiarity or recognition and to acquiremeaning. An IllustrativeExperiment2 In orderto providea clearexampleof suchlearning,we studiedthe development of a singleidentifyingresponse . The5 waspresented with a visualitem consisting of a nonsense "scribble " ; his recognitionof it wastestedwhenit wasinterspersed

298

J. J. Gibson& E. J. Gibson

Figure 17.1

Nonsenseitems differing in three dimensions of variation.

in a series of similar scribbles, and then the single showing and the multiple presentationwere repeateduntil the item could be identified. We devised a set of 17 scribblesintended to be indistinguishablefrom the critical item on the first trial, and another set of 12 items intended to be distinguishablefrom the critical item on the first trial. The items which had to be differentiated are shown in Fig 17.1. The critical item, a four-coil scribble, is in the center and 16 other items are arrangedoutward from it . The eighteenth item (a reversal of the critical item) is not shown. It may be noted that there are three dimensionsof variation from the critical item: (a) number of coils- three, four, or five, (b) horizontal compression or stretching, and (c) oripntation or riQht-Ieft reversal. The latter two kinds of variation were produced '-" by photographic transformation. There are three degreesof coil frequency, three degrees of compression, and two types of orientation, which yields 18 items. Since one of these is the critical item, 17 remain for use in the experiment. The reader may observe that when these differences are verbally specified and the figures are displayed for immediatecomparison, as in Fig. 1, they are clearly distinguishable. The Ss of the experiment, however, saw the items only in succession . The 12 additional items presented on each recognition trial are shown in Fig. 17.2. Each differs from every other and from all of the set of 18. The differences

PerceptualLearning: Differentiation or Enrichment?

tC)

f

@

~

~

~

299

(fj)

(@ )

'-' Nonsenseitems differing in many dimensions of variation.

from the scribbleswereintendedto be sufficientto makethemappeardifferentat the outsetto 5s with a normalamountof experience with drawnforms. The 30 items(12 plus 18) wereprintedphotographicallyon stiff 2 in. X 4 in. cardswith blackborders , andmadeinto a pack. The materialavailablefor anyone learning trial consistedof the criticalitem plusa shuffledpackof cardsamongwhichwere interspersed four replicasof the criticalitem. The 5 wasshownthe criticalitem for about5 sec. and told that someof the itemsin the packwould be exactlylike the oneshown. The seriesof 34 wasthen presentedeachwith a 3-sec. exposureand 5 wasaskedto report which of them werethe samefigure. The identifyingresponserecordedwasany report suchas "that's it" or "this is the one I saw before." The 5 was nevertold whetheran identificationwascorrector incorrect . A recordwaskeptnot only of the identifying responses , but alsoof any spontaneous descriptions offeredby 5, which were laterclassified asnamingresponses andqualifyingresponses . At the endof the first trial the criticalfigurewaspresenteda secondtime and anothershuffledpackwasrun through. The procedureof examininga figureand thentrying to identify it whenmixedwith a seriesincludingfiguresof both great andlittle similaritywascontinueduntil 5 madeonly thefour correctidentifications in onetrial. Threegroupstook part in the experiment : 12adults, 10 olderchildren (8! to 11 years), and 10 youngerchildren(6 to 8 years). Results In this experiment, learning is taken to be an increasein the specificity of an identifying responseor, in other words, a decreasein the sizeof the class of items that will elicit the response. The data therefore consist of the

300

J. J. Gibson & E. J. Gibson

Table 17.1 Increasein Specificity J. ~ of an Identifying - - Response for Three Age Groups Older Adults children Variable (N = 12) (N = 10)

Meannumber ofundifferentiated items onfirsttrial Meannumber of trialsrequired for completely ----C J specific J. response . Percentage oferroneous recognitions foritems differing inonequality Percentage oferroneous recognitions foritems differing in twoqualities Percentage of erroneousrecognitions for itemsdiffering '-' in threequalities "

Younger Children (N = 10)

3.0

7.9

13.4

3.1

4.7

6.7.

17

27

53

2

7

35

0.7

2

28

.. Only two of the youngerchildrenachieveda completelyspecificidentification . Themean numberof undifferentiated itemson the lasttrial wasstill 3.9.

number of items (out of a probable maximum of 17) reacted to as if they were the critical figure . As will be evident , this class of undifferentiated items was reduced as a result of repetition . The three groups of 5s, how ever, began to learn at very different levels and learned at very different rates. The results are given in Table 17.1. For adults, the class of undifferentiated items at the outset was small (Mean = 3.0), and only a few trials were needed before this class was reduced to the critical item alone (Mean = 3.1). Two of these adults were able to make no other than correct identifying responses on the first trial . Both were psychologists who could have had previous acquaintance with nonsense figures . The learning task was so easy for this group that not much information about the learning process could be obtained . At the other extreme , however , the younger children " recognized " nearly all of the scribbles on the first trial (Mean = 13.4), which is to say, that the class of undifferentiated items was large. The number of trials needed to reduce this class to the correct item was so great that most of the Ss could not be required to complete the experiment . Two out of 10 reached the criterion , but for the remainder the trials had to be stopped for reasons of fatigue . After an average of 6.7 trials the mean number of undifferentiated items was still 3.9. One child had so much difficulty with the task that E finally gave differential reinforcement by saying I'right " or "wrong " after each presentation of a card. Although this procedure helped, wholly specific identifications were never achieved. The failures of the younger children to discriminate did not seem to be due merely to " inattention " ; they understood that they were to select only the figures which were exactly the same as the critical figure .

PerceptualLearning:Differentiation or Enrichment? 301

Fortheolderchildren (between 8 and11yearsofage)theresultswere intermediate between theseextremes. Forthemtheparticular taskandthe particular itemswereneither toohardnortooeasy.Theaverage number of

undifferentiated itemsonthefirsttrialwas7.9,andallchildrensucceeded in reducingthis to a singleitemaftera meanof 4.7 trials.

Table17.1alsoindicates foreachgroupan important factaboutthe

unspecific responses:they tend to occurmoreoftenas the differences

between thetestitemandthecritical itembecome fewer. AsFig.17.1

shows, a givenscribble maydifferinonequality or dimension (thickness, coilfrequency, ororientation), orintwoofthesequalities, orinallthreeof them.Fiveofthescribbles differinonefeature, eightdifferintwofeatures,

and fourdifferin threefeatures.It willbe recalledthat the 12 additional

forms shown inFig.17.2differed fromthecritical itemwithrespect tomore thanthreefeatures. Amountofdifference canbeusefully statedasnumber of differing qualitiesor, conversely, amountof sameness as the fewnessof

differing qualities. 3 ThelowerhalfofTable 17.1givesthepercentage of occurrence offalserecognitions inthecaseofscribbles withonequality

different, withtwoqualities different, andwiththreequalities different. These percentages arebasedonthenumber oftimestheitemsinquestion werepresented during thewholeseries oftrials. Thedissimilar figures, which hadmanyqualities different, yielded a zeropercentage offalserec-

ognitions exceptfora fewscattered instances amongtheyoungerchildren. Discussion

Theresults showclearly thatthekindofperceptual learning hypothesized hasoccurred in thisexperiment. A stimulus itemstartsoutby being

indistinguishable froma wholeclassof itemsin the stimulusuniverse tested,andendsby beingdistinguishable fromallof them.Theevidence

forthisassertion is thatthespecificity of Ss identifying response has

increased. Whathashappenedto producethisresult?

TheSswereencouraged todescribe alltheitemsofeachseries asthey

werepresented,anda specialeffortwasmadeto obtainand recordthese

spontaneous verbalresponses forsevenof theolderchildren. In general

theytended tofallintotwotypes, either naming responses orqualifying

responses.Consideringonly the responsesto the 17 scribbles,the record

showed thatthefrequency ofthelattertypeincreased during theprogress oflearning. Examples oftheformer arenouns likefigure 6,curl,spiral, scroll.

Examples ofthelatterareadjectival phrasesliketoothin,rounder, reversed. It is notablethat the latterareresponsesnot to the itemas suchbut to the

relation between it andthecritical item.Theyareanalogous to differential judgments ina psychophysical experiment. Anadjective, ingeneral, isa response whichisspecific nottoanobjectbuttoa property oftwoormore

302

J. J. Gibson & E. J. Gibson

objects. It is likely, then, that the development of a specificresponseto an item is correlatedwith the development of specificresponsesto the qualities, dimensions, or variablesthat relate it to other items. The implication is that , for a child to identify an object , he must be able to identify the differences between it and other objects, or at least that when he can identify an object he also can identify its properties . The verbal reactions of the children to the 17 scribbles, both naming and

qualifying, could be categorizedby E as specificor nonspecificto the item in question . These judgments were necessarily subjective , but they were

carried out with the usual precautions. Although a single adjective cannot be specificto a single item, a combination of adjectivescanbe. An example of a nonspecificreaction is IIanother curlicue," and of a specificreaction is " this one is thinner

and rounder ." The latter

sort may be considered

a

spontaneouslydeveloping, identifying reaction, not of the "that's it" type, it is true, but neverthelessfulfilling our definition. The mean number of such verbal reactions on the first trial was 7.7 out of 17, or 45 percent. The mean number of such reactions on the last trial was 16.5. or 97 percent .

This suggeststhat, as a single identifying responsebecomesincreasingly specific to one member of a group of similar items, verbal identifying reSDonsesalso tend to become specific to the other members of the group . "-

As the class of indistinguishable items which will elicit one responseis diminished , the number Increases

Other

of responses

which

can be made to the class

.

evidence

Another sourceof experimentalevidenceabout perceptuallearning comes

from psychophysics . Contraryto what might be expected , psychophysical experimentersover the years have shown a lively interest in perceptual learning, or at least in the bettering of perceptualjudgments with practice. One of the authors has recently surveyed this neglected literature insofar asit concernsimprovement of perceptionor increasein perceptualskills (5). There is a great quantity of evidence about progressive change in acuity ,

variability, and accuracyof perception, including both relative judgments and absolute judgments. It proves beyond a shadow of doubt that the notion

of fixed thresholds

for a certain set of innate sensory dimensions

is oversimplified. Discrimination gets better with practice, both with and without knowledge of results. An example may be taken from the twopoint threshold on the skin. As long ago as 1858 it was discovered that there is a certain distance at which two points are felt double by a blindfolded subject that is characteristic of the area of the skin tested. At the same time , it was found that only a few hours of practice in this discrimination had the effect of reducing the

PerceptualLearning: Differentiation or Enrichment? distance

to

half

lowering instance

,

it

skin

. It

,

on

the

even

the

is

clear

these

criteria

be the

theory

had

any

We

in

it

. The

period

scale

from

of

supposedly

percept

of

seemed

the

to

"

of

to

" far

is

complex

differentiated

put

What ,

- order

, "

oneness

writer .

that apart

of

one

doubleness

higher

depend

absence

. As

stimulation

discover

than

experimental

sensations

data

criteria

) . The

in

distinct

the

becomes

,

. areas

symmetrical

together

these

finer

to

18

for

training

thresholds

. It

or

" close

from

that

,

;

untrained

lower

disuse ,

of

bilaterally

( 12 of

of

continues

the

the

trials

weeks other

much

testing

judgment

and

suggest

observer

a

to

that of

four

for

the

showed

thousands

. during

very

of

support

different

?

mm

had

after

sort

for

complete

of

a

experiments

transferred

beginning lost

any

adopts

mulation

the

correction

never

that

5

subjects

largely

that

. to

nearly

of

observer

and

at

or

twoness

) . Later slowly

mm

blind

development

( 5 ) . It

and

( 17

discrimination

that

was

confirmation

30

being

found

improvement

been continued

improved

subjects

on

had

from

transfers

was

seeing

it

threshold go

, the

the

area

what

the might

Moreover of

of

of

303

it ,

might

not

simple

variables

,

of

sti

,

nor

-

.

Conclusion

There is

is

no

there

the

evidence

theory ,

experience

that

The more

,

is

or

to the

This

perception

it

or

approach

for that

concerned

does

the

,

man

instead

for

man theory

but

of

is

all

.

It

misperception .

more

.

the

Perhaps

.

the learning

dependence on

of

stimulation

the

practical

is

not or

clues

to

question of

the

be

opened

admitted veridical

nothing

about

obviously

useful

personality

,

motives

whether around up

, has

for

says an

his

of world

is

must accounts . It

It

behavior

experiment

array

, it

are

perception

.

imagines

perceptual

perception

strength

thinking

abnormal

he

.

of for

external

discriminates

Perhaps

the . The

by

to

perceptualleaming

wishful

and

.

past

experiments

because he

of

past

by of

these

stimulus

basic

by

memories

not

incidental

require

development

in

because

to

enriched

determined

be

the

merely

points

the ' s

may

dependence

perceptions

with a

,

account or of

's

support

after

as

not

study

a

favorably field

the

to well

fantasy

less

no

stimuli

approach as

and

more

the

,

enriched

not

incorporates

less

this

section

is

one

finds , but

is

and

which

it

learning

last

for

that

variables

principles

, but

convinced

the

differentiate

weakness

imagination

assumes

learning

theoretical of

more

perceptual the

is

necessary

proof is

memories

contradictory

points

productive

or

to

to on

not

,

up

ability

no

progresses

sensitive

perception

is

is

one

percept

practice

hears

on in

is

accurate

there

and

summon

literature reported

perception

more

more

ability

affect

or

sees infers

this

percept

less

learning

observer

are

a

, but

as

of

experiment

accurate

learned

stimulation

and

an

whereas

percept

notion

all

the

. Repetition

improved

is

in that

experience

He

in

evidence

.

.

if

one

But

if

training him

is one

can ,

a

very

304

J. J. Gibson & B. 1. Gibson

Notes

1.Thispaperisa revision, withaddedexperimental materiaL ofonegiveninMay1953at a symposium onthepsychology oflearning basicto problems ofmilitary training (8) conducted bythePanelonTraining andTraining Devices oftheResearch andDevelopment Board, Washington, D.C.

2.Thisexperiment wasfirstreported atthemeeting oftheAmerican Psychological Associa-

tioninSeptember 1950ina paperreadbyEleanor 1.Gibson, andanabstract hasbeen

published (6).

3.Experiments onprimary stimulus generalization haveusually varied themagnitude ofa single difference, notthenumber ofdifferences, between thecritical stimulus andthe undifferentiated stimulus.However,our methodof quantifyingamount of difference has much to recommend

it.

References

1. Bruner, J. S. Personality dynamics and the processof perceiving. In R. R. Blake& C. V.Ramsey (Eds.), Perception: anapproach topersonality. NewYork:Ronald, 1951. Pp. 121147.

2. Brunswik, E. The conceptual framework of psychology. mt. Encyci. unif.Sci.,1952,1, No.

10.

3. Cantril, H.,Ames,A.,Jr.,Hastorf, A.H.,& Ittelson, W.H. Psychology andscientific research. Science,1949, 110, 461464,

491197,

517522.

4.Gibson, Eleanor J.A systematic application oftheconcepts ofgeneralization anddifferentiation to verbal learning. Psycho!.Rev.,1940, 47, 196229.

5. Gibson,Eleanor, J. Improvement in perceptual judgments as a functionof controlled practiceor training.Psycho!. Bull.,1953,50, 401431.

6.Gibson, Eleanor 1. &Gibson, J.1.Theidentifying response: a studyofa neglected form of learning.Amer.Psychologist, 1950,7,276.(Abstract) 7 Gibson,J. J. (Ed.)Motionpicturetestingand research. Washington, D.C.:Government

Printing Office, 1947.(AAF Aviat.Psycho!. Program Res.Rep., No.7.) 8.Gibson,J.J.,&Gibson,Eleanor J.Perceptual learning inrelationto trainingIn Symposium

onpsychology oflearning basictomilitary training problems. Dept.ofDefense, HR-HTD201-1,

1953, 151159.

9. Hebb,D. 0. Theorganization of behavior. NewYork:Wiley,1949.

10.Hilgard, E.R.Theroleoflearning inperception. InR.R.Blake &G.V.Ramsey (Eds.), Perception: anapproach topersonality. NewYork:Ronald,1951.Pp.95120. 11.Ittelson,W.H.Theconstancies in perceptual theory.Psycho!. Rev.1951,58,285294. 12.Jastrow,1.Psychological noteson HelenKeller.Psycho!. Rev.,1894,1, 356362.

13.Keller, F.S.Studies in International MorseCode.I. A newmethodof teaching code reception.J. app!.Psycho!.,1943, 27, 407415.

14.Kilpatrick, F.P.Human behavior fromthetransactional pointofview. Hanover, N.H.: Institute for Associated Research, 1952.

15.Leeper, R.Astudy ofaneglected portion ofthefield oflearningthe development of sensoryorganization.J. genet.Psycho!., 1935,46, 4175.

16.Postman, L.Toward a general theoryofcognition. InJ.H.Rohrer&M.Shenf(Eds.), Social psychology at thecrossroad5. NewYork:Harper,1951.Pp.242272. 17.Volkmann, A. UeberdenEinfluss derUebungaufdasErkennen raumlicher Distanzen. Ber.d. Sachs.Ges.d. Wiss., math. phys. Abth., 1858, 10, 3869.

18.Whipple, G.M.Manual ofmental andphysical tests. PartI.Simpler processes. Baltimore Warwick

and York, 1924.

PerceptualLearning EleanorJ. Gibson

Thechoice ofa review paperforreprinting mayseemdubious, sincereviews are

often merely annotated bibliographies, inevitably outofdate.This paper isnotthe usual review, however, since itwasanattempt toprovide structure fora new field. Italsopoints outsome timely events thatarenowmemorable ashaving a certain historical significance, suchasa book onperceptual development bya French psychologist (Frances, 1962) anda symposium atanInternational Congress that brought aninteresting anddiverse group ofpeople together. Ofmore general interest is thestriking influence ofartificial intelligence people andprograms already apparent inthetheories ofperceptual learning thatwere justbeginning to

surface. Morethana quartercenturyago,the contrastbetween modelsthat

emphasized feature listsandmodels thatemphasized matching toprototypes was clearly established. Thelong-term effects arestillwithus,forexample feature lists

ina recent model ofvisualidentification ofKosslyns (Kosslyn 1989), andmatchingtoprototypes ina bookaboutcategories inperception (Hamad 1987). Sources ofmyowntheory werealsointheliteratureJakobson andHalleon

distinctive features, andinformation theory. The fourareas thatI singled outfor reviews ofresearch inmybook were allrepresented andgavemethebasis forthe classification thatI adopted forthebook: improvement ofperceptual skills with practice, perceptual learning withimposed transformation ofthestimulus array, intermodal transfer ofperceptual learning, andthestudy ofperceptual developmentby meansof controlled rearing. There wasquitea lotofworkonpredifferentiation, relating tomydoctoral dissertation andsometimes taking offfromit,andresearch onacquired distinctiveness (acounter theory) thatwasgristforthemillasI putmybook together andweighed theories. Especially useful inthisrespect wasa quantity of research purporting tobeoneffects ofreinforcement onperceptual learningI Reproduced, withpermission, fromtheAnnual Review ofPsychology, Vol.14,295. ' 19O3by Annual Reviews Inc.

322

E. J. Gibson

summed it upin theconclusion thatextrinsic reinforcement wasnota necessary condition ofperceptual learningan understatement.

It isfuntolookbackonsome otherconclusionsthat attention hasmade a

comeback, andtheimportance ofthelearning-performance distinction thatTolmanhademphasized earlier. It wasrelevant then,andit stillis.I woundup pointing outtheneedfora theoryandtheprediction thatmorespecific cognitive theories ofperceptual learning areontheway. I waswrongtherethe cognitive psychologists haveseldom concerned themselves withperceptual learning .

It is onlya matterof threeyearssinceperceptual learning acquired the statusof an areaworthyof separatereview. In 1960this fieldwas re-

viewed byDrever(24),whoproposed a classification oftheories andsome

landmarks for summarizing research.The presentreviewwill makea furthereffortto clarifyandexposetheissueswhicharespecifically relevant

to theproblems anddefinition ofperceptual learning. Thedefinition ofthe

area,stilla newoneforpresent-day psychology, is not easyandrequired the reviewerto exerciseher own discretionin the selectionof material

whichis trulyrelevant,andbetterincluded herethanundersomeother

aegis.A systematic discussion of whatperceptual learning is doesnot belongina surveyofliterature, butsomeboundaries mustbeset. Theobviousrequirement fortheoriesandexperiments whichshouldbe

consideredis that the process studied be perception (not imaginationor

hallucination or guessing) andthatthekindof modification calledlearning be involved. Anyrelatively permanent andconsistent changein theper-

ception ofa stimulus array, following practice orexperience withthisarray, willbeconsidered perceptual learning. It couldbearguedthatonlychanges whicharebiologically adaptivewhichresultinincreased correspondence or

fidelity ofperception todimensions ofstimulationare relevant. Since this pointis arguable (whether onlyimprovement in thesenseof increased accuracy or veridicalityis learning) it willnot be a criterion forthe present review.

The definitionchosen does excludecertain phenomenawhich are con-

ceivably related. Anchoring effects andsensoryadaptation areexcluded sincetheyarenotpermanent modifications. Perceptual negative aftereffects andfigural aftereffects [Prentice (109)]areexcluded on similar grounds. Sensory deprivation, intheopinion ofsomewriters [Freedman (39);Held (60);Riesen(111);Teuber(124)], is a phenomenon relatedto perceptual learning. Butthechanges produced bytemporary deprivation ofstimulationin the adult(sensoryisolation)seemto be randomandbizarre,arein the natureof the casenot consistently relatedto presentstimulation, and

do not appearto be indicative of permanent modification. Hebb(58),in

commenting onthedistinction, says,So it seemsto methatwemaybe

PerceptualLearning 323

dealing withtwokinds ofprocess here, bothofwhich perhaps canbecalled learning, butoneofwhich ismorefundamental andlong-lasting, theother being a moresuperficial andtransient effect super-imposed onif (p.42). Another phenomenon moreclearly related to perceptual learning is

imprinting. However, the experiments on imprinting haveattainedsuch

numbers in thelastthreeyearsthatthistopicwillbe excluded on the grounds thatifisbetterreviewed separately. Animal studies ofdiscrimina-

tionlearning arealsoexcluded, except when theybearspecifically onone oftheissues critical fortheories ofperceptual learning. Illusions willbe

arbitrarily excluded, onthegrounds thattheirstatus withrespect tolearn-

ingis so inconsistent as to makethemuninstructive as a class,andon the

somewhat bettergrounds thata comprehensive studyof illusions and

development isavailable [Piagef (108)]. Perceptual development (growth

or ontogeny ofperception in theindividual) mightsurelybejustified for inclusion, butif willbeomittedasa topicbecause somerecent,extensive

surveys areavailable [Gibson&Olum(46); Wohlwill (134)]. Piagefs new

book(108) summarizes allthework oftheGeneva laboratory onperception,andthemajoremphasis, ofcourse, isdevelopmental. Thisverycon-

venientsummary includes theworkonillusions, theeffects ofcenfration,

theconsfancies, perception of causality, andperception of movement,

speed, andtime. Inthree concluding theoretical chapters, Piaget compares perception andintelligence, theorigins (perceptual ornot)ofintelligence, anddiscusses theepistemology ofperception. Thebookismarred bythe

totalabsence ofanysubject orauthorindex, buttheinformation isthereif

one has the patienceto findit.

Ingeneral, thematerial included herewillbeorganized around theories of perceptual learning, andsomemorespecific issuesandcontroversies

which have beenthefocus ofresearch. Representative experimental paradigms (reference experiments) areasyetsofewinthisareathatpara-

metric research isinfrequent, although ifcouldbeexploited inthecaseof some perceptual skills. General Treatments

Forthefirsttime, twobooks devoted entirely toproblems ofperceptual learning haveappeared. Theearlier ofthese, bySolley &Murphy (121),

summarizes a considerable amountof literature, but if is orientedtoward presenting the authorstheoretical pointof view.Theydescribe them-

selves, afonepoint, aseclectic; butthebook, which began withdiscussions ina seminar aftheMenninger Foundation, considers primarily theroleof

affect inperception. Detailed accounts ofresearch performed attheFoun-

dationarereported. Theotherbook,by R.Frances (38),hasa broader

outlook andwide coverage (there are408references, from French, English,

324

E. J. Gibson

Italian,German, andRussian sources). Thebookis lesscloselytiedto a

theoreticalbiasthanthe first-mentioned, but the authormakessomeuseful distinctions.Forinstance,he makesa sharpdivisionbetweenidentification and discrimination. For discrimination learning,biologicaladaptivenessincreaseswith fidelityto the objectivesituation;but for recognition,or

identification, hesaysthisis notthecase;particulars areneglected forthe sakeof categorizing, classifying intoa type.Topicscoveredin thebook includeimprovement of thresholds and sensorycapacities withpractice, impoverishment experiments, frequency andthemechanisms bywhichit mayinfluence perception, perceptual modifications inobjective characteristicsof content(constancies, etc).,the effectsof context,andthe effectsof motivation and individual habits and attitudes.

Bevan(11)hassummarized a numberofexperiments inperceptual learn-

ingandpresented hisownviewthat perceptual development depends on the evolutionof a systemof perceptual scales,througha processof differentiation. Dembers(22)textbookon perceptionincludesa chapteron

perceptual learning inwhich Dember acknowledges particularly theinfluenceof Hebb.KochsVolume4 includesa chapterby Attneavecalled

Perception andRelated Areas(5),buttheportiondevotedto learning andperception is disappointingly brief.A recentbookby Hunt(71)includesa chapterontheeffects ofearlyexperience onperception ofform and objects.

Perceptual learning wasoneof thethemes of the 16thInternational Congress ofPsychology heldat Bonnin 1960.Besides a number ofcontributed papersintheareaa symposium wasorganized byJ.J.Gibson (49). Thespeakers wereG.DeMontpelier, J. Piaget,E.J. Gibson, andA.V. Zhaporozets; discussants were1.Drever, M.D.Vernon, andI.Kohler. Theories

Drever(24)classifiedtheoriesof perceptuallearninginto four types

judgmental, stimulation, association, andadaptation theories. Rigid classificationwillbe eschewedhere.But,in consideringpossibilities of classifying,

someotherstronglydivisive principles emerge.Onecanthinkof perceptionas activelystructuringthe environment. Oneauthor(39),forinstance,speaksof imposingconstancies andstabilities on theperceived world. Or one mayconsiderperceptual learningas mainlya taskof discovering structure, order,andstability in theenvironment. Thereare

manywaysofhandling theproblem, whichever lineischosen. Anexample of the secondis an articleby Bruner,Wallach& Galanter(16)on the identification of recurrentregularity.They considerthat theoriesof re-

inforcementare not usefulfor this type of learning;rather, the answer

probably liesinhowtheorganism learns tousetechniques ofweighing the

PerceptualLearning 325

relevancy ofdifferent features ofinputforregularity-to-be-discovere (p.208). Whether reinforcement isgiven a roleornot,a discovery theory could beeither astimulation theory [e.g.,E.J.Gibson inBonn Symposium

(49)]or anassociation theory(a Brunswikian probabilistic type,forinstance), to useDreversterms.A Brunswikian experiment byFraisse (35)

illustrates theprobabilistic approach; proximity groupings ofonetypeof material exposed toSsduring a sortoftraining series wereweighted more heavily thananother type,andtheSseventually applied thegrouping

principle differentially to impoverished teststimuli.

Another distinction canbemade between theories ofperceptual learning

builtona classification or categorizing principle, andthoseof a feature

detection type.Itseems tothereviewer extremely interesting thatpro-

gramsforpattern-recognition bymachines [seeSeifridge &Neisser (117), fora review ofthese] alsodivide intothesetwotypes.Rosenblatt (113) describes theperceptronas performing by comparing newstimulus patterns withpreviously recorded prototypes andplacing eachstimulus in anappropriate class(p.12),andarriving spontaneously at perceptual concepts.Minsky (93)hasa clearstatement of thisapproach andthe

contrastingone. These pattern-recognition methodsmust extractthe

heuristically significant features oftheobjects inquestion. Thesimplest methods simply match theobjects against standards orprototypes. More

powerful property-list methods subject eachobjectto a seriesof tests,

each detecting some property ofheuristic importance. These properties have tobeinvariant under commonly encountered forms ofdistortion (p.10). Thefirstmethod employs whatMinsky refers toasprototype-derived patterns.Identification mayproceed by a normalization andtemplatematching process. Thetemplate matching scheme islimited ascompared

withtheproperty-list system.A property isa binaryfunction andmust

beinvariant under equivalence transformations. Minsky discusses theproblemofgenerating useful properties andalsoarticulationaprocess whichseemsroughlyanalogous to attention. Thematching toprototype model, analogous toa schema orcategorizing theory ofperceptual learning, hasalways beenpopular withpsycho-

logists.Vernon(129),longan advocateof the roleof the schemain

perception, hasagainpresented herideas, relating themnowtoaspects of

information analysis. Bruner andhisco-workers (15)haveemployed the ambiguiter, a lensmechanism which impoverishes thestimulus bypoor focusing, to studyperceptual identification. Theyemphasize thejudgmental nature ofSsbehaviorhypotheses, testing, categorizing, etc.The method ofimpoverishment isthetypical experimental situation usedby

schema theorists forstudying recognition. Experiments byBinder &tFeldman(12)withrecognition of ambiguous stimuli (briefpresentation and

reduction ofcues) investigated theroleoffrequency, reward, andpunish-

32

B. J. Gibson

ment.The S had to learnto respondwith the relevantidentification to test

stimulipossessingonly cueswhichhad beencommonto a previously

presented set.Recognition ofteststimuli withtheresponse madeto the commoncueswas relatedto previousfrequencyof occurrenceof those cues.Thetaskrequiredof S wasa classification, actually,to a prototype

whichpresumably emergedas the set of stimulipossessing the common featureswas presented.Uhr,Vossler& Uleman(125)had Ss learnto

respond withthepropernamefora patternovera setofvariant examples ofthepattern.TheSswere,infact,required to buildupa patternclass. Theirperformance wascompared withthatof a computer modelforpat-

ternrecognition. Thesimulatorout-performed thehumanSsin recognizing membersof a classdespitedistortion.Marx,Murphy& Brownstein(90) were likewiseinterestedin the recognitionof complexvisualstimulioccur-

ringin a varietyofdistorted forms. TheytrainedtheirSswithabstracted versions ofthepatternandthenpresented themwithtestitemswhichhad beenrandomly impoverished or hadrandomnoise added.The5, in this situation, is presented witha prototyperatherthanhavingto abstractit himself. Onetypeofabstraction (asimplegeometric design) wasmarkedly effective in thetrainingprocedure. Another, a linedrawing, wasnot.The question ofwhyonekindofabstraction wasbestisnotanswered. Perhaps thisquestion leadsto theotherapproachdetectingheuristicproperties or distinctive

features.

A distinctivefeature theoryis exemplified by Jakobson& Halles(74) treatmentof the distinctivefeaturesof phonemes.The emphasisin sucha

theoryisnotontheclass.Theclassisdefined by properties. Whatmatters is thedifferences in featurepatternswhichpermituniqueness of onephonemecompared to theothers.Theauthorspropose12featureoppositions whichare sufficient to give minimaldistinctions betweenall phonemes. Distinctivefeaturesare invariantunder certaintransformations, and are relationalproperties.Theymaybe orderedor stratifiedso as to suggest

theirdevelopment by a processofdifferentiation, or progressive splitting. Suchanelegantsystemhasnot,asyet,beenachieved by anypsychologist, butanattemptto applythenotionofdistinctive features to thedevelopment of letter-discrimination in the child has been made by Gibson (45).

Thediscussion proposesthatcertaindistinctive featureswhichdifferentiate objectsoftheworldaretransferred to letter-discrimination andthatother

specified features, previously notcritical butneeded to differentiate letters, are detectedafterschoolage whengraphicmaterialis encountered. It wouldbe naiveto supposethat the momentis at handfor makinga decisionbetweena prototypetheoryanda distinctivefeaturetheory,since

neitherhasas yet beenspelledoutin sufficient detail.An experiment by

Vurpillot &Brault (130)provides anexample. Theyshowed children of

different ages(fivethroughnine)familiar objectsona turntable, whichwas

PerceptualLearning 327

rotatedsothatS sawtheobjectfromallangles. Thenthechildwasshown eightphotographs of the objectin varyingorientations andaskedto choose theonemostlike it.Changes withageweretoward choice ofthe

mostinformative view. Theauthors concluded thatthechild develops a schema ofanobject. Butthedatasupport equally welltheideathatthe childdetects betterandbetterthefeatures whichdistinguish anobject

uniquely fromotherobjects. Itwouldbeunwise, however, to conclude that

thetwoconcepts, schema anddistinctive features, areperfectly parallel.

Progress toward moreadequate theories mightdepend onrecognizing twoseparate problems to be accounted for.One,therecognition of a

patternunderits transformations, hasalreadyreceivedconsiderable atten-

tion.Theother,howcritical features emerge or how,to useMinskys terms, a goodproperty-list isgenerated, needs work. Anexperiment by

Kossov(81)wasdesigned to studymethods of bringingoutessential

features ofperceived objects invisual perception. Heusedtwotechniques

of differentiation andconcluded thattransformation of a featureintoa dominantcomponentis betterachieved by variationof the essential featurethanby variation ofthenon-essential one.Moreworkof thiskind

should beuseful, eventhough critical experiments awaitprogress intheo-

retical models.

There havebeenanumber ofothercontributions totheories ofpercep-

tuallearning in thepastthreeyears,notnecessarily relevant to theissues

justexamined. Solley &Murphys book(121) isone.Thepointofviewis

wellcharacterized bySantos &Murphy (115): Weassume thatthisproactivity and,likemotor activity, itcanbemodified bylearning. Perception should therefore obeytheprinciples relevant toresponse learning (p.7). Conditioning isobviously oneofthelatter. Reward value andpunishment value maytherefore beconditioned tostimuli. Solley &Murphy divide cess(perception) hassomepropertieswhichare similarto thoseof motor

perception intostages:expectancy, attending, reception, trial-and-check,

andpercept. There isalooptoautonomic andproprioceptive arousal atthe

trialandcheckstage.Thisis a judgmental theory,to useDreversterm.

Trial-and-check istheanalyzing andsynthesizing process bywhich tentativeassumptions andsensory inputarestructured intopercepts(p.221). It is alsoa schematheory: As thesensory samples arememorically

accumulated, cognitive frames ofreference orschemata aredeveloped and

newsensorydataarematchedwiththe storedsamples(p.172).Of

course,it is associationistic too, sinceattentionis saidto be conditioned.

Butitismostofalldynamic. Forexample, setsdetermine which perceptualactswillbelearned (p.169); withyoung children a simple pleasurepainprinciple predicts whatwillbeorganized asfigure ina figure-ground learning experiment (p.142); andit isonlythrough theimpact ofsocietys negative reinforcement ofautistic perceptions thatveridical perception is

328

E. J. Gibson

ever achieved, even in part " (p. 78). Order is created: " Unless men were motivated to create an orderly universe in perception , there would be no functional utility in perceiving " (p. 222). The point of view of the book is

rather similar to the transactionalists , whose work is quoted.

The transactionalgroup also has producedseveralbooks in the past three years. There is Ittleson 's book on visual space perception (72), his chapter in the fourth Koch volume (73), and a volume of essays edited by

Kilpatrick (76). Most of the 21 papersin the last are reprints of previous journal publications by Ames, Cantril, Ittleson, and Kilpatrick. Ittleson again assertsthe view that perception rests on the "chaotic and fragmentary data of present experience" (73, p. 701). Perception involves a con-

stant checking and re-evaluating procedure, and "Perceiving is a creative process in which

the individual

constructs

for himself

his own world

of

experiences " (73, p. 697). There is not really a theory of perceptualleaming here, but the transactionalview puts great weight on past experience. Trace theories of perceptual learning do not seem to be much in evidence these days. But Epstein & Rock (29) and Epstein & DeShazo (28)

have treated perceptuallearning in terms of a gestalt type of trace theory, askinghow a recenttraceinfluencesthe current perception. Under marginal perceiving conditions, they think, there is oscillation of the "perceptual alternatives " and when the alternative appears which is congruent with the

recently implanted trace, arousalof the trace occurson the basisof distinctive similarity. This hypothesis is contrasted with an expectancy type of theory .

Finally, somecontributionsof informationtheoryto perceptuallearning should be mentioned . Information

theorists do not have a full -blown

model

of perceptuallearningbut they haveconceptswhicharepertinentandcan bel in the reviewer' s opinion, a healthy influence. The emphasison measurement of input-output relationships obviously fits a correspondence theory , rather than a creative process theory . Ways of accounting for increase or decrease in channel capacity , and information transmitted , may

eventuallyhelp clarify the presentvaguenessin theoriesof perceptual learning. Gamer's new book (41) considersa number of relevant problems, such as the effect of redundancy on pattern discrimination . The effect turns

out to be a very complex one, involving specification of amount of redundancy , kind of redundancy , and the requirements of the task. Frances

(38), as well, points out the importance of considering the perceptualtask in using information analysis. He discusses the nature of " repertorys " and

their formation, andhasreportedan experiment(37) in which perceptual learning of two kinds, (a) learning the set of alternatives and (b) adjustments of the receptor channel, had a greater effect on a recognition task than on

a discrimination task. Broadbent (13) has applied his filter hypothesis to discrimination learning and makes the interesting suggestion that "the

PerceptualLearning

329

mechanism which attaches outputs to inputs , stimulus to response, acts only on filtered information " (p. 266). Barlow (9) suggests that reduction of

redundancy, or "compression," is an essentialmechanismin the coding of sensory messages. He says that the features of the sensory input which are discriminated

from one another are the very ones which enable the sensory

messageto be compressed . "The idea of removing redundancy and so compressingthe sensory messageinto the smallestpossible channel was brought forward becauseit seemedthat the subsequentstep of learning to discriminate

different

sensory

stimuli

would

be easier if the sensory

in -

formation was presentedin this compressedform" (p. 359). This idea seems to link up very nicely with a distinctive feature theory of perceptual learning. Facallssuesfar Research Whatever their relation to theoretical positions , a limited number of issues

are the focus of most researchin perceptualleaming. The questionsasked are seldom capable of an immediate decisive answer, but the phrasing of them is becoming sharper. Nativism -empiricism The nature-nurture question is one of those often supposed to be unanswerable. Nevertheless , sorting out the relative contributions of genetic and environmental factors continues to be a fascinat-

ing pursuit. Whether the abilities to perceive the spatial aspectsof our environment

must be learned or are more or less built -in is of interest

to

psychologists as much today as ever. One finds eloquent backersof nativism [Pastore (103)] as well as of the more popular empiricist position. Hochberg (67) haswritten a full discussionof the historical background, the

logic , and the present status of the Question. .&

The classictechnique for investigation has always been study of the effectsof deprivationof somenormally presentenvironmentalcontribution, suchas stimulation by patterned light . The "born blind" patients corrected (or somewhat corrected ) by surgery are the most dramatic cases. The cases

summarizedby von Sendenhave appearedin an English translation (118), but shedno more light than they ever did, sinceall the confounding effects of nystagmus , emotional upset, interference from old habits, and unreliable testing procedures are still inevitably present in the reports . London (86) has summarized a Russian report of six cases of "postoperatively newly seeing persons." (Choice of a proper label for such research is difficult , for

thesepeople were not born blind but developed cataractsat various fairly early ages after birth .) Progress to " fuller vision " seemed to be related to

age when blindnessoccurred and to length of visual deprivation; but still one never knows how much of the slow progress is occasioned by reliance

330

E. J. Gibson

on habits of recognition by touch and auditory infonnation

necessarily

utilized during the blind period. Gregory & Wallace (54) have reported a casewhich they themselvesstudied at some length. Visual discrimination and recognition seemedto be far more adequatein this casethan in most of those reported. Whether the patient was more intelligent, the testing procedureswere better, or something elseis hard to say. Axelrod (6) studied the effects of early blindness on perceptualperformance in modalities other than vision by comparing blind and sighted children . Differences

between

the groups were not consistent

in direction .

The early-blind Sshad lower two-point limens than sighted Sson the right index finger (presumably due to practice), but they were not superior in light-touch sensitivity. The early-blind Ss were inferior in tasks which requiredabstractionand in transferof a principle of solution acrossmodalities (haotic to auraLfor instance). The taskswere said not to be spatial. The reasonfor this deficit is not clear, though it seemsto indicate the important role of visual experiencein cognitive development in general. Deprivation of a lessdrastic type was exploited by Hudson (70) to study perception of pictorial cues to depth, in sub-cultural groups in Africa (groups often thought to be pictorially naive). He manipulatedobjectsize, ..

superimposition , and perspective , in the conventional artistic manner, to

createpictures in a seriesroughly increasingin redundancyof thesepictorial cues. His results seem to indicate a difference between school-going

and non school-going populations, the former more often reporting threedimensionality . Hochberg & Brooks (68) reared their infant son in an

environment pictorially impoverisheduntil he was 19 months old. Though pictures were occasionally in his field of view , he was never given a name

or specificresponsefor one (i.e., his vocabulary was taught entirely with objects). Despite this, tests with pictures at 19 months revealed that they

were recognizedand labeled. Pictorial recognition, in other words, was not dependent on associationbetween the picture and either the represented object or a naming response.

Although the problemsof control and interpretation are formidablewith natural casesof deprivation, experimentaldeprivation with animalsis feasible and often undertaken. Much recent experimental work with rats makesclear the negligible effect of dark-rearing on visual discriminationin this animal. Gibson, Walk & Tighe (48) found that dark-rearedrats learned a triangle-circle discrimination at about the same rate as ordinary laboratory reared animals . Dark -reared rats tested on a " visual cliff " [Gibson

&

Walk (47); Walk & Gibson (132)] performed like their light-reared litter mates, consistently choosing a shallow over a deep visual drop-off. Nealey & Edwards (97) confirmed this result , with added controls . Young rats

(relatively inexperiencedvisually) avoided the drop-off as consistently as mature ones (132). Results of experiments involving cue isolation (132)

PerceptualLearning

331

suggested that some cues may be "built -in" whereas others are a function of previous visual experience (the texture cue in isolation was effective in

light -reared but not in dark-reared animals). Walk & Gibson (132) also found that chicks less than a day old avoided the visual cliff with perfect consistency. This result has been confirmed by Tallarico (122) with newly hatched chicks. Very young goats have been shown to avoid the cliff (132).

Early appearanceof various perceptualachievementsin ducklings has been reported by Pastore (102).

On the other hand, Riesen(111) has suggestedthat with encephalization of the visual system, patterned visual stimulation becomes essential for

ontogenetic development. Dark-reared kittens did not develop a normal visual placing response. Riesen& Aarons (112) found that kittens reared without patterned-light stimulation or with patterned light for one hour, but with restraint

of movement

in either condition , failed to learn a visual

movement discrimination in twice the time required by controls. Gibson & Walk (47, 132) found , likewise , that dark-reared kittens tested at 28 days

had not developed a visual placing responseand did not avoid the deep side of a visual cliff, unlike their normally reared controls. However, discriminative behavior on the cliff recoveredafter the kittens were brought into the light, without differential reinforcement, suggesting that maturation in the light is a sufficient condition for development of this behavior. Rhesus monkeys tested on the cliff at 10 days (132) could be coaxed to

the deep side, though locomotion differed on the two sides. At 18 days, they could not be coaxed to cross the deep side. Ganz & Riesen (40), comparing dark- and light-reared Macaques, found evidence for modified generalizationgradients to hue in light-reared animals, which they interpreted as indicating the importanceof sensorylearning. Human infants, as soon as they can be tested on a visual cliff, show adequatediscrimination of visual depth (47, 131,132). But testing is not possible before locomotion is possible. Fantz (32) has used a preference method for testing shapeand pattern discrimination in human infants, as well as other young animals, and found convincing evidence for differential

responseto pattern, suggesting that at least some degree of visual form perception is innate even in the human infant . Enhancement

or " enrichment " experiments

have been done , in contrast

to deprivation ones, to see if the enhanced experience facilitates later perceptualperformance. Sometimesresults are positive and sometimesnot [Gibson, Walk & Tighe (48)]. Meier & McGee (91) obtained positive results with rats in an " enriched" rearing experiment and concluded that pro -

gressive differentiation of the objects in the environment had taken place during rearing, thus speedingup a later discrimination task. Positive results such as these do not bear precisely on the nature-nurture problem. They show that perceptualleaming may occur, but not that all perception de-

332

E. J. Gibson

rends on learning, as a radical empiristic view would hold. Fowler (34), surveying cognitive learning in infancy and early childhood , summarized

casesof early cognitive learning (language, reading, drawing, etc.) and emphasizedthe role of IIenrichment" and learning, as opposed to genetic factors

.

Cross-modal transfer

One of the questions most often asked about the

"born blind" casesis whether or not acquaintancethrough one sensory modality will transfer without further practice to a different one. Could the

patient with newly restored sight recognize visually something which he

had previouslyknown only tactually? Gregory's case(54) provided a remarkableexample of such transfer with upper-caseletters. The patient had learned to recognize by touch raised upper -case (but not lower -case) letters . After

surgery

, he could

name

the

upper -case letters , but

not

the

lower-caseones, when they were presentedvisually. This evidenceis the more convincing , because the lower -case letters , never identified by touch ,

provide a control. A number of Russian experiments on transfer from haptic to visual recognition have been briefly summarized by Zinchenko &

Lomov (137). Patternedfigures were preparedfor both tactual and visual presentationand crossmodaltransfer was studied in both directions, with children as subjects. Acquaintance by one modality did transfer to some

extent to recognition by the other. Zinchenko & Lomov emphasizedthe role of exploratory movements (both tactual and visual). "They construct an image of it, take a cast or a copy" (p. 16). There is not perfect isomorphism, however, for different fingers describe different paths, and eye movements vary with restriction of the field . These authors suggest that

touch and vision are "parallel senses " and that there is a genetic connection between

them .

Ettlinger (31), on the other hand, found no cross-modal transfer in

monkeysin discriminatinga spherefrom a pyramid, either tactual to visual, or vice versa. Is this because cross-modal transfer is verbally mediated? An experiment by Hermelin & O 'Connor (65), again on tactual and visual

shaperecognition, makesthis unlikely. The Sswere normal and subnormal (imbecile) children. Five stimulusobjects (Russianletters) were inspectedor manipulated(but never named) and later selectedfrom a group containing five new ones. Better than chance cross-modal recognition occurred for both groups .

Rudel (114) studied the decrement of the Muller -lyer illusion in adult Ss

with repeated exposure in both the visual and tactual modalities, and measured transfer from one to the other . There was cross -modal transfer of

the decrementin both directions, slightly greater from haptic to visual. It seems unlikely in this case as well that there is verbal mediation of transfer.

Caviness&: Gibson (18) comparedthe effects of two kinds of training on

Perceptual Learning 333 cross-modal transfer in adult Ss. Solid, unfamiliar "nonsense" objects were presentedhaptically and matchedvisually. Training, both by visual-tactual associationand by tactual comparisonalone, resulted in transfer as compared to a control group without training. One experiment attempting auditory-visual transfer has been reported by Chorover, Cole &: Ettlinger (20). Resultswere negative. Why cross-modal transfer is found in some cases , but not others, and what IIcarries" the transfer when it occursare clearly questionsfor further study. Compensation for transformedstimulus-arrays Experimentson optical distortion produced by spectaclesor goggles comprised of mirrors, rightangleprisms, wedge prisms, or half-colored glasseshave continued, follow ing the earlier work of Kohler (78). The experiments have usually been performedwith the view of discoveringhow perceptionof spaceis learned. It is assumedthat the compensationor rehabituation which takes place while wearing the spectaclesis analogousto the original development of perception in the child. This reasoningis dubious, but the experiments are ~ instructive for other reasons. What to call the changein stimulus-array is important, sincethe wrong term can prejudicethinking. All the casesso far investigatedare transformations, rather than pennutations, of spatial order, so that the new stimulus order is still in systematiccorrespondencewith the old one (right-left reversal, for instance, is still perfectly correlatedwith the normal order). The term IIdistortion" was criticized by Held (60), who suggestedthe terms "rearrangement " and "disarrangement ." Rearrange ment would apply to the transformations so far studied, like reversals. Disarrangementwould apply to a random change no longer in correspondencewith the normal order. Cohen & Held (21) attempted to create such a situation with rotary prisms of continually varying power, which produceda continually varying image of S's hand ashe viewed and moved it. Dispersion of target markings (test for compensation) increased , rather than decreased , indicating "degraded"perforrnance. In Kohler's (80) recent summary of his goggle experiments, he comments on the complexity of the adjustment demandedby some of the transformations, such as that causedby two-color glasseswhen the eyes move from left to right . Adjustment to this change "involves a process more complex than all previously known processesof visual adaptation" (p. 67). Yet it occurs, as long as there is ordered correspondence . Kohler comments that "in all caseswhere adaptation occurs the eye is provided with systematicclues as to the nature of the distortion" (p. 72). The fact that adjustmentor compensationtakesplacewith ordered stimulus-change but seeminglynot with random change(seealso 60, 61, 124) fits well with the view that perceptual learning is essentially detection of order and

334

E. J. Gibson

regularity rather than creation of it . Kohler (79) has made this point in a theoretical article contrasting the classicgestalt organization theory (internal organization ) with an ordered stimulus theory (external organization ).

Held and his co-workers (21, 62, 63) have performeda numberof experimentsin this field. These papers emphasizeparticularly the role of spontaneousor "self-produced" movements made by the subject and the feed-back stimuli dependent on these movements for the occurrenceor nonoccurrenceof compensationfor errors produced by prisms. Order in the stimulation coming from the environment, plus self-produced movements and the resultant feedback are held to be essential for normal sensory

development as well as for adaptation to "rearrangement " (59, 62). "Reafferent stimulation

is the source of ordered contact with

the environment

which is responsible for the stability, under typical conditions, and the adaptability, to certain atypical conditions, of visual-spatial performance" (62, p. 37). In the Cohen & Held experiment (21) mentioned above, the

authors proposed that self-produced movement was necessaryfor degrading performanceunder a varying transformation as well as for compensation for invariant transformations .

Differentiationand acquireddistinctivenessThat originally confusableobjects of stimulation may become discriminable with practice, or that the

original level of discriminability may be enhanced, is generally acceptedas

a fact of perceptuallearning.But underwhat conditionsdiscriminabilityis increased, and by what mechanism or process, are far from settled . Two

opposing views have been the basesof hypothesesfor most of the research to be reported. These were called by Gibson & Gibson (50) the "differentiation " view and the " enrichment " view . The differentiation

view holds

that practiceservesto reducegeneralizationamong the stimuli, to increase precision of discrimination of variables actually present in stimulation , and

to detect relevant variablesor distinctive featuresnot previously detected. The position holds that the effective stimuli for perception are changedby learning , and it requires a new concept of stimulation . The enrichment

view, on the other hand, emphasizesaddition to the stimulus impressions by an associative process. An older form of this view , Titchener 's context theory , would assume that stimulus -impressions become more distinctive

by the addition of meaningfulassociations . A relatedview more in keeping with behavior theory was originally proposed by William Jamesand later developedby Miller & Dollard (92). The processis generally referredto as "acquireddistinctiveness." It holds that responseslearnedto the stimuli add response-produced cues to the original stimuli. The more distinctive the responseslearned, the more effective the response-produced stimulation will be in increasing distinctiveness of the original cues. Language responses(labels) are often employed in research. The theory also demands

PerceptualLearning 335

acquiredequivalence (greaterconfusability of cues)withtheattachment

of the same label to different stimuli.

Thedifficulty ofdeciding between thesetwohypotheses maybeillustratedbyanelegant experiment byLiberman et al.(85).Theymeasured

discriminability ofacoustic differences within andbetween phoneme boun-

daries,andcompared the datawithcomparable acoustic differences in

sounds notperceived as speech. Discrimination wasconsiderably better across a phoneme boundary thaninthemiddle ofa phoneme category.

Discrimination datawiththe controlstimulirevealedno increasein dis-

criminability intheregion corresponding to thephoneme boundary. The authors conclude thatsharpening ofdiscrimination atthephoneme boundaryis aneffect oflearning, andmorespecifically is anacquired dis-

tinctiveness dueto longpractice in attaching phoneme labels.Thatthe sharpened discrimination isaneffectoflearning seemsclear,buttheroleof

labeling isnotsoclear, sinceacquired equivalence (higher thresholds) did

not occurwithinthe phonemecategory,relativeto the controls.The

heightened discrimination could,moreover, equally wellhavebeendueto finerdifferentiation of thecomplex variable.

An experimentby Chistovitch,Kiass& Alekin(19)corroboratesthe

difference indiscriminability ofspeech andnonspeech sounds. Theytaped

soundsequences ofthreeRussian vowels orthreepuretonesandaskedSs

eitherto identifythe sequences by namingor to makea same-different discrimination. Sequences of vowels(named by vowellabels)wereboth betteridentified and betterdiscriminated as same-or-different thanwere

sequences of tones(labeled high,medium, or low).Butwhenhightones

similar tocertain vowel sounds wereidentified byvowel labels, asopposed to the wordshigh, medium,or low, moreinformation wastransmitted. Thisresult suggests thatthelabelusedforidentification mayplaya rolein

theidentification taskitself,butnotinacquisition ofdistinctiveness.

Lane &Moore (83)trained anaphasic patient todiscriminate twophone-

micpatternspreviously indiscriminable by conditioning withlabels.The aphasic pressedoneor theotheroftwosuitably labeled buttons,anda flash

oflightfollowed ifthecorrect onewaspressed. Differential responses were

quickly acquired, andaccuracy transferred to an ABXdiscrimination task. Thiswasa dramatic caseof acquireddistinctiveness. Butthe dramatic

change occurred sorapidly (after a fewminutes) thatonewonders again

what the role of the label,as such,was.

Anexperiment byPhaup&Caidwell (107)wasaimedagainst thedifferentiation hypothesis, ratherthantowardtestingacquired distinctiveness. Theyrepeated anearlierexperiment by Gibson&Gibson(50)butadded

familiarizing training(looking threetimesat alltheitemswhichwerelater to be discriminated fromthe standard). Because familiarization facilitated

336

E. J. Gibson

the later achievementof discrimination, they concludedthat enrichment, as well as differentiation, must be involved. But the familiarization could, of course, simply have provided an opportunity for differentiation of those features which made the items distinct. E. J. Gibson [in (49)] discussed whether one could "have it both ways" and proposed that differentiation must precedeassociation, even with stimuli suchas written symbols. The meansof investigating these opposed processesexperimentally is the so-called "predifferentiation" experiment. In this paradigm, a set of stimuli are first tested for discriminability or identifiability. Training of some type is then given, and finally discriminability or identifiability is retested. Ideally, a control group or, at least, control stimuli are included. The effectivenessof the training is measuredby transfer to the final discriminability test. There are obviously a large number of possiblevariables in this type of experiment. The critical one for the two opposedtheoriesis the type of training introduced in stage two. It might be a matter of learning different labels or other specific responses ; or it might be equivalent exposureto the stimulusmaterial with a nonassociativetask suchas comparing one item with another. But other parametersmay be important in determining whether, or how much, transferoccurs. Two of theseare the kind of stimulus material and the nature of the criterion task. Several recent experiments have compared types of training in the predifferentiation experiment. DeRivera (23) had Ss learn letters as responsesto fingerprint-patterns (training), and afterward learn numbers as responsesto them (transfer stage). One group learned a specificletter for each print. Two other groups learned only two letters, half the prints having one letter and half the other as a common response. One of these two groups was told to learn the responseto each print individually; the other was told to look for something common to the prints having the same letter-response. Transfer to the second task (numbers) was significantly poorer for the group told to look for common characteristics , but was equivalent for the other two groups. It was concludedthat the labels themselveswere irrelevant and merely forced 5 to searchfor discriminable aspectsof the stimuli. An experiment by Pfafflin (106) was run with three conditions of pretraining: (a) relevant labels learned to visual forms (of three levels of meaningfulness ), (b) irrelevant labelslearnedto the sameforms, (c) observation of forms. A discriminationtask followed (5 pressedone of two buttons for eachform). All kinds of pretraining resultedin a decreasein errors, over all, compared to a control group. Observation, over all, resulted in the greatesterror decreaseand was the only condition which producedfacilitation with all levels of meaningfulness . With highly meaningful forms, learning relevant labelsproduced interference. With very low meaningful-

PerceptualLearning

337

ness , relevantlabelswerefacilitating, probably, pfafflin suggests , because they helped S to selectdistinctive featuresof theseitems. An experiment by Ellis et al. (26) tested the hypothesis of acquired distinctiveness and acquired equivalence , employing a recognition task with tactual shapes. The three pretraining conditions were: (a) learning relevant labels, one to each"prototype" shape; (b) learning one label to half the shapesand another to the other half, both relevant; (c) observation by tactual inspection. Group a was not superior to Group c in the criterion (recognition) task. Group b was poorer, probably becausethe 5s disregarded all distinctive features save the one suggestedby the label. The addition of response-producedcueshere did not facilitate later recognition, beyond the instruction to inspect and differentiate shapes. A little further evidence is provided indirectly in an experiment by Rasmussen& Archer (110), who tested the effect of languagepretraining (learning relevant labels to nonsenseshapes ) on a concept identification task. It was hypothesized that pretraining with labels would enhancediscrimination and thereby facilitate concept identification . It did not , but

another group which judged the shapesfor aesthetic qualities instead of learning responsesdid perform better on the criterion task, presumably because they had attended to and discriminated among the several dimen -

sionsof the stimulusfigures. To summarize,none of thesefour experiments yields decisivesupporting evidencefor acquireddistinctivenessin the form of cues generatedby specific labels or responseslearned to the stimulus items. But pretraining of any kind that allows detection of distinctive features appears to be facilitatin ~ .

Another parameterof the predifferentiation experiment is the nature of the stimulus items, i.e., their complexity, meaningfulness , and the original degreeof confusability with other items of the set. Obviously, there is no point in this kind of experimentif the items are already easily discriminated or identified . Neither is there if they are so nearly identical as to be beyond

the basiclimits of the discriminatory capacityof the sensorysystem. Albert (1) had 5s repeat labels (given by E) while hefting weights previously shown to be indiscriminable, on the theory that learned cue-producing responses would

facilitate

later discrimination

. It did not . The author sug -

gestedthat it did not becausethe weights were literally below the physiological limit of discriminability. Vanderplas& Garvin (127) studied the role of complexity and associationvalue in shaperecognition following labeling practice, testing the hypothesis that the initial difficulty of the material determinesin part the effect of practice on recognition. Complexity turned out to be the only main variable of significance, and there was a significant complexity by practiceinteraction. Labeling practiceas suchwas not significantly related to either correct recognition or correct rejection, as might

338

E. J. Gibson

have been expected from an unelaborated acquired distinctiveness (response-produced cue) hypothesis. The third parameter of importance in determining whether transfer has

resulted from predifferentiation training is the criterion selectedfor measurement. Yarczower (135) showed that predifferentiation training was effective in lowering generalization (conditioned galvanic skin response) even though similar training in another experiment did not appear to facilitate recognition. Lenneberg's (84) study of the effect of color-naming habits on discrimination gives further evidence that the criterion task must

be specified. He found that hue discrimination (differentiallimens for simultaneouslypresentedstimuli) was not affectedby naming habits but that color recognition was, to the degree that memory was taxed and that S was forced to search for anchoring points . He commented that " semantic habits

provide no absolute, invariable meansof distinguishing stimuli, but serve as a device for classificationor articulation of a continuum and thus help us in many situations to find points of reference, 'anchorage,' for judgments, whatever they may be" (p. 382). Hayes, Robinson & Brown (57), in a

replication of an earlier experiment by Henle (64), found that normally oriented

letters were identified

better

than disoriented

ones , due presum -

ably to past experience.But when a sameness -differencetest procedurewas employed, there was no significant difference in thresholds for the two types of orientation .

Vanderplas (126) has suggestedthat five kinds of perceptual tasks be distinguished (detection, discrimination, recognition, identification, and judgment), since transfer may depend on similarity of operations in the training and the transfer task. Vanderplas, Sanderson& Vanderplas(128) reported an experiment in which transferwas determinedfor severalkinds of training (observing and discriminating, or labeling random shapes) to several kinds of criterion task (discriminating, recognizing, labeling, or differential switch -pressing). Transfer of training (relative to a control

group), by both a discrimination and a recognition test, was equal for the several kinds of pretraining . By an identification measure (learning labels),

observing and discriminatingyielded sometransfer, but not asmuch asprevious labeling practice. With the motor task (differential switch-pressing) as criterion, transfer occurred for all groups but slightly more for one group which learned labels (girl's names were better than nonsensesyllables). There was, thus, over -all transfer from all three kinds of training to all four

criterion tasks, but the amount dependedto some extent upon the relation of the two

tasks .

Future progresswith regard to this issuewill depend, in the reviewer's opinion, on more creative handling of the hypotheses, and on careful

selectionof experimentaldesign with regard to the parametersjust discussed .

PerceptualLearning

339

Educationof attention In considering how cues become distinctive, or to put it the other way , how distinctive features are selected and differentiated

from the potential stimulus input, it is useful to talk about attention and its education . The word attention , for years suspect, has made a come-back,

though somebehaviortheoristsprefer to speakof observingresponses [Atkinson (4) for instance, uses"observing responses " as criteria for what might well be selectiveattention]. Broadbent's filter theory (13) is another way of dealing with attention. Solley &: Murphy (121) speakof the "conditioning of attention," which puts them in head-on conflict with Broadbent, who hypothesizes that only already - filtered stimuli can be conditioned. Broadbent's hypothesis is similar to one proposed by Zeaman & House (136), who developedan attention theory of discriminationlearning in retarded children. Their theory is a two -stage one which assumesthat attention to the relevant dimensionmust take place before any differential

cue-learningcanoccur. That is, acquisitionof two responses is required : (a) attending to the relevant stimulus dimension and (b) approaching the correct cue of that dimension. The retardates' difficulty is with (a) not (b), since"reverse" learning curvesall looked alike at the end. There is transfer within stage (a); an intradimensional shift was very fast, but an extradimen sional shift was difficult . Furthermore , if an easy discrimination within a dimension is run first , it will transfer to a hard one . A discrimination

problem is hard for retarded children when stimuli are patterns painted on a background, whereas the same patterns as cut-out three-dimensional objects are easily discriminated . House & Zeaman (69) found that the

difficult pattern discrimination could be accomplishedif it were preceded by the cut-out discrimination (68 percent of such a group learned in 10 days compared to only 10 percent of a group without transfer from cutouts). Walk et ale(133) explained in a similar way the facilitating effect of prolonged exposureto metal cut-out patterns on later discrimination learning in rats, as opposed to no facilitation from similar exposureto painted

patterns .

Karn & Gregg (75) reported an experimentwhich seemsto the reviewer to be concerned

with the education

of attention . The 5s were shown three

circlessimultaneously on a screen , tachistoscopically , and askedto report presence or absence of a dot within

the circles . The circles were so " loaded "

with the dots that there were always two stable conditions and one unstable . If 5 detected

these conditions , attention

could be directed

to the

unstable condition and accuracyimproved. With repetition it was found that accuracy did improve , without reinforcement of any kind . Improve -

ment was not over all a suddenshift, but was very rapid for targets which were consistently positively loaded. It seemsimpossible to describe this type of learning as conditioning, but it is certainly perceptual. Kossov's experiment (81) is also relevant here.

E. J. Gibson

340

Learning

to

education

disregard

of

tactual

discrimination

board

.

The

which

task

5

was

was

brought

the

years

)

not

a

be

)

.

Pre

was

of

(

7

"

8

)

.

or

younger

children

,

on

,

but

,

pretraining

age

,

group

(

therefore

,

on

under

here

.

9

-

to

10

be

a

Education

difficult

Zeaman

a

about

task

older

going

as

a

tacks

noninterference

seems

features

make

scattered

groups

There

of

to

raised

tacks

an

learning

dimension

with

age

with

aspect

5s

by

the

two

of

another

required

"

with

years

)

interfering

of

kind

be

53

made

only

the

relevant

(

training

5s

-

may

patterns

effective

in

input

Gollin

with

with

the

-

different

one

slower

stimulus

by

disregard

younger

to

retardates

conditions

found

it

to

be

with

.

Response

bias

scopic

There

is

recognition

of

Psycholinguistics

plague

for

contention

of

learning

function

ant

of

influence

.

It

is

of

.

It

a

That

shown

by

"

method

of

L reported

But

on

the

response

other

bias

bias

;

.

Brown

effect

an

level

more

interpreting

& :

words

;

and

Schiff

found

that

.

It

116

)

)

,

comparing

the

14

experimental

,

of

then

that

a

forced

,

Does

it

going

to

]

.

role

in

&

of

response

experiments

a

is

that

-

blank

slides

with

significantly

to

but

frequency

subliminal

factor

;

sensory

concluded

correctly

a

.

response

presumably

guessed

)

reduced

as

word

shown

at

96

count

that

,

S

(

word

well

the

group

bias

a

in

to

set

stimuli

)

Humbert

found

producing

.

119

presented

perceptual

as

forms

(

than

-

method

&

recognition

of

experi

choice

was

a

realm

Kanfer

Moscovici

auditory

control

the

usage

-

stimulus

)

in

of

experiment

-

group

,

tachistoscopic

set

,

.

are

still

that

word

using

a

a

limits

frequency

predictor

another

) ,

negligible

we

Smock

found

in

-

to

guessing

if

issue

by

better

Kanfer

a

oper

held

of

with

no

a

high

stimulus

must

a

( 51

some

shown

seems

results

99

(

least

It

)

which

of

at

group

often

& :

Rubenstein

(

experimental

,

Smock

52

and

Hawkins

determinant

have

)

was

(

a

98

a

sometimes

method

the

(

in

&

Newbigging

?

of

is

encourages

in

Ratliff

(

,

all

The

effects

recognition

the

a

.

results

&

frequency

was

stimulus

affect

experiments

hand

of

is

.

be

of

a

leave

can

Goldstein

emitted

information

experiments

Goldiamond

that

at

one

Newbigging

pseudoperceptual

rthe

card

of

that

is

by

perceptual

bias

to

,

role

obtained

stimulus

~

the

"

learning

appear

situation

are

number

bv

make

-

under

bias

perceptual

availability

to

tachisto

response

manifestations

tachistoscopic

a

"

instance

only

on

recently

called

in

Response

as

the

the

shown

was

much

response

was

-

on

-

what

( for

really

.

experiment

Results

theory

are

literature

reviewed

so

that

words

the

was

of

is

thresholds

the

it

interested

)

that

all

since

issue

psychologist

so

true

what

consider

the

of

report

review

,

But

responses

when

matter

.

perception

verbal

is

especially

)

frequency

the

any

to

material

psychologists

word

level

occasion

113a

the

the

of

(

many

on

no

verbal

"

special

"

-

difference

may

ment - ~

same

"

experiment

task

attention

if

(

noisy

improvement

developmental

of

An

presented

to

interference

,

.

also

told

about

with

"

"

attention

that

be

considered

the

instructions

in

Perceptual Learning 341 given 5 (to guess, or to try for accuracy of perception) and the stimuli presented are also effective. That the number of response alternatives affectsrecognition threshold was recently demonstratedby Fraisse& Blancheteau(36). It is clear that the size of the set would influenceguessing. That this question of response bias is not simply a red herring becomesclear in the interpretation of many experiments. Newbigging & Hay (100) made predictions for recognition thresholds on the basis of the differentiation-enrichment issuein perceptuallearning. The differentiation hypothesis should predict, they contended, that as trials progress, "more" of the stimulus word will be discriminatedat a given exposure(responses should show increasingsimilarity to the actual word). They then test this hypothesis by examining 5' s successiveresponsesto a particular stimulus item as he works up to threshold and conclude that the prediction is not confirmed. But to what extent perception is involved is a seriousquestion sinceS was instructed to "makea guess" after eachpresentationas to what the word was. There is no good reason to suppose that the laws of guessingare the sameas the laws of perception. The problem of response bias is also a vexing one for interpretation of experimentswith ambiguous figures, some of which are describedbelow.

Reinforcement in perceptual learning To whatextentis reinforcement , either theeffectsof rewardandpunishment or knowledgeof results , necessary for perceptuallearning7 Solley& Murphy (121) emphasized the role of reward and punishmentin perceptuallearning , as noted earlier, and quoted a numberof pertinentexperiments asevidence . Threequestionsfor defining theirrolesmaybeasked : (a) Is the observedeffectof rewardor punishment a functionof the criteriontest? (b) Is the effectof rewardandpunishment differential ? (c) Is their role morethaninformational ? As for the first question , thereseemsto be evidencethat rewardinga categoricalresponseto ambiguousmaterial , suchasinkblots, increases the frequencyof theresponse [Essman (30)]. Butheretheresponse biasrearsits ugly head. Is this effectmerelyverbalconditioning ? Solley& Engel(120) attemptedto eliminatethis possibilityby training an S (with rewardor punishment ) to nameone or anotherprofile in an ambiguousfigure and latertestingthe effectof the trainingby havingS merelypoint to indicate which facehe saw (i.e., makea responsewhich had not beenspecifically reinforced ). Sincethe rewardedprofile was more often pointedto under these conditions , they concludedthat a genuineperceptualshift had occurred . The differentialeffectof rewardand punishmentwas alsoobservedin Solley& Engel's experiment(120). Money rewardandwithdrawalserved as reward and punishment . They concludedthat differentialeffectsoccurred, but changeddevelopmentally , the youngerchildrenrevealingan

342

E. J. Gibson

IIautistic" effect of punishment, whereas the older children revealed its effectsas "attention- getting." This finding seemsto shedmore light on the role of punishment (and its interaction with development) in the motiva-

tion of actionthanin learning.In anotherexperimental with profilesusing money reward and withdrawal , Beatty , Dameron &: Greene (10) found both ineffective as reinforcing agents. No differential effect was observed be-

tween rewarded and punished stimulus alternatives. A different criterion, the threshold

for correct

identification

, was used . McNamara

(87 ) found

that reward and punishmentwere effective in producing constant errors of over- or underestimationof lines, direction of error being differential in the predicted directions (e.g., reward for overestimation produced errors of overestimation; punishmentof overestimationled to underestimation). The effect transferred to different figure. The shift in constant error here is comparableto that producedby knowledge of results. Figure-ground shifts produced by punishment with electric shock have been studied by Mangan (88, 89) with three degreesof shock intensity. The principal effect with weak or medium shock was a vigilance reaction rather than a defensereaction; that is, responsechangesemphasizedthe shocked material . With strong shock, the effect was the opposite . The

vigilance effect was retained for sometime. The role of the shock does not seem to be informational here, but the interpretation

of the differential

effectsof different intensitiesneed not be in terms of learning. A survey of knowledge of results, including its role in perceptuallearning, was prepared by Annett (3). A helpful distinction is made between intrinsic and extrinsic knowledge of results. The point is emphasizedthat the information . rather than the motivation , content of knowledge of

results is important to learning; whereasmotivation is an important variable in performance.An earlierstudy by Annett (2) demonstratedthe value of intrinsic information for learning and retention. The task was reproducing a preciseamount of pressure.Extrinsic feedbackfrom a supplementary cue made Ss more accurate at the time , but immediately less accurate on withdrawal of it . Annett 's results underline the importance of the percep-

tual aspectof a task and intrinsic feedbackin learning; merely repeating a precise response was insufficient for learning even if S were informed at once that it was correct . Baker & Young (8) considered effect of knowl -

edge of results (and withdrawal of knowledge) on constant and variable error in a Thorndike -type line-drawing experiment . The sign of the con-

stant error was learned early in training, whereas subsequentlearning increasedimprovement by gradual reduction of variable error. One cannot

conclude

from these studies that reinforcement

or knowl -

edge of results is a necessarycondition for perceptual learning. Experiments by Pearson &: Hauty (104, 105) provide evidence that extrinsic knowledge, at least, is not necessary . Their 5s set themselvesto the vertical

PerceptualLearning

343

from an offset position in a lateral tilt -chair. Error was compared under conditions of visual knowledge of results (a visual error scale on a disc opposite to 5), postural anchoring (E returned 5 to correct setting), and no extrinsic knowledge. The 5s improved in all conditions. When no knowledge of results or anchoring was given by E, alternation through the vertical favored progressive error reduction. A control experiment ruled out adaptationas a factor and strengthenedthe interpretation that passing through the vertical gave 5 unique proprioceptive cues. This seemsto be a kind of intrinsic knowledge of results, allowing S to checkhis last setting against the unique feeling of the vertical as he passesthrough it . Action and feedback All feedbackfrom action is in a sensepotential information. The question can be raised then whether action and the feedback from it is essentialto all perceptualdevelopment and learning. There are in fact psychologists who strongly hold this view. Drever made an eloquent plea for it, referring particularly to Piaget's view that IIawareness of spaceis basedupon action in space" (25, p. 4). The role of exploratory movements of both the hand and the eye perception has been studied - in ~ ~ by Russianpsychologists (137). There are, they proposed, two types, (a) searchingand directing, and (b) pursuit or "gnostic." The latter movements have a constructive function; that is, there is a correspondencebetween the form of the tracing movements and the contour of the object perceived. With maturity and familiarity with objects the movementsare presumably reduced. Ghent (42) and Ghent & Bernstein(43), in explaining someresults on recognition of figures in different orientations by young children, proposed (a la Hebb) that in the early stagesof development the child scansa form, looking first at one portion and then another, and that this sequential processproceedsin a top-to-bottom direction. On the other hand, Mooney (94, 95) reasonedthat if eye movements are essentialin learning to see visually presentedconfigurations, then recognition of meaninglessnovel configurations should be markedly more effective when initially observed by visual inspection than by single brief fixations. His experiments indicated that suchwas not the case. The role of gross postural movements, as well as exploratory and tracking movements of the senseorgans has been held to be important (60, 61, 111, 112). Held and his colleagues, mentioned earlier, consider cues from self-producedmovement critical not only for compensationfor transformed stimulus orders, but also for perceptualdevelopment. Hein & Held (59) reared kittens either with self-produced or passive movements. The "active" animals developed effective responseto visual spatial situations, but the "passive" animalsdid not. Many questionsremain to be answered. Did the passivekittens, pulled round and round, get dizzy, for instance? Researchin this areawill undoubtedly multiply .

344

E. J. Gibson

PerceptualSkills

Heightenedsensitivity Very little applied researchon perceptual skills is to be found in psychologicaljournals, though one supposesit must exist in trade journals. The delicate palates of wine and tea tasters, so often assumed, are worth experimentalstudy if only for the importanceof grading in industry. Both methods of testing skill and methods of scaling or categorizing stimuli are probably important. Engen (27) measuredolfactory thresholdswith a forced-choice method of limits for relatively pure odorants in weak solution. Thresholds were appreciably lowered by practice and, especially, by changes in 5' s criterion of discrimination. Absolute identification does not, typically, show a large improvement with practice bv~ information measures . However, one subject [Hanes & Rhoades(55)] practicing with Munsell color chips for five months increasedthe number of absolutely identifiable colors from 15 to 50. Errors increasedmarkedly after a three-month period of no practice. It seemslikely that the more complex stimuli emanating from real objects of the world yield more discriminablecategories, for absolute identification, and that perception is highly educablewith respectto them. Code learning A factor study, of successivestages of learning Morse code, by Fleishman& Fruchter (33) revealedthat factor loadings changed with stage of training, indicating (to the reviewer) a two-stage learning process. In the early stages, Auditory Rhythm Discrimination and Auditory PerceptualSpeedwere the only factors with high loadings. But at a later stage, they played a smallerrole with Speedof Closure highest. Does this not suggestthat S first hasto learn to differentiate the items, and that later, stimuli are organized into higher-order units for perception? These stages are characteristicof reading, as well. Reading Most of the voluminous literature on reading skill is not pertinent to perceptualleaming. But perceptuallearning is a part of the skill of reading. One aspectof reading is the acquiring of the directional scanning habit. That this habit involves some perceptuallearning, as well as motor, is suggestedby a number of tachistoscopicexperiments in which letters are exposed simultaneouslyon both sides of a fixation point. Following Heron's experiment (66), recognition has been found to be better for material to the left than to the right [Kimura (77); L'Abate (82)]. Sometimes nonalphabeticalmaterial has not yielded the same differential effect as letters [Terrace (123); Bryden (17)], but in other experiments the results were similar to those with verbal material [Ayres & Harcum (7); Harcum & Dyer (56); Kimura (77)]. A number of factors, suchas successionor simulta-

PerceptualLearning 345

neityofpresentation, spacing, andorderofreporting, havebeenfoundto affectresultsin theseexperiments. Interpretation of the resultsis not

self-evident, butitisgenerally agreed thathabits acquired through reading

experience develop a kindofperceptual primacy phenomenon.

Thatperceptual learningis involvedin letterdifferentiation is the conclusionof a developmental studyof the discrimination of letter-like forms [Gibson(45);Osser& Gibson(101)].Standardformswereconstructed

according to ruleswhichdescribe lettersand12specified transformations

of eachof thoseprepared.Thetaskwasto matcha standardwitha form

froma setcontaining thestandard and12transformations ofit.Developmental curves ofprogressive accuracy varied withtransformation type, suggestingan explanationbased on learningof distinctivefeaturesof

letters.

Aswiththelearning ofMorse code, thereisa second stageofperceptual learning inreading, wherein theletterunits, nowdiscriminable, areorganizedintohigher-order units,sothatmoreisperceived ata glance. Familiarity andmeaning havealways beenassumed tobetheorganizing factors

at this stage,but reasonhas been found[Gibson (44)1to believethat

grapheme-phoneme correspondence rules,complex astheymaybeforthe

Englishlanguage, generatehigher-order unitsfor skilledreaders.Pseudo-

words(meaningless andunfamiliar) wereperceived moreaccurately with

tachistoscopic presentation whentheyfollowedtheserules. In Conclusion

Thepicture whichemerges oftheareaofperceptual learning isthatofa

healthily growing one,intermsofinterest andamount ofresearch being

produced. Thisgrowthmakes especially apparent thelooseness oftheoreticaldevelopment. Anytightlyworked theories thatmightbeconsidered at

allrelevant tendto becouched inresponse termsandtherefore failto take intoaccount someofthemostinteresting problems. Response biasis still

abiasofmanyexperimenters. Butmoreattention isbeingpaidtoattention. People whoworkoncomputer models forpatternrecognition arenot afraid tousecognitive terms. Theirexample should shake upsomeofthe response-biased theory makers. Itispredicted thatmorespecific cognitive theories ofperceptual learning areontheway,andthattheparameters

of the paradigmatic experiments willget morecarefulconsideration and

control.

1.Thesurvey oftheliterature pertaining tothisreview wasconcluded AprilI, 1962.

346

E. J. Gibson

References 1. Albert, R. s. The function of verbal labelsin the discriminationof subtle stimulus differences . J. Genet . Psychol ., 94, 287- 96 (1959) 2. Annett, J. Learninga pressure underconditionsof immediateanddelayedknowledgeof results.Quart. J. Psycho I., 11, 3- 15 (1959) 3. Annett, J. TheRoleof Knowledge of Results in Learning : A Survey(Tech. Rept: NAVTRA DEVCEN342- 3, PortWashington , N.Y., 53 pp., 1961) 4. Atkinson, R. C. The observingresponsein discriminationlearning.J. Exptl. Psychol ., 62, 253- 62 (1961) 5. Attneave, F. Perceptionand relatedareas . In Psychology : A Studyof a Science . Vol. 4: Biologically OrientedFields : TheirPlacein Psychology and in Biological Science , 619- 59 (Koch, S., Ed., McGraw-Hill BookCo., New York, N.Y., 731pp., 1962) 6. Axelrod, S. Effects of EarlyBlindness : Perfonnance of BlindandSighted Childrenon Tactileand AuditoryTasks(AmericanFoundationfor the Blind, New York, N.Y., 83 pp., 1959) 7. Ayres, J. J., and Harcum , E. R. Directionalresponse -bias in reproducingbrief visual patterns . Perceptual Motor Skills, 14, 155- 65 (1962) 8. Baker , C. H., andYoung, P. Feedback duringtrainingandretentionof motor skills. Can. J. Psychol ., 14, 257- 64 (1960) 9. Barlow, H. B. The codingof sensorymessages . In CurrentProblems in AnimalBehavior , 331- 60 (Thorpe, W. H., and Zangwill, O. L., Eds., CambridgeUniv. Press , London, England , 424 pp., 1961) 10. Beatty, F. 5., Dameron , L. E., andGreene , J. E. An investigationof the effectsof reward andpunishmenton visualperception . J. Psychol ., 47, 267- 76 (1959) 11. Bevan , W. Perceptual learning:anoverview. J. Gen.Psychol ., 64, 69- 99 (1961) 12. Binder, A., andFeldman , S. E. The effectsof experimentally controlledexperience upon recognitionresponses . Psychol . Monographs , 74 (9), 43 pp. (1960) 13. Broadbent , D. E. Humanperceptionandanimallearning.In CurrentProblems in Animal Behavior , 248- 272 (Thorpe, W. H., andZangwill, O. L., Eds., CambridgeUniv. Press , London, England , 424pp., 1961) 14. Brown, C. R., andRubenstein , H. Test of response biasexplanationof word-frequency effect. Science , 133, 280- 81 (1961) 15. Bruner,J. S. (Chainnan ). Specialsessionon Processes of Cognitive Growth(EasternPsychol . Assoc., 33rd meeting,Atlantic City, N.J., April 26- 28, 1962) 16. Bruner, J. S., Wallach , M . A., and Galanter , E. H. The identificationof recurrent regularity. Am. J. Psycho /., 72, 200- 9 (1959) 17. Bryden, M . P. Tachistoscopic recognitionof non-alphabetical material . Can. J. Psycho I., 14, 78- 86 (1960) 18. Caviness , J. A., andGibson, J. J. TheEquivalence of VisualandTactualStimulation for the Perception of SolidForms(Presented at EasternPsychol . Assoc., 33rd meeting , Atlantic City, N.J., April 26- 28, 1962) 19. Chistovitch , L. A., Klass , Yu. A., andAlekin, R. O. a Znacheniiimitatsiidlya raspozna vaniyazvukovykhposledovatel ' nostei(On the significance of imitationfor the discriminationof soundsequences .) Voprosy Psikh % gii, 7 (5), 173- 82 (1961). 20. Chorover,S. L., Cole, M., andEttlinger, G. ApparentLackof Positive Cross -modalTransfer ". Englishtranslationwasannounced in Technical Translations , issuedby the Officeof TechnicalServices , U. S. Departmentof Commerce , andwasmadeavailableby the Photoduplica tion Service , Libraryof Congress , andby the SLA TranslationCenterat the JohnCrerar Library, Chicago , Illinois.

PerceptualLearning

347

ofTraining withAuditory andVisual Rhythms inMan(Presented atEastern Psychol . Assoc ., 33rdmeeting , Atlantic City,N.]., April26- 28,1962 ) 21. Cohen , M., andHeld , R. Degrading Eye -hand Coordination byExposure toDisordered Re -afferent Stimulation (Presented atEastern Psychol . Assoc ., 31stmeeting , NewYork , N.Y., April15- 16,1960 ) 22.Dember , W. N. ThePsychology ofPerception (Henry HoltandCo., NewYork , N.Y., 402 pp., 1960 ) 23.DeRivera , ]. Some conditions governing theuseofthecue -producing response asan explanatory device .]. ExptI . Psychol ., 57,299 - 304(1959 ) 24.Drever , ]. D. Perceptual learning .Ann . Rev . Psychol ., 11,131 - 60(1960 ) 25.Drever , ]. D. Perception andaction . Bull . Brit.Psychol . Soc ., 45,1- 9(1961 ) 26.Ellis , H. C., Bessemer , D. W., Devine , ]. V., andTrafton , C. L. Recognition ofrandom tactual shapes following predifferentiation training . Perceptual MotorSkills , 10,99102(1962 ) 27. Engen , T. Effect ofpractice andinstruction onolfactory thresholds . Perceptual Motor Skills , 10,195 - 98(1960 ) 28.Epstein , W., andDeShazo , D. Recency asa function ofperceptual oscillation . Am.J. Psychol ., 74,215 - 23(1961 ) 29.Epstein , W., andRock , I. Perceptual setasanartifact ofrecency . Am.J. Psychol ., 73, 214 - 28(1960 ) 30. Essman , W. B. I'Learning without awareness " of responses to perceptual andverbal stimuli asafunction ofreinforcement schedule . Perceptual Motor Skills , 9, 15-25(1959 ) 31.Ettlinger , E. Cross -modal transfer oftraining inmonkeys . Behavior , 16,56- 65(1960 ) 32.Fantz , R. L. Theorigin offormperception . Sci .Am., 204 , 66- 72(1961 ) 33.Fleishman , E.A., andFruchter , B. Factor structure andpredictability ofsuccessive stages oflearning Morse code .J.Appl . Psychol ., 44,97- 101(1960 ) 34.Fowler , W. Cognitive learning ininfancy andearly childhood . Psychol . Bull ., 59,116 - 52 (1962 ) 35.Fraisse , P. Miseenevidence d'unapprentissage perceptif desgroupements parprox imite .Annee psychol ., 59,374 - 80(1959 ) 36.Fraisse , P., andBlancheteau , M. Theinfluence ofnumber ofalternatives onthepercep tualrecognition threshold . Quart .J. Psychol ., 14,52- 55(1962 ) 37.Frances , R. Information etajustment dans l'apprentissage perceptif . InProceedings 16th Intern . Con gr.ofPsycho 1.Bonn , Germany , July31-Aug.6, 1960 ,331 - 2(North Holland Publishing Co., Amsterdam , 944pp., 1962 ) 38.Frances , R.LeDeveloppement Perceptif (Presses Universitaires deFrance , Paris , France , 275 pp., 1962 ) 39.Freedman , S. ]. Perceptual changes in sensory deprivation : suggestions foraconative theory .]. Nervous Mental Disease , 132 , 17-21(1961 ) 40.Ganz , L., andRiesen , A. H. Stimulus generalization tohueinthedark -reared Macaque . J. Compo Physiol . Psycho I., 55,92- 99(1962 ) 41.Gamer , W.R. Uncertainty and Structure asPsychological Concepts Oohn Wiley &Sons , Inc., NewYork , N.Y., 369pp., 1962 ) 42.Ghent , L. Recognition bychildren ofrealistic figures presented invarious orientations . Can .J. Psycho I., 14,249 - 56(1960 ) 43. Ghent , L., andBernstein , L. Influence of theorientation ofgeometric forms ontheir recognition bychildren . Perceptual Motor Skills , 12,95- 101(1961 ) 44.Gibson , E.]. The Role ofGrapheme -phoneme Co " espondence inWord Recognition (Presented atEastern Psychol . Assoc ., 33rdmeeting , Atlantic City,N.J., April26- 28,1962 ) 45.Gibson , E. J. TheDevelopment ofPerception : Discrimination ofDepth Compared withthe Discrimination ofGraphic Symbols (Presented atConference onBasic Cognitive Processes

348

E. J. Gibson

in Children , Minneapolis , Minn. April, 1961) Monographs Soc . Research ChildDevelop . (In press ) 46. Gibson,E. J., andalum, V. Experimental methodsof studyingperceptionin children.In Handbook of Methodsin ChildDevelopment , 311- 373(Mussen , P. H., Ed., John - Research Wiley & Sons,Inc., New York, N.Y., 1061pp., 1960). 47. Gibson, E. J., andWalk, R. D. The "Visual cliff." Sri. Am., 202, 64- 71 (1960) 48. Gibson, E. J., Walk, R. D., and Tighe, T. J. Enhancement and deprivationof visual stimulationduringrearingasfactorsin visualdiscrimination learning.]. Camp . Physiol . Psycho !., 52, 74- 81 (1959) 49. Gibson, J. J., et al. Symposiumon perceptualleaming . In Proceedings 16thIntern. Congr . of Psychol ., Bonn, Germany , July 31- Aug. 6, 1960, 317- 330 (North HollandPublish ing Co., Amsterdam , 944 pp., 1962) 50. Gibson,J. J., andGibson,E. J. Perceptualleaming : differentiationor enrichment ? Psycho I. Rev., 62, 32- 41 (1955) 51. Goldiamond , I., andHawkins , W. F. Vexierversuch : the log relationshipbetweenwordfrequencyandrecognitionobtainedin the absence of stimuluswords. J. & pti. Psycho I., 56, 457- 63 (1958) 52. Goldstein , M . J., and Ratliff, J. Relationshipbetweenfrequencyof usageand easeof recognitionwith response biascontrolled.Perceptual Motor Skills, 13, 171- 77 (1961) 53. Gollin, E. S. Tactualform discrimination : developmentaldifferencesin the effect of trainingunderconditionsof spacialinterference . J. Psycho I., 51, 131- 40 (1961) 54. Gregory, R. L., andWallace , J. G. Recoveryfrom earlyblindness - a casestudy, Quart. J. Exptl. Psycho I., Monograph Supple 2 (1962) (In press ) 55. Hanes , R. M., andRhoades , M. V. Coloridentificationasa functionof extendedpractice . J. Opt. Soc . Am., 49, 1060- 64 (1959) 56. Harcum , E. R., andDyer, D. W. Monocularandbinocularreproductionof binarystimuli appearingright andleft of fixation. Am. J. Psychol ., 75, 56- 65 (1962) 57. Hayes,W. N., Robinson , J. S., andBrown, L. T. An Effectof PastExperience on Perception : An Artifact(Presented at AmericanPsychol . Assoc., 69th meeting, New York, N.Y., Aug. 31-Sept. 6, 1961) (Abstr. in Am. Psychol ., 16, 420, 1961) 58. Hebb, D. O. Discussion (Symposium on sensorydeprivation : factsin searchof a theory). J. Nervous MentalDisease , 132, 40- 43 (1961). 59. Hein, A. V., andHeld, R. Movement -produced Stimulation in theDevelopment of Visually guidedBehavior {Presented at EasternPsychol . Assoc., 33rdmeeting,AtlanticCity, N.J., April 26- 28, 1962) 60. Held, R. Exposure -historyasa factorin maintainingstabilityof perceptionandcoordina tion. J. Nervous MentalDisease , 132, 26- 32 (1961) 61. Held, R. Adaptationto rearrangement and visual-spatialaftereffects . Psycho I. Beit., 6, 439- 50 (1962) 62. Held, R., andBossom , J. Neonataldeprivationandadultrearrangement : complementary techniquesfor analyzingplasticsensory -motor coordinations . J. Camp . Physiol . Psychol., 54, 33- 37 (1961) 63. Held, R., and Schlank , M. Adaptationto dissarranged eye-hand coordinationin the distance -dimension . Am. J. Psychol ., 72, 603- 5 (1959) 64. Henle, M. An experimental investigationof pastexperience asa determinantof visual form perception . J. & ptl. Psychol ., 30, 1- 22 (1942) 65. Hermelin , B., and O'Connor, N. Recognitionof shapesby nonnal and subnormal children. Brit. J. Psychol ., 52, 281- 84 (1961) 66. Heron, W. Perceptionas a functionof retinallocusand attention. Am. J. Psycho /., 70, 38- 48 (1957)

PerceptualLearning

349

67. Hochberg , J. Natureandnurturein perception . In Psychology in theMaking(Postman , L., Ed., Knopf, New York, N .Y., in press) 68. Hochberg, J., and Brooks, V . Pictorial recognition as an unlearnedability : a study of one child's performance. Am. ] . Psycho !. (In press) 69. House, B. J., and Zeaman, D. Transfer of a discrimination from objects to patterns. J. Expt!. Psycho!., 59 , 298 - 302 (1960 )

70. Hudson,W. Pictorialdepthperceptionin sub-culturalgroupsin Africa.]. SocialPsychol ., 52 , 183 - 208 ( 1960 )

71. Hunt, J. McV . Intelligenceand Experience (The Ronald PressCo., New York, N .Y., 416 pp ., 1961 )

72. Ittleson, W . H. Visual SpacePerception(Springer Publishing Co., New York, N .Y., 212 pp ., 1960 )

73. Ittleson, W . H. Perception and transactional psychology. In Psychology : A Study of a Science . Vol . 4: BiologicallyOrientedFields: Their Placein Psychologyand in Biological Science, 660 - 704 (Koch , S., Ed ., McGraw 1962 )

- Hill

Book

Co ., New

York , N .Y ., 837

pp .,

74. Jakobson, R., and Halle, M . Fundamentalsof Language(Mouton & Co., The Hague, Netherlands

, 87 pp ., 1956 )

75. Karo, H. W ., and Gregg, L. W . Acquisition of perceptual responsesas a function of loading, location, and repetition. J. Exptl. Psycho /., 62, 62- 69 (1961) 76. Kilpatrick, F. P. Explorationsin TransactionalPsychology(New York Univ . Press, New York , N .Y ., 405 pp ., 1961 )

77. Kimura , D . The effect of letter position on recognition . Can. J. Psychol., 13, 1- 10 (1959 )

78. Kohler, I. ObeyAufbau und Wandlungender Wahrnemungswelt(Rohrer, Vienna, Austria, 118 pp ., 1951 )

79. Kohler, I. Interne und externe organization in der Warnehmung. Psycho /. Reit., 6, 426- 38 (1962 )

80. Kohler, I. Experimentswith goggles. Sci. Am., 206, 62- 72 (1962) 81. Kossov, B. B. Some methods for bringing out essentialfeatures in perceived objects. Problems of Psycho/., 1, 71- 81 (1960 )

82. L'Abate, L. Recognition of paired trigrams as a function of associative value and associative strength . Science, 131 , 984 - 85 (1960 )

83. Lane, H. L., and Moore , D. J. Operant reconditioning of a consonant discrimination in an aphasic. Progr. Rept. No. 1 & ptl. Analysisof Controlof Speech Productionand Perception (Contract No . SAE-9265, U.S. Office of Education, Washington, D.C., 1961) 84. Lenneberg, E. Color naming, color recognition, color discrimination: a reappraisal. Perceptual Motor Skills, 12, 375- 82 (1961) 85 . Liberman , A . M ., Harris , K . S., Kinney , J., and Lane, H . The discrimination

of relative

onset-time of the components of certain speech and nonspeech patterns. J. Exptl. Psychol ., 61, 379- 88 (1961) 86. London, I. D . A Russianreport on the postoperative newly seeing. Am. J. Psychol ., 73, 478 - 82 (1960 )

87. McNamara, H. J. Nonveridical perception as a function of rewards and punishments. Perceptual Motor Skills, 9, 67- 80 (1959) 88. Mangan, G. L. The role of punishment in figure-ground reorganization. J. ExptI. Psychol ., 58, 369 - 75 (1959 )

89. Mangan, G. L. Retention of figure-ground reorganization occurring under electric-shock punishment. Perceptual Motor Skills, 12, 151- 54 (1961) 90. Marx , M . H., Murphy , W . W ., and Brownstein, A . J. Recognition of complex visual stimuli as a function of training with abstractedpatterns. J. Exptl. Psycho I., 62, 456- 60 ( 1961 )

350

E. J. Gibson

91. Meier, G. W., and McGee, R. K. A re-evaluationof the effect of early perceptual experience on discriminationperformance duringadulthood]. Camp . Physiol . Psychol ., 52, 390- 95 (1959) 92. Miller, N. E., andDollard, J. SocialLearning andImitation(YaleUniv. Press , New Haven, Conn. 341 pp., 1941) 93. Minsky, M. Stepstowardartificialintelligence . Proc . Inst. RadioEngrs ., 49, 8- 30 (1961) 94. Moonev. C. M. Recognitionof symmetricaland non-symmetricalinkblots with and without eyemovements . Can.J. Psychol ., 13, 11- 19 (1959). 95. Mooney, C. M. Recognitionof ambiguous andunambiguous visualconfigurations with shortandlongerexposures . Brit. J. Psychol ., 51, 119- 25 (1960) 96. Moscovici, S., and Humbert, C. Usageet disponibilitecommefacteursdeterminantla dureede stimuliverbaux.Bull. Psychol ., 13, 406- 12 (1960) 97. Nealey, S. M., and Edwards , B. J. "Depth perception " in rats without pattern-vision experience . J. Compo Physiol . Psychol ., 53, 468- 69 (1960) 98. Newbigging, P. L. Personalvaluesand responsestrengthof value-relatedwords as measured in a pseudoperceptual task. Can.J. Psycho /. 14, 38- 44 (1960) 99. Newbiggin.g, P. L. Theperceptual redintegrationof frequentandinfrequentwords. Can. J. Psychol . 15, 123- 32 (1961) 100. Newbigging, P. L., andHay, J. M. ThePractice Effectin Recognition Threshold Determina tionsasa Function of WordFrequency andLength(Presented at EasternPsychol . Assoc., 33rdmeeting,Atlantic City, N.]., April 26- 28, 1962) 101. asser, H., and Gibson, E. J. A Developmental Studyof theDiscrimination of Letter -like Forms(Presentedat EasternPsychol . Assoc., 32nd meeting, Philadelphia , Pa., April 7- 8, 1961) 102. Pastore , N. Perceptual functioningin theducking.J. Genet . Psychol ., 95, 157- 69 (1959) 103. Pastore , N. Perceivingasinnatelydetermined . J. Genet . Psychol ., 96, 93- 99 (1960) 104. Pearson , R. G., and Hauty, G. T. Adaptive processesdeterminingproprioceptive perceptionof verticality. J. Exptl. Psychol ., 57, 367- 71 (1959) 105. Pearson , R. G., and Hauty, G. T. Role of posturalexperiences in proprioceptive perceptionof verticality. J. Exptl. Psychol ., 59, 425- 28 (1960) 106. Pfafflin, S. M. Stimulusmeaningin stimuluspredifferentiation . J. Exptl. Psychol ., 59, 269- 74 (1960) 107. Phaup , M. R., andCaldwell,W. E. Perceptualleaming : differentiationandenrichment of pastexperience . J. Gen.Psycho I., 60, 137- 47 (1959) 108. Piaget,J. LesMecanismes Perceptifs (Presses Universitaires de France , Paris,France , 457 pp., 1961) 109. Prentice , W. C. H. Aftereffectsin perception . Sci. Am., 206, 44- 49 (1962) 110. Rasmussen , E. A., and Archer, E. J. Conceptidentificationas a functionof language pretrainingandtaskcomplexity.J. Erptl. Psychol ., 61, 437- 41 (1961) 111. Riesen , A. H. Studyingperceptualdevelopmentusingthe techniqueof sensorydeprivation. J. Nervous MentalDisease , 132, 21- 25 (1961) 112. Riesen , A. H., andAarons,L. Visualmovementandintensitydiscrimination in catsafter earlydeprivationof patternvision. J. Compo Physiol . Psychol ., 52, 142- 49 (1959) 113. Rosenblatt , F. The designof an intelligentautomaton . Research Reviews (ONR), 5- 13, (October, 1958) 113a. Rubenstein , H., and Aborn, M. Psycholinguistics . Ann. Rev. Psychol ., 11, 291- 322 (1960) 114. Rudel, R. Decrement of theMuller-LyerIllusion : A Studyof Intermodal Transfer (Presented at EasternPsychol . Assoc., 31stmeeting,New York, N.Y., April 15- 16, 1960) 115. Santos , J. F., andMurphy, G. An Odysseyin perceptual learning.Bull. Menninger Clin., 24, 6- 17 (1960)

PerceptualLearning

351

116. Schiff, W . The effect of subliminal stimuli on guessing-accuracy. Am. ] . Psychol ., 74, 54 - 60 (1961 )

117. Selfridge, O . G., and Neisser, U. Pattern recognition by machine. Sci. Am., 203, 60- 68 (1960 )

118. Senden, M . Yon. Spaceand Sight (Free Press, Glencoe, Ill ., 348 pp., 1960) 119. Smock, C. D., and Kanfer, F. H. Responsebias and perception. J. Expt/. Psycho /., 62, 158 - 63 ( 1961 )

120. Solley, C. M ., and Engel, M . Perceptual autism in children: the effects of reward, punishment, and neutral conditions upon perceptual learning. ] . Genet. Psycho /., 97, 77 - 91 ( 1960 )

121. Solley, C. M ., and Murphy , G. Developmentof the PerceptualWorld (Basic Books, Inc., New

York , N .Y ., 353 pp ., 1960 )

122. Tallarico, R. B. Studies of visual depth perception: III. Choice behavior of newly hatched chicks on a visual cliff. Perceptual Motor Skills, 12, 259- 62 (1961) 123 . Terrace , H . S. The effects of retinal locus and attention on the perception of words .

] . Exptl. Psychol ., 58, 382- 85 (1959) 124. Teuber, H. L. Sensory deprivation, sensory suppression and agnosia: notes for a neurologic theory. ] . NervousMental Disease , 132, 32- 40 (1961) 125. Uhf, L., Vossler, C., and Uleman, J. Pattern recognition over distortions, by human subjectsand by a computer simulation of a model for human form perception. J. Exptl. Psychol ., 63, 227- 34 (1962) 126. Vanderplas, J. M . SomeRelationsof Learningto Form Perception (Presentedat meeting of the Southern Soc. for Phil. and Psychol., Biloxi, Miss., April , 14- 16, 1960) 127. Vanderplas, J. M ., and Garvin, E. A . Complexity, association value, and practice as factors in shaperecognition following paired-associatestraining. J. Exptl. Psycho /., 57, 155 - 63 ( 1959 )

128. Vanderplas, J. M ., Sanderson, W . A ., and Vanderplas, J. N . SomeTask-relatedDeterminantsof Transferin PerceptualLearning(Presentedat meeting of Southern Soc. for Phil. and Psychol., Memphis, Tenn., April , 19- 21, 1962) 129. Vernon, M . D. Perception, attention and consciousnesss . Advanc. Sci., 62, 111- 23 (1959 )

130. Vurpillot , E., and Brault, H. Etude experimentale sur la formation des schemesempiriques . Annee psycho/., 59, 381 - 94 (1959 )

13I . Walk, R. D. A Study of SomeFactorsInfluencingthe Depth Perceptionof Human Infants (Presented at Eastern Psychol. Assoc., 32nd meeting, Philadelphia, Pa., April , 7- 8, 196 I .)

132. Walk, R. D ., and Gibson, E. J. A comparative and analytical study of visual depth perception. Psycho /. Monographs, 75 (15), 44 pp. (1961) 133. Walk, R. D., Gibson, E. J., Pick, H. L., Jr., and Tighe, T. J. The effectivenessof prolonged exposure to cutouts vs. painted patterns for facilitation of discrimination. ] . Compo Physio/. Psycho/., 52, 5 19- 2 I (1959 )

134. Wohlwill , J. F. Developmental studies of perception. Psychol . Bull., 57, 249- 88 (1960) 135. Yarczower, M . Conditioning test of stimulus-predifferentiation. Am. ] . Psycho /., 72, 572 - 76 ( 1959 )

136. Zeaman, D., and House, B. J. An attention theory of retardate discrimination learning. Prog. Rept. No . 3 (Research Grant M - I099 , Natl . Inst . Mental Health , Bethesda , Md ., November 137 . Zinchenko

, 1962 ). , V . P ., and Lomov , B . F . The functions

of hand

processof perception. ProblemsPsychol ., 1, 12- 26 (1960)

and eye movements

in the

Perceptual Development andthe Reduction of Uncertainty

EleanorJ. Gibson

Thereader willfindthispaper dated, tobesure,butI thinkina ratherinstructive

fashion. The1960s wasa time fora revolution ofideas inthecolleges ofthe country, notonlyamong theundergraduates butalsoamong theleaders ofthe disciplines theywere taught. Psychology wasnoexception, andideas from outsidefromneuroscience, computer science, andeven engineeringwere eagerly received. Single fiberrecording, information theory, artificial intelligence, and

cybernetics allhada hearing andwere enthusiastically incorporated into psychol-

ogy, frequently rather uncritically. I wasnoexception, asthispaper makes clear. Thepaper waswritten foranInternational Congress ofPsychology inMoscow ofwhich I have mixed memories, some very pleasant such asa dinner given by A.V.Zaporozhets inhisthree-room apartment forAmerican psychologists who

hadbeen invited toprepare papers forhissymposium. The titleofthesymposium

wasTheDevelopment ofPerception andActivity. Psychologists included were myself, Julian Hochberg, Richard Held, A.V.Zaporozhets, V.P.Zinchenko, M.I. Lissina, R.D.Walk,Herbert Pick, JohnHay,0. W.andP.C.Smith,

J.B.Gippenreiter, Elaine Vurpillot, L.A.Venger, Allen Hem anda number of Soviet psychologists whom I cannot remember. TheSoviet psychologists were all students orfollowers ofVigotsky andLeontiev, a second generation ofclever

theorists andresearchers. Unfortunately, concern with perception inRussian psychology hasdeclined since then, although atranslation ofJ.J.Gibsons Ecological

Approach toVisual Perception hasrecently appeared (Logvinenko 1988). A Soviet psychologist I remember particularly wellfromthissymposium is Zinchenko, whose talkimpressed megreatly. Hispointofviewwas(tosome

extent) a Sovietpartyline viewandinvolved constructing an internal model through action, butatthesametimeheincorporated ideas thatwere rather similar

to my own.Hereis a sample:

InProceedings ofthe18th International Congress ofPsychology, Symposium 30,Perception and

Achon, 717, Moscow, 1966.

354

E. J. Gibson During

the

first

acquaintance

stages

.

informative

to

memory

process

we

have

such

.

This

accompanied

of

the

the

.

eliminating

.

and

The

,

one

as

two

the

were

is

, "

70

a

self

although

I

features

into

-

alphabet

the

"

,

char

expense

of

and

basic

structural

" unit

is

of

object

the

critical

consolidated

compatible

.

So

- regulating

would

course

given

selecting

,

more

the

But

at

the

" unit

features

replaced

is

"

very

by

collating

of

my

)

very

uncertainty

the

.

the

features

the

of

reduced

separate

in

object

,

in

eye

subject

the

longer

learning

of

the

the

in

and

,

several

,

recorded

the

hand

radically

such

collating

1966

of

action

information

of

sentences

reduction

perceptual

action

process

of

,

,

created

information

by

selected

observation

with

is

groups

lasts

model

the

changes

objects

( Zinchenko

It

that

selecting

features

repeated

of

collating

is

etalon

essential

its

and

redundant

of

the

.

last

,

and

it

.

and

like

and

the

elements

object

of

transforming

feature

The

process

When

of

detailed

substantially

discovering

with

movements

identification

rational

.

number

process

The

task

into

a

numerous

excessive

,

ones

the

identification

objects

features

in

includes

by

acter

divided

as

object

great

is

features

acCluaintance ,

the

the

a

is

actions

to

are

collating

recognition

such

identify

there

of

of

see

adequate

and

When

process

also

,

collate

.

the

paper

we

content

begins

of

the

Here

with

was

his

process

reword

of

it

to

Y

onas

the

theme

definition

of

search

read

for

" search

own

perception

and

:

" A

processing

for

information

his

Ph

of

to

guide

. "

An

experiment

Cornell

performed

( Yonas

reduction

and

of

reaction

designated

) .

as

displayed

,

member

of

the

they

.

each

one

that

feature

from

.

relative

principle

operating

and ~ ~ . ~- -

,

-discovery - -- - -

but

~

it

."

of .I

one

common

involves

invariance

the

,

feature

self

negative

subjects a

no

drive

is

the

.

.

did

This

and

essence

sense

of -

a

neat

like

.

economy

Cognitive

's

finding

,

the

search

for

only

one

in of

three

of

an

that .

Few

economy

.

reduction

economy

,

to

latencies

fashion

Hull

these

other

with

case

the

of

distinguish

only

set

a

three

Both

the

to

condition

adaptive

bit

in

reduce

us

. one

used

positive

to

highly

a

his

the

indeed

in

In

but

be

needed

of

selected

seemed

sound

in

Subjects

level

randomly

- regulatory

may

the

,

for

differed

.

be

only

positive

rule

were

might

way

set

could

of

which

letter

set

useful

that

set

to

Any

positive

of

disjunctive

of

certain

designated

the

any

.

a

at

principle

experiments

some

positive

was

to

to

,

set

key

feature

time

with

uncertainty

negative

to

dissertation

the

the

owing

the

.

utilized

letters

reaction

letter

distinctive

practice

a

one

design

nine

of

according

of

in

of

only

. D

of

time

of

the

the

The

the

set

assigned

a

reaction

the

the

Reduction

,

selected

Given

at

some

letters

of

illustration

membership

one

their

to

noticed

In

part

.

a

press

The

member

reduce

member

condition

.

as

excellent

vogue

and

shared

every

to

positive

subjects

not

numbers

an

given

to

three

were

positive

stimulus

set

had

conditions

three

set

was

conditions conditions

in

was

subject

positive

is

learning

much

subject

positive

the

)

perceptual

,

The

the

but

experimental other

in

experiment

( 1967

Albert

1967

information

time

Sternberg

by

Gibson

of

makes

order

in

change

a

drive

sense

.

,

Perceptual Development andtheReduction ofUncertainty355 Introduction

Inthe1940s, psychologists wereverysecure andverysmug.Theyhada choice oftwobrandsofS-Rreinforcement theory, andiftheydidntlikeit, there was S-S association theory. Almost no one bothered defining theSs and the Rs.

Butinthefifties a greatblowfell.Information theoryhaditsimpact on

psychology. It wassuddenly obviousthatyoucant talkaboutanS or an R.Youhaveto talkabouta setof them.Youhaveto knowwhatthe set of alternatives is andhowbigit is.Notonlythat,youhaveto knowhow

members ofthesetdiffer fromoneanother. Andstillworse, youhaveto

knowwhatthetaskis in relationto thesedifferences (isit discrimination? recognition? free recall?)becausepotentialdistinctivefeatureswithina

set,it beginsto beclear,aredetermined bytheirtaskutility.

Asifthiswerenotenough, thesixties havebrought themostshattering strokeofall.It hasbecome obvious thattheoldconcept ofa stimulus asa pointin spaceor an instant of energycannotpossibly be right.To be a

stimulus, theenergy mustcontain information andthatmeans, at thevery

least,inhomogeneity andchange. A fewpsychologists, amongthemKurt Koffka, myhusband, andWendell Garner, havebeenpointing thisoutfora

longtime.Theywere,I fear,voicescryingin the wilderness untilsome very respectableneurophysiologists (Maturanaet al. 1960;Hubeland

Wiesel 1962,1963) camealongandshowed thatevensingle nervefibers in

theopticsystem arefiredbyinvariant patterned stimulationby edges,

convexities, linesat different orientations, andterminal endingsoflines

inotherwords, bystructural features ofstimulation, however simple. These

factsholdfornewbornanesthetized cats.SoGodonlyknowswhatcomplexities ofstimulus information canbepickedupbya mature, unanesthe-

tized, educated man.

Ontheotherhand,weknowthatthereisa limitto thecapacities of

information pickupof eventhis God-likecreature.Thereis too much information, as a rule,in the totalstimulus flux.He cannothandleit all.

Hemust,therefore, develop somestrategies ofinformation processing forexample,selection, adaptation, andinhibition.

Andsonow,atthispointinhistory, thequestion ofperceptual develop-

mentcomesinto its own.Naturedoesnot turn out an infantwith knowl-

edgeandoperations ready-made in alltheserespects. Wearefacedwith

thefollowing questions (andmore).HowarethesetsofSs andRs sorted out and assembled? Howdoesthe personlearnhow membersof a set differ? Howdoeshelearnto utilizedistinctive features ofthesetinrelation

to thetask? Arehigher orders ofstimulus information picked upincreasinglywithdevelopment? Whatstrategies ofinformation processing are

correlatedwithdevelopment?

356

E. J. Gibson

The perception psychologist is not responsiblefor the answers to all thesequestions, but he must facehis shareof them. So let me try . A Twentieth -CenturyTheoryof Perceptual Learning Theoriesof perceptualleaminghave existed sincethe day of Bishop Berkeley. Perhaps, though, I should say a theory, for one theory persistedwith a stranglehold even into the thirties and in fact gained new strength with Titchener and even with the functionalistsof Carr's laboratories. That was, in essence , the notion that sensory elementsare welded together by association to form oercevtions , and that sensationsare enriched, or added to ~ ~ through associativeprocesses , by memory images or by experiencesin another modality (e.g., vision was thought to gain meaning from association with touch). The naive simplicity of this notion was rejected by the Behaviorists but was supplanted with something equally indefensible, again an IIenrichment" or additive theory, but with the difference that responses were thought to be added, rather than mental contexts. My husband and I debated this view with Leo Postman in the Psychological Reviewin 1955 (Gibson and Gibson 1955; Postman 1955). I think all three of us have changedour minds a bit in the meantime (see Gibson 1959, and Postman 1963), but there are still current severalversions of an additive mediation theory of perceptuallearning. They can be classifiedin various ways. The main division of additive mediationtheories, as I shall call them, is into cognitively oriented and response-oriented theories. Theories which assumeunconsciousinference (or a kind of submergedproblem-solving process) are of the cognitively oriented type. They can be traced back to Helmholtz, but they have their modem advocatesas well: notably, Brunswik, Bruner, and the transactionalists. Another form of cognitively oriented theory assumesthat the underlying mediational process is not inference, but rather accruing of a representation or schemato which input can be matched. M . D. Vemon seemssometimesto advocatethis position, as often does Piaget (though in his casethere is more emphasis on activity than passiveaccrual). A number of computer-basedmodels of perceptuallearning and recognition are perhaps the purest examplesof accountingfor one aspectof perceptualleaming by storage of a representationand matching it, template fashion. The response-oriented theories also can be divided into two groups. I call one the "motor copy" theory. This theory holds that objects in the world are reflected in perception by action- that is, by responseswhich trace them or embody them in such a way as to recreate their form or structure. The process is analogous to schemaformation, but the "rep-

Perceptual Development andtheReduction ofUncertainty357

resentation ismotorratherthansomekindofcognitive mediation. Soviet

theories of perception especially havefavored thistheory(cf.Leontiev 1957).Typical examples of its application areexperiments thatstudy

hand-tracing ofobjects, andtheeyemovements made inscanning pictured

forms, relating themtoperceptual development (cf.Pick1963). InAmerica, Hebbstheoryofdevelopment offormperception is somewhat similar. Buttheperceptual learning model favored bymostAmerican responseoriented psychologists doesnotstresscopyorreflection ofobjects inthe world,but ratheremphasizes discrimination learning basedon additive

mediation of individual responses, andgoesby thenameofacquired

distinctiveness ofcues.William James wastheoriginator oftheessential ideathatimprovement indiscrimination takesplacebylearning distinctive

associates to originally barelydiscriminable pairsofstimuli. First,theterms

whosedifferences cometo befeltcontract disparate associates andthese helptodragthemapart....Theeffect ofpractice inincreasing discriminationmustthen,inpart,bedueto thereinforcing effect, uponanoriginal

slightdifference betweenthe terms,of additional differences betweenthe

diverseassociates whichtheyseverally affect(James1890,510511). MillerandDollard(1941)castthisideaintoits modernS-Rform,the

associates ofcourse beingresponses. Distinctiveness isaddedbylearning a distinctive response; equivalence isadded, ontheotherhand, bylearning a

responsealreadyassociatedwithsomethingelse.

Tome,alltheseadditive theories seemwrong. Theydo notsquare

withthepoints I madeaboutthenature ofstimuli andresponsesthat

information, thedegreeof uncertainty in both,mustbe considered, and

furthermore thatstimulation always hassomestructure. I couldquote experimental evidence against them,aswell,butI thinkthestrongest

argument istheirfailure totakeaccount oftwentieth century developments

in our knowledgeof howperceptionworks. Letme,therefore,presentanotherpointof view,whichI shallreferto

as thespecificity view.Thisviewholdsthatstimulation is veryrichin

potential variables andinformation which mayormaynotberesponded

to. Discrimination learning proceeds notby addingimagesor distinctive responses, norbybuilding motorcopiesor cognitive representations, but ratherby discovering distinctive featuresof objectsand invariants of eventsinstimulation. Thefeatures (distinguishing isperhaps a betterterm thandistinctive) andtheinvariants overtimearethereto bediscovered, notaddedonbyattaching responses orimages oranything. Thecriterion

ofperceptual learning isthusincreased differentiation ofthestimulus array,

greaterspecificity and correlationbetweenstimulation and discriminative

response. Theoperations by whichrelevantdistinguishing featuresare

discovered doindeed include responsesexploratory responses suchas

358

E. J. Gibson

scanning, feeling, fixating, andsearching. Butit isnottheaddition ofthese responses whichexplainsdifferentiation, fortheyarenot themselves specific to the differences to be discovered.

Objectsare discriminated by theirdimensions of difference. Thereare setsor classesof objectswhichsharecertainfeatures.Butinsofaras they are discriminable, each has a patternof featureswhichis unique.This

uniquepatternis notcreated by themind,butis discovered by thein-

dividual as he develops, or whenhe is exposedto a newclassof objects.

When I firstarrivedat Cornell,I was invitedto work at the BehaviorFarm

whereProfessorLiddellhad a largeexperimentalpopulationof goatsa

herdof onehundredor more.I wasassignedsomegoatsas subjectsand hadthedailyproblem ofextracting mysubjects fromtheherd.I hadnever

seena goatup closebefore,andfor a fewweeksI spentmostof my working dayjustfinding mygoat.Butwithdailyexposure andplentyof searching, I eventually learned torecognize individual goatsandtoidentify new onesalmostimmediately. I had learnedthe distinctivefeaturesof

goats.Though I couldnottellyouwhattheywere,members oftheset

cameto possess,forme,uniquepatternsof thesefeatures. Themostelegantexample I knowofisthesystemofdistinctive features workedout forphonemes by RomanJakobson. Jakobson andHalle(1956) enumerateda set of twelvedistinctivefeaturesfor phonemessuchthat any

phoneme is distinguishable fromanyotherbyitsbundleofattributes. All

of the featureshavea plusand a minusvaluethat is, forma binary

opposition. Theopposition mayrequire a choice between polarqualities of

the samedimension,or presenceor absenceof a given attribute. The choiceof these featuresis both intuitiveand empirical.Objective

criteriafor choosinga featuretableare the requirements that the table mustprovidea unique bundleforeachmember of the set,andthatthe

properties chosenmustbe invariant overcertaintransformations (forin-

stance,differentspeakers,shouting,and whispering). To quoteJakobson and Halle,A distinctivefeatureis a relational propertyso that the minimum same of a feature in its combinationwith various other concurrent or

successivefeaturesliesin the essentiallyidenticalrelationbetweenthe two opposite alternatives (p. 14).

Thisfeaturedescription of phonemes constitutes, presumably, a set of ruleswhichthespeakermustobey,eventhoughhe isunawareofthem(as hesurelyis).Thepsychologist willwonder,at thispoint,whetherwehave any convincing experimental evidencethat theserulesactuallyoperate, thattheyhavepsychological aswellaslogicalvalidity. I amafraidthereis verylittle.Forthisreason,I havebeeninterested inworking outa feature table for a set of materialthat I could,myself,manipulateexperimentally.

Perceptual Development andtheReduction ofUncertainty 359 Research withDistinctive Features ofLetters

I choseas mysetof itemsprinted lettersa setof simplified Roman capitals. Developmental research on thediscrimination of letterlike forms

hadconvinced methatthereis progressive learning of contrastive rela-

tional features which permits theforms tobedistinguished. Theproblem was,first,to specify thesefeatures ina tablesoasto provide a unique

patternfor eachletter.Thefeatures,furthermore, mustbe invariantunder

various transformationsspecifiable transformations suchassize,perspectiveandbrightness changes, andotherssuchasthepersonal idiosyncracies

of differentwriters.Havingchosena list on a rationalbasis,the next problem wasto checkit by anyexperimental meansI coulddevise.

A listof features thatI selected for the testis arranged in a chart

comparable toa phoneme feature chart. Each letterhasa unique pattern of

features. Forsomeletterpairsthedistinction is minimal (onefeature), for

somegreat.Because of thisvariance, a goodtestof thevalidityof thelist couldbe obtainedin a confusion matrix.Forif discrimination of letters depends onthedetection ofcertaincritical differences betweenthem,then

therateofconfusion errorsmadebetween letterpairsshould depend on

the number of feature contrasts between them.

Accordingly, an experiment wasrunto obtainthe confusion matrixfor

everyletterpairedwitheveryother.A discrimination taskrequired the subject to matcha standard letterwitha listof sixletters,randomly

selected butalways containing a correct match. Every letterhadanequal chance to beconfused witheveryother,overa largenumber ofsubjects

andtrials.Thesubjectswerefour-year-old children.

Correlations wereobtainedbetweenpercentfeaturedifferences and number ofconfusions, oneforeveryletter.Ofthetwenty-six correlations, twelvewerepositive, onlya fairbattingaverage. Anotherexperiment wasruncomparing two conditions of discrimina-

tionone a set of lettermatching trialswithhigh,confusion context, anotherwitha low-confusion context.Thehigh-andlow-confusion valueswerepredictedfromthe featuretable.Errorsand reactiontimes

wererecorded, andresultsforthetwoexperimental conditions differed as

predicted.

Thefeature listsofartestedrequires revision. Hintsforrevising it can be obtained, I believe, fromthekindof multidimensional analysis that Dr.Warren Torgerson workswith.Dr.Torgerson has,infact,analyzed my confusion matrix withresults bothgratifying andpuzzling. Twodimensions cameoutwhichaccord beautifully withthefeature listcurve-straight,

anddiagonality. Beyond that,itwasnotclear whatwehad.Ournextstep

willbecollection ofdatafora bettererrormatrix, usingadultsubjects, and

360

E. J. Gibson

a furtherattemptat analysis . The featurelist canthen be amendedand a newexperimental testrun. Besidesaskingwhether a list of distinctivefeaturesof letters exists whichis reallyutilizedin letter discrimination , it is interestingto askhow discriminationtakesplace- how the featuresare processed in the act of discrimination : and how the list is learnedand its processingstrategy developed . Thesequestionsleadme backto my originalthesis , how perceptuallearninganddevelopmentactuallyoccur. TheRoleof Activity I havesaidthat what is learnedin perceptuallearninganddevelopmentis the distinctivefeaturesof objectsand the invariant featuresof events overtime. Whatcanbe saidabouttheprocesses involved?Thetopicof this symposiumis the role of activity in perceptualdevelopment . And hereI canbegin. Perceptionis active. It is not registration , or assimilation , or compositephotography , or automatichitchingof responses to stimuli, but rather active explorationand searchfor critical featuresand invariants . Search , it will be noted, takesplaceover time andthis is important, implying againthat perceptionis nevercorrelatedwith an isolatedmomentary stimulus . Theideathat perceptioncruciallyinvolvesexploratoryactivityhasbeen sressed by my husband , who distinguishes betweenexploratoryand performatoryactivity. Perceptiondoesnot incorporateperformatoryactivity of the kind that doesthingsto the environmentandits furnishings ; rather it exploresand searches out informationfrom the environmentand the stimulationit provides. The total potential stimulationreachinghis receptorsis neverperceivedby any man. He samplesfrom this vast pool; only part of the potentialstimulationbecomeseffective.To perceivemost selectively , strategiesof activeexplorationmustbe developed . Theexploratoryandactiverole of perceptionhasbeenstressed by other psychologists , Pavlovfor one. Pavlovspokeof the "investigatoryreflex," whichhe contrastedwith the conditionedreflex. I believehe meantby this termwhatI shouldpreferto callattention,the activeexploratory,selective aspectof perception . I have said that somestimulusinformationis selectedfrom the total stimulation , andthat the strategyof perceptualsearchandselectioncanbe good or bad. This implies, to me, the necessityof perceptuallearning. Optimal strategiesdiffer for differenttasksand differentsetsof potential stimuli, so learningtakesplacein the exploratoryactivity itself. At once, the questionarises , how is it determinedwhat gets selected whenlearningoccursin a given situationbetweenone occasionand the next? Do we want to invokereinforcement ? I think not in the traditional

Perceptual Development andtheReduction ofUncertainty361 senseof an external rewardor confirmation, butinsteadin thesenseof

taskutility. Theultimate goalofperception, I think, isdifferentiation, the reducing ofuncertainty; ortoputit another way,theextraction ofdistinguishing features andinvariants froma stimulus fluxwhereinformation is

verygreat.Whereit isnt great,no learningis needed.Babiesalmost

certainly distinguish contrasts oflines andedges presented onhomoge-

neous backgrounds without anylearning atall.Butfaces, withtheirvariety offeatures andexpressions, present toomuchinformation andtaketime

and practiceto differentiate.

Reduction of Uncertainty

Theprinciple ofreinforcement inperceptual learning isutility forthereductionofuncertainty. Thatis justexactly whatdistinguishing features and

invariants of eventshave;theyserveto reduceuncertainty in a world

otherwise toofullofinformation. Features thatarenotdistinguishing and variables thatdonotserve tocutdown uncertainty must beignored. The

constancies areexamples ofinvariance inperception, casesofreduction of

uncertainty by attendingto therightfeaturesof the stimulus. Whena car

drives toward me,itsprojected size, proximal-stimulus-wise, keeps increasing.ButIdonotseeaninfinite number ofdifferent carsoflarger andlarger

size.Thereis an invariant aspectofthestimulation, overtime,whichcan

be and is attended to.

Nowit willbeclear whyI amagainst additive theories ofperceptual learning. Itisbecause Ithink thesecret ofitisjusttheopposite, reduction.

Perceptual learning is takingout,notaddingon.Theeffective stimulus which activeandeducated perception picksoutisa reduced stimulus. It is extracted, filtered out,whereas otherstimulus information whichhasno

utilityfordifferentiation is ignoredby theeducated attention.

Someone mayaskwhatmotivates thiskindoflearning. I amcontent withthenotionthatthereis a built-in needfordifferentiation, forthe

reduction ofuncertainty. Others (Berlyne 1957) have spoken ofperceptual alsotostimulus conditions characterized bynovelty, complexity, surprisingness, incongruity, anduncertainty. These properties aresaidtoincrease exploratory activity. Measurements ofperceptual curiosity, inBerlynes senseoftheterm,dotendto showlesssuchcuriosity inretardates thanin equal M.A.andC.A.normal male subjects (Hoats, Miller, andSpitz 1963). Butwhatever onecallsthemotive, moreactive exploration would be expected toeffect greater perceptual learning andthuscognitive growth. Iwould liketobring inalittle support from information theory, specificuriosity. Thisconstruct hasbeenrelated, by Berlyne, to arousaland

callyfromtheworkofGarner andClement (1963). Garner andClement

showed thatgoodness ofpattern iscorrelated withitsuncertainty, the

362

E. J. Gibson

sizeof the set to whichit belongs.To quotethem,A goodpattern... is

onewhich isperceived asstable, asnoteasilychanged, andashaving few alternatives...,the best possiblepattern is that perceivedas unique.

(p.452)Thisstatement delights methe good pattern isinvariant and

unique. These attributes, I believe, aretheendsofperceptual development; thusfollowsthe reductionof the set and the trendtowardspecificity. I doubtthat the correlationbetweenperceivedsizeof set andgoodnessis a

causalone.Theyareboth,I think,resultsof discrimination of stimulus attributesof the setandpatternsofattributesof membersof the set.Some attributes, suchas linearityandsymmetry, evidently areeasilypickedup,

evenby youngchildren andretardates. Othernonlinear onesarenot so obvious, and this is where development comes in.

Letmeciteonemoreexampleof a developmental trendtowardperception of higher-dependency constraints. Perception of languageprovides

manyexamples ofchunking,touseGeorge Millers term;thatis,process-

ingbigunitsinstead oflittleones. Long ago,Cattell (1885) showed that

onlyfourunconnected letterscouldbeperceived ina verybriefexposure,

buttwelveor morecouldwhenthelettersformeda word,andstillmoreif subunitsof wordsformeda sentence. BryanandHarter(1899),alsolong

ago,showed thatreception ofMorse codeshifts during learning toward decoding bylarger units,syllables, words andsoon,ratherthandecoding thesymbols forsingleletters.Obviously, theinternal dependencies which structurethe superordinate unitsare learnedin both thesecases.

I myself havestudied thedevelopment ofperception ofspelling patterns

inreading. Children inlearning to readmoveratherquickly fromsingle letterdiscrimination to discrimination of longerstringsof letterswhich

havepredictable correspondences withspeech sounds. Thesespelling-tosoundcorrespondences formunitsand,asthechilddevelops skillhepro-

gressively handles unitsof increased lengthanddependency order.He perceives theseunits,ordered byinternal regularities, although hecould not verbalize the rules for you. Conclusion

Apoetonceremarked thatthe worldissofullofa number ofthings.

It is, but that doesnot makeus happyas kings,it is onlybewildering.

Being happy asa king, cognitively speaking, isbeing abletoselect outof

the massivestimulusfluxinformationthat is relevantand that is invariant

overtime,and thusto reduceuncertainty. Therest is thrownaway,ig-

nored,by a smartadult.Hehandles asmuchashe can,learning as he growsupto picktheattributes whichhavegreatest utilityforreducing uncertainty.

Perceptual Development andtheReduction ofUncertainty363 References

Berlyne, D.E.Conflict andinformation-theory variables asdeterminants ofhuman perceptualcuriosity.J. Exp.Psycho!., 1957,53,399404. Cattell, J.McK. Ueber dieZeitderErkennung undBenennung vonSchiftzeichen Bildem und Farken.Phil.Stud.,1885, 2, 635650.

Bryan, W.L.,andHarter, N.Studies onthetelegraphic language. Psycho!. Rev., 1899, 6, 345375.

Garner, W.R.Toperceive istoknow. Amer. Psycho!., 1965, 20,569(Address given as recipient ofDistinguished Scientific Contribution Award).

Garner, W.R.,andClement, D.E.Goodness ofpattern andpattern uncertainty. I.verb, learn. verb.behav.,1963, 2, 446452.

Gibson, 1.1. Perception asVol. afunction ofstimulation. Pp.456501 inS.Koch (Ed.), Psychology: A Studyofa Science, I.NewYork:McGraw-Hill, 1959. Gibson, 1.J.andGibson, E.J.Perceptual learning: differentiation orenrichment? Psycho!. Rev., 1955, 62, 3244. Hoats, D.L.,Miller, M.B.,andSpitz, H.H.Experiments onperceptual curiosity inmental retardates andnormals. Amer. J.Ment.Defic., 1963,68,386395.

Hubel, D.H.,andWiesel, 1. N.Receptive fields, binocular interaction andfunctional architecture inthecatsvisualcortex. J.Physio!., 1962,160,106154.

Hubel, D.H., andWiesel, T. N.Receptive fields of1963, cells instriate cortex ofveryyoung, visually inexperienced kiltens. J.Neurophysio!., 26,9941002.

Jakobson, 1956. K., and Halle, M.Fundamentals ofLanguage. The Hague: Mouton and Company, James, W.ThePrinciples ofPsychology, Vol.I.NewYork: Henry HoltandCo.,1890.

Leontiev, A.N.Thenature andformation ofhuman psychic properties. Pp.226-232 in B.Simon (Ed.), Psychology intheSoviet Union. Stanford: Stanford Univ. Press, 1957.

Maturana, H.R.,Lettvin, J.Y.,McCulloch, W.S.,&Pitts, W.H.Anatomy andphysiology ofvisioninthefrog(Rana pipiens). J.gen.Physiol., 1960,43,129175.

Postman, L.(1955). Association theory andperceptual learning. Psycho!. Rev., 62,438446.

Postman, L.(1963). Perception andlearning. InS.Koch (Ed.), Psychology: Astudy ofascience. Vol.5.NewYork:McGraw-Hill, 1963,Pp.30113.

Pick, H.(1963). Some Soviet research onlearning andperception inchildren. In1.C.Wright

& J.Kagan (Eds.), Basic cognitive processes inchildren. Monogr. Soc. Res. Child Develpm., 1963, 28, No. 86. 185190. Miller, N.E.&Dollard, J.(1941). Social learning andimitation. NewHaven: YaleUniv. Press.

TrendsinPercep ua1 Developmenf EleanorJ. Gibson

I have chosen toreprint partsofthislastchapter from mybook onperceptual learning forseveral reasons. Itstands alone fairly wellasa kindoflastword or

summing up.A version ofit served as mypresidential address to theEastern

Psychological Association in1969. This wasa memorable occasion formethe book hadjustcome out,ithadwontheCentury Prize, I hadbeen made afaculty

member at Cornell at last,andthere were allmyfriends intheGrand Ballroom oftheSheraton-Park Hotelin Washington. Notonlythat,therewasa serious

political confrontation atthebusiness meeting, which justpreceded thepresidents address. Adrenalin flowed, especially mine. Nonmembers occupied theplatform andaddressed theassociation infavor ofa vote forsome highly politicized issueof theday.I cant remember whatthe issuewas,but I remember the

turbulance oftheaudience. Itallturned outhappily, intheend,andwecelebrated

thatnight ina suite donated bythehotelvery grand andgaudy. Intheaddress I triedtodealwithsome age-old problems thathaddominated

workinperceptual development fordecades andI believe stilldotosome extent

forexample, which isprior, thewhole orthepart.Thatissue wasa major oneas lateas1981, when I wasdiscussant ata conference held atDartmouth Colleges Holderness Conference Center (Tighe andShepp1983). There wasrecurrent

discussion forseveral days astowhether development progresses fromhIhlt toanalytic. I found myself arguing allover again thatperception isalways

unified, however differentiated. Thelayout oftheworld andevents, especially

events, aretypically embedded, unitswithin units. Weperceive wholestheunits

canbeofanysize. Ashift toperceiving larger units isasnatural astheother way round, because order intheworld isembedded andperceptual learning yields detection ofboth finer texture andmore encompassing structure, another example

oftheduality ofperception. It hasbeen suggested, occasionally, thatthisisan example ofthephilosophical issue ofthe oneandthemany.ButI thinkthe

Excerpts fromChapter 20ofPrinciples ofPerceptual Learning andDevelopment, ' 19O9,

pp.450472.Reprinted bypermission ofPrentice-Hall, Inc.,Englewood Cliffs, NJ.

E. J. Gibson

366

world we perceiveis alwaysone, howeverdeeplynestedthe informationwe can obtain .

Theotherissuespersistamongsomedevelopmental psychologists , but as they wane, newonescometo take their place. Thereis little talk of stages , nowadays , nor of development from raw perceptionto inference . Insteadthere is a new

argument aboutinference - whetherit alreadyexistsas a rathersophisticated form of logicin neonates (see , for example , Bower1989). Someclassicissuesthat persistdespitechangingfashionsin theoryarediscussed in a recentpaperof mine (E. Gibson 1987). Heredity vs. environment, continuity vs. discontinuity, mecha-

nism vs. functionare hardyperennials . Thethreetrendsobserved in perceptualdevelopment still seemvalid- increase in specificity , optimizationof attention, and increasingeconomyof information pickup- althoughtheycoulddo with somerevision. Oneof thetrends,optimization of attention, needsa lot of revision. I have come to doubt that perception

becomes moreactive, or moreexploratory , or moreselective with development . We know much more about perceptionin infants than we did in 1969, and there is a

wealth of evidencethat perceptionis active, exploratory , and selective from the beginning . Mechanismsof explorationdevelopand efficiencyincreases as new action systemsbecome available. By this time, onecouldwrite a muchfuller story aboutdiscoveryof structure - therole(or nonrole ) of Gestaltprinciplesin percep tual development for instance . Thebig gap, what was missingin all the evidence summarized for thesetrends (most of which I omit here), is the absenceof any

researchon the role of dynamicfactors in perceptualdevelopment . No research existedin 1969on perception of eventsor theroleof motionin pickupof structure in earlydevelopment . Now thereis a wealthof it to put to work in an up-to-date view of perceptual development . nowadays is thetight linkageof Thebig newemphasis for manypsychologists perceptionand action in a systems theory. Provocative, consideringwhat was in the air in the sixties , is the hint that that was on the way . The idea was around

in many forms, in the Russiandogmas, in Piagetian thinking, and in J. J. Gibson's

perceptual systems , which emphasized activesearchand exploration . Theseideas wouldberefashioned many timesin thenexttwentyyearsand cometo fruition in the ecologIcal . approachto perception(Gibson- 1979) . 1- , 1-' "

"" " v

" v6

. "' ~ ~ ~

r ~rr

' _ PO_ OF

- -

r

-

,

.

Last Thoughtson SomeClassicFalseIssues

Learning orMaturation ? This issuewasexamined in chapter12, wherewe considered rearing experiments andthelighttheythrowonperceptual development . It seems clearthatasa dilemma it canbeburied , for concurrent processes of both maturation andlearningcanbe demonstrated . As for learning , children (andanimals ) learnto attendto distinctive features of things , to invariants

TrendsinPerceptual Development 367

thatleadtoperceptual constancy andpermanence, andtohigher order

structures andrules.Thelattermaybebestobserved inhumanskillssuch

asspeech andreading, butperception ofinvariance overtransformations,

suchasa ballbeing seentorollbehind a screen andoutagain, canbe

demonstrated to developin a cat.

Itwasalsodemonstrated thatalltheprogress witnessed intheseaccom-

plishments isnotsolely amatter oflearning. Experiments onthevisual cliff withanimals of different ages,species, andrearing histories showthat

maturation hasa roleindepth perception. Theeffects ofspecial rearing ecologically normal forthespecies, andthatabilities already mature require itfortheirmaintenance. Thetrends inperceptual development emerge as the product of both experience with an environment and the maturing powersof an individual.Thereis no either-orissue. conditions alsomadeit clearthatmaturation assumes an environment

Perceptionand Production

Thequestion ofwhich comes first, perception oraction, islikethequestion ofwhich comes first, thehenortheegg.Thecurrently fashionable predilec-

tionformotortheories ofperception would seemtoimply thataction has priority, butweknow, ontheotherhand,thatproduction oftenfollows

after discriminatory achievements. Weconsidered thisfactinthechapter

onperception of symbolic stimuli, forspeech is a casein pointwhere

discrimination oftenprecedes production. Another interesting caseisthe

drawing ofadiagonal line(see chapter 8).Achild candraw anacceptable circle atagethree, a square atfour, a triangle atfive, buta diamond only

laterstill,at seven.Thisis an interesting progression in itself,butit becomes stillmoreinteresting when onelearns thatthediagonals ofthe triangle andthediamond canbediscriminated considerably earlier than theycanbedrawn. Itisnotthemotor performance assuch thatiswanting,

fortraining inexecuting themovement does notfacilitate accurate produc-

tion, whiletrainingin discriminationdoes.

Leaving visible tracings onpaperwitha tool,tracemaking, is anin-

teresting eventto children, asJ.J.Gibson andPatricia Yonas haveshown,

whereas themotor actwithout thetracings, aswitha pencil thatleaves no

marks, isnotinteresting. Even watching thetracings being made, although thechildhimself is notthedirectexecutor, is interesting. I observed children atanexhibition ofkinetic sculpture bytheartist Tingueley. One oftheexhibits wasa complicated machine which helda pencil. When activated thepencil made random marks ona piece ofpaperscribbles. Themachine wasprovided withpadsofpaper andcolored marking pens

which achild could putinplace. Crowds ofchildren were waiting foraturn

at thismachine. Sothereisa strongmotivation merely to observe marks

368

E. J. Gibson

being madeonpaper, anditisnotthemotoractassuch, thefeelofit,that isresponsible. Thereis noparticular reason, therefore, whytheabilityto produce a givenpatternandtheabilityto discriminate it should emerge at

the same time. Motor schematathere may be, and response-produced

stimulationas well.But perceptionis an activityin its own rightan

exploratory activity, nota performatory activity. Performatory actshavea

developmental history, butsodoexploratory acts,andthehistories arenot the same.

Whichcomesfirst,perceptionor production? The questionis a red

herring thatleadsusoffthescent. Perception andproduction servedifferentpurposes. Adaptiveness of behavior is servedby exploratory perceptualactivitythat providesinformation aboutthe environment priorto performance. Exploratory activitybeginsearly,yielding informative stimulationwith the firsteye movements.We wouldnot call these eye move-

mentsproduction, but they are a kindof activity.Performatory action,

likeseizing a profferred rattle,orbanging a spoonona tinplate,appears laterthanlooking attheobject. Butwhenit comes about,thereisinformative feedbackfrom this action,too, whichtakes on a role in monitoring behavior.

The Part or the Whole?

Anotherwhich comesfirst questionthat has bewildereddevelopmental

psychologists isthepart-whole controversy. Ithasbeenthecontention of manychildpsychologists thatchildren beginbyperceiving globally, and progressively analyze ordifferentiate outthedetails. Butjustasmanyhave takentheopposite viewthat theyoungchildbeginsby noticing only details in isolationand gradually,through a learningprocess,integrates them into a whole.

Thisissuewasconsideredin chapter16,on the perceptionof objects.It

is anotherfalseissue.Children learnto perceive distinctive featuresof

objects, andsoonemightbeinclined to saythatthesearepartsandthat theyarebeingdifferentiated froma whole. Butwhatkindofwhole? Nota

wholewith intricatestructure,certainly.It is equallytrue that the pickup of structurecharacterizes development, in factprogressively higherorder levelsof structure,andthissoundslikeprogresstowarda whole.Butthe

verynotionof partsandwholesin perception is mistaken; objectsare differentiated by distinctive featureswhichmustbe discriminated, and objectsarealsocharacterized by structure. Higherorderstructurecreates

newunitsbygrouping subordinate units,enabling moreinformation tobe

handledwhilereducing uncertainty. Enlarging the chunksmightbe con-

sideredintegration of partsintowholes, but we mustnot forgetthat thereisalsoprogress towarddiscovery ofthemosteconomical andcritical set of distinctive features. Whichof thesetwoequallyadaptivekindsof

Trendsin Perceptual Development 369

changecharacterizes behavior willdependuponthetaskandthestimulus

information,and both may even occurat once. FromPerceptionto Inference

It isoftenasserted thattheyoungchildisstimulus bound, enslaved bythe

surrounding milieu,dependenton the presentsensoryinformation, and thatperceptual development is a processof liberation fromthe constraints

of stimulation. Anotherwayof puttingthishasbeento saythathe is

misledby perceptual factorsandmustmakeinferences. A relatedstatement

is thatcognitive development consists in going beyondtheinformation

given (Bruner 1957).

Onecannotdoubtthatthechildsconceptual lifeexpands ashematures

andgainsexperience. Concepts andgeneralized rulescanhavea guiding

anddirectingeffecton perception, justaslabelsandverbalinstructions can

directattention to distinctive partsofa display, andconcepts, especially

rulelike ones,canhelpto revealstructure noteasilyor automatically detectedperceptually. Butthisisnotto saythatperception isleftbehindin favorof inference as wegrowup,noris it evento saythatperception

develops by making use of inference.

Whatis wrongwithsayingthattheyoungchildis stimulus bound,and

thatcognitive development isa liberation fromthesebondsbytheoperationsofintelligence? ThisisPiagefs opinion. Onemustadmititspopular-

ity and its persuasiveness, for a neonates attentiondoes seem to be

captured bya fewkindsofeventsinitsenvironment. Butthedevelop-

mentalchangeis not oneof doingwithoutstimulus information; it is one

of seekingstimulus information in a directed, systematic fashion. Does perception mislead us,whereas conception andgeneralization leadto truth andreality?Conceptsandgeneralizations cansometimes themselves mis-

leadus;wespeakofbiasedobservation. It is thennotperception thatis

misleading. We are misledby the failureto graspthe invariants in the

stimulus flux.Whentheinvariants occurovera temporal sequence of

transformations, asinaneventlikethepouringofwaterfromonecontainer

to another,they may or may not be detected.A still shot of a filled

container at onemomentof stimulation canbe misleading, likea single

framefroma motionpicture,for invariantsoccurthat can be discovered only over variation and transformation.

Does the childneed more informationin stimulation,or more redun-

dancyin stimulation, thanadultsdo because hisconcepts areimmature? Onekindofevidence citedforthisconclusion isdrawnfromexperiments withincomplete figures. Children canfillinthefigures andrecognize them betteras ageincreases. To quoteWohlwill, Comparedto theadultthe

youngchildrequires moreredundancy ina patternto perceive it correctly; thusbothincomplete andverycomplex patternswillbe difficult forhim

370

E. J. Gibson

( 1960

, p . 281

younger

) . They

child

handle

more

using

redundancy

objects

Bruner

,

"

of

also

,

.

As

the

model

matching

to

variants

in

and

in

number

that

cepts

our

. We

would

the

be

world

Stages

us

brated

,

the

possibly and

a

question

.

the

the

time

some

? point

with but

behavior

biological

but

this

one

rep

-

input

it

to

is fill

in

interpretation not

of

effective

upon

it

.

of the

can

be

pickup

older

-

not

our . If

is

in

does

of

what

,

this

it

more

about

-

concepts

,

reflections

conceptualize

in

useful

of our

grow

more

process

extracting

Although

we

of

a

stimulation

more

and

con

we

-

did

going

on

, in

difficult ,

can

is

to

emphasizes

a

trend

identity If

so

the

,

.

But imply

while how

do

- cut a

of

it

is

answer

.

describe

or

are new

to

this in

adapting

gradual

to growth

up

,

of

as

if

in

psycho

transition

exhibiting we

, quite

continuous

identifying

is

metamorphosis

speeded

the

" lean

there

arbitrarily

a

of

change

a

,

fashion

manner

cases if

the

are

and

is

-

could

and

stages

clear

and

only

,

abrupt

there

point

lover

contrasts

no

cele

infant

continuous

structure

trends

mewling

is

for

Shakespeare

different in

where

in

with

?

these

there

.

, the

transition

insects

Does

the

these

proceeds

also

that

.

Are

spectacular

maintaining

adaptation

.

the

shift

one

as

problem

alike

how

,

radical

fascinating

schoolboy

man

the

a

biologist

dramatic

Biologically

concerned

is

,

really

to

are

the

wondering

when

picture is

time

and

of

is

motion

trends of

a ,

chapter

logical

ages

transformation

- lapse This

grow

on

the uses

is

as

whining et

that

can

environment

Y

sequence -

follows

,

where

with

information

,

the ."

the

the

One

development

we

recognition based

child

Perception

we

over

man

sequence

in

organization

which

in

more

getting

with

over

times

is

against

,

because

,

pantaloon

from

there

of

as

perception

complexity

change

of

identified

identity

chosen

of

.

from

that

head

on

archaeologist ages

slippered

some

kind

the

seven be

features knowledge

needed

the

and

? of

dramatist

seeing

pictures is

,

cannot

.

manifestations

the

at

evidence

Constraints

become

for

adept

" the

He

. Distinctive

disagree

the

, and

maladaptive

Transitions

I

dependence

less

."

reality

.

.

perceive

around

or

The

not

increased

suggests

depends

percepts

do

he

,

in

, richness

, because

.

now

less

an

the

development

than

less

present and

is

,

information

far

redundancy

of

richer

development

increase

is

suggest

incomplete

stimulation

by

is

again

reasoning

in

, I

information

recognizing " less

representation

rather

and

,

model

knows

adaptive

follow

The

not

surplus

it

) , using

grows

stimulus

formation

in

23

perceptual a

,

however

aid

,

given

reader and

,

constructed

not

adult

that

1966 .

a

are

perception

it

on

this

that

As

.

pictures ,

details

al

but of

when

learned

claimed

et

incomplete

an

to

has

,

resentations tested

be

expected

( Bruner

difficult amount

stimulation

to

be

more

greater than

in have

would

older

be a

information

do

them

may

requires

progress the

changed

a

in

a

beha

-

toward

Trendsin Perceptual Development 371

viorandstillrecognize theidentity? Ordoesa trendimply notgradual transitions butstagesofbehavior, withabruptchanges in organization

comparable to metamorphosis in insects?

Human growth, afterbirth,doesnotdisplay metamorphosis. Theclosest

thingto it is thespeeded-up change at adolescence, andperhaps the heightened responsiveness occurring at so-called critical periods during infancy. Weareaware ofthegreatdocility ofthehuman organism, andthe

extentto whichitsbehavior canbe shapedby programmed schedules of

training. Thecontribution oftheenvironment issogreatthattheappearance of stages might result from the program that theculture hasprovided for the childs education. I wantto lookfortrendsin development, butI amverydubiousabout

stages. Instead ofthechildstudypsychologists offiftyyearsago,who

thoughtofthechildasalmosta distinctspecies, or eventhemorerecent

experimental child psychologists, letustrytobedevelopmental psychologists.Letusexamine developmental studies ofperceptual activity, hoping to discovergeneralizations that willrevealthe lawsof behaviorand its

adaptation to ongoing events. Welookfora progressive sequence which spansthe activityfrombirthto maturity.

It isa hazardous undertaking, butI amgoingtopropose, andsummarize evidence for,certain trendsinperceptual development. Torepeat, trends donotimply stages ineachofwhich a radically newprocess emerges, nor do they imply maturation in which a new direction exclusive of learning is created. ThreeTrends in Perceptual Development IncreasingSpecificityof Discrimination

Inearlierchapters, it wasshownthatperceptual learning is characterized

by a progressive increasein the specificity of discrimination to stimulus

information. In chapter9, I described theeffectsof practice on acuity judgments, on differential limens, andon absolute estimations alonga stimulus dimension; allshowed thiskindofchangea narrowing of the

bandof stimulusvalueselicitingany givenresponsevaluewithinthe continuum. Thesametrend,thoughcomplicated byspecies-specific methodsofadaptation, appearsin phylogenetic development. Withevolution,

specialized receptors proliferated, alongwithdistinct modesofsensitivity

to differenttypesof energychange. Theevidence citedin preceding chaptersshowedthatthehumaninfant

differentiates properties ofthingsandevents to somedegree quiteearly.

Nevertheless, a developmental trendtowardspecificity canbe demonstrated in the ways describedbelow.

372

E. J. Gibson

Decrease

in

is

stimulus

called

the

class

Stimulus

of

various

age

from

; the by

seven

and

then

,

group

generalized

creased

with

were

eighteen for

distinguish

followed

a

to

for

the

of

, had

an

not

, if

group

these

the

the

yet

effectively

the

conditioned

is

an

, ranging

a

con

three

youngest in -

oider to

groups all

. Properties

words

that

differentiated

and

response

. ( Other

-

synonym

synonyms

more

words

a

infancy

. The to

only

which

response

generalized of

of

, and

to

words

tradition

groups

antonym

of at

example

age

electrodermal

, an

size

to

. One

four

what

response

between

. Generalization

other

elicited

steepen

generalization

youngest

were

stimulus

decreases

a half

the

same

- response a

fact

subjects

relationship

word

in

homophones

particular

words of

to

the

stimulus

a homophone

tested

. But

appearance

and

the

around

does

of

comparing

elicit

strict

which

which

, relative

the

the

of

will

generalization

) , in

most

compared of

of

were

age

regardless

in

associated

( 1946

to

a word

that

gradient

amount

to

existed

a method

been

Riess

years

ditioned

is

measurement

a method

events

generalization has

maturity

experiment

or

. It

The

experimental

affords

stimulus

levels .

response

and

The

generalization

different

psychology given

Generalization

,

relate

merely

the

examples

.)

Reduction

of

reduction

with

Variability

ment

. Decrease

with

age

in

. The

provide

with

the

height

This

is

of ,

. six

of

Piaget

. The

it

is

years

width

times

as

with

age

of

a

his

rods

great . ( Other

the as

group

varying area the

asymptote

constant

over

as

more

uncertainty

mean

constant followed

the

size

twelve

each

as

error

in

half the

the variable

, and

it

con

the

the

height

to of

error narrows

-

same to

height

. by

and on

-

( for

, not

compared

equivalent

con

years

child

, plotted

than

size

matches

precision

little

accepts

in

about

by

error

of

examples

at

made in

perception

change of

-

development on

) . The

remarkably

showing experi

of

variability

increase

but

age

in an

cited

psychophysical

collaborators

1946

judgments

genuine

age

of

to

of

were

a

characteristic

and

. The

with youngest

a group

The

five

cases in

also

decrease

discrimination

changes

is

( Lambercier

variability

so

many error

error

of

) from

,

9

variable

a striking

the

their

rod

standard

gressively

in

error

standard

about

a rod

also

variable

kind

shows

children

sistency graph

age

drop

different

this

chapter the

demonstrations

of a

of

experiments

many

stancy

In

practice

a

the pro

is -

.)

Reductionof Discrimination Time Psychologists have known for many years that the reaction time required for making different responsesto different objectsthat are suddenlypresentedincreaseswith the similarity of the objects. The latency corresponds, in other words, with their discriminability. It is interesting, therefore, to find that the reaction time for discriminating between objects decreaseswith age.

Trends in PerceptualDevelopment

373

The disjunctive reaction time experiment by Yonas and Gibson , de-

scribed in chapter 7, is an example. The subject pushed a lever in one direction if a letter exposed on a small screen before him was a letter that

had been designatedby the experimenter; he pushedit in the other direction

if it was any

ot .her letter . The class of letters

used was limited .

Children from secondgrade, from fourth grade, and college studentstook part in the experiment. There were huge differences between the age groups

.

Could it be that the younger children simply do not understandthe task, or are not always paying attention? The latter explanationis possiblebut it is not sufficient , for latencies of reaction

reflect the influence of similarities

and differences between stimulus displays much as do the adults'. In an

experiment with same-different judgments, a confusion matrix for a set of letters was obtained (seechapter 5). The latenciesfor discrimination correspond to the tendency to confuse the letters very well; for example,

discriminatingE and F hasan extremelylong latencyand the lettersare frequentlyconfused , whilediscriminating C andM hasa very muchshorter one and they are almost never confused. These differences appear in the

data of both seven-year-old children and adults, although the children's reaction times for any given pair are up to twice as long as the adults'. There may be a shift with age in the set of distinctive features used in discrimination that accounts for the shorter time ; either a reduction of

the set to a more economicalone, or a pickup of tied features(higher order structures) that reduces the number of comparisons required for discrimina -

tion as same or different .

The Optimization of Attention

The secondtrend in perceptualdevelopment I shall call the optimization of attention . We refer to the selective aspect of perception as attention ,

implying activity on the part of the perceiver. Now I want to point out that there are changesin the strategy of this exploratory activity, which are in fact correlatedwith other developmentalchangesin perception. They are, first, the tendency for attention to become more exploratory and less captive; second, the tendency for the exploratory searchto becomemore systematic and less random ; third , the tendency for attention to become

more selective; and fourth, the inverse tendency for attention to become more exclusive .

From Capture to Activity The aim of perception is to get information from the environment . But from the vast array of potential stimulation arriving at receptor

surfaces at any instant , selection must occur . There seems to be

some mechanism for selection very early in life . Precocial animals at birth

will follow instinctively a moving, shining, clucking object; they will in fact

374

E. J. Gibson

becomeimprinted by it . A humaninfant is incapableof running bodily after a moving thing, but he will pursue it with his eyes, demonstrating his attention by prolonged fixation. As Ames and Silfen (1965) pointed out, the young infant is not so much himself selectingthe bright comer or the dancing sunbeamto attend to, as it is capturing his attention. Others have referred to this kind of attention as "obligatory" or "involuntary." William James , who was not afraid to use the word "voluntary," said: "This reflex and passivecharacterof the attention which, as a Frenchwriter says, makes the child seemto belong lessto himself than to every object which happens to catch his notice, is the first thing which the teacher must overcome" (1890, p. 417). Along the same lines, Piaget speaksof centration, meaning that the young child is caught by what he calls Gestalt-like structuralfeaturesof the stimulus array, from which he is later able to releasehimself by letting his intelligence take over. I do not like to make this distinction between perception and intelligence, for I think there is intelligent perception. I cannot doubt, however, that perception becomesmore active. Periods of fixation becomeshorter and visual exploration of the world increases . How attention becomesvoluntary, we cannot say, but we can describe someways in which it changesfrom being capturedto being exploratory; from being wandering and mobile to being sustained; and from being random and repetitive to being systematic; in short, toward optimizing the active searchfor information in the world of stimulation. Strategies of Search How is the searchfor information optimized? We have noted some characteristicsof the early development of visual exploration in the chapter on object perception- how the infant reachesout to the world visually, with stationary fixations and pursuit fixations of his eyes, long before he can grasp objects with his hands. Zinchenko et ale(1963) studied the child's method of visually familiarizing himself with an object and the way the method changesfrom three years of age to six. They photographed the eye movements of a child while he was examining an unfamiliar design to be remembered. The three year olds kept their eyes fixed on a single spot longer than did the older children, did not seekout distinctive features, and stayed within the area of the figure rather than following the contour. The visual exploration of the six year old was quite different. Fixed '-' ~azeswere of briefer duration, the number of movements was much greater, there were movements along contour, and there was orientation to distinctive features. This was not due to mere loosening up of the eye muscles, for Zinchenko showed that children of three could pursue a lighted moving target following the same contour very accurately. What was lacking at three yearswas locating the distinctive features and singling them out from the total display. Zinchenko also comparedthe

Trendsin Perceptual Development 375

eyemovements during thelateractofrecognition ofthepicture previously givenforfamiliarization. Thethreeyearoldswhohadnotadequately searched werepoorat recognition andnowmademanymoreeyemove-

mentsthan they had originally; but the six year oldsshoweda different strategy.Therewasa diminution of thenumberofmovements, aneconom-

icaleye-movement trajectory, examination of a fewkeyfeatures, anda

passingover of the redundantinformation,such as much of the outline. (Other examplesfollowed.)

Attention develops, then,froma relatively fixedorforcedactivity to an

exploratory activity, which inturnprogresses toward selective, systematic, andflexible patterns ofsearch thatareadapted to thetask.Thisactivity cannotbe described in termsof a chainof stimuli andresponses, or in

termsof a sequence of sensations. In writingaboutactivetouch,Gibson

(1962) pointedoutthecomplexity ofthestimulation to whichthesubject intentionally exposes himself whenhemoves hisfingers overtheobjectin

a varietyof ways;he doesso in orderto get the information fromthe flux

ofstimulation. Heisactively selecting theinformation ormaximizing it. Selective Pickup ofInformation Selective attention, focusing onthewanted

information, seems to mature developmentally. People working withdis-

advantaged children whomakelittleprogress in schooldescribe theirlack ofattentiveness astheoutstanding qualitymarking themofffromchildren whomakesatisfactory progress. Butuntilrecently wehavehadlittleother

thananecdotal evidence to showthattheability to attendselectively ina sustained anddirected manner grows withage.Wenowhavesome experimental evidence. Eleanor Maccoby hasmodified Broadbents technique of selectivelisteningto askwhetherchildrensabilityto attendto onekind

ofmessage andfilter itoutfromthetotalauditory inputincreases withage

(Maccoby 1967).Themethodinvolves presenting the subjectwithtwo messages concurrently, coming overa loudspeaker or through earphones.

Whenchildren listened to a mansvoiceanda womans voicespeaking

wordsat thesametime,withinstructions to reportwhatonlyoneof the voiceswassaying, thenumber ofcorrect reportsofthewordspoken bythe specified voiceincreased withage.Conversely, the numberof intrusive erors,thatis,reportsofwordsspoken bythenonspecified voice,decreased progressively. Therewasimprovement in theabilityto selectthewanted stimulus information.

Doesthegrowingchildreallyimproveinhisabilityto filtertherelevant fromtheirrelevant, to selectwhatis wantedanddiscardwhatis unwanted

before finalperception occurs? Ordoeshe,perhaps, sensebothmessages equally andonlyremember betterthemessage thatwasdesignated as relevant? Thisisa debatable question atpresent. Experiments witha pre-

paratoryset (tellingthe childbeforethe messageswhichvoiceto listen

376 E. J. Gibson for, rather than after) indicated that perception is truly selective, for a prior set to listen to one voice improved the report as comparedto an immediately subsequentinstruction about which voice to report. The advantage was present , Maccoby found, to as great an extent in the younger subjects ~ as the older. (Other examplesfollowed.) These experiments, especially the one investigating the role of preparatory set, demonstratethat six-year-old children do have the ability to select wanted information from a complex stimulus display, though they do so lessefficiently than older children. Ignoring IrrelevantInformation If a component is to be selectedfrom a complex of ongoing stimulation, then the rest- all that is irrelevantmust be discarded. The opposite of attention to somethingis inattention to something else. Is it the ability to ignore what is unwanted that develops with age? Is it, perhaps, the ability not to attend that increases ? The so-called incidental learning experiment seemsto be an appropriate paradigm for this question. Do both incidental learning and intentional learning become more effective as a child grows older or, on the other hand, does the task set progressively enablea child to shut out stimulation that does not aid him in achieving the assignedend? In an unpublished experimenton learning to identify letters, SharonShepelaand I found that five-year-old preschoolchildren noticed and rememberedcharacteristicsof the letters which did not distinguish them uniquely and were not needed for the identification. The task was simply to learn the namesof nine roman capital letters of the alphabet. The letters presentedwere colored, three of them red, three blue, and three yellow. After a number of practicesessions , the children were askedto identify the samenine forms but uncolored; they were now black. As expected, if a child had learned to identify correctly (say) five or six of the colored letters, he could do about as well with the black letters. We then asked the child if he rememberedwhat color each letter had been, showing him the black letters one at a time. To our surprise, these children rememberedcorrectly as many or more colors as they did letter names, although the former were incidental and irrelevant. Correct responsesof the two sorts did not necessarily coincide. Confusion errors during learning were influencedby sharing a color, further evidencethat the colors had been noticed and remembered. We now asked whether older subjectswould do the samething. We repeated the above experiment as nearly as possible with nine-year-old children, but used unfamiliar artificial graphemesinsteadof familiar letters. Arbitrary namesof seven or eight of the colored graphemeswere learned in about six trials, and the namestransferredto the black copies. But the number of colors correctly rememberedwas at chance level. Some nine

Trendsin Perceptual Development 377 year olds did not even rememberwhat colors had been present. In short, there was no incidentalleaming. (Other examplesfollowed.) It can be concludedthat changesin ability to focus on wanted information and shut out the irrelevant do occur with age. As the child matures, he is better able to selectout and usethoseproperties that serveto distinguish things from one another and are adapted to his task, and to disregard nonessentialproperties. This improved filtering ability is itself evidenceof increasingeconomy in information pickup, my third trend. IncreasingEconomyof InformationPickup, and the Searchfor Invariance The third trend in perceptualdevelopmentthat I identified at the outset of the chapter is progressive economy in the extraction of information from stimulation. I think of extraction asusually being a searchfor and discovery of invariants in the stimulus flux. Ways in which economy is achievedare by the detection of distinctive features of things, by the extraction of invariants over time, and by the processingof larger units of structure. DistinctiveFeatures It is highly economicalto discriminatethe objects of the world by means of the minimal set of features that will serve to distinguish them. As my husband put it, "Those features of a thing are noticed which distinguish it from other things that it is not- but not all the features that distinguish it from everythingthat it is not" (Gibson 1966, 286). Many illustrations, both anecdotaland experimental, were given in the preceding chaptersto show that children do learn such sets, and learn to assign priorities to critical featuresin the course of development. Caricature in art is a caseof enhancingthat which servesto distinguish and omitting that which doesnot. There is a minimal featureset that makesthe caricatured object unique, and the good cartoonist has left out of his drawing that which is redundantor nondistinguishing. Children appreciate caricaturesastonishinglyearly, asall publishersof comic books know. They can learn to perceive objects on the basisof sparsedetails, as long as they are essentialdetails. In chapter 7 I describedan experiment by Albert Yonas and myself in which both children and adults learnedto abstracta single feature in order to discriminate two subsetsof stimulus figures. With practice there was increasing use of an optimal strategy. While the reaction times of the children were very much longer than the adults', the same trend was found. Discrimination time was faster with only one letter in the positive set. But when a single feature could be extracted from a positive set of three letters and used for differentiation, the curve fell faster with practice than in a control condition. An interesting exampleof economicalextraction of a single differentiating featureoccursin a developmentalstudy by Vurpillot et al. (1966). The

378

E. J. Gibson

experiment isostensibly a studyofconcept identification, sincetheaimwas to comparetwo theoriesof howconceptsare identified whencommon dimensions areperceptually presentin the instancesgiven.Whenpositive

instances presented to a childcontainseveralcommon features, doesthe childlearna conceptby abstractingall thesefeaturesand representing themschematically, or doeshe simplylearnto differentiate the positive

fromthe negativeinstances on the basisof the fewestcriteriathatwill suffice?

Vurpillot designed a transfer experiment to testthisidea,usingchildren agedsixanda halfto ten,andadults.Thepositive instances of the concept allcontained fourspecified values offeatures ofa cartoon drawing of a bird(a roundtail,linedeye,two chevronson the wing,and spotson the chest),whilethenegativeinstances containeddifferentvaluesof these features(e.g.,squaretail,threechevronson the wing,etc.).Theywere

assigned, however, so that a singlefeaturedifference wouldsufficeto separate thetwosets.Aftersortingtheexamples to learnwhichoneswere positive instances andhadthenamedesignated fortheconcept, thesubjectsweregivena numberof newinstancesin a transferseriesto see what was generalized as a positive instance.

The new instancespresentedin the transfertaskwerechosenso that

thesubjectcouldgeneralize to onlythoseinstances carrying allfourofthe

commonfeatures,or to all the instancescarryinga single feature that sufficedto differentiatethe sets in Task I. In the transfertask, no subject

choseonlythecardswhichcontained allfourof thecommon features of theconcept. Theychose,instead, allthecardswhichhadjustonefeature that sufficed,the one they had selectedas criterialfor sorting.

The interesting thinghereis that fourfeatures,all commonto the

positive instances, wereuniformly reinforced, butwhatwaslearned and generalized wasonlya singledistinctive featurethat wassufficient for differentiation.This demonstratesremarkableeconomy in learning,and it

isespecially interesting to notethatthechildren behaved inessentially the same way as the adults.

liwariants The search for an invariantthe

relation that remains con-

stant over changeis the essenceof objectperception.The stimulus invariantthatkeepsits identitydespitethe transformations of stimulation

causedbya motionoftheobjectora movement oftheobserver isthebasis forperception of thatobject.Letmegivean examplefroman experiment

byGibson andGibson (1957).(Theexample givenwastheexperiment de-

scribedinpaper15onperceiving anobjectviaperspective transformations) Doesthe childprogressfromexperiencing a succession of ever-changing retinalimagesat birthtowardtheperceiving ofa movingobjectofthissort as he growsolder?Suchhas been the assumptionof the empiricistsin

Trendsin Perceptual Development 379

philosophy andpsychology forcenturies, but thereis no evidence forit.

Whatkindofprogress mightthechildmake intheperceiving ofobjects? I

believeit is progressin the searchforinvariants undertransformations. If

thisistrue,some oftheinvariant properties inthestimulus fluxarepicked

upbytheinfant, andothermorecomplex invariants areonlydetected later, butatnotimedoesthechild, youngorold,experience thestimulus fluxas

aflowofchanging sensations. Observation ofthestimulus flux, perception

over time,is necessaryfor progressin detectinginvariantsbecauseit

permits theinvariant s tobeextracted, notbecause it permits associating of

separate sensations. Thechildseemsto havebuilt-in propensities forteaching himselfaboutthe permanentaspectsof the world.One deviceis

observing hisownhands, bringing themtogether, rotating them,moving themtoward orawayfromhimself, andlatergrasping objects andmanipulatingthem.Theinvariants overtheresulting perspective transformations givehimtheinformation forshape andsizeconstancy. Another propensity forself-teaching isdropping things, orthrowing themfromacriborahigh chair, orspilling milk. These arewonderful waystolearnabout properties

likerigidity, elasticity, andfluidity, andthechildtriesmostofthem.(Other examples followed.)

Higher Order Structure Doesthespanof perception stretchwithage? Doesthe abilityto processa lot of thingssimultaneously increase? The

lengthofa perceptual spandoesseemto distinguish theperception of adultsfromthat of children, but I believeit is the abilityto findthe

structure, theembedded relations constituting subordinate unitsor clusters,

thatmakes thedifference. Itisnotjusta stretching outbuta making ofone

unit out of many smaller units.

Thevisual grouping principles embodied inWertheimers lawsoffigural

perception, the lawof proximity, of similarity, of goodcontinuation, and

of common fate,arerelations thatstructure unitsforperception. There

is reasonto believethat the law of similaritydoesnot functionas wellto structurelargerunitsearlyin development as it doeslater,or in backward

children compared to brightones.Honkavaara (1958) gavea sorting task to backward andbrightchildren fiveto eightyearsold.Inonetaskthey

wereshowntwenty-four cardsof fourdifferent shapesscattered overa tableandwereaskedto sortthemintopiles.Thenormalolderchildren instantly sortedthemintofourpiles,buttheyounger andretarded children

put themin rowswithoutmatching,or left themin disorder.In another

test,theyweregivencardsonwhichweredrawndotpatterns embedded amongdifferent shapes. Whenaskedto finda patternmatching a model, theolderchildren surpassed theyounger andthebrightsurpassed the

backward children of thesameage.Rush(1937) madedevelopmental

comparisons oftheeffectiveness ofseveral Gestalt principles onperceptual

380

E. J. Gibson

grouping within drawn patternsand found that the effectivenessof similarity increasedwith age. (Other examplesof detecting order based on pattern regularitiesfollowed.) The examplesjust quoted were chosento illustrate the pickup of structure considered as a perceptual economy that develops with age and experience. The abstracting of common featuresor dimensionsover time also increasesdevelopmentally. The literature of discrimination learning provides examplesof the latter. One of these is the oddity problem. The requirement is as follows. Over a seriesof single problems with varying stimulus objects, or varying properties of similar objects, the subject must learn to choosethe odd object from a set of three, whatever the nature of the variation. Then he must perform without error on a new problem. Insteadof treating eachproblem as a new and independentone, in which a specificS-R connection must be learned, the subject must operate with a rule which coversall the problems, despitechangein absolutepropertiesof stimuli. A rule does not meana verbalizedrule, for a monkey can solve the general oddity problem (Meyer and Harlow 1949, Moon and Harlow 1955). Children may verbalizethe rule, but they neednot, as House (1964) has shown. In any case, learning of a three-position oddity problem by human children improves from kindergarten to third grade (Lipsitt and Serunian1963). Perceptualenhancementof the odd item by increasingthe number of identical items facilitates solution for kindergarten children, but does not help preschool children, who still fail to reach criterion (Gollin, Saravo, and Salten 1967). These are less complex than the order found in spelling ----- - relationshios -Jpatterns. In the last chapter . , I discussed ~ perceptionof higher order structure in letter strings, and presentedevidencethat patternsand clustersof letters within words are perceived as units by adults, and that there is develop~ mental progress in detecting the kind of structure found in orthography and in spelling-to-sound correspondences . Rulescome to be used as learning progresses , despitethe fact that the usercould not explain the rules and that the order in the constraints has only recently been penetrated by trained linguists. Psychology has not gone far enough in investigating the growth of ability to detect regularity, order, and structure. This ability is basic for cognition. The detection of similarity, equality, symmetry, transitivity , and congruenceis essentialfor learning mathematics, and the good teacher does his best to make them perceptibleby clearing away the superfluous detailsand baring the skeleton. Here is a last example, taken from Wertheimer's book on ProductiveThinking (1945). He taught a child of 5-1/ 2 how to find the area of a rectangle by drawing it for him, filling it with small squares , and counting them in various ways. Then he presented him with a parallelogramand asked if he could find its area. After some

Trends in PerceptualDevelopment

381

momentsof puzzledstaring at the new figure, the child asked, "Do we have a pair of scissors ?" They were producedand he proceededto snip a triangle off one end of the parallelogram and fit it neatly on the other, thus converting it into a rectangle. Then he said, "Now I can do it.I' He had literally perceived the solution. (It is notable that researchin recent years has disclosed that Wertheimer 's laws of organization , except for common

fate, which involves motion, are not functional in infancy before about 7 months [Schmidt and Spelke 1984]. Order in static displays is not easily detected in early years, at least in artificial cases such as dot patterns . It may be that natural cases, where some affordance

for action exists , would

be more easily grasped.) References Ames, E. W . & Silfen, C. K. (1965). Methodological issuesin the study of age differencesin infants' attention to stimuli varying in movement and complexity. Paper presentedat the meeting of the Society for Researchin Child Development, Minneapolis, Minn ., 1965

.

Bruner , J. S. (1957 ). On perceptual readiness . Psychol. Rev., 64 , 123 - 152 . Bruner , J. 5., Olver , R. R., and Greenfield , P. M ., et al. (1966 ). Studies in cognitive growth . New

York: Wiley .

Gollin, E. 5., Saravo, A ., and Salten, C. (1967). Perceptual distinctiveness and oddity problem solving in children. Journalof ExperimentalChild Psychology , 5, 586- 596. Honkavaara, S. (1958). Organization processesin perception as a measureof intelligence. Journal of Psycholo~ , 46, 3 - 12.

House, B. ] . ( 1964). Oddity performancein retardates. I. Size discrimination functions from oddity and verbal methods. Child development , 35, 645- 651. Lambercier, M . (1946). Recherchesur Ie developpement des perceptions: VI . La constance de grandeurs en comparaisonsseriales. Arch. Psychol . Geneve , 31, 79- 282. Lippsitt, L. P. and Serunian, S. A . (1963). Oddity -problem learning in young children. Child Development , 34, 201- 206. Maccoby, E. E. (1967). Selective auditory attention in children. In L. P. Lipsitt and C. C. Spiker (Eds.) Recentadvancesin child developmentand behavior. Vol . 3. New York: Academic

Press , pp . 99 - 124 .

Meyer, D. R. and Harlow, H. F. (1949). The development of transfer of response to patterning by monkeys. Journal of comparativeand physiologicalpsychology , 42, 454462 .

Moon , L. E., and Harlow, H. F. (1955). Analysis of oddity learning by rhesus monkeys. Journalof comparativeand physiologicalpsychology , 48, 188- 195.

Riess , B. F. (1946). Geneticchanges in semanticconditioning . Journalof Experimental Psychol ogy , 36 , 143 - 152 .

Rush, G. P. (1937). Visual grouping in relation to age. Archives of Psychology , N .Y., 31, Whole

No . 217 .

Vurpillot , E., Lacoursiere, A ., de 5chonen, 5., and Werck, C. (1966). Apprentissage de conceptset differenciacion. Bulletinde Psychologie , 252, XX, 1- 7. Wertheimer, M . (1945). Productivethinking. New York: Harper. Zinchenko, V. P., van Chzhi-Tsin, and Tarakanov, V . V. (1963). The formation and development of perceptual activity . Sovietpsychologyand psychiatry, 2, 3- 12.

Retrospectand Prospect: The Coming of Age of PerceptualDevelopment

Ten yearsafter the appearanceof Principlesof Perceptual Learningand Development , it was named as a licitation

classic " (Current Contents , 1979 ), and I

was askedto write a few paragraphsabout why a number of people should be disposedto cite the book. I quote the paragraphsbelow. Their rather optimistic concluding suggestion about application doesn't seem too opti mistic

to

me

now

.

Learningby perceiving the world around us, its permanentproperties, its furnishings and ongoing events has always been of interest to philosophers, and deservedlyso. Where elsehas an adult acquiredthe information about his environment that permits him to act adaptively in it and upon it? Yet modem experimentalpsychologists generally ignored the problem, although their interests for many years were dominated by learning. Motor learning, verbal learning, affective learning, and simple contingency learning were studied intensively, but comprehensivebooks on learning never mentioned perceptual learning. Developmentalpsychology, a younger branchof the science than experimentalpsychology, was almost equally negligent, but for better

reasons -

no one had devised

feasible , reliable

methods

for

studying early perceptualdevelopment. It is gratifying, therefore, to seethat the problemsdiscussedin this book and the attempt to provide a framework for understandingthem have had an impact . The book alone, however , was not responsible

for the progress that has taken place since its publication in our knowledge of perceptualdevelopment. There is always an elementof luck in the successof a book or a theory. There has to be an audience ready to listen and experimental progress dependson concomitant advancesin technology. Fortunately, the theory and these factors appearedtogether at the right time. Psychologistswere dissatisfied with S-R learning theory and were ready to pay attention to a theory

384

E. J. Gibson of perceptuallearning. At the sametime, new methods of studying perceptionin infants were being worked out and a whole new field of researchopenedup. A third factor explains why this book is widely cited. It has an important field of application. While I was writing the book, I was conductinQ - - ---- ~ : researchon processesinvolved in learning to read. ReadU . ingJ was making a comeback as an area for scientific-study. Granting agencieswere generous with funds, and my own work made the connection between a theory of perceptuallearning and learning to read. It is interesting to consider progress in the book's field since its publication. The theory of perceptual learning in adults has progressedscarcelyat all. Work on reading has burgeoned. There has been a surge of researchon perception in infants and very young children, accountingfor many of the citations. I have recently directed my own researchto this area, and I find that the theory generatingmy experimentsis reflected more and more in work of others on similar problems. As we discover more about early development, it will be possible to refine the theory of perceptualdevelopment and to provide guidelinesfor applied work.

Now, twenty years later, it is striking to see how the trends noted in the last paragraphwere borne out. There is still little or nothing new on the theory of perceptualleaming. Researchon reading is firmly ensconced in the information processing camp, prolific but going nowhere, in my opinion. But researchon perceptualdevelopmentin infants hasproliferated vigorously and is leading us to new insights. How did it happen? When I wrote the 1969 book, there was little researchon perception in preverbal children, mainly becauseof inadequatemethodology. But since that time we have had a kind of technologicalexplosion. If you can't ask a preverbal child to tell you what she seesin a neat psychophysicalparadigm, how do you find out what information sheis picking up? The answer was right there, waiting for us. Babiesare motivated to obtain information; thev from the start. They are endowed with percep., are active oerceivers ~ tual systemsthat include exploratory activity - looking, mouthing, listening. We have learned to use these exploratory activities as indicators of what is perceived. Even newborns can look at and listen to events and explore substances put in their mouths. Looking preferentially at one of two displays can tell u.<;whether or not the baby - differentiatesthem, and on what basis, if the displays are designed and ordered strategically. A baby will look preferentially at new information, so habituation and familiarization have becomethe basisof an elaboratetechnology. And a baby will explore the

v

Years ofSignificance : Research onReading (1965 - 1977 )

Introduction to Part V

During the year at the Princeton Institute, when I was thinking about perceptual learning, I was approachedby two Cornell colleagues , Harry Levin and Alfred Baldwin, and urged to join them in an about-to-be-formed interdisciplinary consortium devoted to basic researchon the reading process. This incident occurred in the early sixties, when a kind of age of relevanc ~ was in vogue , so federal agencies offered

encouragement

and

lavish support for the project . It was to embrace the talents of experimental

and developmentalpsychologists and linguists as well. I was hesitant to start a brand new enterprise, since I was determined to devote my best

energiesto perceptual learning, but I was eventually persuadedthat the two projects could be carriedon simultaneously. So, I becamea memberof a group dubbed I'Project Literacy" and began to think about how children learn to read. The project lasted for over a decade, culminating in a book, the Psychology of Reading , which Harry Levin and I coauthored. My thinking about reading (as well asperceptuallearning) went through many changesduring that time. I wrote a number of general papers, the first appearing in 1965 in Scienceand the last in 1977 in a symposium on

reading and cognition. The scene, as well as my ideas, had changedgreatly in the courseof those twelve years. When Project Literacy began, reading was a defunct topic in experimentalpsychology. The last time it was so much as mentioned in that context was in Woodworth 's classic Experimental Psychology, published in 1938. The book had later editions , none of

them of equal stature, and reading did not appearas a topic again. Like the

otherchaptersin that book, the readingchapterwasgood, beginningwith a short history of writing and getting into the still-persisting questionsof the relationship between speechand reading and the role (or nonrole) of eye movements. Good experimental researchperformed before 1900 by such time-honored pioneer psychologists as Cattell and Dodge was reported. Alas, the topic had fallen into disrepute as psychology becamea

390

PartV

"respectable " science . But the effort to bring it backinto the fold worked, because by 1975 the climate was right and there was plenty of research.

The topic had become a popular one with information processers , as it still is .

A splendidbook about readingwas publishedby EdmundBurkeHuey in 1908, the Psychologyand Pedagogyof Reading . Except for Woodworth 's classicchapter, this book, with an astute analysis of the reading process, was a kind of swan song. In psychology, stimulus-responsetheory appeared, and in education, where the topic might have been kept alive, curriculum research became the fashion . Classroom questions (shall we

teachphonicsor whole word?) preoccupiededucators , and theory-based research, on the kind of learning process engaged in as reading skill is acquired, disappeared . The cognitive revolution has brought it back, but the information processingapproachneglects learning and also generally neglectsthe fact that skilled reading varies with the kind of text and what sort of information the readeris seeking. Readingis an active process, like perception , and it has many affordances.

On the other hand, all readersshareone very generalpurpose- getting information from printed text- and cross-language comparisonsdo not show as drastic differencesas one might at first expect. One of the most intriguing questions having to do with learning to read concerns the influenceof the writing system. Does an alphabeticsystemhave an advantage over an ideographic system? Does a languagethat has a strict one-toone phonetic correlation with its alphabethave an advantage? Answers to these questionshave been eagerly pursuedin recent years, but they seem to have yielded little knowledge about the way children learn. My last exercise relating to reading occurred during a sojourn of six weeks in China, in 1982, when I was invited to give a seminar for a group of Chinese

psychologistswho were catching up after ten years out of businessduring the cultural

revolution

. Some of the members

of the group

wanted

to

perform an experiment, one that could be completed quickly, during my stay there. I thought we might try a kind of Stroop experiment, with pictures and Chinesecharacters , similar to ones performed with children in this country using English spelling (Rosinski, Golinkoff, and Kukish 1975). The pictures were drawn, appropriately labeled with Chinesecharacters, and three groups of subjectswere run through the experiment in five weeks. The result -

that the Stroop effect is present even in very early

readers(secondgrade, in fact) and does not increasein more skilled readers- was the sameas the result found with readersof English (Rosinski 1977). Details of processing may vary with the task, but the cognitive

process of reading appears to be similar from one language or writing system to another, even in rather early stages. Writing (and spelling)

Years of Significance: Researchon Reading (1965- 1977)

391

systemshave different constraints, but they all have them. And whatever the rules, the search for meaning is universal .

Experimentsexaminingthe developmentof a Stroop effect were of particular interest for questionsaskedby information-processingtheories

of reading. One such theory assertedthat beginningreadersattended primarily to the decoding aspect of converting the graphic characters

to soundand only secondarily , after determiningthe phoneticmapping , searcheda phonetic representationfor meaning. As skill in decoding increased , the decoding task should becomeautomatic and meaning should be accessed quickly and thus become a strong competitor in any kind of

interferencetest, suchasa Stroop task. In that case, interferenceon a Stroop task should increasewith reading skill. But many experimentalresultshave shown that pickup of meaning from words is very direct, even in beginning readers (Gibson and Levin 1975, 279 ff .). Information -processing

approachessometimesinterpret this result in terms of sharedor nonshared " semantic stores" or semantic representations , which compete for attention at a response level . Such arguments about processing seem to me trivial . What is really impressive is the strong indication that reading affords pick -

up of information aboutsomething, as does direct perception of eventsand objects in the world, and that learning to read is, and should be, a search for meaning from the start.

Discrimination of written charactersmay present more problems in very early stageswith somesystems(Hebrew and Arabic, for example), but the process of finding meaning and discovering rules that reduce the information seems to be universal . The papers I have chosen to reprint

show a pilgrim's progressfrom an early analysisof the process, through a phase of thinking close to information processing, to a developmental emphasis , and finally to a functional, ecological approach. The book on reading, especiallythe section on learning to read, has been translatedinto other languagesand still gets some use, so the effort seemsto have been worthwhile

.

The reading project gave me the pleasure of a number of top-notch research assistants, such as Anne Pick, Harry Osser, Albert Yonas, Carol

Bishop, and Richard Rosinski and several good graduate students who performed theses on related topics . Rather than reprint research papers that we produced , I have chosen primarily more general papers that at-

tempt to analyzethe readingprocess, summarizethe research , and illustrate the progress in thinking along the way. These essays were originally preparedfor oral presentation to some group, one way of spreading the word that reading was a good topic for seriousresearch.

21

Learningto Read EleanorJ. Gibson

This was my first general paper on reading, preparedfor a conferencesupported by the Social ScienceResearch Council (SSRC) and held at the Center for Advanced Study in the Behavioral Sciencesin Palo Alto . I was a fellow therefor the year 1964- 1965, member of a "cutting-edge" group that also included Lee Cronbach, Richard Atkinson, and Jack Wohlwill . We were supposed to be

plottingabouteducational research , dearto theheartof Ralph Tyler, thendirector of the center. This conferencewas one of our projects.

We wereaskedby SSRCto considerespeciallysomethingcalledthe Initial TeachingAlphabet(ITA), which was beingpushedby JohnDowning, a British educatorwho wantedto selltheideato plannersof readingcurricula.It wasa kind of alphabetsemireform , designed for teachingbeginningreadingas a one-to-one letter-sound code, later to begiven up for conventionalspelling. It seemedto usjust

anotherfad for restylinga curriculumand psychologically quite unsound . Why train studentsin somethingsoonto bedropped ?And, moreimportant, is reading reallya matterof deciphering a code ? Or are there,perhaps , rulesfor readingtext that transcend a decodingapproachand, onceinternalized , providea moreecono mical way of extractinginformationfrom text? At that time, I thought of economical readingasdecoding"higherorderunits," as this essayproposes . Over the years, I changedthis view to one moreakin to the ecologicalapproachto perception. That view emerged only gradually : the notion that text contains information in the form of invariants (both syntactic and semantic) that we learn

(via perceptuallearning ) to extractfor thesakeof economical pickupof what the

text affords.

This first essaywas a rather long way from that idea. I tried to analyze the

readingprocess(not unlike what an informationprocessormight do), and I came up with several stages. The first stage confronting children, I thought, was

learningto discriminatethegraphicsymbolsusedby their nativewriting system . @ AAAS. Science , 1965, 148, 1066- 1072.

394

E. J. Gibson

The reader can imagine how easily that idea was come by, in a direct line from

the theoreticalnotionsproposedin my Ph.D. thesis . Thefirst research I planned (with the helpof Anne Pickand Harry asser) was an experimenton differentia tion of what we called "letterlike forms." The experiment, comparing children four

througheightyearsof age, is summarizedin thepaperthat follows. We thought that thechildrenlearnedoverthis .periodto differentiatethecharacters by learning the dimensionsby which they differed , discoveringa set of distinctivefeatures , rather than learning a kind of prototype for each one. Anne Pick's thesis (1965)

later investigatedthis question , showing that dimensionsof differencecan be discoveredin perceptuallearningand will be generalizedto a new group of itemsif theyaredistinguished by the samefeatureset. Theseexperiments had two results . One was theoretical , my (perhapspremature) conclusionthat the principal thing learnedin perceptuallearningwas distinctive featuresof objectsand later of representationsof them and of items such

as graphic symbols . The other was a searchfor the distinctivefeaturesthat characterize the lettersin our written alphabet . Therewereseveralattemptsto do this. Themethodwas to obtaina confusionmatrix and thenanalyzeit searching for the features underlying confusionsof one letter with another. The first attempt at this, basedon errors made when matching letters by four -year-old children, is

summarizedin this paper. A more sophisticatedone is presentedin a later paper .

Thesecondstagein learningto readproposedin this paperwasdecoding from graphicpresentations to sound. This presumedprocesswas typical of the way many peoplethoughtaboutreadingat the time. Note Bloomfield 's (1942) statement, "In orderto readalphabeticwriting onemust havean ingrainedhabit of producingthe soundsof one's languagewhen oneseesthe written markswhich conventionallyrepresent the phonemes (p. 128)." This statementsoundsas if the processof learning to read is like a simple S-R learningparadigm. Indeed , Bloomfieldwas a hard-corebehaviorist . We did not think of the matteras being that simpleand werelookingfor rulesthat would permit transferand waysfor them to be taught. Carol Bishop's master's thesis(1964) was an early effort to investigatesimplerule transferand is summarizedhere. The tricky part was, of course , that Englishdoesnot haverulesof oneletterto onesoundcorrespondence . In an effortto supersede this apparentdifficulty, we divertedmuchof our effort to a searchfor mappingrules that would involve larger, "higher-order" units. Thesolutionwe cameup with at the time wasa functionalunit that we referred to as a "spellingpattern"- a setof lettersconstrained as to positionin a word that had a consistent mappingto sound. We performeda numberof experiments with what we dubbedpseudowords to show that theseunits werebeingused.A pseudoword like GLURCK wascompared with its reversalCKURGL, whichdid notfollow themappingrules, by variousmethods , mostoftenusingtachistoscopic presentations . Thepronounceable nonsense wordswerealwayseasierto read. This

Learning to Read

395

method hasbeen used ininnumerable experiments since thattime byinformation-

processing psychologists. I stillfavorthenotion thatchildren mustinsome way internalize theconstraints provided in thespelling system in orderto become

efficient readers, butwedidnotatthattime appreciate sufficiently thatthespelling

ofEnglish ismorphophonemic, notjustletter-sound mapping ofgreater orless complexity. Themainchange thattookplace eventually, however, wasthediscarding ofthenotion thatreading isessentially a decoding process .

Experimental psychologists examine theprocess by whicha fundamental

intellectual skillis acquired.

Educators andthepublic haveexhibited akeeninterest intheteaching of ofordespite theirinterest, thismostimportant subject hasbeenremarkably susceptible totheinfluence offadsandfashions andcuriously unaffected by disciplined experimental andtheoretical psychology. Thepsychologists havetraditionally pursued thestudyofverbal learning bymeans ofexperireadingeversincefreepubliceducation became a fact(I).Eitherbecause

mentswithnonsense syllables andthelikethat is,materials carefully

divested ofusefulinformation. Andtheeducators, whofoundlittleinthis workthatseemed relevant to theclassroom, havestayedwiththeclass-

room; when theyperformed experiments, themethod wasapttobeagross

comparison ofclasses privileged andunprivileged withrespect tothelatest

fad.Theresult hasbeentwocultures: thepurescientists inthelaboratory,

andthepractical teachers ignorant oftheprogress thathasbeenmadein thetheoryofhuman learning andinmethods ofstudying it. Thatthissplitwasunfortunate is clearenough. True,mostchildren do learnto read.Butsomelearnto readbadly, sothatschool systems must provide remedial clinics; anda small proportion (butstilla largenumber of futurecitizens) remainfunctional illiterates. Thefashionswhichhaveled

toclassroom experiments, suchasthewholewordmethod, emphasis on context andpictures formeaning,theflash method, speedreading,

revised alphabets, thereturn to phonics,andsoon,havedonelittleto

change the situation.

Yeta systematic approach to theunderstanding ofreading skillispos-

sible. Thepsychologist hasonlytotreatreading asa learning problem, to

applyingenuity in theoryconstruction andexperimental designto this fundamental activity onwhich therestofmanseducation depends. A beginning hasrecently beenmade inthisdirection, anditcanbeexpected thata number oftheoretical andexperimental studies ofreading willbe

forthcoming(2).

396

E. J. Gibson

Analysis of the Reading Process

A prerequisiteto good researchon readingis a psychologicalanalysisof the reading process. What is it that a skilled reader has learned? Knowing

this (or having a pretty good idea of it), one may considerhow the skill is learned, and next how it could best be taught. Hypotheses designed to answer all three of these questions can then be tested by experiment .

There are severalways of characterizingthe behavior we call reading. It is receiving communication; it is making discriminative responsesto graphic symbols; it is decoding graphic symbols to speech; and it is getting meaning from the printed page. A child in the early stages of acquiring reading skill may not be doing all these things, however. Someaspectsof reading must be mastered before others and have an essential function in a sequence of development of the final skill . The average child , when he begins learning to read, has already mastered to a marvelous extent the art

of communication. He can speak and understand his own language in a fairly complex way, employing units of languageorganizedin a hierarchy and with a grammaticalstructure. Sincea writing system must correspond to the spokenone, and sincespeechis prior to writing , the frame-work and unit structure of speechwill determine more or less the structure of the writing system, though the rules of correspondencevary for different languagesand writing systems. Somealphabeticwriting systemshave nearly perfect singie-letter-to-soundcorrespondences , but some, like English, have far more complex correspondencebetween spelling patterns and speech patterns. Whatever the nature of the correspondences , it is vital to a proper analysis of the reading task that they be understood. And it is vital to remember, as well , that the first stage in the child 's mastery of reading is

learning to communicateby meansof spokenlanguage. Once a child begins his progression from spoken language to written language, there are, I think, three phasesof learning to be considered. They present three different kinds of learning tasks, and they are roughly sequential, though there must be considerableoverlapping. These three phasesare: learning to differentiate graphic symbols; learning to decode letters to sounds ("map" the letters into sounds); and using progressively higher-order units of structure. I shall consider these three stagesin order and in some detail and describeexperimentsexploring eachstage. Differentiation of Written Symbols

Making any discriminative responseto printed charactersis consideredby some a kind of reading. A very young child, or even a monkey, can be taught to point to a patch of yellow color, rather than a patch of blue, when the printed charactersYELLOW are presented. Various people, in recent

Learning to Read

397

popular publications, haveseriously suggested teaching infants torespond discriminatively inthiswaytoletter patterns, implying thatthisisteaching themtoread. Such responses arenotreading, however; reading entails decoding to speech. Letters are, essentially, an instruction to produce a given speech sound. Nevertheless, differentiation of writtencharacters fromoneanotheris a

logically preliminary stage todecoding them tospeech. Thelearning problemisoneofdiscriminating andrecognizing a setoflinefigures, allvery similar inanumber ofways(forexample, allaretracings onpaper) andeach differing fromalltheothersinoneor morefeatures (asstraight versus

curved). Thedifferentiating featuresmustremaininvariant undercertain transformations (size,brightness, andperspective transformations andless

easily described onesproduced bydifferent typefaces andhandwriting).

Theymusttherefore be relational, so thatthesetransformations willnot

destroy them.

Itmight bequestioned whether learning isnecessary forthesefigures to bediscriminated fromoneanother. Thisquestion hasbeeninvestigated by Gibson, Gibson, Pick,andOsser(3).Inorderto tracethedevelopment of

letterdifferentiation as it is relatedto thosefeaturesof letterswhichare criticalfor the task,we designedspecified transformations for eachof a

groupofstandard, artificial letter-like formscomparable to printed Roman capitals. Variants wereconstructed fromeachstandard figure to yieldthe following 12transformations foreachone:threedegrees oftransformation

fromlineto curve; fivetransformations ofrotation orreversal; twoperspective transformations; andtwotopological transformations (seeFig.

21.1forexamples). Alloftheseexcepttheperspective transformations we

considered critical fordiscriminating letters. Forexample, contrast v andU; c and U; o and c.

Thediscrimination taskrequired thesubject tomatch a standard figure

against allofitstransformations andsomecopies ofit andto selectonly

identical copies.Anerrorscore(thenumberof timesanitemthatwasnot anidentical copywasselected) wasobtained foreachchild,andtheerrors

wereclassified according to thetypeoftransformation. Thesubjects were children aged4 through 8 years.Aswouldbe expected, thevisualdiscrimination oftheseletter-like formsimproved fromage4 to age8, but theslopes oftheerrorcurves weredifferent, depending onthetransforma-

tionto be discriminated (Fig.21.2).In otherwords.sometransformations

arehardertodiscriminate thanothers, andimprovement occurs atdifferent ratesfordifferent transformations. Eventheyoungest subjects maderela-

tivelyfewerrorsinvolving changes ofbreakor close,andamong the 8-year-olds theseerrors dropped tozero.Errors forperspective transformationswereverynumerous among 4-year-olds andstillnumerous among

8-year-olds. Errors forrotations andreversals started highbutdropped to

E. J. Gibson

398

CLOSE

STANDARD LINE TO CURVE 1800ROTATION SLANT BACK

6 STANDARD LINE TO CURVE

BREAK

450 ROTATION SLANT RIGHT

-

-

Figure21.1 Examples of letterlikefiguresillustratingdifferenttypesof transformation .

80

PERSPECTIVE

TRANSFORMATIONS 70

60 (j )

a: : 0 a: :

50

a: :

W

LL 0 W

ROTATION

40

<. 9
AND

REVERSAL

, TRANSFORMATIONS \ , \ \ \

30

\

~

,

a::

~

LINE TO CURVE

TRANSFORMATIONS

L. U a. .

......

20

" " " ...... ..... ... .. .. .. ... ..

" -

......

- ~

10

......

BREA~AN-6-c ~ 0"SE ' - --:::.>-...- -

" - -

-

~ -

~

-

-

0 4

5

6 AGE

IN

7

8

YEARS

Figure 21.2 Errorcurves showing rateofimprovement in discriminating fourtypesoftransformation .

Learning to Read

399

nearly zeroby8years. Errors forchanges from linetocurve were relatively numerous (depending onthenumber ofchanges) among theyoungest children andshowed a rapiddropamong theolderalmosttozeroforthe

8-year-olds.

Theexperiment wasreplicated withthe sametransformations of real lettersonthe5-year-old group.Thecorrelation between confusions ofthe

same transformations forrealletters andfortheletter-like forms wasvery high(r +.87),sotheeffect ofa given transformation hasgenerality (is

not specificto a givenform).

What happens, intheyears from 4to8,toproduce orhamper improvethefeatures ordimensions ofdifference which arecritical fordifferentiating mentin discrimination? Ourresultssuggestthatthechildren havelearned

letters.Somedifferences arecritical, suchasbreakversusclose,lineversus

curve, androtations andreversals; butsome, suchastheperspective trans-

formations, arenot,andmustinfactbetolerated. Thechildof4 doesnot

startcold uponthistask,because someofhisprevious experience with

distinctive features ofobjects andpictures willtransfer to letterdifferentiation.Butthe set of lettershasa uniquefeaturepatternfor eachof its

members, solearning ofthedistinctive features goesonduring theperiod

we investigated.

Ifthisinterpretation iscorrect, it wouldbeuseful toknowjustwhatthe

distinctive featuresof lettersare.Whatdimensions of difference musta

childlearnto detect inorderto perceive eachletterasunique? Gibson,

Osser,Schiff, andSmith(4)investigated thisquestion. Ourmethod wasto drawupa chartofthefeatures ofa givensetofletters(5),testto seewhich

oftheseletters weremostfrequently confused byprereading children, and compare the errors in the resulting confusion matrix with those predicted by the feature chart. A set of distinctive featuresfor lettersmustbe relationalin the sense thateachfeaturepresents a contrast whichisinvariant undercertaintransformations, andit mustyielda unique patternforeachletter.Thesetmust

alsobereasonably economical. Twofeature listswhichsatisfy thesere-

quirements fora specified typefaceweretriedoutagainst theresults ofa

confusion matrix obtained withthesame type(simplified Roman capitals

available on a sign-typewriter).

Eachofthefeatures inthelistinFig.21.3is or isnota characteristic of

eachofthe26letters. Regarding eachletter oneasks, forexample, Is there

a curvedsegment?andgetsa yesor no answer. A filled-in featurechart

givesa unique pattern foreachletter.However, thenumber ofpotential features forletter-shapes isverylarge, andwould varyfrom onealphabet andtypefontto another. Whether ornotwehavetherightsetcanbe

testedwitha confusion matrix. Children should confuse withgreatest

401

Learning to Read Table

21 .1

Number of Errors Made in Transfer Stage by Groups with Three Types of Training Type of training -

-

..

J

C

-

-

-

-

-

-

-

-

- -

' 0

Group

Standards

Transformations

E1

Same

Different

69

E2

Different

Same

39

C

Different

Different

=

-

Errors

- - --

_ &&"' .. ..,

101 ~

,

,

~

That it is perceptual learning and need not be verbalized is probable (though teachersdo often call attention to contrastsbetween letter shapes.) An experiment by Anne D. Pick (6) was designed to compare two hypothesesabout how this type of discrimination develops. One might be called a "schema" or "prototype" hypothesis, and is basedon the supposition that the child builds up a kind of model or memory image of each

letter by repeatedexperienceof visual presentationsof the letter; perceptual theories which propose that discrimination occurs by matching sensory experience to a previously stored concept or categorical model are of

this kind. In the other hypothesis it is assumedthat the child learns by discovering how the forms differ, and then easily transfersthis knowledge to new letter-like figures. Pick employed a transfer design in which subjects were presented in step 1 with initially confusable stimuli (letter -like forms ) and trained to discriminate between them . For step 2 (the transfer stage) the subjects were divided into three groups . One experimental group was given sets of stimuli

to discriminate

which

varied in new dimensions

from the same

standards discriminated in stage 1. A second experimental group was given sets of stimuli

which

deviated

from new standards , but in the same dimen -

sions of difference discriminated in stage 1. A control group was given both new standards

and new dimensions

of difference

to discriminate

in

stage2. Better performanceby the first experimentalgroup would suggest that discriminationlearning proceededby construction of a model or memory image of the standards against which the variants could be matched.

Conversely, better performanceby the secondexperimentalgroup would suggest that dimensions

of difference

had been detected .

The subjects were kindergarten children. The stimuli were letter-like forms of the type described earlier. There were six standard forms and six transformations

of each of them . The transformations

consisted

of two

changes of line to curve, a right-left reversal, a 45-degree rotation, a perspective transformation , and a size transformation . Table 2 I .I gives the errors of discrimination for all three groups in stage 2. Both experimental

groups performed significantly better than the control group, but the group

402

E. J. Gibson

that had familiar

transformations

of new standards performed

significantly

better than the group given new transformations of old standards . We infer from these results that , while children probably do learn proto types of letter shapes , the prototypes for differentiation . The most relevant

themselves are not the original basis kind of training for discrimination is

practice which provides experience with the characteristic differences distinguish the set of items . Features which are actually distinctive letters could be emphasized by presenting letters in contrast pairs .

that for

Decoding Letters to Sounds When

the graphemes

are reasonably

discriminable

from one another , the

decoding process becomes possible . This process , common sense and many psychologists would tell us, is simply a matter of associating a graphic stimulus with the appropriate spoken response - that is to say , it is the traditional stimulus -response paradigm , a kind of paired -associate learning . Obvious as this description seems , problems arise when one takes a closer look . Here are just a few . The graphic code is related to the speech code by rules of correspondence

. If these rules are known , decoding

of new

items is predictable . Do we want to build up , one by one , automatically cued responses , or do we want to teach with transfer in mind ? If we want to teach for transfer , how do we do it ? Should the child be aware that this is a code game with rules ? Or will induction of the rules be automatic ? What units of both codes should we start with ? Should we start with single letters , in the hope that knowledge of single -letter -to -sound relationships will vield the most transfer ? Or should we start with whole words , in the hope that component relationships will be induced ? Carol Bishop ( 7) investigated the question of the significance

of knowl -

edge of component letter -sound relationships in reading new words . In her experiment , the child ' s process of learning to read was simulated by teach ing adult subjects to read some Arabic words . The purpose was to de termine the transfer value of training with individual letters as opposed to whole words , and to investigate the role of component ciations in transfer to learning new words .

letter -sound asso -

A three -stage transfer design was employed . The letters were 12 Arabic characters , each with a one -to -one letter -sound correspondence . There were eight consonants and four vowels , which were combined to form two sets of eight Arabic words . The 12 letters appeared at least once in both sets of words . A native speaker of the language recorded on tape the 12 letter -sounds and the two sets of words . The graphic form of each letter or word was printed on a card . The subjects were divided into three groups - the letter training group (L), the whole -word training group (W ), and a control group (C ). Stage I of

Learning to Read

403

the experiment was identical for all groups . The subjects learned to pro -

SO'dOM .L:)3 'd'dOJ :::fa 'd3811 \JnN

nounce the set of words (transfer set) which would appear in stage 3 by listening to the recording and repeatingthe words. Stage2 varied. Group L listened to and repeatedthe 12 letter-soundsand then learnedto associate the individual graphic shapeswith their correct sounds. Group W followed the same procedure, except that eight words were given them to learn, rather than letters. Learning time was equal for the two groups. Group C spent the sametime-interval on an unrelatedtask. Stage3 was the samefor the three groups. All subjectslearned to read the set of words they had heard in stage I , responding to the presentation of a word on a card by

pronouncing it . This was the transfer stage on which the three groups were compared .

At the close of stage 3, all subjectswere tested on their ability to give the correct letter-sound following the presentation of each printed letter. They were askedafterward to explain how they tried to learn the transfer

words .

Figure 21.4 shows that learning took place in fewest trials for the letter group and next fewest for the word group . That is, letter training had more

transfer value than word training, but word training did produce some transfer . The subjects of group L also knew , on the average, a greater

number of component letter-sound correspondences , but some subjectsin group W had learned all 12. Most of the subjects in group L reported that

they had tried to learn by using knowledge of component correspon-

8 7

6 5

GR0UPL/ '

/

/

0

V

0:

0

0/ 0/

C

C ~~~...--.-.

C/

/

~--_....---~~

/ ....."",............

4 1 I ./ C-~--;'~ 3 2 r / 0C/ [J/ 1 // 0

12

14

16

18

20

22

24

TRIALS

Figure 21.4 Learning curves on transfer task for group trained originally with whole words (W), group trained with single letters (L), and control group (C).

404

E. J. Gibson

dences.But so did 12 of the 20 subjectsin group W , and the scoresof these 12 subjectson the transfer task were similar to those of the letter-trained group. The subjectswho had learned by whole words and had not used individual correspondencesperformed no better on the task than the control subjects. It is possible, then, to learn to read words without learning the component letter-sound correspondences . But transfer to new words dependson use of them, whatever the method of original training. Word training was as good as letter training if the subject had analyzed for himself the component relationships. LearningVariableand ConstantComponentCorrespondences In Bishop's experiment, the component letter-sound relationships were regular and consistent. It has often been pointed out, especiallyby advocates of spelling reform and revised alphabets(8), that in English this is not the case. Bloomfield (9) suggestedthat the beginning reader should, therefore, be presentedwith material carefully programmed for teaching those orthographic-phonic regularities which exist in English, and should be introduced later and only gradually to the complexities of English spelling and to the fact that singie-letter-to-sound relationships are often variable. But actually, there has been no hard evidence to suggest that transfer, later, to reading spelling-patterns with more variable component correspondencewill be facilitated by beginning with only constant ones. Although variable ones may be harder to learn in the beginning, the original difficulty may be compensatedfor by facilitating later learning. A series of experiments directed by Harry Levin (10) dealt with the effect of learning variable as opposed to constant letter-sound relationships, on transfer to learning new letter-sound relationships. In one experiment, the learning materialwas short lists of paired-associates , with a word written in artificial charactersas stimulus and a triphoneme familiar English word as response. Subjects(third-grade children) in one group were given a list which contained constant graph-to-sound relationships (one-to-one component correspondence ) followed by a list in which this correspondencewas variablewith respectto the medial vowel sound. Another group started with a similarly constructed variable list and followed it with a secondone. The group that learnedlists with a variable componentin both stageswas superior to the other group in the second stage. The results suggestthat initiating the task with a variable list createdan expectationor learning set for variability of correspondencewhich was transferredto the secondlist and facilitated learning it. In a second experiment, the constant or variable graph-sound relation occurredon the first letter. Again, the group with original variable training

Learning

performed

better

native

on

speakers

were

of

paired

with

variable

first

one

.

letter

A

"

set

many

true

letter

-

For

and

ent

.

kinds

units

granted

,

system

,

to

What

Exactly

how

units

units

the

across

the

picture

is

of

into

the

reader

of

,

.

print

is

correct

nor

is

perception

It

it

is

what

occurs

saccadic

jumps

exposure

,

a

from

skilled

learning

of

.

,

But

because

.

other

While

levels

of

can

"

of

a

does

.

course

( 11

to

perceive

"

the

to

readers

they

and

analysis

see

of

what

the

right

,

the

the

kinds

of

.

single

grapheme

,

by

letter

not

,

With

unconnected

this

of

con

the

letters

sequentially

not

a

across

goal

60

and

,

that

the

nearly

,

,

letter

map

names

letters

.

are

?

analysis

graphemes

is

showed

"

unequivocally

is

there

experimentally

single

name

are

correspondences

predict

is

that

four

link

logical

logical

and

next

writing

which

chunks

a

skilled

to

fixations

the

-

be

the

units

by

left

that

)

of

rules

what

the

the

to

but

during

fixation

can

,

) ,

Dodge

only

In

able

once

,

the

con

criterion

units

of

reading

-

word

reading

one

at

-

of

-

a

predictions

for

morphemes

,

out

If

our

language

The

differ

definition

such

and

from

assert

possible

?

be

unit

.

of

.

check

the

If

levels

behavior

should

to

.

units

of

recodability

questions

reading

English as

What

systems the

we

is

of

units

mapping

units

sequentially

we

reader

?

question

.

these

at

,

by

structure

speech

that

he

in

at

with

the

unit

units

answer

look

though

a

,

linguistic

to

the ,

code

units

and

spell

,

of

during

only

( "

regular

solved

,

with

words

and can

the

speech

alphabet

line

Spanish

letters

variable

of

levels

to

try

we

But

For

be

single

structure

another

graphic

proceeds

.

the

to

to

indeed

linguistic

all

suggests

.

a

variable

highly

necessarily

criterion

other

writing

reader

page

sistently

at

processed

false

for

a

relationships

remain

not

accepted

mapped

sense

that

,

processed

actually

,

transfer

concerned

One

in

true

rules

are

from

to

variable

hierarchical

parallel

,

first

the

are

Common

a

must

Then

correspondence

a

to

defined

they

.

transfer

constant

contains

learned

been

.

the

of

graphs

.

translation

must

them

whole

facilitate

are

generally

code

are

We

properties

between

is

units

are

is

are

discovered

has

,

speech

?

be

reached

be

how

perceived

the

contain

English

have

levels

then

a

adult

artificial

Units

or

the

The

more

from

405

experiment

. was

on

,

code

in

linguists

graphic

third

language

therefore

code

language

should

them

,

the

- Order

easily

mapping

and

graphic

That

not

sistent

of

how

the

may

and

is

may

relations

years

language

,

which

about

Higher

many

in

"

regularity

Lower

a

Read

.

- sound

structure

In

there

than

native

correspondences

of

Again

difficult

their

.

compared

list

more

diversity

units

single

.

second

were

- sound

questions

the

list

were

words

because

for

variable

Spanish

variable

relationships

letter

,

and

a

perhaps

- sound

new

to

lists

,

second

nonsense

list

Variable

speakers

the

English

to

years

at

fast

of

all

a

skilled

ago

,

that

during

the

tachistoscopic

letters

,

a

very

-

406

E. J. Gibson

long word, and four or more words if they form a sentence(12). Even first graderscan read three-letter words exposedfor only 40 milliseconds, too short a time for sequentialeye-movementsto occur. Broadbent (13) has pointed out that speech, although it consists of a temporal sequenceof stimuli, is respondedto at the end of a sequence . That is, it is normal

for a whole

sequence to be delivered

before a response is

made. For instance, the sentence 'Would you give me your

7"

might end with any of a large number of words, suchas "name" or "wallet" or "wife." The responsedependson the total message . The fact that the component stimuli for speechand reading are spreadover time does not mean that the phonemesor letters or words are processedone at a time, with each stimulus decoded to a separate response. The fact that 0 is

pronounced differently in BOATand BOMBis not a hideous peculiarity of English which must consequently be reformed. The 0 is read only in context and is never responded to in isolation. It is part of a sequence which containsconstraintsof two kinds, one morphological and the other the spelling patternswhich are characteristicof English. If any doubt remains as to the unlikelihood of sequential processing letter by letter , there is recent evidence of Newman (14) and of Kolers (15)

on sequentialexposureof letters. When letters forming a familiar word are exposedsequentiallyin the sameplace, it is almost impossibleto read the word . With

an exposure of 100 milliseconds

per letter , words of six letters

are read with only 20 percent probability of accuracy; and with an exposure of 375 millisecondsper letter, the probability is still well under 100 percent . But that is more than 2 seconds to perceive a short , well -known

word! ~ e can concludethat, however graphemesare processedperceptually in reading, it is not a letter-by-letter sequenceof acts. If the single grapheme does not map consistently to a phoneme, and furthermore, if perception normally takes in bigger "chunks" of graphic stimuli in a single fixation, what are the smallest graphic units consistently coded into phonemic patterns? Must they be whole words? Are there different levels of units? Are they achieved at different stages of

development? SpellingPatterns It is my belief that the smallest component units in written English are spelling patterns (16). By a spelling pattern, I mean a cluster of graphemes in a given environment which has an invariant pronunciation according to the rules of English. Theserules are the regularitieswhich appearwhen, for instance, any vowel or consonantor cluster is shown to correspondwith a given pronunciation in an initial, medial, or final position in the spelling of a word. This kind of regularity is not merely "frequency" (bigram fre-

Learning to Read

407

quency, trigram frequency, and so on), for it implies that frequencycounts are relevant for establishing rules only if the right units and the right relationshipsare counted. The relevant graphic unit is a functional unit of one or more letters, in a given position within the word, which is in correspondencewith a specifiedpronunciation (17). If potential regularities exist within words- the spelling patterns that occur in regular correspondencewith speechpatterns- one may hypothesizethat thesecorrespondenceshave been assimilatedby the skilled reader of English(whether or not he canverbalizethe rules) and have the effect of organizing units for perception. It follows that strings of letters which are generatedby the rules will be perceivedmore easily than ones which are not, even when they are unfamiliar words or not words at all. Severalexperimentstesting this prediction were performed by Gibson, Pick, Osser, and Hammond (18). The basic design was to compare the perceptibility (with a very short tachistoscopicexposure) of two sets of letterstrings, all nonsenseor pseudowords, which differed in their spellingto-sound correlation. One list, called the "pronounceable" list, contained words with a high spelling-to-sound correlation. Eachof them had an initial consonant-spelling with a single, regular pronunciation; a final consonantspelling having a single regular pronunciation; and a vowel-spelling, placed between them, having a single regular pronunciation when it follows and is followed by the given initial and final consonantspellings, respectively - for example, GL/ UR/CK. The words in the second list, called the "unpronounceable " list, had a low spelling-to-sound correlation. They were constructedfrom the words in the first list by reversing the initial and final consonant spellings; The medial vowel spelling was not changed. For example, GLURCK becameCKURGL . There were 25 suchpseudowords in each list, varying in length from four to eight letters. The pronounceability of the resulting lists was validated in two ways, first by ratings, and second by obtaining the number of variations when the pseudo words were actually pronounced. The words were projected on a screenin random order, in five succes sive presentationswith an exposuretime beginning at 50 millisecondsand progressing up to 250 milliseconds. The subjects(college students) were instructed to write eachword as it was projected. The meanpercentageof pronounceable words correctly perceived was consistently and significantly greater at all exposuretimes. The experimentwas later repeatedwith the samematerial but a different judgment. After the pseudo word was exposed, it was followed by a multiple-choicelist of four items, one of the correct one and the other three the most common errors producedin the previous experiment. The subject chosethe word he thought he had seenfrom the choicelist and recordeda number (its order in the list). Again the mean of pronounceablepseudo

408

E. J. Gibson

words correctly perceived significantly exceededthat of their unpronounceable counter-parts. We conclude from these experiments that skilled readersmore easily perceiveas a unit pseudowords which follow the rules of English spelling-to-sound correspondence ; that spelling patterns which have invariant relations to sound patterns function as a unit, thus facilitating the decoding process. In another experiment, Gibson, asser, and Pick (19) studied the development of perception of graphemephonemecorrespondences . We wanted to know how early, in learning to read, children begin to respondto spellingpatterns as units. The experiment was designedto comparechildren at the end of the first grade and at the end of the third grade in ability to recognize familiar three-letter words, pronounceabletrigrams, and unpronounceabletrigrarns. The three-letter words were taken from the first-grade reading list: eachword chosencould be rearrangedinto a meaninglessbut pronounceabletrigram and a meaninglessand unpronounceableone (for example, RAN , NAR , RNA ). Somelonger pseudowords (four and five letters) taken from the previous experimentswere included aswell. The words and pseudowords were exposedtachistoscopicallyto individual children, who were requiredto spell them orally. The first-gradersread (spelledout) most accuratelythe familiar three-letter words, but read the pronounceabletri grams significantly better than the unpronounceableones. The longer pseudowords were seldomread accuratelyand were not differentiatedby pronounceability. The third-grade girls read all three-letter combinations with high and about equal accuracy, but differentiated the longer pseudo words:. that is. the vronounceablefour- and five-letter pseudo words were A - correctly than their unpronounceablecounterparts. more often perceived These results suggest that a child in the first stages of reading skill typically readsin short units, but has already generalizedcertain regularities of spelling-to-sound correspondence , so that three-letter pseudowords which fit the rules are more easily read as units. As skill develops, span increases , and a similar difference can be observed for longer items. The longer items involve more complex conditional rules and longer clusters, so that the generalizationsmust increasein complexity. The fact that a child can begin very early to perceive regularities of correspondencebetween the printed and spokenpatterns, and transfer them to the reading of unfamiliar items as units, suggeststhat the opportunities for discovering the correspondencesbetween patterns might well be enhancedin programming reading materials. I have referred several times to levelsof units. The last experiment showed that the size and complexity of the spelling patterns which can be perceived as units increasewith development of reading skill. That other levels of structure, both syntactic and semantic, contain units as large as and larger than the word, and that perception of skilled readers will be

Learning to Read

409

found,in suitable experiments, to bea function ofthesefactors is almost axiomatic. As yet we havelittledirectevidencebetterthanCattellsor-

iginal discovery (12)thatwhenwords arestructured intoa sentence, more

letters canbeaccurately perceived at aglance. Developmental studies of perceptual chunking in relation to structural complexity may be very instructive. Wheredoesmeaning comein?Within theimmediate spanof visual

perception, meaningis lesseffectivein structuring writtenmaterialthan

goodspelling-to-sound correspondence, as Gibson, Bishop, Schiff, and Smith (20)haveshown. Realwordswhich arebothmeaningful and,as strings ofletters, structured inaccordance withEnglish spelling patterns are

moreeasilyperceived thannonword pronounceable strings ofletters; but

thelatter aremore easily perceived thanmeaningful butunpronounceable letter-strings (forexample, BIM isperceived accurately, withtachistoscopic exposure, faster thanIBM). Theroleofmeaning inthevisual perception of wordsprobably increases aslonger strings ofwords(more thanone)are dealt with. Asentence hastwokinds ofconstraint, semantic andsyntactic, which make itintelligible (easily heard) andmemorable (21). Itisimportant thatthechilddevelop reading habitswhich utilize allthetypesofcon-

straintpresentin the stimulus, sincetheyconstitute structureandare,

therefore, unit-formers. Theskills which thechild should acquire inreading arehabitsofutilizing theconstraints inletterstrings (thespelling and

morphemic patterns) andinwordstrings (thesyntactic andsemantic patterns). Wecould goontoconsider stillsuperordinate ones,perhaps, but theproblem oftheunit,oflevels ofunits, andmapping rules from writing to speech hasjustbegun to beexplored withexperimental techniques.

Further research onthedefinition andprocessing ofunitsshould leadto

newinsightsaboutthenatureofreadingskillanditsattainment. Summary

Reading begins withthechildsacquisition ofspoken language. Later he

learnsto differentiate thegraphic symbols fromoneanother andto decode

theseto familiar speech sounds. Ashelearns thecode,hemustproto attaintheskilled performance which is characterized byprocessing of higher-order unitsthe spelling and morphological patterns of the language. gressively utilizethestructural constraints whicharebuiltintoit in order

Because ofmyfirmconviction thatgoodpedagogy isbased ona deep understanding ofthediscipline tobetaught andthenature ofthelearning process involved, I havetriedtoshowthatthepsychology ofreading can benefit from aprogram oftheoretical analysis andexperiment. Ananalysis ofthereading taskits discriminatory anddecoding aspects aswellasthe

410

E. 1. Gibson

semantic andsyntactical aspectstells uswhatmustbe learned. Ananalysisofthelearningprocesstellsushow.Theconsideration offormalinstructioncomesonlyafterthesesteps,anditsprecepts shouldfollowfromthem. Referencesand Notes

1.SeeC.C.Fries, Linguistics andReading (Holt, Rinehart, andWinston, NewYork,1963), for anexcellent chapteronpastpractice andtheoryintheteaching ofreading. 2.In1959,Cornell University wasawarded a grantfora BasicResearch ProjectonReading

bytheCooperative Research Program oftheOffice ofEducation, U.S. Department of Health,Education, andWelfare. Mostof the workreportedin thisarticlewassup-

portedbythisgrant.TheOffice ofEducation hasrecently organized Project Literacy,which willpromote research onreading ina number oflaboratories, aswellas encourage mutual understanding between experimentalists andteachers ofreading.

3. E.J.Gibson,1.1.Gibson,A.D. Pick,H.Osser,J. Comp. Physiol. Psycho!. 55,897(1962).

4. E.1.Gibson,H. Osser,W. Schiff, J. Smithin A BasicResearch Program on Reading, FinalReporton Cooperative Research Project No.639to theOfficeof Education, Departmentof Health,Education, andWelfare.

5.Themethod wasgreatly influenced bytheanalysis ofdistinctive features ofphonemes by Jakobsen andM.Halle,presented in Fundamentals ofLanguage (Mouton, TheHague, 1956). A tableof12features, eachinbinary opposition, yields a unique pattern for

allphonemes, sothatanyoneisdistinguishable fromanyotherbyitspattern of attributes. A pairofphonemes maydiffer by anynumber offeatures, theminimal distinction beingonefeature opposition. Thefeatures mustbeinvariant undercertain transformations andessentially relational, soasto remaindistinctive overa widerange of speakers,intonations,and so on.

6. A. D. Pick, J. E.xp.Psycho!.,in press.

7.C.H.Bishop, J. Verbal Learning Verbal Behav. 3, 215(1964). 8. Currentadvocatesof a revisedalphabetwho emphasize the low letter-sound cone-

spondence inEnglish areSirJames Pitman andJohnA.Downing. Pitmans revised alphabet, called theInitial Teaching Alphabet, consists of43characters, sometraditionalandsomenew.It is designed forinstruction ofthebeginning reader,wholater transfersto traditionalEnglishspelling.SeeI. 1. Pitman,J. Roy.Soc.Arts109,149

(1961); J.A.Downing, Brit. J.Educ. Psycho!. 32,166(1962); , Experiments with Pitmansinitialteaching alphabet in British schools,paperpresented at theEighth AnnualConferenceof International ReadingAssociation, Miami,Fla.,May 1963.

9. L. Bloomfield,Elem.Engl.Rev. 19, 125, 183 (1942).

10.Seeresearch reports ofH.Levin andJ.Watson, andH.Levin, E.Baum, andS.Bostwick, in A BasicResearchProgramon Reading(see 4).

11. R. Dodge, Psychol.Bull.2, 193 (1905). 12. J. McK.Cattell, Phil.Studies2, 635 (1885).

13.D.E.Broadbent, Perception andCommunication (Pergamon, NewYork,1958). 14. E. Newman, Am. J. Psycho!.,in press.

15.P.A.Kolers andM.T.Katzman, paperpresented beforethePsychonomic Society, Aug. 1963, Bryn Mawr, Pa.

16.Spelling patterns inEnglish havebeendiscussed byC.C.FriesinLinguistics and Reading (Holt,Rinehart, andWinston, NewYork,1963), p. 169if.C.F.Hockett, in A Basic Research Program onReading (see4),hasmadeananalysis ofEnglish graphic

monosyllables whichpresentsregularities ofspelling patternsinrelationto pronuncia-

tion.Thisstudywascontinued byR.Venezky (thesis, Cornell Univ.,1962), whowrote

Learning to Read

411

a computer program for obtaining the regularities of English spelling-to-sound correspondence. The data obtained by means of the computer permit one to look up any vowel or consonant cluster of up to five letters and find its pronunciation in initial , medial , and final positions in a word . Letter environments as well have now been

included in the analysis. Seealso R. H. Weir, Formulationof Grapheme -PhonemeCorrespondence Rules to Aid in the Teachingof Reading , Report on Cooperative Research Project No . 5-039 to the Office of Education, Department of Health, Education and Welfare

.

17. For example, the cluster GHmay lawfully be pronounced as an F at the end of a word,

but neverat the beginning.The vowel clusterE1GH , pronouncedIAI (/ej/ ), may occur in initial, medial, and final positions, and does so with nearly equal distribution . These casesaccount for all but two occurrencesof the cluster in English orthography. A good exampleof regularity influencedby environment is [c} in a medial position before 1plus a vowel. It is always pronounced 151 (social, ancient,judicious). 18. E. J. Gibson, A . D. Pick, H. asser, M . Hammond, Am. ] . Psychol . 75, 554 (1962). 19. E. J. Gibson , H . asser , A . D . Pick, J. Verbal Learning Verbal Behav. 2, 142 (1963 ).

20. E. J. Gibson, C. H. Bishop, W . Schiff, J. Smith, ] . & p. Psychol ., 67, 173 (1964). 21 . G . A . Miller and S. Isard , ] . Verbal Learning Verbal Behav. 2, 217 (1963 ); also L. E. Marks and G . A . Miller , ibid . 3, 1 (1964 ).

22

A

Developmental

Study

Eleanor

]

-

.

Gibson

seems

the

only

fair

to

information

Search

Behavior

Yonas

to

presumed

of

feature

even

that

on

inquired

to

reprinted

here

letter

's

,

there

scanning

no

in

acoustic

ability

did

This

task

not

issue

-

-

was

students

no

think

a

,

that

approach

set

then

,

that

of

verbal

the

secret

,

Gibson

,

the

.

adults

But

children

,

tasks

' s

except

like

of

turning

.

is

for

little

in

"

,

described

the

the

target

scan

.

performed

a

Three

another

context

unlike

,

The

and

of

high

graphic

or

confus

-

)

to

one

a

.

after

letter

they

skilled

,

,

be

enter

or

school

readers

deaf

we

found

drawn

from

consider

for

task

better

mature

but

to

- search

For

reading

with

1970

conclusion

latencies

or

performed

profitable

in

soon

scanning

Yonas

very

into

a

one

main

a

longer

in

notably

performance

all

)

a

earphones

task

with

1966

placed

processing

most

not

adults

them

to

experiment

the

of

,

Shurcliff

stage

question

.

.

"

experiments

(

(

was

all

in

a

relation

experiment

that

accuracy

confusability

at

In

rhymed

Shurcliff

letter

rate

decoding

that

reading

like

target

phonological

other

phonetic

is

and

.

to

,

letters

an

over

did

or

in

distinguish

pronounced

rate

,

translation

)

names

time

paper

"

in

1966

they

the

,

to

models

letter

acoustic

-

College

understanding

tially

the

of

in

for

I

role

Yonas

The

scanning

Gallaudet

evidence

,

the

pursued

at

this

which

the

the

with

,

.

slow

when

Kaplan

confusability

,

Even

interference

,

experiment

low

.

served

being

one

this

involved

way

at

include

models

Yonas

names

same

I

processing

characters

the

,

end

that

and

letter

in

attracted

that

processing

Gibson

heard

students

head

about

(

letters

was

the

was

To

"

alphabetic

here

graduate

in

children

reprinted

name

)

,

I

.

discussed

question

processing

for

paper

( then

)

much

reading

that

reading

of

searched

the

search

further

whether

they

how

to

somewhere

a

mode

( not

letter

lists

went

phonological

as

just

approach

study

We

in

acknowledge

- processing

developmental

is

essen

worse

.

,

Why

-

they

is

still

it

,

eludes

?

Reprinted 169

Albert

Visual

-

It

us

,

of

by -

171

.

permission

of

Psychonomic

Society

,

Inc

.

Perception

and

Psychophysics

,

1966

,

1

,

414

E. J. Gibson & A . Yonas Children

in

second

compared

on

conditions

. ! . 2

letters

of

sought

It

than

searching

time

at

has

long

,

skill

in

what

has

a

of

task

specified

target items

should

show

)

many

or

there

tures

,

of

the

make

matching

might

be

context

age

of

,

of

were

single

target

time

was

de

no

-

harder

search

and

) .

quantitative

there

is

is

,

irrelevant

.

and

an

in

vary

a

requires

a

age

search

of

over

for

more

age

when

.

and

searching

context

differences

which

with

employing

task

in

unwanted

activity

should

,

selecting

Systematically

distracting

The

increasing

environment

interactions

and

Ss

with

can

items

the

a

or

groups

less

,

suitable

present

these

parameters

,

be

after

long

loss

to

in

the

differentiating

those

the

a

number

of

search

for

with

of

targets

of

distinctive

.

for

features

we

On

the

context

.

by

Such

a

target

and

.

Method

Letters

were

strings

1

/

could

4

in

chosen

,

one

.

high

be

inserted

as

string

.

to

The

a

list

in

stimuli

line

was

the

.

.

fixed

display

They

were

The

typed

to

apparatus

typed

list

a

in

was

longer

10

cardboard

.

The

lists

apparatus

of

in

.

long

30

four

,

the

mounting

was

-

which

adults

shared

,

fea

confusable

than

as

,

age

properties

highly

ale

processing

increases

number

children

of

et

parallel

learning

) ,

for

,

number

Neisser

can

If

contrasting

possible

difficult

.

once

increase

.,

,

speed

at

involves

( i .e

.

practice

of

attended

presented

,

confusability

little

with

maximize

experiment

contextual

adult

once

more

items

list

increased

that

the

embedded

attention

and

background

( 1964

,

interaction

relatively would

a

a

Search

targets

;

question

in

time

that

set

,

.

epitomize

this

variable

that

selective

III

information

manipulated

at

if

letters

context

ignoring

to

qualitative

and

one

an

hand

in

target

.

things

expect

other

while

us

letter

are

targets

of

should

a

age

shown

number

,

Neisser

as

were

ten

sought

qualitative

what

processing

manipulated

have

as

was

two

behavior

ignoring

to

by

such

at

both

from

target

as

sought

were

experimental

two

Condition

visual

are

investigated

If

,

confusability

for

attentive

seemed

devised

are

in

and

and

variables

targets

( 1963

task

,

up

In

confusable

information

the

,

.

high

Searching

there

,

designated

.

task

Two

highly

that

experiment

similar

A

II

sophomores

three

letter

list

of

.

under

Condition

given

tasks

college

.

selecting

scanning

the

.

a

and

task target

In

letters

three

one

context

present

of

all

of

incoming

for

strategies

in

list

assumed

- ground

The

in

for

single

.

developmentally

utility

searching

back

a

levels

filtering

a

appeared

for

age

been

differences

I ,

in

age

all

grades

scanning

confusability

one

with

sixth

and

Condition

sought

creased

and

search

visual

only

was

fourth

visual

In

low

but

letter

,

a

- letter

letters

which

a

light

- tight

Developmental Study of Visual SearchBehavior

415

Figure22.1 Three-quarterview of the scanningapparatus .

box with a slantingglasswindow (seeFig. 22.1). The glasswaspartially silveredsothat thelist wasvisibleonly whentwo lumilinelampsinsidethe box were turnedon. Above the glasswindow, a smallcoveredbulb was positionedso that a dot of light appeared on the mirror at the samepoint in spaceasthe top line of the list whenthe lampswereon. Thisallowed5 to accommodate properlyandstarthis scanpreciselyat the top of the list. The 5's eyeswereapproximately14 in. from the mirror. The slopeof the mirror(andlist beneathit) weresuchasto keepthisdistanceapproximately constantasthe scanprogressed downward. The 5 wasinstructedto startat the top of the list and scandownward until he found the designatedtargetletter, proceedingas rapidly as possible. He wasurgedto scandownwardin order, without skippinganylines. He helda pushbuttonin his hand. Whenhe wasreadyto beginscanning , he pressedthe button which causedthe list to appearand alsostarteda Hunter clock-counter(calibratedin .001 sec.). The 5 immediatelystarted scanningdown the list. On findingthe targetletter, he pressedthe button again. This stoppedthe clockand turnedoff the lamps. He was askedto indicatethe approximate spoton themirror wherethe targethadappeared

416

E. J. Gibson & A . Yonas

in order to discourageguessing. The E then recorded the time and placed another list in the display box. If S had failed to find the target when he reached

the bottom

of the list , the list was set aside and rerun

later .

Five practice trials precededthe experimentproper. The letters chosen were simplified capitals typed with an IBM signtypewriter , because these are easy for a child to read and because this type

had previously been analyzed for distinctive features and a confusion matrix was available(Gibson et ale1964). The experimentaldesign included three conditions .

ConditionI

was a single target search, with low contextual confusability.

The S was told to search for the letter G, which would appear only once on a list . The target letter remained the same for 20 trials . Its position in

the 20 lists was randomized, with the restriction that it appearedequally often in four quadrantsof the list (from top to bottom), never in the same line, and never in the first or the last line . In the practice trials , the target

item did appearonce in eachof these lines to insure that a complete scan would be made . The six context letters in Condition I (L, K , V , M , X, A ) were ones which had seldom or never been confused with G in the confu -

sion matrix. They were randomly placed across and down the list, but appearedan equalnumber of times. Condition II

was a two -target search. The context was low -confusion

againl but S was told to look for either of two letters , G or R, only one of which would appear on any given list . He was not told which one. Each

target appeared10 times, randomly ordered through the 20 trials. ConditionIII

was again a single target task, hut the context was highly

confusable with the target letter G (the letters H, Q , C, J, 5,; and R, which

ranked highest in confusability with G in the confusion matrix). There were 20 lists, with target position ordered as before .

The 5s were 72 children, 24 from each of three grades(second, fourth, and sixth), and a group of 12 college sophomoreswho were fulfilling a requirement of the introductory course in psychology. There was an ap-

proximatelyequalnumberof eachsexin eachagegroup. Each5 took part in two of the three conditions of the experiment , Condition I and either Condition II or Condition III . Condition I was always run first . 5s within a

gradewere paired by age, with one of the pair assignedto Condition II, the other

to Condition

III . Two

rather

than all three

conditions

were

run

because the children 's time was limited to one-half hour . The experiment was run at a school3, during school hours . The college 5s were run in the laboratory .

Developmental Study of Visual SearchBehavior

417

Results In treating the results, the latency of detecting the target for each trial was transformedto take account of its position in the list. Theoretically, a linear relationshipmight be expectedbetweentarget position and scanning rate. The expectation did not hold true for the first few list positions, however. Sincea similar initial lag appearedfor all three conditions, and for all age groups, all the transformed scoreswere included in the analysis. Means for the transformed scoresare plotted by age group for the three conditions in Fig. 22.2. As expected, it is evident that speed of scan increasedfrom secondgrade to college sophomoreson all three conditions. Three analysesof variancewere performed on the transformed scores, one for the high- and low-confusion conditions, one for the one- and two-target conditions, and one for the two-target and high-confusion conditions. Grade differenceswere significant at the .05 level of significanceor better in all conditions. The differencebetweenthe high- and low-confusion conditions was significant (p < .001), but the grade by conditions interaction was not. Context confusability was, it appears,equally damagingat all age levels. The differencebetween the one- and two-target conditions was not significant, nor was the interaction with grade. Two targets were searchedfor as efficiently as one, at all age levels. The high-confusion condition was significantly more difficult than the two-target condition (p < .001).

1 ..0"".""... ~ B ' . . , . High confusion , I target c -. 8 7 ' . . OJ . ---- -Z (.c-G /)06 ""'"'", .0 E 5 Low confusion , 2 targets ..i0 ~ . c . = 4 1 . -(tG 3 . . 0 . . u " ' . " ):.0 2 . Low confusion , I torget I2 4 6 Soph . Grode

Effect

of Target

Number and Low and Hi9h - Confusion Context on Sconnino Rote .

Figure 22 .2

Mean latenciesfor the three conditions of the experiment plotted by age group.

418

E. J. Gibson &: A . Yonas

Individual differencesbetweenSswere marked, especiallyin the younger age groups. Thesedifferencesinteractedsignificantly with conditions, suggesting that strategies for different Ss vary with the task. This strategy difference was especially noticeable when younger Ss began the highconfusion task; some lowered the pace of scanningalmost at once, while others were slower to do this and traded time for errors (missing the target). In general, however, the omissionsreflectedthe sametrends as the latencies. Discussion Our expectation that increasing the number of things sought for at one time would be relatively more difficult for the younger 5s was not confirmed. The fact that the double-target search was no harder than the single-target searchconfirms the finding of Neisser et ale(1963), though searchingfor two targets is surely not as difficult as searchingfor ten. The 5s had the benefit of practice on Condition I in the double target task, so performancemight have looked poorer if practice had been balanced. But it seemsclear that any difference, with two targets at least, would be negligible. One need not infer, however, that parallel processingis going on. Another explanation of the good performancein Condition II stems from analysisof distinctive features. The two target letters, G and R,had in common a curve, but none of the context letters did. Optimal strategy, by feature processing, would be searching for one feature, a curve, which would differentiate the target from context letters . This possibility is at presentbeing explored further in another experiment. The effect of context on searchspeed is impressive. When the target letter is embeddedin a context of letters containing a high percentageof common distinctive features (as these did) the search task is rendered far more difficult. However the featuresmay be processed , sequentiallyor in parallel, discovery of the target takesmuch longer. It is not clear exactly how a simultaneousprocessingmodel would handle this finding, although a seQuential model of visual discrimination, when a matching or discrimi. Jnating . judgment is required, would predict it . If discrimination involves tests of a featurelist (the distinctive featuresof letters), then moving down the list until a differenceis found would take much longer when the number of featurescommon to the target and the context items is increased. That younger children should take longer than older ones or adults might be expected, sincethe feature set might be lesswell assembledand a possible hierarchicalordering of tests for greatestefficiencynot yet optimized. The lack of an age-condition interaction, however, fails to support the possibility of a qualitative processchangeduring development.

419

Developmental Study of Visual SearchBehavior Notes 1 . This

research

was

2 . We

wish

to

express

subjects

.

running 3 . Thanks children

are

due

Mr

supported

by

NIH

gratitude

. Pastre

Grant

to

MH

- O7226

Mr . Arthur

, principal

of

the

- 02 .

McCaffrey

for

valueable

, the

Cayuga

Heights

School

, J . An

analysis

of

assistance

teachers

in

, and

the

.

References

Gibson

, E . J ., Osser tested with

Neisser Neisser

by the

, H ., Schiff

a confusion Office

, V . Visual

of search

, W ., &

matrix Education

. In , A

Smith Final

basic

. Scient . American

Report research

, 1964

, V ., Novick , R ., & Lazar , R . Searching Skills , 1963 , 17 , 955 - 961 .

, Cooperative program

critical Research

on reading

features Project

of

letters

No

,

. 639

.

, 210 , 94 - 101 . for

ten

targets

simultaneously

. Percept . mot .

23 ConfusionMatricesfor GraphicPatternsObtained with a Latency Measure

Eleanor]. Gibson , FrankSchapiro , Albert Yonas Thispaperwasneverpublishedand was in a sensea kind of terminationof a line of thought.I remember FrankSchapirosayingto meoneday as theveryambitious experimentwas nearingcompletion , "This is terminal research , isn't it?" I was surprisedand wonderedfor a while what he meant. But it was an astute comment. I was becominglessand lessattracted by an information-processingapproach, and this experimentwas a prime exampleof it . It had begun to seemto me that it was

either leadingnowhere , or worse, to the wrong place. I had beenthinking of distinctive featuresas bits of information lodgedin static wedgesof print , and that didn't capture what goes on when we really read at all .

Nevertheless , this was a kind of super-experiment . A vast numberof subjects took part (72 adults and 84 children), each one giving us a large number of

judgments , "same" and "different" responses , and reactiontimesin milliseconds . Theonly reasonwecouldhandlethemammothamountof resultingdatawasthat the experimentwas "automated." The subject presseda key for same or different

and, whenready, operateda foot pedalto bring up the nextpresentation (a pair of lettersor artificialgraphemes ). Thejudgmentand its latencywereautomatically recordedon an IBM card. We had renteda keypunchand amassed files full of IBM cardsthat could be dealt with by the computerdirectly. Such technical wonders! We felt we were at the forefront of the art. But it was only a couple of years beforethe whole procedurewas outmodedand cardswerea thing of the past. Other psychologistswere interestedin our results, however, and we had a great many requestsfor the data. The results were written up for circulation, and they are reprinted here as I sent them out. They are useful for a data bank and have

beenreplicatedin otherlaboratories(seeGarner1979), but I lost interestin them and the kind of "processing" talk that this approach led to.

Developmentally , thedatado havea certaininteresthowever . Theremarkably straightforward partitioning of the children's judgments in the cluster analysesled

me to think that, whethertheycouldtell you so or not, the childrenwerehaving no troubledifferentiatinglettersand were not held up by tendencies to global,

422

E

.

J

nonanalytic

likely

.

Gibson

,

F

perception

to

be

found

at

.

Schapiro

.

Any

this

,

&

A

.

Yonas

difficulties

level

,

I

children

think

face

in

learning

to

read

are

not

.

One of the more difficult but essential tasks for a child in learning to

read is the identification of the letters of the alphabet. This achievement involves not only the attaching of a unique nameto eachletter; there must be a one-to-one matching of the graphic pattern to the name, and that accomplishmentrequires visual differentiation of each of the graphic patterns as uniquely different from the others. How are these patterns differentiated? It has been suggested(Gibson et al. 1963, Gibson 1969) that this is accomplishednot by a processof matching the total form of a letter to a stored representationthat is independentof other letters, but rather by a more economical process of discriminating a smaller set of relational distinctive features that yields a unique pattern for each letter . The features

are distinguishingwith respectto the set, not componentsfitted together to build the letter .

What might these features be, and how can we investigate such a

hypothesis? It is apparentthat a set of distinctive featuressmallerthan the total set of letters requires that some letters share certain features with others . Uniqueness for a given letter can be achieved only by a pattern of features that is distinct

from the others . It follows

from this that two letters

that share a number of features and differ from one another minimally by only one or two should be harder to discriminate than two letters that

differ from one another by many features. We can test this prediction by determining experimentally the actual confusability of the letters with one another .

This program was followed by Gibson et al. (1963), and a confusion

matrix was obtained for the 26 Roman capitals, using errors in a matching task as the measure.Subjectswere 87 preschoolchildren. Sinceevery letter had to be matched against every other with equal opportunities for error, a very large number of comparisonswere required. Anyone child could go through only a small block of the design before becoming fatigued. Since we did not want errors due to inattention or misunderstandingof the task, there was a very small yield of errors, despite a large investment of experimental time. The error matrix contained many "holes" where no errors occurred . Those that did occur, however , showed a low but sig-

nificant relationship to an intuitively generatedf.eaturelist. While the correlations between degree of feature-difference and confusion errors were not large, it seemed worthwhile to make a new attempt to collect more satisfactory data on confusability of letters . It was decided to use a latency measure, rather than errors, since there

would then be a yield of information on every trial and no empty spacesin

ConfusionMatrices for Graphic PatternsObtained with a LatencyMeasure

423

the final matrix. The old literature on disjunctive reaction time (Cattell 1902; Lemmon 1927; Woodworth 1938) suggestedthat responselatency is increasedwhen highly similar items must be discriminated. Latency should, therefore, yield a good measureof confusability. It was possible to check this expectationby correlating latencieswith sucherrors as occurred. The judgment chosen was "same" or "different." A pair of items was presentedsimultaneouslyand the subject respondedby pressinga button with one hand if they were the sameand by pressinganother button with the other hand if they were different. We were interestedin this judgment becausethe comparisonof meanlatenciesfor "same" responseswith mean latencies for "different" responsesis pertinent to the question of what processesgo on in discrimination, as well as is the distribution of latencies for "different" pairs (Egeth 1966). Two setsof nine letters eachwere chosenfor test. Eachset was chosen to include a range from very different to very similar pairs, as predicted by feature differences. This would give us two 9 X 9 confusion matrices to checkon our predictions. We did not attempt to obtain the 26 x 26 matrix of all the letters, since running only once through the simplest design for this would require 1,300 judgments. We also devised a set of nine artificial graphemes , constructedso that the featureswe thought might distinguish letters would also distinguish them. This would permit another check on the predictability of features, one with unfamiliar letterlike oattems. We were also interestedin comparingmean"same" and "different' ; latenciesfor these unfamiliar patterns, which might lack codability or "Gestalt-like" properties characterizingreal letters. It was possible that our rather low correlations between feature differencesand errors in the earlier experiment had to do with the age of our subjects(a mean around four years) who were as yet rather inexperienced in discriminating letters. We therefore ran subjects of two age groups: adults and seven-year-old children. Whether the samefeaturesare selected for discriminationat all developmentallevels is an interesting question, and it is possible that adults, after long experience, achieve some higher-order more economicalmeansof processingvisual letter patterns. Sincethere are so many unknown possibilitiesfor processingdifferencesdependenton age and material, it was decided to analyze the data by a cluster method in order to compareit with our a priori feature analysis.

Method

Material Both the lettersand artificialgraphicstimuliwereprepared photographically onslides . Thefirstsetof lettersrunwasC, E, F, G, M, N, P, R, W. ThesecondsetwasA, 0, H, K, 0 , Q, 5, T, X. Thetypewas

424

E. J. Gi'bson, F. Schapiro, & A . Yonas

E3 - G r--,

~

-::l

9

L

~

~

Figure23.1 Artificialgraphemes . simplified Roman capitals . The same " master " copies were photographed for all the slides . The artificial graphemes are sketched in Figure 23 .1. A master copy was prepared for each one , black on white , with the same line thickness as the letters . The artificial graphemes differed from one another in such features as curve , straight , diagonality , intersection , and others deemed characteristic of letters . We attempted to include items that were very similar and others that differed by many features , on an a priori basis . Apparatus The subject sat in a chair with his eyes approximately 5- 1/ 2 feet from a display apparatus . His foot was on a starting pedal and the index finger of each hand on a response button . The slides were presented to him simultaneously , via a projector . They were projected in two small windows of a Foringer display apparatus . The windows were 2 x 2 in . and were separated by 1- 1/ 8 in ., one to the right of the other at the same height . They could be observed without shifting fixation . The slides were presented automatically , paced at a 3 to 4 second interval by a pro gramming device . If the pace became fatiguing for the subject , he could remove his foot from a pedal and stop the displays . He could start them again when he was ready by pressing the pedal . A pair of slides presented for the same -different judgment was exposed until the subject made his decision , when

he pressed one of the two

buttons

" different " judgment . When the subject pressed the button , a shutter

for a " same " or a

closed , terminating

the

display , and the subject ' s response time from the beginning of the display was measured and recorded automatically by an IBM keypunch . A green light went on when the subject responded he made an error . Procedure following

correctly . A red light came on if

The adult subjects made 360 judgments

of same or different ,

a brief practice period to become accustomed

of the presentations control for response Each of 36 different was randomized on

to the set -up . Half

were " same " pairs and half were " different " pairs , to bias . Each of the 9 same pairs was presented 20 times . pairs was presented 5 times . Order of presentation replications within and between subjects . Only one

arrangement for a given different pair was used (e.g ., either AB or BA but not both ), but a given letter , throughout , appeared equally often on left and right . Procedure was similar with the artificial

characters .

Confusion Matrices forGraphic Patterns Obtained witha Latency Measure425

Itwasessentially thesamewiththechildren, butthenumber ofjudgments wasreduced to tworeplications foreachdifferentpairandeight foreachsamepair.Thechildren wereallowed tochoose atoywhen they cameinandwerepresented withit at thecloseoftheexperiment. The

experiment wasconducted in thelaboratory withadultsubjects butin a mobilelaboratoryat the schoolwith the children. Instructions were as follows:

Every threeorfourseconds twoletters (orforms) willappear onthe screenbeforeyou.Youare to decideif thesetwo are the sameor

different. Iftheyareidentical press thebutton marked same withyour

indexfinger. (Keep yourfingeronthatbuttonsothatyoucanrespond rapidly.) Iftheyaredifferent, pressthebuttonmarked different

with the indexfingerof your otherhand. Focus between the two windows.

Solongasyourfootispressing thefootpedal, theexperiment will continue. Ifyouneedto stop,liftyourfoot;thisstopstheprojector

until you are ready to continue.

Youareto respondasquickly asyoucanwithoutmaking mistakes.

We willhave3 practicetrials.Try to learnwhichhandis sameand

whichis different. Thegreenlightmeansyouwerecorrect; thered

light indicatesthat you were in error.

A subjectwas askedwhichwas his dominanthand.The dominant

handwasassigned to thesamekeyor thedifferent keywithalternate

subjects.

Twenty-four adultsubjects wererunin thefirstexperiment withthe firstsetofletters. Thisexperiment wasthenreplicated withanother 24

adult subjects, since wewanted acheck onreliability ofourmeasure. Sixty 7-year-old children wererunon thissetof letters. Twenty-four adult

subjects wererunontheartificial graphemes, and24moreonthesecond

setofletters. Twenty-four children wererunontheartificial graphemes,

butthetimecostwassogreatthatthisphasewasthendiscontinued.

SubjectsTheadultsubjects werecollegestudentsobtainedfromthe subjectpoolof the introductory psychology course.Thechildren were

obtainedfroma summer daycampat a school. Results

LatencyData for Letter Pairs

AdultsTheresultsforthetwogroupsof24adultsubjects runon the firstsetofletterswereverysimilar; thecorrelation between thetwowas

426

E. J. Gibson, F. Schapiro , & A. Yonas

Table23.1 MeanLatencies (in msec .) for a Response of "Same " or "Different " to Pairsof Nine Roman A'~~~'--- Capital - -r---- Letters - ------ (48 , AdultSubjects I ) C E F G M N P R W C 467 500 495 552 481 465 496 483 467 E 475 560 479 502 495 504 490 485 F 488 464 488 491 495 495 482 G 496 463 470 501 481 458 M 497 545 477 481 538 N 500 463 486 510 P 466 571 461 487 489

+ .82. The datawerethereforecombined and are presented in Table 23.1. Thelatencies for responding to a pairof differentlettersby pressing the appropriate keyvariedfrom458msec . (GW) to 571msec . (PR), a rangeof over100msec . Thedifference in latencyof responding to thesetwo pairs , theextremes of therange , wasverysignificant (P < .0001). In fact, differencesaslow as30 msec . aresignificant at .05 or better.Thismethodof testingconfusability thusgivesa usefulspread of responses and, judging fromtheconfirmatory resultsof thereplication , is reliable . Thecorrelation is particularly satisfactory , sincethe replication wasrun by a different ~ experimenter in a differentplace . Thereis alsoa rangeof latencies for judging"same " to identicalpairs (downthediagonal ), from466msec . (PP) to 500msec . (NN). Therangeis by no means asgreatasfor differentpairs , but theextremes of therange arenevertheless significantly different(P < .01). DecidingthatNN or MM isa "same " pairrequires a longertimethandoesPPor CC. Something that is peculiarto theletteris influentialin evensucha simpleperception as sameness . Letterscontaining diagonality seemto beassociated withlonger latencies , buttheobservation maynotholdupandthereason for it, in any case , is not clear . It maybe thatspeedof a "same " judgmentdepends on thecomposition of theseries to someextent , suchaswhetherotherletters highlyconfusable withit arepartof thesetbeingtested . Theresultsfor the secondsetof lettersrun on 24 adultsubjects are presented in Table23.2. Therangeof latencies for "different " judgments extends from472msec . (QT) to 593msec . (OQ), a slightlylongerrange thanthefirstset. Themeanfor all thejudgments is alsoa little higherfor thisset, whetherdueto thesample of lettersor to thesample of subjects is notclear . In anycase , thereis a widespread of latencies andthedifference

Confusion Matrices for Graphic Patterns Obtained witha Latency Measure 427

Table 23 .2 (Subjects inmsec .))for aResponse of'/Same "or"Different "toNine Pairs ofLetters (Latencies 24 Adult . A

D

H

K

a

Q

S

T

476

509

534

521

505

489

535

484

X

520

494

497

510

548

521

507

511

491

A D H 514

580

505

484

499

540

542

526

504

493

496

492

588

K a 491

593

512

490

490

555

506

472

490

486

514

501

487

524

Table 23.3 Latencies (inrosec .) foraResponse of "Same " or"Different " toNinePairs of Letters (60Seven -year -oldSubjects ) C E F G M N P R W C 1047 1097 1112 1252 1130 1093 1128 1107 1110 E 1033 1229 1107 1112 1124 1054 1114 1144 F 1072 1108 1139 1132 1092 1148 1106 G 1082 1127 1148 1099 1091 1042 M 1043 1279 1063 1057 1203 N 1075 1084 1112 1155 p 1030 1138 1070 1050 1036

between extremes is highlysignificant (P < .0001 ). Differences over30 msec . aresignificant at.05orbetter . There isagain avariation in "same " judgments , from476msec . (AA) to 555msec . (QQ). Thisis a considerably longerrangeof "same " latencies thanfor thefirstset. Thedifference between extremes is verysignificant (P < .0001 ). Letters containing diagonals donotgenerally takelonger . The Q, whichtakes longest , hasa veryconfusable letter , 0, in theset,butthe 00 pairis judged"same " comparatively rapidly . Theremaybe some structural feature of the"same " pairasa whole , likesymmetry , thatinfluences speed of processing . We shall return to this speculation in a later section . ChildrenLatency datafor the60seven -year -oldsarepresented in Table 23.3. Latencies aremuch longer thanthose ofadultsubjects , butagain there

428

E. J. Gibson , F. Schapiro, & A . Yonas

is a range of latencies. The shortest latency was 1036 msec. (RW) and the longest 1279 msec. (MN ), a range of 243 msec. The extremedifferencesare satisfactorilysignificant(P < .001), although the children are quite variable. Artificial Graphemes

.~

]

Adults The latenciesfor judging pairs of artificial graphemes"same" or 'Idifferent" are presentedin Table 23.4. Judgmentsof "different" vary from 459 msec. (r-lvs . 8- ) to 527 msec. (G vs. ~ ), a range of 68 msec. This is not as long as the range of IIdifferent" latenciesfor either set of letters. Is this becausethe set of artificial graphemesis not as well differentiated as the very familiar letters? If that were the caseone would expect the overall mean latenciesfor the artificial graphemesto be considerablylonger than those for letters, but they are not. They are slightly shorter, in fact, than the overall meansfor the secondset of letters. We concludethat although the artificial graphemesare not familiar in the senseof figures identified or codable as wholes, neverthelesstheir distinguishing features (curve-

~

Graphemes -~ -r-'-~ -(,-2-Adult -94 Subjects .G )' 479 472 488

Table23.4 Latencies (in msec.) for a Response of "Same " or "Different" to Pairsof Nine Artificial

456

r--i

Jl e-

~

~ L

.--i:-

e-

(::::>J

~

L

493

497

477

507

496

477

470

487

515

467

569

454

459

476

462

467

527

481

475

455

474

459

463

482

492

477

474

488

516

503

481

496

468

498

479

472

498

487

514 484

Matrices

straight

,

readily

detected

we

,

.

simply

didn

graphemes

is

The

rance

here

figures

Or

.

is

But

a

"

"

.

there

There

more

of

"

pair

are

as

"

a

made

.

same

on

(

for

rrrr

)

than

thus

fact

that

OQ

.

the

for

The

artificial

processed

.

-

3

)

?

We

igno

the

-

two

other

order

,

,

the

than

-

3

Again

Why

in

higher

-

.

our

lines

one

a

(

. 001

and

in

of

487

<

others

of

basis

to

general

number

the

are

the

,

at

in

be

to

instance

significant

pairs

the

and

due

for

429

.

regularities

to

whole

,

msec

some

features

same

)

is

of

letters

latencies

455

for

Measure

be

longest

from

easier

of

as

difference

detection

manifest

the

vary

Latency

likely

similar

the

a

those

. 001

extreme

is

judgment

characterizing

as

very

as

and

same

The

about

are

pair

shortest

replication

is

same

significant

.

little

with

may

a

the

msec

of

the

construct

of

very

are

Obtained

range

nevertheless

32

detection

know

. )

shorter

' t

judgments

of

Patterns

etc

The

between

range

Graphic

diagonality

difference

a

for

. 11111111111

Confusion

?

feature

?

Table 23 . 5 (.imsec n~ .)for Discrimina Pairs of Artificia Graph as Diffe (Latencies N =24 Children 11295 91267 G1305 r~1332 -:l 1244 e-1304 ~125 ~~ 1216 13 Children

Latency

sented

per

in

child

data

Table

)

and

23

the

. 5

.

variability

for

The

artificial

number

graphemes

of

between

for

judgments

for

subjects

very

24

each

children

are

pair

great

is

,

so

pre

small

few

(

of

-

two

the

~

9 G r -,

~ e-

~ ~

1183

1105

1177

1086

1299

1223

1213

1383

1271

1269

1355

1245

1277

1178

1137

1240

1139

1255

1252

1257

1239

1178

1385

1228

1231

1175

1144

1220

1234

1317

1208

430

E. 1. Gibson, F. Schapiro, & A. Yonas

latency differences between pairsarereliable, andtheydo notcorrelate very well with those of the adults.

Latencyof Same and Different Judgments

If the latencyof judgingdifferent increases as the similarity of two

itemsincreases, shouldit not takelongestof allto judgethat twoidentical items are the same?If one conceivesof a sequentialfeature-testingprocess

thatcompares twoitemsto finda match,theprocess mightbeexpected to

takelongerifanexhaustive search ofeverysingle feature mustbecarried out,as a judgment of same wouldpresumably require. A decision of different couldbe reachedas soonas any difference wasfoundand the

searchcouldstop.Therangeof latencies forjudgments of different supports sucha notion; themoretwoitemsdiffer, thesooner a difference shouldbe found, and that is exactlyhow the latenciesfor different pairs look.

Butwhenone looksat the meanlatenciesfor same as comparedwith

different, this reasoningbreaksdown.The overallmeansfor same

judgments areshorter thanthemeans fordifferent,asTable23.6shows. Thistrendispresentintheadultsjudgments ofbothsetsofletters,inthe

childrens judgments ofletters,andinthejudgments ofartificial graphemes for both childrenand adults.The trend was particularlystriking(and

surprising) inthechildren. Ofthe60children runonletters, 53hadshorter

meanlatenciesfor same judgments.Of the 24 run on artificialgraphemes, 21 had shorter mean latencies for sames.

Howis thisto be interpreted? Doesit meanthata feature-testing model of discrimination mustbe wrong?Woulda template-matching modeldo

anybetter?It mightforsuccessive discrimination, for,asSorenson (1968) hassuggested, a subject couldbeset withanappropriate template fora same judgment. Iftheseconditemmatches thetemplate thatis ready, the decisionis made.If not, a furthersearchof somesort must go on. But

whatsort?it it werea comparison with a set of templates,why should therebe sucha systematic rangeof latenciesfor differents?Indeed,why Table

23.6

Mean La encies

of Same and Different Judgments in msec.

Adults, first set of letters

Adults,secondsetofletters Children,

letters

Adults, artificial graphemes Children, artificial graphemes

Different 493

514

Same 485

505

1121

1052

507 1270

496 1186

Confusion Matrices forGraphic Patterns Obtained witha Latency Measure 431

shouldtherebe a significant difference in latency between someof the

same judgments?

One simpleexplanation that cannotbe easilydismissed is that the

same pairsarerepeated moreoften. Thisprocedure wasnecessary to

control forresponse biasinourexperiment, sincetherewere36different

pairsandonly9sames.Itseems reasonable thatrepetition should reduce

latency. Thefinding thatsamesareshorter, ontheaverage, hasbeen

reported byothersbefore us(Nickerson 1965,Egeth1966), butsohas the opposite finding(Bindra, Williams, andWise1965,Chananie and

Tikofsky 1968). Bindra, Donderi, andNishisato (1968) compared same

withdifferentlatencies asa function ofseveral variables (stimulus modality,simultaneous versus successive presentation, intertrial interval, etc.).All

pairsofstimuli werepresented equally often, which appears todispose of

repetition asthesoleexplanation ofdiscrepancies. Theyconcluded thatthe discrepant findings weredueto codability (orlackof it) of thestimulus items.Letters, forinstance, arereadilycodable (identifiable by name)and

yielded shorter latencies forsames.Pairs oftonesdiffering inpitchor

pairsoflinesofdifferent lengtharenoteasyto identify absolutely, and theseitemsshowed longerlatencies forsames.Thisobservation, interestingasit is,seems unsatisfactory asanexplanation, because it givesno inkling ofwhysamesshould beshorter orlonger depending ontheir codability. It alsodoesnotexplain therangewithinsame distributions. Certainly oneletterisjustas codable as another, andyetdecisions are madesignificantly fasteronsomepairsthanonothers(andnotevenonthe

mostfrequent ones). Finally, theartificial graphemes inourexperiment

werefarlesscodable thantheletters, butthesame relationship heldaswith letters.

These rather baffling discrepancies suggest thattheprocess ofdeciding thattwothings arethesameisnotlikethatofdeciding thattheyare

different; thata simple modelthatsayscheckoutallthefeatures, either sequentially orinparallel, isjustnotappropriate forboth.It seemsmore likelythatundercertaincircumstances, a decision ofsame is a direct perception ofreplication; ofa structural property ofthepairasa wholethat requires nofurther lookatsubordinate features. Replication would bemore

readily detectable under some circumstances thanothersperhaps when simplicity andsymmetry arepresentbut theconditions forit arecertainly notwellunderstood. Inanycase, perception ofreplication canbea fastshortcut todeciding thattwothingsarethesame,ascontrasted witha feature-by-feature check. Thisinterpretation isbornoutbyanexperiment ofSekuler andAbrams (1968). Recognition ofidentity ofa pairintheir

experiment wasverymuchfasterthanfinding similarity, eventhough the similarity decision involved finding a same feature (anyfeature) intwo

pairs.It is thusnot thejudgmentof same as suchthatis faster,but the

432

E. J. Gibson, F. Schapiro, & A. Yonas

opportunity forimmediate apprehension of replicationGestaltprocessing as Sekularand Abramsput it. ClusterAnalysis of Different Pairs

Thequestion ofgreatestinterestto usiswhatislinkedwithwhatinpairs of differentitemsso as to makethemconfusableto differentdegrees.Can

wepulloutfromourdatasomeindication ofwhatthebasisfordifferentiation or confusionis? A hierarchicalclusteranalysis(Johnson)seemedto

offera promising method.Thisis, looselyspeaking, a methodof progressively clustering the set of letters.If we findsystematic differentiationin ourlettersetsandartificial graphemes, perhapswe canidentifythe featuresthataccountforclustering. If the sameonesturnup in replications of the sameletterset and withdifferentlettersets,we are in luck. Tables23.7athrough23.7cshowtheresultsofclusteranalysis oflatency matricesfor both sets of letters.ConsiderTable23.7a,basedon latency dataof 48 adultsubjectsfor the firstset of letters.Resultsof two methods, connectedness and diameter are presented.These are very similar,so

weshalldiscussonlyone,thatby diameter. Theanalysis pullsout in the firstrow the mostcompactand isolablecluster,PR;in the next row,EF

appears. Otherpairsappear progressively, untillongerandlooserclusters areleft,windingup withonlytwoseparated ones:CGEFPR on the one hand,andMNWon the other.Onecanthinkof the analysisthe other

wayaround, asa progression fromthetotalundifferentiated settoward moreandmorespecific clusters. A treestructure showsnicelythecontrasts

thatemerge whentheclusters appear. Atthefirstbranch, alltheletterson Table

23.7a

Cluster Analysis andTreeStructure SetI Letters, Latency (48AdultSubjects ) Connectedness

Diameter

Method

PR PR

CC

CG

CG CG CG

Method

PR EF

EF

PR

PR

PR

EF

EF

EF

PREF

MN

MNW MNW

PR

CG

EF

PR

CG

EF

PR

MN

CG

EF

PR

MNW

CG

EFPR

CGEFPR

PREFMNW CEFGMNPRW

CGEFPR

CG

MNW

EFPR

MNW

/ EF

PR

MN

W

MNW MNW

ConfusionMatrices for GraphicPatternsObtained with a LatencyMeasure

433

the right contain diagonals (MNW ) while those on the left have straight lines and/ or curves. At the next branch, MNW splits into MN versus W (all diagonals); the big cluster CGEFPRsplits into round letters, CG, versus letters with verticality, EFPR. At the next branch the cluster EFPRdifferentiates into a purely vertical -horizontal cluster, EF, versus one with curves and verticals , PRo

The children's latency data (Table 23.7b, 60 subjects) are very straightforward. The first branch is a curve-straight differentiation, all the letters with curvesbunchedtogether. Then the cluster with curves separatesinto the round cluster and the curve and straight cluster (PR). The right branch separatesinto a diagonality cluster and a vertical-horizontal cluster, etc., exactly as with both sets of adult data. The other set of letters yields similar contrasts, as Table 23.7c shows. Latency data for 24 adults are presented. Again the first branch contrasts

curved-straight, branchingagain on the left to round vs. curve and straight (OQ vs. D). On the right the last branch again yields a diagonality vs. vertical-horizontal split. The structure here is not as orderly as for the other set (the AS cluster looks strange), probably because there were only 24 subjects and a few confusions are accidental. There is a hint here of something that does not show up in other letter sets in the T vs. KXHAS

split. One might call it intersection or "information in the middle." Since the data in each caseare based on a sample of only nine letters, a new feature could easily turn up in the second set.

Sincewe do not have the complete matrix of all 26 letters, we cannot expect these analyses to generate all the features that would be necessary Table23.7b ClusterAnalysis - andTreeStructure

Set (60Seven -year -oldSubjects - 1Letters - , Latency - ). Connectedness Method MN MN EF CG MNW CG EF CGMNXW EFR

Diameter Method MN CG EF MN CG EF MNW CG PR EF MNW CG PR EFMNW CGPR EFMNW

CEFGMNPRW ,......--.....,,-- - -.....-..~~~------CGPR EFMNW ","""""""""""""'", "'/ """"-"""""""""~""""" CG PR EF MNW // """""""~""""'" MN W

E. J. Gibson, F. Schapiro, & A . Yonas

434 Table

23 .7c

Cluster Analysis and Tree Structure Sets 2 Letters , Latency (24 Adult Subjects ) u

' - ~ .. .

-

.... _

.

. -

~ -

,

-

-

Connectedness

OQ OQ OQ DOQ DOQ DOQ DOQ

to

vs

-

a

sets

of

J

, -

I

'

-

Diameter

;

pattern

vs

; .

open

attempt

experiments

of

features

but

is

on is

rather

vs

These

are

a

to

us

random

an

orderly

high

the

But

of

one in

;

first .

The of

.

;

.

in

abstract

a

relatively in these

pair

; and

a

the

intuitively within

difference

either

in

: curve

and

chosen replications

for fairly

see

structure

features

idiosyncratic of

easily

tree

intersection

distinguishing

hierarchy

can

the

features

perception

set

or

.

are

analysis

that a

letters

. horizontal

feature

detecting

not

26 that

vertical .

at

indicate depend

all

features

diagonality

earlier

for

certain

)

-

Method

OQ OQ KX OQ KXH OQ KXH AS DOQ KXH AS DOQ KXHAS DOQ TKXHAS

unique

( round

does

- -

shapes

. straight

closed our

-

KX KXH KXH TKXH TKXH AS TKXHAS

provide

two

-

Method

shape

or

properties

of

shapes

that a

the

set

perceiver

.

Conclusions

The

latency

range

for

of

times

between The

the these

pairs

. Two

slightly held children

being

significant

sets

reliably

as

significantly

well

for

nine

mean

both

letters

than

adults

shorter

Not

than

a

of

others

artificial

that

" same

for "

.

The

- year

confuse

that

as

pairs

.

seven to

set

gives

discriminated and

deciding

well all

characters

be

adults

deciding as

.

graphic to

tendency

and for

graphemes as

of pair

the

latency

shorter

artificial

pair

the

, reflect

of

. The

but for

on

, furthermore

trends

a

depending

are

latencies

pair

discriminating ,

- old the

pair

was

was

the had

reason

children

for

.

,

we

was

This ,

latencies

this

out

same

letters

equal

. of

bore the

different

familiar

wide

members

graphemes a

it

a

Differences

trend and ,

think

for some

,

is

Confusion Matrices for

Graphic PatternsObtained with a LatencyMeasure

435

that a judgment of same may be in some casesa direct perception of replication, without further analysisof features. There is reason to think, however, that a judgment of IIdifferent" for these graphic charactersinvolves an analysisof distinguishing features. A pair that differs by many features is seldom confused and the decision is

faster than for a pair sharing many features. Furthermore, hierarchicalcluster analysesof ~he ma~rices yielded tree structuresthat showed an orderly progressive differentiation of features. Latency data for children and adults

on the first set of letters yielded very similar structures. The latency data of the adults for the secondset of letters, suggestthat a very similar hierarchy of features is detected in the discrimination

process for both sets of letters .

We conclude that perceiving a difference between two letters is not a

matter of matching to a Gestalt-like template, or decoding to a name, but involves

de ~ection of distinctive

features .

References

Bindra , D., Donderi , D. C. & Nishisato , S. Decision latencies of "same " and"different " judgments . Perception andPsychophysics , 1968 , 3, 121 - 130 . Bindra , D., Williams , J. &Wise , S.S.Judgments ofsameness anddifference : Experiments on reaction time . Science , 1965 , 150 , 1625 - 26. Cattell , J. McK . The time perception asameasure ofdifferences inintensity . Philosophical Studies , 1902 , 19 , 63of - 68 . Chananie ,J. D. &Tikofsky , R. S.Reaction timeanddistinctive features inspeech discrimina tion . Report No . 49 , in a Program on Development of Language Functions , University ofMichigan , 1968 . Egeth , H. E. Parallel versus serial processes inmulti -dimensional stimulus discrimination . Perception andPsychophysics , 1966 , 1, 243 - 252 . Gibson , E.J,. Principles Learning andDevelopment . NewYork : Appleton -Century Crofts 1969 . ofPerceptual Gibson , E. J., Osser , H., Schiff , W. andSmith , J. Ananalysis ofcritical features ofletters , tested byaconfusion matrix . InFinal Report onA Basic Research Program onReading , Cooperative Project No. 639 , Cornell University andu.S. Office ofEduca tion , 1963 . Research Johnson , S. C. Hierarchical clustering schemes . BellTelephone Labs ., Murray Hill,NJ. Lemmon , V. W. Therelation ofreaction timeto measures of intelligence , memory , and learning .Arch . Psychology , 1927 , 15,No.94. Nickerson ,, R .,S16 . Response for"same " "different " judgments . Perceptual Motor Skills , 1965 20 - 18. times Sekuler , R. W. andAbrams , M. Visual sameness :A choice timeanalysis ofpattern recogni tionprocesses .Journal ofExperimental Psychology , 1968 , 77,232 - 238 . Sorenson , R. T. Guidance ofattention incharacter recognition : Theeffect oflooking for specific letters . Ph.D. dissertation , Cornell University , 1968 . Woodworth , R. S. Experimental Psychology . NewYork : Holt,1938 .

24

The Ontogeny of Reading Eleanor] . Gibson

By the time this paper was written, I had becomevery dissatisfiedwith an approachto readingthat seemedto me artificial and to be leadinginto bypaths that werenot going to take usfurther in discoveringhow peoplebecome skilled readers. It seemeda good idea to look at the way the skill beganand progressed in as natural a way as possible, borrowing a lessonfrom ethology where we could. This paper describesour attempts to do that. The first two studies, the one on

scribblingby ] . ] . Gibsonand P. Yonas, and Linda Lavine's work on children's developingconceptualization of what constituteswriting (Lavine 1977) aregood examples . Bothstudiesshowedthat childrengrowingup in a culturethat provides appropriate material and opportunities discovera great deal about writing , how it is constituted, and what its usesare by themselves . The moral seemedto be the

wisdomof providingyoungchildrenwith an environmentsuppliedwith plentyof literature of the kind one wants them to be curious about, and letting them explore a variety of reading material with adults who are ready to answer questionsand provide role modelsby reading themselvesand to the children. Head Start is in the right direction, but we seemnowadays to give children more opportunities to gaze

at television than at books.

A methodof naturalisticobservationbecomes moredifficult to implementonce childrenhaveembarkedon the routinesof schoolinstruction. We tried to apply the methodto finding what childrenknow, at progressivelevelsthrough the grades , aboutspellingpatterns.Someof thosestudiesaredescribed here.A further large-scalestudy was carriedout for her Ph.D. dissertationby Carol Bishop (1976). Shetestedchildrenin third and sixth gradesfor knowledgeof spelling patterns and correlated those results with several tests of reading progress. She found that children who are progressingnormally through schoolgenerally know somethingabout spelling patterns by late third grade, but individual differencesdo AmericanPsychologist , 1970, 25, 136- 143. Copyright 1970 by the American Psychological Association. Reprinted by permission of the publisher.

438

E. J. Gibson

weightfor overallreading ability. Something moreis not carry a lot of predictive needed.

It beganto seemto me that the crucial ingredientfor learning to read is motivation that drives a search for the information in text. More study of the learning processand better analysis of the information to be extracted were

indicated . An ethological approachisfine, sofar asit goes , but it is not a substitute for a gooddevelopmental theory. Onelessonseemed to standoutfrom theattempt to beethological , however : that onedoesn 't teachreadingto someone , ratherthat personlearns to read. We can only learn to read by engaging in reading, and we

haveto do it for ourselves . Motivating a child to do it (read) is oneplacewhere we oftenfail. -

-

-

-

J

-

-

- -

J

- -

Despite decades of concern on the part of educators, parents, and propo nents of homespun wisdom , we seem to know little more about how

to teach reading than our great grandparentsdid. In fact we do not even know why it has to be taught. Why doesn't it just grow, like language? No one teachesa child to speak. We do not know much about how a child

acquiresspeecheither, but in recentyearsstudiesof the developmental process

have

been

very

instructive

. I think

the

reason

for

this

is that

we

have begun to look at the processas a piece of natural history, somewhat as the ethologist looks at behavior. Observation followed by a careful analysis may be the essentialpreliminary to a good theory. Perhapswe have not really tried it with reading. I would like to consider, as a start, the natural history of the origins of

reading skill, before it is "taught" in a formal sense. Where does its deve-

lopmentalhistorybegin? I think we could agree that it has twin beginnings. One is its linguistic origin- it is a specialform of speech; and one is its origin in writing making marks on a piece of paper. The origin of reading in speech is obvious . Long before the child goes to school he has learned to segment a

sequentialstream of acoustic information; to divide it into valid units of structure ; to discriminate these units by means of an economical set of

distinctive features; to assignsymbolic meaningsto units of an appropriate size; to infer the rules that structurethe units in permissibleways; and even to recombine units in these rulelike ways so as to produce original messages. Surely this massive achievement must transfer in some way to the

perception of written speech, which is also processedsequentially. It, too, must be segmented, discriminated, assigned symbolic meaning, and its combination

rules mastered . That

there is a carry -over

is clear from

a

comparisonof hearing children with deaf children, who must do without this head start .

Ontogeny of Reading

439

I amgoingto leavetheearlyhistory ofspeech to people moreexpert thanI andbeginwiththeearlyhistoryof writingbehavior. It starts,in

James Gibsons opinion, withwhathecallsthefundamental graphic act.

It consists of producing visibletraceson a pieceof paperor someother

surface suchasrockorsand. Thetraces leftonrocksurfaces byprehistoric menaretheearliest known attempt at graphic communication; pictorial, to besure,notsymbolic, buttheevolution ofwritingseemsto havebeen

a gradual transition frompictures to symbols. Pictures havea projective

relationship to whattheystandfor,whilesymbols beara codedrelation-

shipto theirsurrogates. Ininfants, theearliest graphic actis notevena

picture,onlya scribble, but the childsscribbling is remarkable for one

thinghis intense interest inthevisible marks thatheismaking. It seems tomecomparable tobabbling, intheearlystages ofspeech acquisition, and themotivation forbothmaywellbebuilt-in andarisespontaneously when

opportunity is afforded.

Asearlyas12months, according to theCattell Infant Intelligence Tests (Cattell, 1960),a childgivena paperandpencilwillmakemarkson the paper.At 14months, hewillmakea definite scribble, thatis,a progressive, continuous tracing.Thetestersimplysays,Tommywrite. If necessary,

aftera fewminutes hegivesa demonstration. By18months, theaverage

childneedsnodemonstration, butscribbles spontaneously. At27months,

hecandrawa lineas distinctfroma scribble. At 30 months,he candrawa

line as distinct from a circular stroke.

It hasoftenbeenheldthatscribbling yieldssatisfaction simply as a

motoractivityas arm gymnastics,a mereexerciseof kinesthesis.Gib-

sonandYonas (1968) compared thiswiththealternative hypothesis that

whatmotivates thechildis theresulting newsourceof visualstimulation.

Theycompared thescribbling behavior ofchildren givena stylusthatdid

not leavea tracewith that whenthe stylusleft markson a surface.The childrenrangedin agefrom15to 38 monthsandwereobservedat home ina playlikesituation. Eachchildhada sessionwithbothtools,whichwere

identical exceptforthemarking potentialof one.Elimination of the trace

significantly reduced scribbling activity. Mean timeofvoluntary scribbling

was 72 secondswith the tracingtool as against21 secondswith the

nontracing tool.Some ofthechildren stopped immediately ondiscovering thatthe nontracing tool didnt work. One 16-month-old childhadno

previous experience withtracing tools.When giventhemarking stylus, she

wavedit andstruckthepaperuntila fortuitous markresulted, afterwhich

shebeganto scribble withgreatinterest andincreasing control. These children werenotonlyinterested inmaking scribbles, buttheolder subjects

werealsoeagerfortheexperimenter to lookat themanddemanded that

shedoso.Whatthechildcanseeinpracticing thegraphic actarevariable properties oflines, suchascurvestraight, verticalhorizontal, longshort,

E. J. Gibson

440

and featuresof graphic structure, such as intersection, closure, symmetry, and continuation

.

How does the child come to distinguish the important variables of graphic information? I have thought for some time that letters, like the phonemic information that they code to, are discriminable and convey information by virtue of their distinctive features; and that letters share a set of distinctive features that characterizes them as writing , contrasted

with pictures and scribbling. How early does a child begin to recognize "writing " as such? Must he learn to identify and name the letters first? At Cornell University, Linda Lavine is studying preschoolchildren's responses to simple line drawings, geometric figures, letters, numbers, words printed in upper and lower case, cursive writing , scribbling, artificial letters, and characters from strange scripts. The drawings , characters, and scribbles

were inscribed on 4 x 6 inch cards, inked in lines of approximately equal thickness. A few characters(Roman capitalsand some pseudoletters) were a little heavier. Size varied as it might in everyday printing or writing . Figure 24.1 shows samplesof the material. The children were merely told, "I'm going to show you somethings, and I want you to tell me what they II

are

.

Children of three and four from the Cornell Nursery School served as

the first subjects. All the three-year-old children showed that they could separate the pictures from the writing , the numbers, and the scribbles. The children of four could not only do this, they could separate the scribbles from the writing . However , in a cooperative nursery school, few of the

four-year-olds could distinguish scribbling from writing . Children in seven different kindergartens were tested. They varied markedly ; but in the

schoolwith the highest socioeconomicstatus, 75% distinguishedscribbling from cursive writing- , although they could not write yet. They identified all the printed characterscorrectly as letters, although in many casesnone or only a few could be named. They also frequently called the artificial letters and the samplesof other script (e.g., Hebrew script) writing . This is particularly interesting , because it substantiates the idea that there is something

categorical in the structure of writing that distinguishesit from random marks on paper and from pictures even when none or only a few individual characters can be identified . We intend to find out , in future experiments ,

what the visual information (the graphic composition) is that permits this early classification of a set. The distinctive features of individual letters -

those features by which

they differ from one another so that a unique pattern characterizes each

one- have been investigated in a number of experimentsby my students and myself. Gibson, Osser, Schiff, and Smith (1963) obtained a confusion matrix for the 26 Roman capitalsbasedon the errors of 90 four-year-old children. We compared the errors with a list (intuitively derived) of 12

Ontogenyof Reading 441

Figure24.1 Redrawnsamples of materialpresented to preschoolchildrenfor differentiationas"writing."

442

E. J. Gibson

features including verticality, horizontality, diagonality, curvature, opennessor closure, intersection, etc. The straight- curve and diagonality distinctions also emergedfrom a multidimensionalanalysisperformed for us by Warren Torgerson. More recently, Gibson, Schapiro, and Yonas (1968) have obtained confusion matrices for two sets of nine Roman capitals and for a set of nine artificial graphemes. In this experiment, the subjects made a same- different judgment. The data included both errors and latencies. Forty-eight adults and 60 seven-year-old children took part in the experiments. The results seemto me to confirm the hypothesis that letters are distinguished from one another by way of distinctive featuresthat are shared to greater and lesserdegreesby different pairs of letters, but that yield a unique pattern for every member of the set. Mean latenciesfor pairs of differentletters varied very significantly, by more than 100 millisecondsat extremes of the range. Latenciesand errors were highly correlated, and the degreeof confusability for given pairs ranked the samein a replication of the experiment. The artificial grapheme pairs also had a significant spreadof latencies. The meanlatency for discriminatingartificial graphemes was nearly identical with that for letters. This may indicate that although the artificial graphemeswere unfamiliar and had no names, we were successful in building into them the same set of distinctive features that characterizesletters as a set. Each of the matrices was subjected to Johnson's (1967) hierarchical cluster analysisto find out what underliesthe structure of the matrix. This should tell us what featuresare usedby the observerwhen he must decide whether a given pair is the same or different. The analysis looks progressively(in steps) for the most compactand isolableclusters, then for the next most compact and so on until it winds up with loose clusters and finally the whole set. The results can be turned upside down and diagrammedin a tree structure. Figure 24.2 shows the tree resulting from an

CEFGMNPRW ///,/ , , ~ CGEFPR MNW / / / /~ CG EFPR A EF

PR

""" MNW A MN

W

Figure24.2 Treestructureresultingfrom a hierarchical clusteranalysisof latencydatafor discriminating pairsof lettersby adultsubjects .

Ontogeny of Reading

443

analysisof 48 adult subjects' latency data for nine characterspaired in all combinations. The first split separatesthe letters with diagonality from all the others. On the left branch, the "round " letters, C and G, next split off from the others. At the third branch, the squareright-angular letters, E and F, split off from letters differentiated from them by curvature. The error data for theseletters with the samesubjectsreveal an identical structure. Now consider the hierarchical

structure for 60 seven -year -old children in

Figure 24.3. It is similar, but not quite the same. The first split is a clean curve- straight one. On the secondbranch, the round letters are split off from the P and R. The squareletters next split off from those with diagonality . This is very neat, and it suggests to me that children at this stage

may be doing straightforward sequential processing of features, while adults have progressedto a more Gestalt-like processingpicking up higher orders of structuregiven by redundancyand tied relations. This is speculative, but it would be a most adaptive kind of development, achieving the highest level of differentiation with the greatesteconomy.

Differentiatingletters, howeverfundamental , is still a very low-order aspect of reading skill. The heart of the matter is surely the process of decoding the written symbols to speech. We should know a great deal about how the writing -to-speechcode is learned. I am afraid that we do not, but we do know that it is not a simple matter of paired-associate learning, either of a letter to a sound, or of a written word to a spokenone. Sincewriting is alphabetic, why doesn't a child just pair eachletter with a correspondingphonemeand decodethe letters one at a time? For one thing he cannot, becausethe letter-to-sound correspondencein English orthography is not a one-to-one matching of eachletter to a sound element. For another, as researchat the Haskins Laboratory has shown us (Liberman, Cooper, Shankweiler, & Studdert-Kennedy, 1967), phonemesare not recognized by the hearer as invariant over a speech segment smaller than a

syllable, so sucha method would be impossibleeven with "regular" orthoCGPREFMNW ...-/ CGPR

/

A CG

/

" " """, EFMNW

/////"-"'"" PR

EF

MNW

A

MN

W

Figure 24.3

Treestructureresultingfrom a hierarchical clusteranalysisof latencydatafor discriminating

pairs of letters by seven-year-old children.

444

E. J. Gibson

graphy. The presumedIIelements " carry informationabout their context (andviceversa)aswell asaboutthemselves . Thisis truenot only of speech but of written languageaswell. Furthermore , a skilledreaderdoesnot readletterby letter. Cattell(1885) showedlong ago that a word, if not too long or unfamiliar , canbe read with asshortanexposuretimeasa singleletter. (For a five-letterword like START, for instance , one millisecondis enoughif contrastis good.) This findingledmanyeducators afterCattellto concludethat readingshouldbe taughtby the so-calledIlwholeword" method. But difficultiesarosehere, too. Many childrentaughtthis way couldnot recognizenewwords. They did not analyzecomponentrelationshipswithin words (Bowden , 1911). Research from our own laboratory(Bishop, 1964), andfrom others(Jeffrey & Samuels , 1967) hasshownthat knowledgeof componentrelationships within the word is necessary for transfer . I think the solutionto this apparentdilemmais to aim our researchat discoveringthe unitforming principles in readingactivity. Whataretherules, the constraints , the structurefor creatingunits? Only if we know this will we know how to teachso asto optimizeperceptionof the mosteconomi calunitsfor reading. Thereare many possibilities . The white spacesbetweenwords might seemto afford an obviousbasisfor segmentationinto words, at least. Carry-over from principlesof unit formationin speech , as I mentioned earlier, is anotherpossibility, and suggestsat leastthree levelsof rules: regularcorrespondence betweenclustersof graphemes andphonemes (often referredto asI'pronounceability " of a string of letters); morphological rules; andgrammatical syntax. Therealsomaybe orthographic rules , that is, spellingpatternsthat give structurewithout regardto the speechthey decodeto. We havebeguninvestigatingall thesepossibilities , andI would like to mentionbriefly someof the experiments . Considerfirst the white spaces betweenwords. We know that nearlyall kindergartenchildrentoward the end of their fifth year can recognize lettersaslettersevenwhenthey cannotidentifyindividualones. Canthey recognizewordsaswordsbeforeformalinstructionhasbegun?At Cornell, Lorelei Brushand Nancy Tither askedchildrento point out words in varioussamples of print. The childrenwereaboutsix yearsold, just finishing kindergarten . Nearlyhalf of thesechildrencouldnot point out words correctly, evenaftera demonstration . They sometimes ignoredthe spaces betweenwords and sometimeschoselettersas word units. It is not obvious, therefore , that the word is automaticallya unit for a child. Evenif it were, simplesegmentation of this sort would not provide the rules for whichwe arelooking. I havebeenespeciallyattractedto the ideathat thereis a carry-over to readingof unit-formingprinciplesin speech . Clustersof phonemes do map

Ontogenyof Reading 445 with considerableregularity to clustersof letters. Certain combinationsof soundsmay begin a word, for instance, and also are spelledcongruently in a consistent way in this position. These might be called pronounceable combinations, and it could be that pronounceability forms units for the skilled reader. There might be articulatory or acousticalencoding of what is read, so that the unit is formed on a kind of analysis-by-synthesis principle. Early in my researchon reading, I investigated this possibility, that is, that pronounceability of letter strings facilitates reading them becauseof the correspondenceof component clustersof letters with units of

speech . Gibson, Pick, Osser, and Hammond(1962) did tachistoscopic experiments with two kinds of letter strings , one that we called pronounce able and one that we called unpronounceable . The pronounceable ones

beganand endedwith consonantclusterspermissiblein Englishspeechthat also map regularly to spelling in those positions. The medial vowel clusters were also regular in their context. An example would be the nonsense word GLURCK . The unpronounceable control words were made by ex-

changing the initial and final consonantclustersso that the word was no longer a permissible sequence of sounds or letters in English, for example, CKURGL . The pronounceable combinations were read with far greater accuracy with short exposures than were the unpronounceable ones.

Although I replicatedthis finding many times, I cameto have misgivings about its interpretation. Several people suggested to me that the pronounceablewords merely had higher sequentialprobability of the letters, calculatedby summedbigram or trigram frequency. Was the correspondence with heard speech really important ? I decided to run the experiment

with deaf subjects. If it were the pronounceability of the letter strings that facilitated their reading, we should find an interaction with ability to hear and speak them. With the cooperation of the students and faculty at Gallaudet College, this hypothesis was tested (Gibson, Shurcliff, & Yonas, 1970). There was no interaction. The deaf subjectsprofited just as much in reading so-calledpronounceablecombinationsas hearing subjectsdid. The term "pronounceable" now seemsmisleadingto me. I believe the difference between the two types of words can be accounted for by rules of ortho -

graphy, that is, spelling patterns. I do not mean simply statisticalstructure, like transitional probability, but real rules suchas what consonantcombinations can begin a word.! There is a kind of grammar for letter sequences that generatespermissiblecombinations. This grammarmay be analogous to that devised artificially by Miller (1958) for his experiment on memory for redundant strings of letters , but it is natural to the writing system. At Cornell , Richard Rosinski and Kirk Wheeler

have found that most children

can differentiate permissible from nonpermissible combinations of letters

during the third grade.

446

E. J. Gibson

Morphological usage is another possible causeof units, either carried over from speechor found in orthography. Berko (1958) found that morphological transforms, such as the past tense, had been learned so as to transfer to new words or nonsensewords spoken to a child of four years. Shouldnot this be true for reading, too? If endingsthat mark tensesor verb forms carry over as units to reading, we might expect that a child could read a longer word, with a short exposure, when part of its letters are accountedfor by such an ending. A four-letter word is about as long as a third grader can read with a brief exposure. But if we expand the word by adding an ed or ing does this vocally well-known transformation create a larger unit that is easily detected? Lynne Guinet and I have been trying suchan experiment with third-and fifth-grade children. Table 24.1 showssamplesof the words we used. There were three kinds of stem words: real words, pronounceablepseudowordsand unpronounceable ones, both anagramsof the real words. There were three kinds of verb transformationsadded to the words; an 5 to form the third person present, an edto form a past tense, and an ing to form a progressive. We usedstems of varied length (three to six letters). No child was shown more than one form of a stem word. The main findings are very clear. The third graders make many more errors than the fifth graders, for any kind of word. Real words are read with fewer errors than pseudowords, and pronounceable pseudowords with many fewer errors than unpronounceableones. The longer the word, the more the errors. But would this be true if the length were increasedby a regular ending7 To our disappointmentwe found very little evidence that this method of increasing length increasedspan of perception. The evidencewas in this direction for the pronounceablepseudowords, but not for the other two types. We cannot say that there was automatic carry-over of sucha structural principle from speech. There was evidence, however, that the ending itself gains status as a structuralunit. Errorsin respondingto the transformedwords did not occur very frequently in the endings, as comparedwith nontransformedwords of equal length. But when there was an error in the ending, there was a significanttendencyfor it to be a substitution of anotherregular ending, for instance, ing for ed. This substitution of an ending happenedmore often for the transformed words than for the nontransformed ones, and the tendency was more marked for fifth than for third graders. It is as if the subject detected the ending first as a marker or feature added to the stem word, and only then identified it specifically. BecauseI have become convinced that rulelike information in orthography structures units for reading, I want to know how these rules or spelling patternsarelearned. Two hypothesesthat seemedmost reasonable to me were first that there is abstractionof an invariant pattern over many variable contexts; and second that a learning set develops for finding

I

Ontogeny of Reading

448

E. J. Gibson

Table24.2 DistinctiveClustersin InvariantPositionsandVariableContexts,All VariedAcross Problems SONG RING BANG HUNG

TEAM READ SEAL LEAN

CHOP CHIN CHAT CHUM

regular patterns in orthography . A training procedure was designed therefore to provide opportunities of both sorts. Gibson , Farber, and Shepela (1967) constructed a large number of " problems " that required the subject to sort positive from negative instances, like a simple concept-learning experiment . All the problems had four positive instances, each of which contained a cluster of two letters in an invariant position - initial , medial or final . The other letters always varied . Table 24.2 shows sets of the positive instances from three of the problems . The negative instances for each problem were words containing various similarities to the positive ones in order to control for sorting on the basis of a single letter or some other superficial similarity . The child was given one problem at a time . On the first trial , the experimenter laid two cards, one bearing a single positive instance and one a single negative instance, before him , and said, "This is mail for you , and this other [negative instance] is for someone else. You sort the mail and put all yours here [indicating positive card] and everyone else's here." The child then was given a randomly arranged deck of negative and positive instances to sort . The experimenter corrected him when he was wrong . He sorted for one problem four times, and then went on to another whether he had succeeded in a perfect sort or not , following a procedure much like that of a discrimination learning -set experiment . We first tried this training procedure on kindergarten children and first graders, running them on six problems a day for five consecutive days. This task was extremely hard for the kindergarten children . Only lout of 12 developed an indubitable learning set. This child could sort all the new problems correctly on the first sort by the fifth day . The task was somewhat easier for first -grade children ; about half of a first -grade sample showed evidence of developing a learning set to abstract common patterns of orthography . We wondered whether these subjects had really attained a set to search for invariant structure over a series of items, or whether more specific habits already had been learned in first grade that helped on this task. My student , Arlene Amidon , and I therefore tried an experiment comparing

Ontogenyof Reading

449

success onthespelling pattern problems withsuccess onanalogous pro-

blemsinwhichtheelements werecolorchipsinsteadofletters.

Subjects forthisexperiment werefirst-grade andthird-grade children, somerunfirstwithcolors andsomerunfirstwithletters. Forthefirst-grade children, colorandletterpatterns wereequally (andvery)difficult. Ifsuccessoccurred onthecolorproblem, however, it transferred tothespelling

patterns. Forthirdgraders, theletterpatternswerepickedupmuchmore

easily thanthecolorpatterns, andcolorproblems werefarmorelikely to besolvedif theyfollowed solution ofletterproblems thanif theywere

given first.

Weconcluded fromthisthata settolookforstructure canbedeveloped

andcantransferto newproblems, andthatthe abilityto detectstructure in letterpatternsimproves withage andschooling. It seemsthat thisis

morethanspecific learning, sinceit transferred to colorpatterns. Thethird

graders, I think, hadlearned mostly tosearch actively forinvariant spelling patterns. Following a taskin whicha childfoundthem,hecouldpursue

ananalogous searchforcolorpatterns.Findingstructure, I think,isrewarding,andwhenthesechildrenhadfoundit, theyrepeatedtheirsuccessful

strategy.

Howcouldwe assistthe firstgraderstowardsuccess in findingthe

structure,so as to facilitatetransferto new cases?Lowenstein(1969)com-

paredthreeprocedures ina newexperiment withfirstgraders. Onegroup

wasgivenno specialhintsor help,as in thepreceding experiments. One groupwasgivenspecific help.Whena problemwasfirstpresented, the experimentersaid,You willbe ableto findyour own mail,becauseall the

cardswillhavethesetwolettersonthem.Thepairofcommon letterswas

pointedout.Thiswasrepeated foreachproblem. A thirdgroupwas told,You willbeableto findyourownmailbecause yourcardswillhave

the sametwo letterson them. Butthey werenevertold whichletters.

Aftertwodaysofpractice, allthesubjects hada posttest setofproblems without any further instructions.

Thegroupgivenspecific helpmadeveryfewerrorsduringtraining,

evenonthefirstday.Thegroupgivennohelpmademany,although some improvement occurred. Thegroupgiventhegeneralhintmademanymore errorsto beginwiththanthe childrengivenspecific help(a meanof 6 errors), buttheyimproved steadilyandontheposttesthada medianof 0

errors; whereas errorsroseontheposttest toa median of5.5forthegroup thathadreceived specific help.Thiswasmorethanfivetimesasmanyas

theyhadmadeonDay1.Although 70%ofthesubjects inthespecific

groupmadeno errorson Day2, only20%madeno errorson Day3, the

posttest.But in the group given the generalinstructionto look for an invariantletterpattern,40%madeno errorson Day2, and60%madenone

on Day 3.

450

E. J. Gibson

Responselatenciesconfirmed these results. The specificgroup showed no changefrom Day 1 to Day 3, whereasthe group given the generalhint showed a steady and significant decreasein latency over the three days. We concludethen that first-grade children can easily sort words on the basisof presenceor absenceof two specificletters that have been pointed out to them. But it is not this ability as such that leads to detecting common spelling patterns across items. I think there must be a search for an invariant pattern and discovery of suchstructure for transfer of this kind of abstractionto new problems. Subjectsgiven no specialinstructions or hints may eventually accomplishthis on their own- 20% of the subjects in the control group madeno errors on Day 3 (the samepercentageas the "specific" help group on Day 3). But it is clearly better to have attention directed to searchfor invariant featuresin the stimulus array, and finding them is reinforcing; it leadsto repetition of the successfulstrategy and thus to consistently acceleratedperformance. This is the point that I wish to emphasizein concluding. Motivation and reinforcementfor cognitive learning such as speechand reading are internal. Reinforcementis not reduction of a drive, but reduction of uncertainty, specificallythe discovery of structurethat reducesthe information processing and increasescognitive economy. This is perceptuallearning, not just rememberingsomething. I realizethat my evidencefor this conclusionis asyet rather slender, and therefore I am aiming my researchefforts at obtaining better evidence and further clarification of the process. A current experiment is aimed at promoting discovery of semanticstructure and consequenttransfer of the strategy to a new problem. But that is another story, since we have only begun the experiments. I feel confident enough to predict, however, that the searchfor invariants and the discovery of structure are basic forces in cognitive motivation. If this proves true, methods of teaching that would promote efficient strategiesof perceptualsearchand detection of invariant order should be a first concernin instructional programs. Note 1. The results of this study were subjectedto a multiple regressionanalysisin which several summedfrequency counts (all that we could find) were tested as predictors. All turned out to be remarkably unsuccessful .

References Berko,J. The child's learningof Englishmorphology. Word, 1958, 14, 150- 177. Bishop, C. H. Transfereffectsof word and letter training in reading. Journalof Verbal Learning andVerbalBehavior , 1964, 3, 215- 221. Bowden , J. H. Learningto read. Elementary SchoolTeacher , 1911, 12, 21- 33.

Ontogeny of Reading

451

Cattell, J. M . Ueber die Zeit der Erkennungund Benennungvon SchriftzeichenBildem und Farben. Philosophische Studien, 1885, 2, 635- 650. Cattell, P. The measurement of intelligenceof infantsand young children. (Rev. ed.) New York: Psychological 1960. - Corporation, ~ Gibson, E. J., Farber, J., & Shepela, S. Test of a learning set procedure for the abstraction of spelling patterns. ProjectLiteracyReports , 1967, No . 8, Cornell University, Ithaca, New York .

Gibson, E. J., Osser, H., Schiff, W ., & Smith, J. An analysisof critical featuresof letters, tested by a confusion matrix. In, A basicresearch programon reading. (Cornell University and United StatesOffice of Education Cooperative ResearchProject No . 639) Ithaca, N .Y.: Cornell University , 1963. Gibson, E. J.I Pick, A ., Osser, H., & Hammondl M . The role of grapheme-phoneme correspondence in the perception of words. American Journal of PsychologYI1962, 751 554 - 570 .

Gibsonl E. J.I SchapiroI F., & Yonas, A . Confusion matrices for graphic patterns obtained with a latency measure. In, Theanalysisof readingskill: A programof basicand applied research . (Cornell University and United States Office of Education, Final Report, Project No . 5-1213 ) Ithaca , N .Y .: Cornell University , 1968 .

Gibson, E. J., Shurcliff A ., & Yonas, A . Utilization of spelling patterns by deaf and hearing subjects. In H. Levin & J. P. Williams (Eds.). Basicstudieson reading. New York: Harper & Row . 1970 .

Gibson, J. J.. & Yonas, P. M . A new theory of scribbling and drawing in children. In, The analysisof readingskill: A programof basicand appliedresearch . (Cornell University and United States Office of Education Final Report, Project No . 5-1213) Ithaca, N .Y.: Cornell University , 1968. Jeffrey, W . E., & Samuels, S. J. Effect of reading training on initial learning and transfer. Journalof Verballearningand VerbalBehavior , 1967, 6, 354- 358. Johnson, S. C. Hierarchical clustering schemes.Psychometrika , 1967, 32, 241- 254. Liberman, A . M ., Cooper, F. S., Shankweiler, D. P., & Studdert-Kennedy, M . Perception of the speechcode. Psychological Review, 1967, 74, 431- 461. Lowenstein, A . M . Effects of instructions on the abstraction of spelling patterns. Unpublished master's thesis, Cornell University, 1969. Miller , G. A . Free recall of redundant strings of letters. Journal of ExperimentalPsychology , 1958

, 56 , 484 - 491 .

25

PerceptualLearningand the Theory of Word Perception Eleanor] . Gibson

While I wasmakingattemptsat a naturalistic , ethologicalapproachto reading ,I was at the sametime trying to improveon a theoryof learningappropriatefor reading . What it is that must be learnedis the key question , and it has to be answeredbeforethereis any point in askinghow the informationis processed or what the learning mechanismsare. What is the information? Oddly enough, information processorsdo not spend much time on that question.

At the sametime, it is important to rememberthat readinghas a function, morethan oneof course , but like perception itselfreadingis a toolfor learning,for acquiring information. That both complicatesthe questionand simplifiesthe answer. One can analyze text and ask what kinds of information it contains, but

it is oftenthecasethat not all of it is wanted. Certainlyit oftenhappensthat not all of it is obtained , just as we seldompick up all the informationavailablefor perceptionof our surroundings. Reading, like perception, is selective. So, not only must we learn to extract whatever kinds of information are available, but we also must learn to selectwhat the text offers that is useful in any given reading task.

Thispaperis an exampleof the way I tried to combinea theoryof perceptual learningwith investigations of the way we learnto read. I would not now argue very hard for the successof my attempt to combine the two with a conceptof

distinctivefeatures , but lookingfor the kindsof informationin text was the right tack. I think it was alsothe right tack to emphasize that the informationin text has to be based on information in the world, in the final analysis, as in the

relationshipof semanticinformationto eventsoccurringin the world, evena child's world. The structure of text is nested, like events in the world. This paper

doesnotgo asfar asit might havein exploringthepossiblerelationships , but there hasbeenprogressin doingso sincethen. I think the main contribution of this paper is its analysis of the kinds of information in text. My preoccupationwith distinctive features led me to dwell on ~ AcademicPressCognitive Psychology , 1971, 2, 351- 367.

454

E. J. Gibson

features of words rather than text in larger units, but things were moving in the right direction. Earlier we had focusseda lot of researchon graphic and ortho-

graphicaspectsof text, but now we weremovingtowardsemanticand syntactic informationand combinations of them. An experiment on perception of morpholo gicalfeaturesof words(Gibsonand Guinet 1971) is an exampleof this trend. The inJ1uence of information processingis still evident in this paper, however.

It is suggested , for example , that the classesof informationare "processed independently and sequentially ." This ideanow seemsto me patentlywrong- not because "parallelprocessing " is in the air, but because informationis obtained , as opposedto processed , and the searchfor it is directedby the reader 's task. The hypothesis aboutdevelopmental shiftsis toorigid, also. Youngreadersdo lookfor meaning , only lessexpertlyandflexibly than theywill asskill increases . Thetasks that providedevidence for a developmental sequence of word "features " werenot natural reading tasks. But I am content with the emphasis on search for the information that has utility for the task, and the evidencefor it, even when the tasks are rather artificial ones.

I am especially contentwith theevidence presented that subjects , evenchildren , learn in the courseof a task what features have utility and shift toward the most

economical extractionof information. The conclusionin the last paragraphstill holdsgood- that word perception , like otherkindsof perception , is a searchfor relevant information and that the search evolves toward the most economical

strategythat hasutility for the task. Showingthat this principleholdsfor reading providesa bridgefrom everydayperceptionto a more"cognitive " process . ,

. ... "

-

-

.

.

Perceptuallearning involves the learning of distinctive featuresand higherorder invariants, learning progressing actively toward the most economical featuresand structure. Featuresof words are classifiedas phonological, graphic, semanticand syntactic. Featuresof these classesare processedindependently and sequentially. Ordering of priorities changes with development, and shifts strategically with the demandsof the task. Evidence is presented for priority differencesfor eachclassof feature depending on task differences.

This paper is the outcome of two long-time endeavorsof the author- the developmentof a theory of perceptuallearning, and a program of research on reading.! The aim is to try to show how the two are related. First, the theory of perceptual learning will be described, as briefly as possible. It attempts to answer three questions: First, what is learned? Second, how? What are the processesinvolved? Third, what is the motivation and reinforcementfor perceptualleaming?

Theory of Word Perception

455

What is Learned ? I believe that what is learned in perceptuallearning are distinctivefeatures of things, of representationsof things, and of symbolic entities like words; also the invariants of events that occur over time; and finally the economical structuringof both. I think the information for learning these is potentially present in stimulation, to be picked up by the observer given the proper conditions for it. Considersomeexamples. Setsof distinctive featurescharacterizeobjects and entities both natural and artifactual- the furnishingsof the world, such as people, dwelling places, things to eat; and, particularly relevant for the present topic, symbols written on piecesof paper, like letters and words. The set of letters of our alphabet is characterizedby a set of distinctive features, which in different combinationspermit a unique characterization of each one. My students and I have spent much time trying to describe the set of distinctive features that are shared by letters of the Roman alphabet. We have had some success , since confusion matrices obtained experimentally reveal, via cluster analysis, some contrasting featuresthat can be diagrammedin a tree structure (Gibson, Schapiro, & Yonas, 1968). What about invariants of events? These occur over time. The nicest examplesof learning to detect them occur in perceptualdevelopment. The constanciescan be understood as invariants under transformations that occur in an event like the approachingor recedingof an object; or rotation of it . Perceived existenceof an object despite temporary occlusion by a screenof some sort is another (the event of IIgoing in front of" or IIgoing behind"). What has this to do with words? Events like appearance , disappearance , and reappearancehave meaning , and these very meaningsare expressedspontaneously in the child's earliest two-word utterances (Hi Daddy; all-gone ball; more milk; ball again). Theseutterancesappearto be invoked by the event, quite literally mapped to it (Bloom, 1970; Brown, 1970). Detection of invariance, in other words, is prior to symbolic referential meaningsand is reflectedin them. Semanticfeaturesof words generally indicate perceivable features of the world, both events and things. Verbs for instancecan be classifiedas actionor state, an important semantic distinction arising directly from differentiating invariants in the environment. Nouns can be classifiedas countor mass , a distinction that depends on differentiating things that have fixed borders or shapesfrom substances like fluid, sandor water. Finally, what are some structures discovered by perceptual learning? Entities in the world, both natural and artifactual, have structure; that is, relationsbetweenfeatures. Theserelationscanbe subordinateand superordinate. Both superordinateand subordinate structure are progressively discoveredin development for both objects and events- the structure of

456

E. J. Gibson

events was referred to by Brunswik as the "causal texture of the environment." Examplesof the subordinate and superordinatestructure of words are obvious . Words corne in sentences; sentences are parts of para-

graphs; phrasesare parts of sentences . Processes of Perceptual Learning How are these things learned? Not , I believe, by associatinga response of any sort , or an image or a word , to a " stimulus ." Distinctive features

and invariants must be discovered- extracted from the multiplicity of information in the flowing array of stimulation. We have always accepted the notion of abstraction to explain the genesis of concepts. I think a process akin to abstraction -

the dissociation

of an invariant

from

trans -

forming or variable context- happens also at a perceptual level. The phoneme is segregatedfrom the flow of speechheard by the pre-verbal infant and its invariant

distinctive

features are abstracted

from many vary -

ing samplesand over many transformationsproducedby different speakers . The

process

has

to

be

one

of

abstraction

-

there

is no

response

to

be

associated , nor is there an identically repeatedunvarying stimulus. What happens to the variable, irrelevant components of stimulation? When something invariant is abstracted and selected out for attention ,

what happensto the rest? I think that the processof abstractionis accompaniedby a filtering processthat attenuatesand suppresses the irrelevantthis is what happensto the words that aren't heard in the dichotic listening experiments , for instance.

Finally, a very important processin perceptuallearning is the operation of mechanisms of external attention . The sensory systems are all active and

exploratory. "Looking," "listening," and "feeling" are terms that describe the search for information in stimulation . They also underline the fact that

perception and perceptuallearning are activeprocesses . There is improvement and flexibility in attentive strategiesdepending on age and on the business

in hand .

Motivation and Reinforcement

What selectsthe good strategy, the economicalfeature, the structure that most effectively orders the information ? This is the question of moti -

vation and reinforcementin perceptuallearning. I do not think the motivation is to be found

in drives like tissue needs , nor the reinforcement

in

reduction of a metabolic drive, nor in cessationof punishment. I think there is a built -in need to get information about one's environment . One could

call this, as a number of psychologistsdo, "intrinsic cognitive motivation." I can call it the " search for invariance ." This motivation is always related to

Theoryof Word Perception 457 the task in hand, for different information is neededfor different tasks. But active looking for information about the solid, permanentsafeplacesof the world, and the invariant aspectsof events (like the swift approach of an object) is essentialfor behaviors like locomotion and avoidance of predators. A desireto understandwhat others are saying seemsequally basic for learning to comprehendlanguage. Reinforcementof perceptual learning (indeed of all cognitive learning) is, I would contend, the reduction of uncertainty. Discovery of structure, or discovery of the economicaldistinctive feature, or of the rule describing an invariant reducesthe "information load," and leads to permanentperceptual reorganization for the viewer of the world so viewed. This kind of discovery also leads to repetition of the successfulstrategy when a similar occasionoccurs. The evidencefor this is slight at present, but I have been conducting experiments to see whether discoverv - of structure is indeed reinforcing. I think the experimentsI am going to cite as illustrations of changing strategies in selection of word features will show the self-regulating, adaptive characterof the process. How Are WordsPerceived ? N ow we are ready for the question of how words are perceived. I shall concentrateon how they are perceived in reading, since that is where my researchhas been centered. The answer sounds simple, and is not new. Words, like other entitiesin our environment , are perceivedby detectingtheir distinctivefeatures .2 But what are a word's distinctive features? Features of Words A word is part of a vast system of information. The way to identify this information is to refer to what is learned when we learn to hpar and speak, and to read and write. The information dealt with in hearingspeakingis traditionally divided into three aspectsor classes , phonological, semantic and syntactic information. There are three parallel aspectsfor reading- writing ; graphological, semanticand syntactic information. These classesof information tell us what kinds of featuresa word can have. When written, it can have graphologicalfeatures of many kinds. For example, it has a characteristicshapeand length (referred to as "word shape" by reading teachers). Within the word there are letter shapes , themselvesdifferentiated by distinctive features. And then the word has orthographicstructure; letters are combined into words according to a rule system so that given combinations or clusters are permissible only in certain locations and contexts. Q must be followed by U, for instance; Qu can begin a word or a syllable, but it cannot end them (Venezky, 1970).

458

E. J. Gibson

How

do

being

we

know

detected

that

as

produce

errors

can

like

"

features

about

spelling

A

has

of

aware

( 1962

) . We

were

not

letters was

final

that

the

, quite

profoundly

be

a

word

. Then

is

were

presented

poetry

, we

that

vastly

affects

, and

Hammond

began

were

its

they

with

initial

terminated

with

strings

exchanging

the

construction

of

fact

, though

they

, and

by

the

. Homo . The

unpronounceable

ones

example

by

read

, asser

).

perceived

instance

pronounced

so

a rule

confused

, Pick

could

end one

we

pronounceable

them

not

begin

. An

Yonas

,

it

could

were

initial

and

CKURGL

"

from

onomic

. semantic

.

It

antonyms

show

to well

like

. We

know

Osgood word noun

. It

these

in

can

' s semantic 's

syntactic

or

verb

.

or

for

for

are

scales equally

adjective

. It

has

word

like

of

up

within

is ,

various "

word

,

tax

-

or

dis

-

one ,3

and

include

- unpleasantness

nouns to

synonyms

because classes

, from

looming that

that

but

, I ' proper

-

a

,

belonging

picked

of

later , Schurcliff

markers

events

features

are

a

nouns

relations

pleasantness

features

"

to

orthographic found

objects

confusible

Semantic

rated

- differential

or

,

for

due

articulatory

( Gibson

have

mixed

been of

we

of

II animate

stand

contrast

are

good

, for

may

features

they

recall be

It

, like

things

by

features

.

may

and

that "

Words

properties

- up

a

accurately

have

pick

advantage

potential

features

edible

similarity

II cluster .

all

nouns

has

experimentally

tend

or "

influenced

relative

more

might

easier

features

same

are

classes

categories

appearance

the

many

, " mass

or

been

in

version

advantage ;

phonological

showed

classes

nouns

. The

have

) . These

indicating

" count

also

tachistoscopically

pronounceable

structure

from

Ss

has

the

acoustic

apart

1970

word

kinds

read version

of

feature

A

- strings

- up

deaf

different

like

is

also

, 1970

is

, for

sometimes

Gibson

pronounceable

letter

pick

structure

A

could

unpronounceable

. But

on

, are

visually

- words

invariably

structure

as

differently

,

can

discovering

it

. When

that

palindromes

( Gibson

when

rhymes

the

being

infer

. We

, like

the

are

.

the

and

in

tasks in

O . We

cases

, even

up

that

confuses

new

hearing

constructed

not

clusters

these

easier

and

by

similarity

that

, subjects

an

perceived

. We

could

to

features

pseudo

from

When

than

!l ; ical

up

consonant

order

it

set

relational

generalizations

" words

clusters

GLURCK

accurate

presented

child

actually

confused

and

something

transferring

spelled

E

is

to

is shapes

not the

word

been

letter

but

when

a

what

of F

by

which

constructed

and

has

study

features

demonstrated

made

consonant

to

generalizing

phonolo

of

method

then

E

at

are

feature

useful

the

acoustic

of

potential

perceived

is

being

. This

" real

clusters

He

and

of

they

a string

readibility

is

."

potential -

, albeit

that

of

looking

instead

A

confuses

patterns

keenly

phones

and

some

" was by

word

reading

?

structure

and

study

feature

child

graphic

" saw

are

a

a

errors

infer

when

a larger

a

some

. We

noticed

whether

,

can they

values or

ranked

of

speech

,

. obvious a

role

in

. It a

is

sentence

a

part , like

subject

, or

Theory of Word Perception

459

object. And it may possessa morphological marker, like pluralization or tense. This last feature, morphological information, was investigated in an experiment by Gibson and Guinet (1971) . We wondered whether the length of a word correctly perceived tachistoscopicallycould be increasedby adding a well-known inflected ending to a baseword, as compared to an uninflected word of equal length. Are the endings themselves a unitary structure? We addedinflectedverb endingsto three types of basewords; realfamiliar words, pronounceablepseudo-words that were anagramsof them, and unpronounceablepseudo-words, and provided comparisonsof uninflected words of different lengths. Thesewords were shown tachistoscopicallyto subjects from third grade, fifth grade, and the elementary psychology course. The results were not what we had anticipated, but they were very illuminating. Subjectswere not able to perceive a longer word correctly when a base word was expandedby adding an inflected ending. But the endings gave evidence of unitariness, for there were significantly fewer errors in inflected endings than in endings of base words of equivalent length (especially when these were not meaningful or pronounceable). Furthermore, when there were errors in the inflected endings, the errors tended to be substitutions of other inflected endings. There was a confusion within a morphological feature-class. These latter two tendencies increasedfrom third to fifth grade, showing progressivepick-up of syntactic features. Why wasn't a longer word perceptible when the syntactic marker was added? Becausethe subject had to processan extra feature. It is features that must be processed , not elements like letters or syllables, and the processtakes time. No matter how long the baseword, its featuresmust be recognizedand so must the morphological tag. ThreeHypotheses I would like to suggest three hypotheses about how these features are perceived in reading. First, I think the four general classesof features, phonological, graphological, semantic, and syntactic are processedindependently and sequentially, in a kind of hierarchy. If presentation time is cut short, a feature low in the hierarchy may be missed. It may not be deleted, but just not fully detected, so a confusionwithin its feature-class may result (like substituting inKfor ed). Proofreader's errors are a caseof not fully processingan orthographic feature; but as a matter of fact, spelling is generally noticed in reading, when the time is not limited. Somethinglooks odd and we go back after a sentenceor two and verify the mistake. A classicpiece of evidencefor ordered, independentprocessingof word featuresis so-called "semanticsatiation/' or "loss of meaning," where presentation time is exaggeratedlyprolonged, instead of cut short. If I put a

460

E. J. Gibson

printed word in front of you and tell you to stare at it for five minutes, its "meaning" is said to slip away. What happensis that first the semantic featuresgo, then the phonological featuresgo, then one is left finally with the graphic featuresonly, and even these will eventually fragment. There is a very interesting implication here. Meaning, for an adult reader, is embeddedin the word. He doesn't begin by decodingit letter by letter; the concept symbolized by the word "hits him." It is specified for him in stimulus information. The Stroop test is further evidenceof this (Stroop, 1935). In this experiment, two featuresare put in opposition, meaning and a graphic feature, the color of the ink the word is printed in. The subject is shown an array of words and askedto name the color eachword is printed in, going from left to right as in reading- green, red, blue, and so on. But the words themselvesare the namesof colors. When the name and the color of the ink conflict, the subjectis in trouble. The word comesfirst to mind and his performanceis badly slowed up compared to just giving the name of a color patch. There seemsto be less interferencein this task with the very young readers. The meaning isn't yet as firmly embeddedin, or specified by, the word on the page. The second hypothesis is that there is a developmental change with age and schooling in featureanalysisand pick-up. At an early age, phonological features of a word seem to have more control, in the sense of yielding greater generalization, than semanticones. Riess(1946), using a conditioned GSR technique, found greater generalizationat eight years of age to homonyms than to synonyms, but by adolescencethe situation was reversedand semanticsimilarity becamemore effective. Perhapsthe younger Ss simply had less knowledge of similarity of meaning. Rice and DiVesta (1965) controlled for this in an experiment using a paired associates method, making sure that the homonyms, synonyms, and antonyms used were recognized as such by the subjects. In the younger children (third and fifth grades) generalizationoccurred as a result of phonetic but not semantic similarity. But semantic generalization becameincreasingly apparentin older age groups. Felzenand Anisfeld (1970) confirmed these findings using a continuousrecognition method. Children in third and sixth grade listened to a list of words and judged for eachword whether it had appearedbefore on the list. Falserecognition of phonetically related (rhyming) words was more frequent for third graders than false recognition of semanticallyrelated words, but semanticsimilarity was more effective in producing errors for sixth graders. We can find further evidence for developmental shifts by examining reading errors at progressive levels of instruction. Errors in oral reading through the first grade were studiedby Weber (1970) and Biemiller (1970). The earliest errors, at the beginning of first grade, can generally be attri-

Theory of Word Perception

461

buted to meaningful context. When the child reachesa word he does not "know," he usesall the semanticinformation at his disposal(context of the words already decoded, pictures, and so on) and guesses . He produces a word that makessense, both semanticallyand syntactically, but bearsno resemblanceotherwise to the one on the page. A little later the child stops when he reachesan unknown word and simply says nothing. This is a transition to a stagewhere errors becomedeterminedby graphic similarity. The child is engrossedby discriminating letters and by correspondenceof letters and sounds. Semanticfeaturesof the word are temporarily lowered in priority as the child strives to "break the code." (This is the period when he may be chided by the teacherfor "reading without expression," but the stageneverthelessmarksprogress.) Semanticfeatures return as the decoding process becomeseasier and the orthographic features demand a lesser share of the child's attention. But the orthographic and syntactic rule systemsprobably do not operate fully as important structural constraintsuntil later. The influenceof orthographic structure begins to be quite apparent by third grade. Besidesthe evidencefrom tachistoscopicexperiments(Gibson, Osser, & Pick, 1963 ), it has been shown by Rosinski and Wheeler (personal communication) that third graders can judge correctly over 80% of the time which of two pseudo-words is "more like a word." First graders, however, make this judgment at a chancelevel. In the Gibson and Guinet experiment (1971), morphological inflections suchas verb endingswere found to operateas unitary featuresof a written word. This function was more apparentin the fifth graders than the third graders, and most apparentin college students. Finally, syntactic constraintslike phrasestructure and grammaticalconventions- the word's role in the sentence- have been shown in studies with the eye-voice span to operate in reading, and again this rule system increasesin usefulnessas reading skill progresses , being noticeably more functional in fifth than in third graders. The influenceof phraseboundaries, for example, only shows up after grade 2 (Levin & Turner, 1968) The young readerdoes appear, then, to show a developmentalsequencein the pick-up of word-features. This progression is no doubt not as fixed as I may have implied and probably begins before very long to be influenced by the reader's task, in accordancewith my next hypothesis. My third hypothesis holds that the order of pick-up of word-features changeswith the task. To put it a little differently, priorities of pick-up are geared strategically to task utility . I repeat now my earlier argument that perception is an active search for information, that the perceptual strategy that develops will be as economicalas possibleand that ordering of priorities is adaptive and self-regulating. Common sensesuggeststhat this is so- when we are looking for a weather report in the newspaper,

462

E. J. Gibson

we assign a very low priority to graphic features of the words we read. But in addition there exists a large number of experiments which go to prove my point , some of my own and many by others .4 We can influence priorities by instructions , of course. In a tachistoscopic experiment , if a subject is told to guess at words shown him , features like

meaning and word frequencyare evident. If he is told to report only what is literally seengraphic featuresare advancedin priority (Haber, 1965). But quite aside from instructions, different tasks seem to have acquired their own priorities in the cognitive economy either in the course of development or by learning during the task. All a word's features have their importance for one task or another .

Phonological features, we have learned in recent years, have a high priority for short-term memory. If I am trying to hang onto a telephone number and someone speaks to me before I have dialed it , it is lost . Conrad (1964) and others have shown in a number of experiments that

acousticsimilarity producesconfusionsin short-term memory even when the material is presentedvisually (Baddeley, 1968). Graphic and semantic confusionsin short-term memory, on the other hand, are infrequent (Baddeley , 1966). Auditory presentation has an advantage over visual presentation in short -term memory

(Murdock , 1967 ), even when memory

is tested

by recognition (Murdock , 1968). But as lists are made longer (e.g., 14 vs 7 items) acoustic confusability effects disappear entirely (Anderson , 1969); Ss

no longer give the acoustic feature priority , but adopt some other strategy.5 Articulatory similarity may playa role in short-term memory tasks,

as well (Crowder& Morton, 1969). I shallclassthis with phonological information, but it undoubtedly plays its own role, distinct from acoustic similarity , in some tasks.

A word's pronounceability, I suggested earlier, is one of its phonological features, and it strongly facilitates pick-up with tachistoscopicallypresenteddisplays. Does it do so equally for another task, like later recall or recognition? Gibson, Bishop, Schiff, and Smith (1964) tried to separate pronounceability and semantic referenceand to compare their effects in these two kinds of tasks. T rigrams were preparedwhich either rated high

in pronounceability(like MIB), or in referentialmeaning(like the initials IBM), or in neither (like MBI). In one experiment, they were presented tachistoscopicallyand recognition thresholds were obtained. Pronounceability very effectively facilitated accurate perception of the trigram . Mean -

ing helped little .6 In another experiment, the sameitems were presentedto subjects for 2 sec each and 24 hours later recall was tested. This time the

effect of meaning and pronounceability was reversed. Meaning facilitated

recallfar morethanpronounceability andtherewasevidenceof categoriz ing , for some subjects made up sets of initials , like FOR, that had not been in the list .

Theory of Word Perception

463

While phonological features of words dominate pick-up in short-term memory , there is considerable evidence to show that they have low prior -

ity for long-term memory. Acoustic similarity has little or no interfering effect in a paired-associateretroactive-inhibition paradigm (Dale & Baddeley, 1969; Bruce & Murdock, 1968), whereas semantic similarity does. Wickens, Ory , and Graf (1970) found that acoustic similarity had some negative effectsin a transfer paradigm, but the effect was slight compared

to semanticsimilarityin the sameparadigm . Sharingtaxonomiccategory membership wasa powerfulinfluenceon the subject 's ability to recallitems of a list, for either good or ill dependingon task relations. Concrete words are in general superior to those of abstract meaning

in almost any long-term memory task, such as PA learning, free recall, serial recall, or recognition (Tulving & Madigan, 1970, p. 452), but this

is not true for tachistoscopic recognition , wherepronounceability hassuch

a strong advantage(Paivio & O'Neill, 1970). The utility (and utilization) of semantic features of a word for later recall has often been demonstratedby evidenceof clustering in long-term recall. This was brought out cleverly in an experimentby Hyde and Jenkins (1969). In this experiment , the subjects were sometimes asked to do two

tasks at once. They were presentedwith a list of words for later recall. In two conditions, Ss had to extract some graphic information about a word as it was presented- either estimate its length (number of letters), or detect the presence or absence of the letter E. Another

group

had to rate

the word as it was presentedfor pleasantnessor unpleasantness . Compared to a control group with no second task, recall was greatly reduced for the first two groups

and so was the amount

of organization

in recall as measuredby clustering of words in categories. But the task of rating words as pleasantor unpleasantdid not reducerecall nor organization in recall as compared with a control group that had no incidental

task. Whenthe subjectwasperforminga secondtaskthat gavepriority to semantic features of the word , neither recall nor its organization suffered.

But when the second task required attention to a word's graphicfeatures, like detecting E's or estimating word-length, the semanticpick-up which is apparently vital to later recall of words was blocked . The Hyde and Jenkins experiment suggests that features of the same

class, like semantic features of all kinds, are picked up together, while different feature-classesare processedsequentially(though probably overlapping one another). The value of the word- pleasantor unpleasantcould apparently be assessedat the same time as pick-up of semantic categoriesof the kind that operate in clustering. I will consider this again in connection

with a different

task , visual search .

Over the past five years, I have been particularly interested in visual search tasks, partly becauseI am interested in how perceptual search

464

E. J. Gibson

develops in children and also becausethey offer a good opportunity for studying what the subject learns incidentally. A lot of perceptuallearning

goeson duringthis task. Visualsearchalsoprovidesa fineopportunityfor comparing pick-up of different types of word-features. I have made such comparisons in a number of experiments . The task is similar to one used

repeatedlyby Neisser(1963) and involves scanningsystematicallythrough a matrix of letters for a target letter or word . When the subject is asked to search for a target letter embedded in a

context of other letters by scanning down a column of letters arranged five or six to a row, he very quickly sets his priorities for graphic features. Even seven-year-old children do this. Gibson and Yonas (1966) compared the effect of high and low graphic similarity of context letters to the target letter. Graphic similarity slows the scanning rate enormously, both for adults and children . Given an opportunity for practice, the subject wilileam

to scan for a single very economicaldistinctive feature, as Yonas (1969) and Schapiro(1970) have shown in appropriate transfer experiments.

What aboutphonologicalfeaturesof the lettersin this task? They seem to be virtually , if not entirely ignored. Gibson and Yonas (1966) tried to produce interferenceby exposing the subject to a voice pronouncing letters that sounded like the target letter while he scanned the list visually . There was no effect at all on scanning rate , even in children of seven years

who might have been expected to subvocalize while reading. Kaplan, Yonas, and Shurcliff (1966) comparedthe effect of high and low acoustic similarity of context letters to the target . That is, the target was embedded

in a context of letters that rhymed with it (B, V, 0 , and so on) or in a context of letters that sounded unlike it . Acoustic similarity did not slow

scanningrate at all, in contrast to a powerful effect of graphic similarity. An experiment by Krueger (1970) found some effect of acoustic confus-

ability in visual search,but it is possiblethat acousticwas confoundedwith visual confusability (for instance, C and G, and M and N have both types of confusability and were used within the same target -list set) . In the paper

by Kaplan et ale(1966) this factor was controlled. Changing or maintaining target items from trial to trial (Krueger changedthem, while Kaplan et ale did not ) is also an important task variable , since practice with a target or target set to be discriminated visually from another set (Yonas, 1969) is very effective in reducing search to the most economical visual distinctive feature .

Subjectstypically remembercontext letters in the scanning task very poorly, even when they are tested by recognition with the letters presentedto them visually (Schapiro, 1970). I wondered whether introduction of somestructure of a higher order in the context would not bring it to the

fore perceptually . If subjectscan learn to take advantageof the most economicalpossiblesingle graphic featurethat distinguishesthe target, will

TheoryofWordPerception 465

theynotalsodiscover andusesuperordinate orthographic structure ifitis

present?

Gibson, Tenney, Barron, andZaslow (1971) performed anexperiment inletter-strings which, though notmeaningful, wereorthographically posin whichwecompared scanning ratefora lettertarget,eitherembedded

siblewordsandwerepronounceable; or withthetargetembedded in strings ofthesamelettersscrambled soasto beunpronounceable. I had

thought thatorthographic structure might bepicked upalong withgraphic because thesubject could filter theirrelevant context inlarger unitsstrip it offinbiggerchunks, so to speak. Ontheotherhand, ifthesubject subvocally articulated thepronouncestructure, whichis so salientin thistask.If so,it mightfacilitate search,

ablecontext items,scanning rateoughtto beslowed down.(Ihavebeen tryingtofindoutforalongtimewhether redundant orthographic structure andpronounceability arenecessarily functionally tiedmy bŒte noire,in

fact.) Weran76children inthefifthgradeinthisexperiment. There wasno

difference in meanrateof scanbetweenthe twoconditions. Whatwas

happening? It would appear thatthechildren werenotarticulating the pronounceable items. Like lower-order phonological features, pronounceability ofa letter-string initsliteralsenseofpronunciation seems notto

influence thekindofverbalprocessing thatgoesoninthistask.Butwhat abouttheorthographic structure assuch? Doesitgounnoticed orcanitnot be usedwithoutan accompanying act of articulation thatwouldbe uneconomical? I thinkthelattermaybethecasehere.Whenthechildren were

questioned aftertheexperiment theyappeared, when context strings were spontaneously. Butstillit didnotaffect themeanscanning rate.Thechild pronounceable, tohavebeenawareofit.Sometimes theycommented onit

couldof coursehaveprocessed the wholestringas a unit,andthen searched it forthetargetletterasa secondstep.Thiswouldcontradict the

suggestion Imade earlier, thatagivenclass ofwordfeatures getsprocessed simultaneously. It is alsocontradicted by recentexperiments ofReicher

(1969),Wheeler(1970),andKrueger(1970).

Ontheotherhand,thechildren mayhavefoundtheorthographic

structure earlyin the task(sincetheyso oftencommented on it),but learnedto disregardit, perhapsbecausetheycouldnot use it without

articulationa handicap ina scanning taskwhere speed isemphasized. Thiswould bea kindofperceptual learning involving aninhibitory or filtering process, oneofthethreeprocesses forperceptual selectivity thatI hypothesized earlier. Wethought thatadultreaders might usethestructure

without articulation andwould speed uptheirscanning rateeventhough thechildren didnot.Further experiments didnotconfirm thispossibility. Whatwouldhappenif semantic structure wereintroduced in the scan-

ningtask?Wouldit be picked up,whenpresented incidentally? Would

466

B. 1. Gibson

it speedthe search, andif so wouldit transferas a strategythat is,

lead to a search for similarstructure in a second task? These questions

wereinvestigated in anotherexperiment (Gibson, Tenney,andZaslow, 1971).We introducedsemanticstructureby buildingcategorical meaning into the contextto be searchedthroughin lookingfor a targetword.

Contextwords,in one condition,were all the samelengthand all

belonged to thesamesemantic categorye.g.,kindsof fruit.Weused categorical material thathadpreviously beenshownto clusterwellin recallexperiments. Thetargetwordwasthe nameof an animal, andit

variedfromtrialto trial.The S was told to searchfor the nameof an animal,ratherthanfor a specified word,becausewe foundin preliminary work that S searchedfor nothingbut graphicfeaturesif he was given a

specific targetword.Wewantedto facilitate pick-up of thecategorical relationamongcontextwords,if it couldbe done.A controlgrouphad contextwordschosenat randomasregardsmeaning,but equatedwiththe otherconditionfor frequency, length,and as manygraphicattributesas possible.

Theresultsof thisexperiment ruledout unequivocally the possibility

thatcategorical meaning playsanyroleat allin a searchtaskof this sort. Semanticstructureof the kind we introducedis evidently not an economicalfeaturefor a searchtask when the words are presentedin a

list,as unconnected prose.It did not speedup scanningrate and was notusedby theSs.ManySsdidnotevennoticethecategorical relation withinthe contextalthoughtheyweretold to use categorymembership for locatingthe target.Whatthey actuallydid,it turnedout, wasvery economical indeed.Theylooked,aftera few trials,not for a wordof

anyparticular meaning, butsimply foranycombination of lettersthat had not appeared before.

Subjectsdo learnin this task.Theylearnthe strategythat is most

economicalfor the task.The conclusionis inescapablethat semanticfeatures of wordshave littleutilityfor a searchtask of this type and are

ignoredinfavorofgraphicfeatures thatdo.It isnotthatstructure isnever

utilized; graphic structure inthesenseofa redundant graphic feature that

helpsdifferentiate thetargetfromthebackground ispicked upandusedas Schapiro (1970) hasshown. Butperceptual strategy inthistasksetsthe

graphicfeaturesa highpriority,andotherfeaturessemantic, acoustic, even orthographic, low.

What happensin a searchtask if wordsare presentedin a passage

of connected proseandthe S askedto searchfor semantic, or graphic orphonemic targets? Cohen(1970) triedthiswithallthreetypesoftargets andwithcombinations of them.Thesemantictargetwasa wordof a given

category (e.g.,ananimal)infact,10different wordsbelonging to the category wereto becancelled ina meaningful paragraph. Inthissituation,

Theoryof WordPerception 467

search forthesemantic feature wasfaster thansearch fora graphic feature (aletter), andbothweremuch faster thansearching fora phonological

feature. Thesemantic feature hadtheadvantage overtheothertwoof redundancy withsyntactic andtopical predictability, butforadults itisnot surprising ifmeaning isdetected earlyinreading a meaningful connected

passage, especially whenSissearching fora member ofa concept group

andnota specific word.Theeffect ofevensmall changes intaskvariables

in shifting featurepriorities is impressive andiswitness to theremarkable adaptiveness ofhuman linguistic processing. Onemayask,finally, if thereis anytaskwheresyntactic features of wordshavepriority.In laboratory tasksas faras I know,thishasnot

beenthecase,although theyarecertainly picked up(Gibson &Guinet, 1971). A featurelikepart-of-speech (nounvs verbor adjective) is not veryeffective forproducing clustering in recall(Cofer &Bruce, 1965),

generalization inverbal conditioning orinterference inshorttermmemory (Wickens, Clark, Hill,&Wittlinger, 1968). Thatmaybe,onecould object, because syntactical information isgenerally spread overa string ofwords

andis incomplete in one.Butalthough a wordalwaystellsus morein

context, it appears thatitsdifferent features arepicked upindependently anddifferentially. I findthatthereis onereal-life taskin whichI seem, willy-nilly, togivefirstpriority tosyntactic information. Thatis,reading students papers, ora thesis. Asplitinfinitive, ora singular verbfollowing

theworddata distracts mesothatI losethemeaning! Conclusion

Doesperceptual learning occurinwordperception? I thinkifdoesboth

during development andwithin a task, without instruction orevenappar-

ent intention.Wordscontainmanykindsof information, andwe learnto

perceive themasacomplex offeatures. Words should notbethought ofas coded intosomething, butrather asentities possessing information classifiable asphonological, graphic, semantic, andsyntactic features. Allthese madeup of elements of a givenlength,or as bitsof sense-data to be

kinds ofinformation areinwords, I think, inthesense thataword specifies itsinformation. Itisnotconstructed byputting together letters oradding

on associations, but mustbe foundby discovering invariantfeaturesand relations. The perceiver does something, indeed, but he doesnotinventthe information.

Wordperception, likeotherkindsofperception, is active, searching fortherelevant information instimulation. Perceptual learning withwords,

likeotherexamples ofperceptual learning, develops toward thestrategy thatismosteconomical. Thismeans thatpriorities forfeatures shiftadaptively, withpractice ina task,toward those thathavemostutility forif.

468

B. J. Gibson

Notes

1.Invitedaddress forDivision 3,American Psychological Association, 1970.Thisworkwas

supported inpartbyagrant(USOE OEG-2-9-420446-1071-0I0) totheauthor fromthe Office of Education.

2.Thenotionthata wordis essentially a complex of features hasbeensuggested and

explored previously byAnisfeld andKnapp (1968), Bower (1967), Fillenbaum (1969), KatzandFodor(1963), andWickens (1970). Thepresentpapersharesthisnotionwith theseauthors, butdiffers froma number ofthemin conceiving theperception ofwords,

likethatofthings andevents, tobea process ofselective detection ofinformation rather than one of encodingfrombits of sense-data. 3. See Wickens(1970)for a summaryof many relevantexperiments.

4.Thereis somuchevidence forthisthirdhypothesis thatI canonlygiveexamples ofit

here,roughly oneforeachfeature class. Myapologies to theauthors whose results are relevant but not included

5.Strategies which reduce theinformation willbeadopted inshort-term memory, when possible. Baddeley (1971) hasshown thatredundant strings ofletterswillbeorganized ingroups (e.g., B-E-D asthewordbed,rather thanasthreeindependent letters). Effects of acoustic confusability then drop out.

6.Anexperiment byPynteandNoizet (1971) alsofounda strongeffect ofpronounceability ontachistoscopic recognition oftrigrams whilefinding meaningful setsofinitials also facilitating. Butthistypeof meaning waseffective principally whenthetrigram was unpronounceable and had littleeffectwhenit was. References

Anderson, N. S. Theinfluence of acoustic similarity on serialrecallof lettersequences. Quarterly Journalof Experimental Psychology, 1969,21, 248255.

Anisfeld, M.,&Knapp, M.Association, synonymity, anddirectionality infalserecognition. Journalof Experimental Psychology, 1968,77, 171179.

Baddeley, A.D.Shorttermmemory forwordsequences asa function ofacoustic, semantic, andformalsimilarity. Quarterly Journal ofExperimental Psychology, 1966,18,362365.

Baddeley, A.D.Howdoesacoustic similarity influence short-term memory? Quarterly Journalof Experimental Psychology, 1968,20, 249264.

Baddeley, A.D.Language habits, acoustic confusability, andimmediate memory forredundantletter sequences.Psychonomic Science, 1971,22, 120121.

Biemiller, A.Thedevelopment oftheuseofgraphic andcontextual information aschildren learn to read. ReadingResearchQuarterly,1970, 6, 7596.

Bloom, L.Language development: Form andfunction inemerging grammars. Cambridge: MIT Press,

1970.

Bower, G.H.Amulticomponent theoryofthememory trace.InK.W.Spence &J.T.Spence (Eds.), Thepsychology oflearning andmotivation. Vol.1.NewYork:Academic Press, 1967.

Bower, R.Thefirstsentences ofchildandchimpanzee. InR.Brown, Psycholinguistics. New York:

The Free Press,

1970.

Bruce, D.,&Murdock, B.B.Acoustic similarity effects on memory forpairedassociates. Journalof VerbalLearning and VerbalBehavior, 1968,7, 627631.

Cofer,C.N.,&Bruce, D.R.Form-class as thebasisforclustering in therecallof nonassociatedwords.Journalof VerbalLearning and VerbalBehavior, 1965,4, 386389.

Cohen,G. Searchtimesfor combinations of visual,phonemic, andsemantic targetsin readingprose.Perception andPsychophysics, 1970,8, 3703 72.

Theoryof WordPerception 469

Conrad, R.Acoustic confusions inimmediate memory. British Journal ofPsychology, 1964, 55, 7584.

Crowder, R.C.,&Morton, J.Precategorical acoustic storage. Perception andPsychophysics, 1969, 5, 365373. Dale, H.C.A.,&Baddeley, A.D.Acoustic similarity inlong-term paired-associate learning. Psychonomic Science, 1969,16,209211. Feizen, E,,&Anisfeld, M.Semantic andphonetic relations inthefalse recognition ofwords bythirdandsixth-grade children. Developmental Psychology, 1970, 3,163168. Fillenbaum, S. Words asfeature complexes: False recognition ofantonyms andsynonyms. Journal ofExperimental Psychology, 1969,82,400402.

Gibson, Crofts, E.j.1969. Principles ofperceptual learning and development. New York: Appleton-Centu Gibson, E.J.Theontogeny ofreading. American Psychologist, 1970, 25,136143.

Gibson, E.1..Bishop, C.,Schiff, W., &Smith, 1.Comparison ofmeaningfulness andpronunciability asgrouping principles intheperception andretention ofverbal material. Journal

of Experimental Psychology, 1964,67, 173182.

Gibson, E.j.,&Guinet, L.Theperception ofinflections inbrief visual presentations of words. Journal ofVerbal Learning andVerbal Behavior, 1971,10,182189. Gibson, E.1..Pick, A.,Osser, H.,&Hammond, M.Theroleofgrapheme phoneme correspondence intheperception ofwords. American Journal ofPsychology, 1962, 75, 554570. Gibson, E.J.,Osser, H.,&Pick, A.Astudy inthedevelopment ofgrapheme phoneme correspondences. Journal ofVerbal Learning andVerbal Behavior, 1963, 2,142146. Gibson, E.J.,Schapiro, F.,&Yonas, A.Confusion matrices forgraphic patterns obtained withalatency measure. InThe analysis ofreading skill: Aprogram ofbasic andapplied

research. 7696.Final report, Project No. 5-1213, Cornell University &tU.S.O.E., 1968. Pp.

Gibson, E.J.,Shurcliff, A.,SrYonas, A.Utilization ofspelling patterns bydeaf andhearing subjects. InH.Levin &j. P.Williams (Eds.), Basic studies onreading. NewYork: Basic

Books, 1970.

Gibson, E.J.,Tenney, Y.J.,&Zaslow, M.Theeffect ofcategorizable context onscanning for verbaltargets.Mss.CornellUniv.,1971. Gibson, E.J.,Tenney, Y.1.Barron, R.W.,&Zaslow, M.Theeffect oforthographic structure on letter search.Mss.CornellUniv.,1971. Gibson, E.J.,&Yonas, A.Adevelopmental study oftheeffects ofvisual andauditory interference ona visual scanning task.Psychonomic Science, 1966,5,163164. Haber, R.N.Effect ofprior knowledge ofthestimulus onword-recognition processes. JournalofExperimental Psychology, 1965,69,282286. Hyde, T.S.,&Jenkins, J.J.Differential effects ofincidental tasks ontheorganization ofrecall

ofa listofhighly associated words. Journal ofExperimental Psychology, 1969,82, Kaplan, G.,Yonas, A.,&Shurcliff, A.Visual andacoustic confusability inavisual search task. 472481.

Perception andPsychophysics, 1966,1, 172174.

Katz, J.J.,&Fodor, J.A.Thestructure ofsemantic theory. Language, 1963, 39,170-210.

Krueger, L.E.Search timeina redundant visual display. Journal ofExperimental Psychology, 1970, 83, 391399. Krueger, L.E.Theeffect ofacoustic confusability onvisual search. American Journal of Psychology,1970, 83, 389400. Levin, H.,&Turner, E.A.Sentence structure andtheeyevoicespan. InTheanalysis of reading skill: A program of basic and applied research. Final report, Project No. 5-1213, Cornell Univ. Sr U.S.O.E.,1968.

470

E. I. Gibson

Murdock, B.B.,Jr.Auditory andvisual storesin short-term memory. ActaPsychologica, 1967, 27, 316324.

Murdock, B.B.,Jr.Modality effects inshort-term memory: Storage orretrieval? Journal of ExperimentalPsychology,1968, 77, 7986.

Neisser, U.Decision-time withoutreaction-time: Experiments invisualscanning. American Journalof Psychology,1963, 70, 376385.

Paivio, A.,&ONeill,B.J.Visual recognition thresholds anddimensions ofwordmeaning. Perception and Psychophysics, 1970, 8, 273275.

Pynte, J.,&Noizet, G.Trigrammes, syllables etsigles: tudecomparative deIafacilitØ de leurperception. Mimeo, Lab.dePsych.ExpØrimentale, Aix-en-Provence, 1971.

Reicher, G.M.Perceptual recognition asa function ofmeaningfulness ofstimulus material. Journalof Experimental Psychology, 1969,81, 275280.

Rice, U.M.,&DiVesta, F.J.Adevelopmental studyofsemantic andphonetic generalization in paired-associate learning.ChildDevelopment, 1965,36, 721730.

Riess, B.F.Genetic changes in semantic conditioning. Journal ofExperimental Psychology, 1946,

36, 143152.

Schapiro, F.Information extraction andfiltering during perceptual learning invisual search. Unpublished Ph.D.dissertation, Cornell University, Ithaca,NY,1970.

Stroop, J.R.Studies ofinterference inserial verbal reactions. Journal ofExperimental Psychology, 1935, 18, 643661. Tulving. E.,&Madigan, S.A.Memory andverbal learning. Annual Review ofPsychology 1970, 21, 437484.

Venezky, R.L.Thestructure ofEnglish orthography. TheHague:Mouton,1970.

Weber, R.M.Firstgraders useofgrammatical context inreading. InH.Levin &J.Williams (Eds.),Basicstudieson reading.New York:BasicBooks,1970.

Wheeler,D. D. Processes in wordrecognition. Cognitive Psychology, 1970,1,5985.

Wickens, D.D.Encoding categories ofwords: Anempirical approach to meaning. Psychological Review, 1970, 77, 115.

Wickens, D.D.,Clark, S.E.,Hill,F.A.,&Wittlinger, R.P.Grammatical classasanencoding category inshort-term memory. Journal ofExperimental Psychology, 1968,78,559604. Wickens, D.D.,Ory,N.E.,&Graf,S.A.Encoding by taxonomic andacoustic categories in long-term memory. Journal ofExperimental Psychology, 1970,84,462469.

Yonas, A.Theacquisition ofinformation-processing strategies ina time-dependent task. Unpublished Ph.D.dissertation, Cornell University, Ithaca, NY,1969.

HowPerception ReallyDevelops: A Viewfrom outside the Network

EleanorJ. Gibson

This paper waswritten fora summer institute held in1975 attheCenter for Research inHuman Learning attheUniversity ofMinnesota. The psychologists invited togive presentations, except formyself, were zealots ofinformation pro-

cessing. Thebook thatresulted foreshadows thepresent move toward cognitive science andincludes chapters onneurophysiological substrata ofperception and comprehension, andcomputer models oflanguage acquisition. Perceptual models presented (e.g., Esteshierarchical filter model, andJohnsons pattern-unit model) takea constructionist viewofhowunitsarebuiltandhowwords, asa consequence, areperceived inreading. These views donotdealwithperception (asI conceive it)atallbutaremore concerned withentities likeicons,stores,

andbuilding ofrepresentations. ThetitleI chose wasintended toexpress this I quoted some lines from Auden before beginning because I wanted toempha-

feeling.

sizethatthere isa world oflasting objects andthattheinformation forthem is

theretobeobtained. Weessaytoobtain itthe truthbut onedoesnotcon-

struct truth. Ifelt,andstillfeel, thattheview expressed here applies toperceptual

processes inreading aswellasintheworld ofobjects. Thestrong language inthe opening paragraphs did not seem to offend anyone, I suppose because I was a lone voice crying in the wilderness.

Having railed against theelementarism andconstructionism oftheinformation processing approach, I tried tosetforthmyview thatperception isextraction of information about theaffordances ofthings intheworld, likelayout andevents, andthatperceiving textinvolves extraction ofinformation about these things as well. Itisa less direct process, tobesure, buttheaimandfunction arethesame.

Itried tobring inthenotion ofaffordance, aconcept thathaspreoccupied mein

planning a research program eversince.

D.Laberge andS.1.Samuels (eds.), Basic Processes inReading: Perception andComprehension.

Hilisdale, NJ.:L.Eribaum Assoc., 1977. Pp.155173.

472

E. 1. Gibson

I chose toorganize mytalkaround thetrends I identified earlier inlearning and development (see chapter 20).Attheheart ofmyconception ofperceptual learning is thenotionthatits resultis increasing specificity of correspondence between

information instimulation andwhatisperceived. Learning toreadisreplete with examples ofthischange. Butincreasing economy ofinformation pickup isatthe heart ofprogression toward skilled reading. There isa vastamount ofevidence for thistrend,someofwhichis presented here.I argued, ofcourse, thatthisprogression is nottheresultofgradually stringing together longer unitsviaan

associative process, butisinstead theresult ofdiscovering order andrulesthat generate textual units.They existat manylevels (orthographic rulesinwords, syntax inbigger chunks oftext,andnested semantic unitsthatcanbeasgreatas theorganization of a bookandas smallas a metaphor carried in a single word).

Theabilitytoextract selectively thesekindsofinformation exists tosomeextent ineventheyoungest readers, butitgrows asa reader readsonovertheyears. An

experiment bythree ofmygraduate students (Condry, McMahon-Rideout, and Levy 1979) provided a neat,concrete example ofthisdevelopment. Butitshows at thesametimethatsecond gradersarealreadycapable ofseeking information

fromtextanddosoin muchthesamewayas adultreaders do,however inexpertly.

Inthenotes forthelastcolloquium I gaveonreading I findthisconclusion: Maturereadingis markedbydiscovery ofstructure; byuseofstructure andrulesto achieve economy ofperformance; andbyadaptinginformation pickupto the readerstask.

Thedifference between a reader whousestheredundancy givenbyall these kinds ofstructure efficiently andautomatically, anda beginning scholarwhoispresumed tostumble alongdecoding letterbyletterintospeech soundscannotbeexaggerated. Theaccomplishment ofreadingandcompre-

hending thetextofWarandPeaceisaswonderful asreading thescore of a symphony, which does notcome through notebynoteanymore thanthe formercomesthroughletterbyletter.

Cana process sofundamental asperception besubject to thefadsand

fancies induced by a Zeitgeist in themindsof thosewhoseekto understandit? Perception, the mostsolidlybasedof all cognitiveprocesses,

surelycannotbe whipped aboutwiththewindof thetimes,however processes likememory (short or long?) images(reproductions or constructions?), or intelligence (normativeor operational?)maybe conceived variously by ficklethinkers in onedecadeor another. Andyet,

signsandportents assure us(uneasy psychologists) thatitis.Recall unconscious inference, dredged upfromtimeto timeeversinceHelmholtz. What

HowPerception ReallyDevelops 473 about the fact that it wasn't fashionableto mention perception at all during the heyday of S- R theorizing? What about the "new look" in the early Fifties? And now , what about the craze for information -processing models of perception ?

The craze is with us; one has only to consider other chapters in this book . How

did it come about ? The answer lies , I think , in the transition

from S- R theories to cognitive theories via computer simulation . How

respectableit is! If a computer can "model" a cognitive process, there is somethingsubstantialto take hold of. I believe people who would not have dreamedof working on what they think of asperceptionin 1950 arehappy to now becausethey can invent stagesof "processing" that can be duplicated in computer software, if not hardware. But there is something else that makes it acceptableto these people. It is still elementaristic, just as much as S- R theory. One starts by assumingthat "input" comesin pieces or bits . The assumption is seldom defended; it is just made. And then, of

course, if the input is only bits and pieces, a construction process must be devisedto assemblethings in the head, to reconstructreality, becauseit is obvious that we don't perceivebits and pieces, but a world . The goal is how to build "a world of lasting objects," when, in fact, they are already there

in the world

.

I am inclined to think that many present-day cognitive psychologists consider themselvesredeemedbecausethey have foresworn behaviorism but that they are really pursuing the same course because they conceive of the stuff of the world that is available to an organism as meaningless and atomistic . Such a conception requires them to invent "processing mecha-

nisms" to put the world together. Sensationshave traditionally been the bits and pieces- unstructured, although having a few dimensions like intensity by meansof which they can be arrangedin a scale. Now, however , we have a mysterious

construct

known

as a " feature detector " that

implies that (for vision at least) the piecesat some early stage have a wee semblance of form , such as tiny lines in varying orientations . These may, however , have to be " coded " from

the sensations ; and in any case, the

orientation seemsto be with respect to the retina, so it still doesn't help

with

construction

of a world

.

There must in fact, to judge from the models of perception and cognition, be a seriesof "coding" and "decoding" processes(world to stimulus; stimulus to ikon ; ikon to features; features to parsed units of some kind ; those units to short -term memory ; short -term memory

to another

code

where meaning is somehow incorporated , etc.). There is a " processor" at

each of these stages of recoding, and in the end cognitive psychology becomesa branch of cryptology . The tendency to homunculize appears irresistible; there are cryptologists in the brain. But how to put Humpty Dumpty back together again? No answer to this has ever been proposed,

474

E. J. Gibson

except "integration with past experience," or "inferencefrom past experience." But what was the past experiencelike? How was it obtained? How did it becomecomplete and meaningful? Information processorsquite often conclude that there is no such thing asperception, that what we experienceasperceptionis really memory. The conclusionis almost inevitable from this kind of analysis, for indeed there is no frozen instant of time in which a perception may be said to occur. Perceiving goes on continuously; temporal information is not only available but must be extracted. If it were not, we would not perceive a hurled object looming at us on a collision course, or even feel our own arms reachingout toward a target. So eliminating perceptionmay be consistent, but it leaves us where we started. If what we are talking about is really memory, where did we get the meaningful information to remember? Of course, once something gets constructed, the information processor can make progress becausehe can start matching the bits and pieces to the construction, but the original structureof the information and the meaning have still escapedthe processors. Supposeone does not acceptthis view of the way the world comesto u~ in exoerience as atomic bits or even "pictures" frozen in time and .. without spatial organization?! I do not believe it is essential to invent mechanismsfor assemblingwhat we know we actually do perceive: things, a spatial layout, events that take place in it . There is structure in the array, relational information that does not have to be pieced together because , like truth, it is already there. This is the assumptionI want to proceedwith . I do not want a construction theory, with processorsat every stagelike an assemblyline. What kind of theory do I want, then 7 I might call it a "seekand ye shall find" theory (find sometimes, at least). What do we seek? Living organisms searchfor information for invariant properties of the world: of the spatial layout, which must have a stable structure(in fact, an objective one in which not only I but other animalstoo can locate things); of objects , with constant dimensions(not shrinking, expanding, or losing and gaining substanceas we move around) and reliable affordancesthat meet our requirementsfor survival; and of events , in which things happenwith predictablerelations (not randomly or chaotically). This searchis so much a part of man's nature, evolved over millions of years, that it is as ingrained, strong, and unconsciousas the functions of digestion and breathing and much more elaboratelyprovided for. We have many windows on the world : systems for listening, looking, touching, tasting, and accompanyingpatternsof exploration like scanning, palpating, and licking. I think we have been fooled by our own laboratory paradigms into believing that we have to bribe an animal or an infant into learning something with material rewards like food. For human infants, this procedure does not even work very well. It turns out that they learn best if they

How PerceptionReally Develops

475

are allowed to discover an interesting sourceof information or a predictable contingency or a problem to be solved. From this perspective, if I look for any mechanismat all, it will not be an associativeprocessor a construction process, but an abstraction process. What is neededis abstraction(extraction?) of the relevant information from the flood of available information. Abstraction is supposedto be a very high level process. Can babiesdo it, on a perceptuallevel at least? It seems that they must, if only to abstract phonemes, for instance. from a continuous speechflow. That they do, indeed, abstract consonantalphonemes from vowel contexts has been demonstratedby Fodor, Garrett, and Brill (1975). With this view that perception is not constructedfrom sensoryparticles go a few caveats. Does it carry the implication that all the information in stimulation, provided it is availableto an organism, is picked up automatically, regardlessof species , age, or earlier experience? Certainly not. The information has to be extracted (but it is information that is extracted, a relation, not an element). What does extraction dependon? For the sakeof brevity, I note just three important points: 1. What is extracteddependson the species.Every living organismis equippedappropriately to extract the information neededfor survival in its ecologicalniche. All get somekind of information for the spatial layout, the events, and the objects that are relevant for them, but the information differs. Recordingsfrom single cells in the auditory system of monkeys, for instance, have shown that they are tuned so as to make it possibleto hear the range of vocalizationscharacteristicof the species.Similar studieshave shown appropriatetuning in bats and other species. Primates and birds are equipped to get excellent information for the spatial layout by visual means. Bats, on the other hand, have highly specializedsystemsfor getting acoustic information for where they are going. 2. What is extracted depends on developmental maturity. Human infants appear to differentiate and identify stationary objects rather poorly before they are four or five months old. But they seem to locate things quite well very early. Their attention is greatly drawn to events; how soon theseare differentiatedin any fine-grained way we do not yet know, but highly contrasting ones like looming and zooming are probably distinguishedby a month or so. Small events embedded in larger events seem to be differentiated considerably later. 3. What is extracted dependson learning. There is learning common to membersof a species , who sharean ecology and pickup systems, and there is also idiosyncratic learning by individuals. It is very

E. J. Gibson

476

instructive

to

to

how

look

to

1973

)

as

it

.

A

chess

in

board

,

He

the

not

image

of

the

"

an

results

from

But

are

generated

well

,

some

as

things

,

someone

are

-

in

in

,

but

always

case

a

with

him

that

has

followed

but

cases

are

.

are

,

they

meaningful

what

is

his

of

an

because

They

master

schema

?

not

gives

,

this

chess

is

arrangement

ordered

.

the

learned

It

else

as

and

the

he

that

the

abstract

cases

arrangement

yes

saw

in

that

The

or

every

has

.

them

.

)

when

on

What

once

Rules

something

game

image

for

he

the

"

the

But

,

board

game

.

randomly

order

generated

afford

that

a

move

not

brain

.

There

is

chessboard

.

internal

to

matching

and

no

We

per

-

contextual

.

If

I

look

for

that

than

add

on

or

,

over

a

chosen

to

trends

whether

,

I

can

talked

the

be

I

a

"

about

different

same

is

in

percep

-

development

,

individual

consider

.

along

of

be

in

learned

growing

will

to

rather

interested

perceptual

process

.

development

,

They

I

have

with

integration

these

-

its

explained

in

demonstrating

one

end

or

are

.

ago

six

,

of

development

,

.

and

one

con

their

, "

and

.

-

to

to

authenticity

say

on

,

think

have

moves

them

more

than

the

two

of

characterize

toward

spent

the

each

but

that

that

once

at

intervened

begin

it

I

once

had

descriptions

how

can

I

having

I

are

points

What

years

although

They

if

perceptual

development

years

five

still

,

and

and

in

and

two

into

authenticated

trends

learning

combined

them

but

invoke

about

perceptual

quite

were

,

what

the

have

,

primarily

in

in

to

information

not

trends

but

going

.

Minnesota

was

trends

,

latter

to

am

in

in

the

need

I

interested

second

elements

have

on

of

occasion

am

are

abstract

Development

book

University

a

about

Perceptual

before

I

they

that

.

interested

of

from

in

am

fraction

I

struction

perceptual

I

ones

pieces

.

and

,

and

integrate

talk

therefore

relations

mechanisms

not

my

,

with

tualleaming

Trends

mechanisms

deal

processing

describing

.

that

win

"

specific

ones

be

II

rules

perceive

stored

detector

relations

We

rules

to

specific

feature

ceive

them

.

the

they

so

of

knows

Simon

at

play

.

that

an

been

&

arrangement

board

most

has

Chase

cannot

arranged

clues

the

seconds

amateur

quickly

he

them

because

next

a

two

not

particular

set

that

did

the

that

up

abstract

from

make

to

picks

rules

derived

as

than

on

from

are

better

"

He

an

they

no

(

amateur

been

us

of

learning

as

he

give

One

masters

little

An

have

remember

.

as

.

arrangement

advantage

for

chessmen

is

will

.

chess

( although

master

does

look

perfectly

of

that

understood

perceptual

on

can

almost

number

because

be

literature

midgame

it

,

cannot

idiosyncratic

the

master

stood

same

in

in

learned

or

of

us

reconstruct

is

can

examples

provided

Twice

what

itself

impressive

"

at

learning

in

time

explaining

,

to

in

How Perception Really Develops

477

them. I shall still do that, but I shall ask whether the conceptsI have been using- differentiation; searchfor distinctive features, invariants, and order; a process of abstraction; and reduction of uncertainty by finding order - are sufficient for an adequate account of where perception is going developmentally. IncreasingSpecificityof Correspondence betweenWhat Is Perceived and Informationin Stimulation A striking trend in perceptualdevelopment is the increasingspecificity of correspondencebetween information in stimulation in the world and in what is perceived. I think objectsand events and layout are perceivedfrom the start, insofar as anything is, but with little differentiation. I am not persuadedby experimentsthat display line drawings of geometricalforms to neonates that perceiving begins with the pickup of a line, and then gradually more lines, welded together into cell assembliesby eye movements. What do babiesnormally look at? Not line drawings. They certainly look frequently at faces, and researchhas tended to focus, quite properly, on the development of perception of faces. Unfortunately, even this researchdoes not have all the ecologicalvalidity one might wish, becausethe displays used are often flat, motionless, silent, over-simplified cartoons. Realfacesmove and talk. But if we think of the displays used as varied degreesand types of simulation, attempting to isolate critical information, they canperhapsstill tell us something. It seems pretty clear that during the first few months of life infants progressively differentiate the human face as a face; gross outer contour appears to compel attention very early, then internal details like the eyes; but gradually superordinaterelations are noticed and right ones are differentiated from wrong ones. Only later are individual facesdifferentiated as unique, and properly identified. Work in our laboratory by Spelke with motion pictures of people doing things (chewing gum, nodding, yawning, and so on) shows that a person is recognizeddespite different activities by about 5t months. Invariant, enduring properties peculiarto the person are picked up despite varying actions. The truly unique features seem to be discoveredover change. It is impossiblefor me to conceiveof recognition that develops over varying activities as a process of associatingelements or constructing from elementsa schemato be matched. There certainly is no frozen image to be viewed or "processed " by a comparator. Since the chapters that -precededmine all deal with the oerceotion of J . . -words (single words or alphanumericsymbols of somesort), I shall try now to choose my examples from this realm. An early case of developing specificity of correspondenceis the differentiation of graphic symbols. Researchby Lavine (1977) showed that children in our society differentiate "writing " from pictures (simplified line drawings) at age 3. When tested

478

E. J. Gibson

with many samplesof printing, cursive writing , and scribbles simulating writing , they unhesitatingly identified them as writing , although few could identify a single letter by name. Lavine's experiments showed that the youngest children tend to makeuseof someclassfeatures, suchaslinearity, repetition, and recombination of items, for distinguishing writing from pictures. But older children, at ages 5 to 6, were making greater use of distinctive featuresof individual letters. They differentiated one from another (still usually without naming) and discarded scribblesand artificial characters . Many, however, classifiedunfamiliar characters(e.g., Hebrew and Chinese) as writing , pointing out that it was not their kind of writing . Differentiation of global featuresfor writing as a classevidently preceded differentiation of kindsof writing and of individual letters. I have given a good deal of attention, in the past, to an analysis of distinctive features of letters. Does that not set me up as one who calls upon "feature detectors," one who assumesthat featuresare the elements out of which letters are constructed? Our sins do seemto come home to roost, and I have certainly been thought to hold his notion. It was my own doing, I now see, since I suggesteda table of features that others interpreted in this way. But it is not too late to protest. I have always thought of distinctive featuresas relational, as Roman Jakobson defined them for phonemes. They are contrasting properties useful to distinguish one thing from another, not bricks. They are discovered in a IIdifferencing process" and are abstract. Most children seemto have little trouble learning to distinguish letters - even rather confusableones(cf. Gibson & Levin, 1975, pp. 239ff.). When trouble comes, it generally occursin distinguishing sequencesof letters in given orders (e.g., boatfrom baot; trial from trail). Is this not a failure in, or rather a need to develop, the ability to integrate? Again, I doubt it . Perception of order requiresperception of succession . Perceiving successionover time is a particularly good example of differentiation that improves with development. There is considerableevidencethat temporal acuity develops in children; they becomebetter at resolving information given over time, rather than at integrating it . The temporal relations for perception of phi movement provide one example. Younger children perceive movement rather than successionwith a longer interval and a wider gap than do older children and adults. Another example is the effect of masking a visually presentedtarget. The masking threshold- the temporal interval that must intervenebetweena target (a letter, say) and a mask(another letter or some figure like a grid or a dollar sign) in order to identify the target- has generally been found to be longer, the younger the child. I do not think that this is becausehe cannot remember the target. Rather, I think that eventshave to be differentiated, not integrated. Furthermore, perceptionof order requiresability to discriminatesuccession . In other words, differentia-

How PerceptionReally Develops

479

tion of succession - including segmentation- is a prerequisitefor pickup

of order .

There is supporting but indirect evidence for this notion in studies of

children's problems with segmentationwhen they begin to read. Words resist being broken down into sounds. Perhapsthe sameis true of printed letters in a string for younger children. We have recently initiated what I hope will be more focusedexperimentsto study maskingeffectsin secondand fifth -grade children to see whether such effects with letters and words are related to perception of order in words . Is confusion of order within

visually presentedwords related to developing ability to detect temporal order

or succession ? In some sense, information

for succession must be

perceived, whether one readsletter by letter or not. Information processors would

no doubt refer to short -term memory

for letter order , but I would

talk about perceptualpickup of order information- something that has to be differentiated, especially in the internal portions of words. Of course, there is not only order information in words, there is information about other structural

relations too , and I want to turn to this , since it leads to the

second important trend : increasing economy of infonnation pickup .

IncreasingEconomyof Inforn1ationPickup I think increasingeconomy comesabout becausethe developing organism becomesmore efficient in detecting and using relations that are present in the stimulus array . In general, there are two major ways of increasing the economy of extracting relevant information .

One is the detectionand use of the smallestpossibledistinguishing feature that will permit a decision. It may sound as though relational

informationis not involvedin detectingwhat somepeoplethink of asan element. But it must be rememberedthat it is not an elementthat is singled out. The differentiating processrequires generalization and classification. The minimal

information

is relational -

an abstraction

of a common

con -

trasting property that divides two sets. The two sets could be letters or

artificial charactersin a Sternbergtype of task (Barron, 1975; Yonas, 1969) that are separable if some contrastive property is discovered ; or they could be concepts that share many properties but differ critically in just one or two . It is only that one or two that will be retained for future classifica-

tion (Vurpillot et al., 1966). Children tend to behave this way, as well as adults, but adults can accomplisha finer sifting- with word meanings, for

instance

.

The second way of increasing economy is the detection and use of

superordinatestructure that permits what has often been called IIgrouping information in larger units." This is a typical way of increasingeconomy in reading; a reader will deal with the information in the text in the largest units that are relevant for his task, if he is capableof it (Gibson & Levin,

480

E. J. Gibson

1975). These units are not conglomerationsof associatedparts (whatever the parts might be- lines? letters?), but are generatedby the various rule systems that characterize

discourse , in this case written

discourse , at all

levels. There is order in the way the information is presented- within words , within sentences , and within passages of discourse . There has been much talk of economical use of structure

at the level of

the word during the past decadeand indeed at this symposium. I need hardly remind you that it takes as long to read a letter as a one-syllable word; that a word may be recognizedat as short an exposureduration as a letter ; that a letter in a word

can (under some circumstances

at least ) be

identified , with a short exposure, more accurately than the same letter

alone; that it takes longer to identify a target letter in a word than to identify the word itself as a target ; and so on.

Although we have far less researchbearing on this question, there is reason to think that ability to treat words as units develops over a fairly long period while a child is learning to read. The earliest data I know of were collected by Hoffman in 1927 (reported in Woodworth , 1938, and Gibson & Levin, 1975). He compared the increasewith grade level in the number of letters read correctly with tachistoscopic exposure of un-

connected consonants, nonsensesyllables, unfamiliar words, and familiar words. From the first to the eighth grades, there was only a slight rise in the number of randomly connected consonants and nonsensesyllables read, but there was a large increasefor words, especiallyfamiliar ones. The sharpestrise occurred between the first and third grades, a finding consistent with more recent data collected by me, my students, and my col-

leagues(Gibson, asser, & Pick, 1963; Golinkoff, 1974; Rosinski& Wheeler, 1972 ).

N one of our studies has covered a really wide age range, however , or

comparedmore than two or three kinds of items (usually words, pseudowords, and unpronounceableletter strings). I was pleased, therefore, to receive a copy of a very extensive study by Doehring (1976) comparing latencies of response to a number of kinds of verbal material , with different

judgments, over a very broad age range. I have chosena few of Doehring's results(all typical) that makethe point about growth of economicalprocessing. When an auditory- visual match was to be made and latency recorded, children did not improve significantly after the first grade in matching single letters. But when four-letter

words (presentedvisually) were to be matchedto a samplepresented auditorily, latency improved significantly to the middle of the second grade. When four-letter pronounceablesyllables were presented, latency improved up to the ninth grade. The trends for purely visual matching were similar for the three kinds of material . The same trends were particu -

larly strong for oral reading measuredin mean latency in secondsper

How PerceptionReallyDevelops

481

Table 26.1

Examples ofOrthographic Constraints within theEnglish Monosyllable Initial consonant

Final consonant

cluster

cluster

qu

tat, lit, lul, lael

ck

hr scr

tat, Ii !, lul, tact la/, lit, lut, /ae/

pt

gi /1 sfr

tat, lit, Lu!,lad Ia!,lit, tu/, lae/ /al, lit, lul, lae/

rd rich ngth

ng

syllable. Thevarying ratesofprogress forletters, words, andsyllables had

nothing to dowithsimple motorstrategies ofvocalizing, sincepicture

namingoverthesameperiodshowednochangeat allinlatency.

Onefurther resultdeserves mention. Vellutino, Smith, Steger, andKaman(1975) foundthatpoorreaders arefrequently ableto copyandname lettersinwordscorrectly, withtachistoscopic presentation. Butthesame words wereoftenreadincorrectly. Thepoorreaders performance onletter

reproduction andnaming approximated thatofnormal readers, although

theyweretypically inferior to theminreading. Thefactis indubitable thatmostchildren dolearnto extractinformation fromtext in largerunitsthanthe letterboth actualwordsand letter

stringswithstructural constraints havingto do withlegalpositions of

consonantclustersand vocalicseparationof consonantsand clustersof them.Table26.1illustrates theseconstraints. Theinitialclustersarecon-

strained to thatposition ina monosyllabic wordorina syllable, justasthe

finalclusters areto theirposition. Thevowelmustbe present in the

syllable, and if thereare both initialandfinalconsonants, or consonant

clusters, it mustseparate them.Anysucharrangement, asindicated in the

table, maybea word. Evenifit isnota word, it canbepronounced and readlikeone.Now,doI notneedtheconcept ofintegration to explain suchlearning? Donotchildren forgelargerunitsbyassociating lettersinto blocks orchunks fromoriginally smaller elements likesingle letters(some-

timescalledthe bottom-up view)?

If a wordhas beenintegratedfromletters,we musthavelost the

saliency oftheincorporated letters, because a letteris perceived better whenit ispartofa word,ifthewordisexposed quickly. Several people

haveobjected thatthisfindingis dueto redundancy; of courseit is.The

constraints thatcharacterize a wordas a wordconstitutefirstclass,usable

redundancy. Atthesametime,ifoneisasked todecide whether a given

wordor a givenletterispresentin a display, it is fasterto detecta whole

wordthana letter(Johnson, 1975). Theworddoesnotdecompose easily,

482 for

E. J. Gibson

adults

as well

must

segment

have

the

as young

before

children

they

" pieces " to begin

constraints

provide

new

an anagram

ment

(Beilin

a word

word

frequency

following

from

the

has holistic

examples

solution

word

hand , high

- nonword

anagram

or

the

unpronounceable

unit

(Dominowski

probability

, or

summed

bigram

a pretty

effect

high

(and

other

solution

on

How strong gave

with

such

the

, resisting

been

been

very than

arrange

-

, as in

the

with

and fifth

word

nonwords

of

14 years ) . The

probability have where

would

and

anagrams difference

to as

accord . The

and

a little

and

Horn , 1962 ; Domi letter

with

of

would

-

of words

in relation

does

rearrangement

a

have

orthographic

possible

skill

is

be surprising

sample

-

position

and legality

not

and used to the

, of course . a word ? Beilin

to subjects

of four

in

times

solution

(e .g .,

of letters

only

anagram

results

as a

transitional

of such a word

reading

harder

regularity that

counts

of the

the

word ' s unity

&

structure

of

fragmentation

, are

pronounceability

conflicting

development

the

(Beilin

positions

it . But

of either

to a word ' s integrity

to

a count

as frequency

to make

inhibit

, 1968 ) . " Pro -

suggested

found

solution

such

three - , four - , and five - letter

(8 , 10 , 12 , and

often

and at the same time

interact in

are

, is a source

anagram

of English

variables

early

Dominowski

contributes

are confounded

rules

unity

has

may

frequency

, so orthographic

, 1963 ) . Since bigram

frequency

) would

&

they

again

have

ck in the fourth

phonological

(Ekstrand though

frequency

, a correlation

like

solution

word

varying

, 1959 , 1962 ), but transitional

consideration3

bigram

it resists

, 1965 , 1968 ) . High

facilitate

of the anagram despite

frequency

now ski , 1967 ; Stachnik

orthography

may

arrangements

, 1969 ). It

& Tresselt

given

that

in a random

cause cause table table

, even

bigram

no

presented

Solution Word

and meaning

Mayzner sometimes

, the

pseudo - word ,

such

&: Dominowski

sauce ceusa bleat bleta

word

" anagrams

as familiarity

a legal

properties

Anagram

holds

solution

solve well

do not

on anagram solution provide up a word to solve an anagram

word

frequency

effect

nounceable than

. They

is possible

:2

vs .

On the other

reasons . Children

segmentation

the same letters

vs .

the

after

&: Horn , 1962 ; Ekstrand of

different

and constraints

. The word , or even

segmentation ? Numerous experiments handv ~ evidence . It is harder to break to solve

for

order

with . But

structure

has holistic properties . How do I know that

, but

can observe

exhibit ( 1967 )

age groups for

word

as

HowPerception ReallyDevelops 483

compared withnonsense anagrams wasapparent forfive-letter butnotfor

threeorfour-letter anagrams, butitdidnotreach significance untilage12. Along withthefactthatthechildren wereprogressively abletosolveany

kindofanagram fasterwithincreasing age,structural features oftheword continued to become moreprominent. Whatarethesefeatures? In the firstplace,doesa wordhavefeatures?

Whatisit thatresists segmenting? Consider transitional probability first, although thisisa letter-to-letter property ratherthana property ofthe wordasa whole. Whenotherfactors arecontrolled, summed bigram frequency assuch doesnotprohibit theword inferiority effect foranagram solution. Neither doesit account satisfactorily forthewordsuperiority effectfortachistoscopic recognition of letterstrings. Biederman (1966) found a slight effect ofbigram (digram) frequency, taking letterposition intoeffect(MayznerTresselt count,1965),withfive-letter words,allof lowwordfrequency. Somewerewordshavingtypical English orthographic structure (e.g., clang, bigram count 203)andsome werenot(e.g., gnome, bigramcount15),so it mayhavebeena higherlevelof structure thatwasactually effective. Gibson, Shurcliff, andYonas (1970) compared theeffects ofpronounceability andsequential probability (asmeasured by fourcounts) ontachistoscopic recognition ofpronounceable andunpronounceable pseudowords. Pronounceability hada strongpredictive relationship to correct recognitions, butofthefrequency counts, onlythe

Mayznerand Tresselt(1965)bigramcountdid, and it was a far less

effective predictor thanpronounceability ratings.It standsto reasonthat therearestructural variables inwordsthatfrequency counts(sofar)have

not captured.

Whatelsecharacterizes a word?LaBerge(1976)describes a model

involving coalescing offeatures. IfI understand him,heis thinking of

thewordscontour, orsomekindofsuperordinate graphic information. I doubtthatgraphic information isveryimportant inwordcoalescence (I prefer thetermunity orcoherence). Recent experiments byMcClelland (1975) foundevidence forthetachistoscopic wordsuperiority effectde-

spitescrambled typefaces,makingit unlikely thatthevisualoutlineof the

wordisnecessary fortheeffect to occur. I aminagreement withMcClellandthatknowledge of abstract properties offamiliar stimuli, notjust

knowledge of specific featural configurations, canfacilitate perception Theeffectiveness ofstructural properties inanexperiment insome ways analogous to theword/letter experiments wasdemonstrated recently by (p. 42).

Weisstein andHarris(1974). Theyshowed anobject-superiority effect

forthedetection oflinesegments. Asingle target line(oneoffouroblique lines) hadtobedetected whenit appeared aspartofa display containing

484

E. J. Gibson

the sameeight context lines, variously arrangedso as to produce a unified pattern of connectedsquaresin a three-dimensionalrelationship, or a less unified, unclosed arrangement that appearedonly partially or not at all three-dimensional. The target line was detected more accurately (with tachistoscopicexposure) when it was part of a configuration that looked unitary

and three -dimensional . The effect does not contradict

the notion

that a highly coherent whole is hard to break up; it still may be. But detecting a given line in a " muddle " of context lines is much harder when the subject has a very short time to search. Weisstein and Harris

(1974) suggest that "Perhapsrecognition of both words and objects depends on more general processesthat make use of structural rules and meaning to determine perception " (p. 754).4 It is these two relational properties of words - rule-like structure and meaning - that seem to me to be the ones that confer unity and coherence;

that make them easily recognizable; that make them resist fragmentation; and that yet, when time is limited, make them easierground for detecting a subordinate part than would a random arrangement of all those parts.

They are abstractproperties, not concrete, physical ones. The casefor the structural rules is easy to make. English words have constraints that any monosyllabicword or any syllable in a multisyllabic one must obey. I have illustrated this point in Table 26.1. There must be a "vocalic center" (Han-

sen &: Rodgers, 1968; Spoehr&: Smith, 1973); there may be consonantsor consonant clusters in either initial or final position . The letter may be

constrainedas to position. Indeed, nearly all consonantclustersin English are constrained (Fries, 1963). These constraints are a source of redundancy

that contributes to easeof identifying even a pseudo-word, if the rules are

obeyed.The rulesapplyboth phonologicallyandorthographically . Invarianceof the rules over phonology and orthography may also be important, but that question is still unsettled .

Orthography alone provides further useful rules in spelling pattemscontrastive groups of words that cover a large number of cases(seeFries, 1963, pp. 171ff.). An example is the shift in the value of any vowel when a succeedingconsonanthas an e added, as in the following examples: pan - pane dun - dune

con - cone gem - gene

tin - tine

This principle, as well as others illustrated by Fries, makesindividual association of letters as unnecessary as it would be uneconomical

. The structure

is generatedby the rule. I do not need to appealto a constructive process that integrates letter by letter. The reduction of information provided by these rules is as economical and as compulsively useful as the threedimensional structure in the Weisstein and Harris experiment .

HowPerception ReallyDevelops 485

Theorthographic structure ofaword, itseems tome,isnotintegrated; I donotseehowit canbeandgeneralize thewayit doestopreviously

unseen butwell-formed nonwords. Theprocess oflearning it mustbeakin to the waya childlearnsgrammar, a kindof gradualabstraction. The knowledge itselfisabstract, nota specific physical configuration. I amthe

moreconvinced ofthisbya seriesofexperiments inwhichI triedto teach

children aboutorthographic structure. Noneof themworked verywell,

including one in whichwe explained the rules.An abstraction hasto be

discovered foroneself. Wecanputtherelevant information intheway aswedowhena childhearsustalkwhileweprovide theappropriate

situational contextand hopethathisendogenous motiveto attendto newinformation, andtheeconomy principle thatleadshimto reducethe information by abstraction andclassification, willhavetheireffect.Drillat

leastwillgetyounowhere, norwillM&Ms.Appropriately contrastive displays thatprompt thegeneralizations implicit inorthography appear to

be the mostpromisinginstructionalaid. The Meaning of Words

Now,whataboutmeaning? Thisproperty of wordsdistinguishes them frompseudo-words thatareorthographically andphonologically legal, and

manyexperiments havefoundthat it contributes to the unityandcoher-

enceoftheword,givingit anaddedadvantage forrecognition (Barron &

Pittenger, 1974;Gibson,Bishop, Schiff, &Smith,1964;Murrell&Morton, 1974),andanaddeddisadvantage forfragmentation andrecombination in anagrams. Howmeaning lodgesitselfin a wordis stillanybodysguess, butit isa cogentpropertyofa spokenwordveryearly,andmakesdifficult

thebeginning readerstaskofconsidering thewordasa phonetic object.

Myguessisthatthemeaning ofa wordislearned aspartofa situational

contextin whichit is invariantwithan eventof interest.Heres a cookie goes alongwith beingoffereda cookie,with all the affordances thereof.

Themeaning ofa wordisprobably differentiated frombothitslinguistic

and its situational context.

I havelatelybecome unhappy withtalkaboutmeanings of individual words,becausetheirderivation seemsto me so obviously contextdependent. ButI shalldescribe brieflyonemoreexperiment inwhichthree studentsof mine 5 studiedthe developmental courseof abstraction of fea-

turesofprinted words, including meaning. Thisexperiment wasoriginally designed to illustrate whatI havethought ofasa thirdtrendindevelopment: theoptimization ofattention. Itseems tomenowthatthephenomenathatpersuaded meatonetimeto consider sucha trendasunique are

wellsubsumed underthetrendtowardeconomy in extraction of informa-

486

E. J. Gibson

tion . I shall not argue for a third trend , therefore , but rather point out the cognitive economy and adaptiveness of increasing ability to attend readily

and flexibly to those aspects of presented information that have most utility for the perceiverin his role as performer. In several experiments with young children (Pick, Christy , & Frankel, 1972; Pick & Frankel, 1973) the subjects have been given tasks that re-

quired them to select one aspect of the information presentedby some object , to the exclusion of the rest. The ability to abstract information in

this way, and to change what is abstracted flexibly from one aspect to another as required, seemsto increasewith age. The experiment to be described made use of such a task, but words rather than objects con-

stituted the displays from which several kinds of information- graphic, phonetic, or semantic- were to be abstracted. The subjectswere children from the third and fifth gradesand adults. Supposing that children have the necessaryknowledge of the graphic features; sound, and meaning of a word , do they nevertheless show pro gress in abstracting one of these characteristics from a given word when asked to compare it with another one? How efficiently can they select the

aspectthat is relevant and adapt themselvesanew when askedto selecta different aspect? Reading is always for a purpose, and we have generally assumed that active selection of wanted infonnation is an important part of

the development of reading skill. In this experiment, the selectionprocess with words was investigated.

Slideswerepresentedto the subjectsportrayingthreewords(projected on a small screen). One, a standard, was typed alone at the top center. Below it, one on the right and one on the left, were two other words. One of the other words resembled the standard by either looking like it , sounding like it (rhyming ), or having a similar meaning. The second word was chosen

asa possibledistractor or elsewas neutral with respectto the standard. The subjectwas told before the slide appearedthat he was to choosethe word that looked (or sounded or meant) most nearly the same as the standard,

and to pressone of two buttons (on his right or his left, correspondingwith the choice words ) as soon as he had made his selection .

For example, a slide might appearin one of the following arrangements : cry

high

near

weep (A )

bear

near

close (B)

bear

deer (C )

The subject might be told before the slide (e.g., arrangement A ) was projected, "Choosethe word that means(or sounds) most like the top one"; or he might be told (e.g., arrangementB), "Choosethe word that means(or looks ) most nearly the same as the top one" ; or (arrangement C) " Choose the word that looks (or sounds) most like the top one." A distractor was

HowPerceptionReallyDevelops 487

provided thatlooked or sounded likethestandard ifthesubject wasto

selectformeaning (cf.A orB);thatlookedor meantmuchthesameif the subjectwasto selectforsound(cf.A andC);andthatsoundedor meant

muchthesameifthesubject wasto select forlooks(cf.BandC).Foreach of the threetasks(looks,sounds,means), slideswithneutralratherthan

distractor alternatives werealsoprovided. Asubject made judgments with

respectto allthesetypesof slides(ninein all:threetaskslooks,sounds, or meansand threetypesof distractor foreachtask). Halfthesubjects ineachagegroupwereshowntheslidesforthethree

tasksin a mixedrandomorder.Halfwereshownallthe slidesforeach

taskin a block.Wethoughtthatthelatterprocedure, whichdidnot

requireswitching to a different aspectof the wordfromtrialto trial,

wouldbeeasier, particularly fortheyounger readers. Flexibility in this sensehasincreased withagein at leastoneotherexperiment (Pick&

Frankel, 1973), although thematerials usedweredifferent andtheage Theresults ofthisexperiment, asregards main effects, wereasexpected. Thesecond-grade children mademosterrors andwereslower onevery range younger.

comparison thanthe two oldergroups,althougherrorswerefew over-

all.Thefifth-grade subjects wereslower thantheadults. Allthesubjects werefaster whenthetaskswereblocked, but,somewhat tooursurprise therewasa negligible interaction ofblocking withage.Everyone found switching troublesomebut ontheotherhandeveryone, including the

second-graders, coulddoit.Thelooks taskwaseasiestandthemeans

taskhardest. Judging synonymy isespecially difficult forsecond-graders.

Theinteraction of age x taskwassignificant whenmeans wascom-

paredwithlooks (p< .01)andwhenit wascompared withsounds (p < .01).

Thedistractor items,compared withneutralitems,increased decision

times(e.g.,it is harderto decidewhichwordmeansthe sameas the

standard whena thirdwordispresent thatlooksor sounds likeit,than

whenthethirdworddoesnotresemble it in anyobvious way).The

distractor effect heldatallagelevels. Theyounger children, however, were

moredistracted bya lookslike distractor onthesounds taskthanwere

theolder subjects. Oddly enough, theeffect ofdistractors wasnotespe-

cially strongforthemonthemeanstask,probably because thetaskwas so hardfor themin general.

Allin all,thisexperiment supports thehypotheses thatonekindof information inaprinted wordcanbeabstracted andcompared when atask requires it,andthatthedecision becomes moreefficient withage.Random switching ofthetaskwasharder fortheyoungest subjects, butit washard

foreveryone (even theadults commented onit).Themoststriking age

E. J. Gibson

488

differencesshowed up in length of decisiontime and in making the "means alike" decision. The second-grade children knew what the words meant, in the usual sense, and would have no trouble understanding them in an appropriate sentencecontext. The judgment of synonymy of two single words

is difficult

for them , however , both with

and without

a distractor

present.

Skillful deployment of attention in extracting relevant verbal information evidently improves with age (as well as reading experience) and contributes

to the economy of information pickup. I seeno way of understandingthis fact better , however , in the context of an information -processing stage analysis. I see instead the need for reexamining the concept of attention ,

and weighting heavily the role of the task assigned(either by an experimenter or by oneself). Perhapsattention is simply perceiving that information which is coincident with task demands. Skill then will vary greatly with the task, as it did in this experiment . Perceiving that something has utility for the task certainly does improve with age, but I doubt that we can understand this improvement better by inserting another stage in a pro -

cessing chain and calling it "attention." The danger of homunculizing looms up, and with it comesthe dangerof pushing the problem further into obscurity rather than solving it . Conclusion

And now, to come to a conclusion, or to put it more accurately, simply an end, am I not, in repulsing information processing, preventing myself from theorizing ? Will my intention of describing what is learned in percep-

tual learning, and what the trends are in perceptual development, render me a mere I doubt it . aversion to as scientists

collector of facts , a file cabinet for assorted odds and ends ? It is the themes of atomism and construction that I have an . But I have a theme too , in Holton ' s (1975 ) sense . We all do , we are as born and bred to seek them as Brer Rabbit the briar

patch. What we have to fear is that those ideasthat are easily tagged will deteriorate into "vogue words " and "vogue concepts" (Merton , 1975). Hardening them by mathematicizing them will only make things worse if

they are wrong, sinceit will frighten the innocent into an uncritical respect for them .

My themes are that there is information in stimulation and that what we

experiencedoes not come in bits and particles (a theme I have obviously borrowed from JamesJ. Gibson); that we should speak of learning and development in perception as differentiation rather than construction; that the relevant processes are more akin to discovery and abstraction than to association and integration ; and that man evolved as a seeker of inforrna -

489

How Perception Really Develops

tion , not an intellectual pauper who must build something to believe in, because there would be no truth otherwise .

Notes

1

.

Piaget

,

Basic

]

.

The

,

2

.

Taken

from

3

.

Letter

position

Tresselt

4

.

The

for

cube

of

5

.

1965

rules

P

.

( is

reality

and

of

course

when

many

phonetic

,

is

) ;

( Translation

by

Margaret

Cook

) .

New

York

:

) .

considered

in

McMahon

pictured

it

of

a

one

;

has

real

,

as

its

moves

because

a

frequency

count

by

Mayzner

and

parts

because

it

but

has

it

no

means

unity

,

I

together

texture

of

,

by

move

continuity

object

are

having

still

and

hangs

,

when

and

contours

the

think

together

same

as

because

it

the

it

moves

substance

rules

takes

,

up

and

stay

different

surfaces

.

A

better

from

drawing

than

of

the

a

array

.

,

and

a

perceived

because

observer

and

or

object

fragments

,

are

object

An

properties

discontinuous

. ,

child

( 1968

length

an

.

the

;

S

the

Dominowski

- dimensional

background

,

in

.

word

unity

of

three

lacks

4

) .

for

,

Condry

graphic

.

and

words

together

of

1954

Ekstrand

(

space

the

construction

Books

M

semantic

. ,

&

Levy

information

,

A

.

A

developmental

from

investigation

words

.

( submitted

of

for

publication

extraction

of

)

References Barron , R. W . Locus of the effect of a distinguishing Memory and Cognition , 1975 , 3, 302 - 310 .

feature in a memory

search task .

Barron , R. W ., & Pittenger , J. B. The effect of orthographic structure and lexical meaning on " same- different " judgments . Quarterly Journal of Experimental Psychology, 1974 , 26, 566 - 581 .

Beilin, H. Developmental determinants of word andnonsense anagramsolution.Journalof VerbalLearning andVerbalBehavior , 1967, 6, 523- 527. Beilin, H., & Horn, R. Transitionprobabilityin anagramproblemsolving. Journalof Experi mentalPsycholo ~y, 1962, 63, 514- 518. Biederman , G. B. Supplementary report: Recognitionof tachistoscopically presentedfiveletter words as a functionof digramfrequency . Journalof VerbalLearning and Verbal Behavior , 1966, 5, 208- 209. Chase , W. G., & Simon, H. A. The mind's eye in chess . In Chase , W. G. (Ed.), Visual infonnation processing . New York: AcademicPress , 1973. Pp. 215- 281. Doehring, D. G. The acquisitionof rapid readingresponses . Monographs of theSociety for Research in ChildDevelopment . 1976, 41, SerialNo. 165. Dominowski , R. L. Anagramsolving as a function of bigramrank and word frequency . Journalof & perimental Psychology , 1967, 75, 299- 306. Dominowski , R. L. The effectof pronunciationpracticeon anagramdifficulty. Psychonomic Science , 1969, 16, 99- 100. Ekstrand , B. R., &: Dominowski , R. L. Solvingwordsasanagrams . Psychonomic Science , 1965, 2, 239- 240. Ekstrand , B. R., &: Dominowski , R. L. Solvingwordsasanagrams : II. A clarification . Journal of & perimental Psychology , 1968, 77, 552- 558. Fodor, J. A., Garrett, M. F., &: Brill, S. L. Pi ka pu: The perceptionof speechsoundsby prelinguisticinfants. Perception andPsychophysics , 1975, 18, 74- 78. Fries,C. C. Linguistics andreading . New York: Holt, Rinehart& Winston, 1963.

490

E. J. Gibson

Gibson, E. J., Bishop, C. H., Schiff, W ., &: Smith, J. Comparison of meaningfulnessand pronounceability as grouping principles in the perception and retention of verbal material. Journalof ExperimentalPsychology , 1964, 67, 173- 182. Gibson, E. J., & Levin, H. Thepsychologyof reading. Cambridge, Mass.: MIT Press, 1975. Gibson, E. J., asser, H., & Pick, A . D . A study in the development of grapheme- phoneme correspondences . Journalof VerbalLearningand VerbalBehavior, 1963, 2, 142- 146. Gibson, E. J., Shurcliff, A ., & Yonas, A . Utilization of spelling patterns by deaf and hearing subjects. In H. Levin & J. P. Williams (Eds.), Basicstudieson reading. New York: Basic Books , 1970 . Pp 57 - 73 .

Golinkoff, R. M . Children's discrimination of English spelling patterns with redundant auditory infonnation. Paper presentedat American Educational ResearchAssociation, Feb ., 1974 , New

Orleans , La .

Hansen, D ., &: Rodgers, T . S. An exploration of psycholinguistic units in initial reading. In K. S. Goodman (Ed.), Thepsycholinguisticnatureof the readingprocess . Detroit : Wayne State University Press, 1968 . Pp . 59 - 102 . Holton , G . On the role of themata in scientific thought . Science, 1975 , 188, 328 - 334 .

Johnson, N . F. On the function of letters in word identification: Somedata and a preliminary model . Journal of Verbal Learning and Verbal Behavior, 1975 , 14, 17- 29 .

LaBerge, D . PerceptualLeaming and Attention : In W . K. Estes (Ed.), Handbookof Learning and Cognitive Processes , Vol . 4. Hillsdale , N . J.: Lawrence Erlbaum Associates , 1976 .

Lavine, L. a . Differentiation of letterlike fonns in prereading children. Developmental Psychol ogy . 1977

, 13 , 89 - 94 .

Mayzner, M . S., &: T resselt, M . E. Anagram solution times: A function of transition probabilities. Journalof Psychology , 1959, 47, 117- 125. Mayzner , M . S., &: T resselt, M . E. Anagram solution times : A function of word transition

probabilities. Journalof & perimentalPsychology , 1962, 63, 510- 513. Mayzner, M . S., &: Tresselt, M . E. Tables of single-letter and digram frequency counts for various word -length and letter-position combinations. PsychonomicMonographs (Suppl.), 1965, 1, 13- 32. McClelland , J. Preliminary

letter identification

in the perception of words and nonwords .

Tech . Report No . 49 , Center for Human Infonnation Processing , University of Cali fornia at San Diego , 1975 . Merton , R. K . Thematic analysis in science: Notes on Holton ' s concept . Science, 1975 , 188 , 335 - 338 .

Murrell , G. A ., &: Morton , J. Word recognition and morphemic structure. Journalof & perimentalPsychology , 1974, 102, 963- 968. Pick, A . D., Christy, M . D., &: Frankel, G. W . A developmental study of visual selective attention. Journalof ExperimentalChild Psychology , 1972, 14, 165- 176. Pick, A . D ., &: Frankel, G. W . A study of strategies of visual attention in children. DevelopmentalPsychology , 1973, 9, 348- 357. Rosinski, R. R., &: Wheeler, K. E. Children's use of orthographic structure in word discrimination . Psychonomic Science , 1972, 26, 97- 98. Spelke, E. Infants' recognition of moving faces. Paperpresentedat meeting of the American Psychological Association, Chicago, 111 ., September, 1975. Spoehr, K. T ., &: Smith, E. E. The role of syllables in perceptual processing. Cognitive Psychology , 1973, 5, 71- 89. Stachnik, T. Transitional probability in anagram solution in a group setting. Journal of Psychology , 1963, 55, 259- 261. Vellutino , F. R., Smith, H., Steger, J. A ., &: Kaman, M . Reading disability: Age differencesand the perceptual deficit hypothesis. Child Development , 1975, 46, 487- 493.

HowPerception ReallyDevelops 491 Vurpillot , E., Lacoursiere, A ., de Schonen, S., &: Werck, C. Apprentissage de concepts et differenciation. Bulletin de Psychologie , 1966, 252(XX), 1- 7. Weisstein, N ., & Harris, C. S. Visual detection of line segments: An object-superiority effect. Science, 1974 , 186 , 752 - 755 .

Woodworth , R. S. E.rperimentalpsychology . New York: Holt , 1938. Yonas, A . The acquisition of information-processing strategies in a time-dependent task. Unpublished doctoral dissertation, Cornell University , Ithaca, N.Y., 1969.

Readingin Retrospect:Perception, Cognition, or Both ?

Often nowadayswhen I talk to an audienceabout the ecologicalapproach to perception, I am askedwhether this approachhasanything to say about cognition , or whether there must be a firm line drawn between perception

and cognition with different principles applying. The first answer to this questionis that perception is cognitive. Cognition has to do with knowing. The number one definition of cognition in my favorite dictionary (Random House) is "The act or process of knowing: perception." Many psychologists think of cognition exclusively as problem solving, reasoning, remembering, and so on, however. I like to point out that these processes begin with and dependon knowledge that is obtained through perception, which extracts information from arrays of stimulation that specify the events, layout, and objects of the world . The ecological approach holds that this processis a direct one, in that the information is picked up without the intermediary of secondarysources, like inferencefrom past experience or from premises that are somehow inherited .

The reading processstandsin an interesting relation to perception with respect to this question . Like perception , it has the function of extraction of

information, and the information doesspecify something in the world. But reading is indirect in a way becausethat information is carried in text rather than in optic or acousticarrays of stimulation specifying layout and events. Does that make it different , more cognitive ?

Whether it does or not, we can look to see whether principles that characterizeperception also characterizereading. The ecological approach to perception has three important principles that set it off from other theories of perception. We can ask whether theseprinciples also apply to reading. The first of them has to do with information: there is information in arrays of energy available to our perceptualsystemsthat specifies events, layout, and objects in the world. The information may specify

494

E. J. Gibson

permanentproperties of the world, in which caseit is invariant over transformations, and spread over time. This principle is quite distinct from theories that characterizestimulation as impoverished, nonspecific, and requiring supplementation. Perhapswe can carry this principle over to reading. Information is carried in text. It specifieslanguage, which in turn specifiesevents, places, and so on in the world . There are indeed invariants in language. Although the rules for them are not the sameas those that specify, for perception, something like the persisting properties of a place or an obiect, they can substitute for them and they are most certainly spread sequentially over time. One of the principal tasks of a perception psychologist is to describethe information specifying events in the world (seeJ. J. Gibson 1979, chapter on information). How is information in an ambient array able to specify the environment? A comparabletask exists for the reading psychologist- what kinds of information are carriedin text and how do they specifythe world of places, objects, and events? There are more levels of correspondenceto be described. The secondprinciple of the ecologicalapproachis that information must be obtained by an actively seeking perceiver (J. J. Gibson 1966). Any teacher will agree that this principle applies to reading; mere mechanical decoding, converting a letter to sound, is worthless. The reader must be searchingfor what the text specifies,what it carriesinformation about. One looks aheadfor information that is foreshadowed. What the text specifiescan be likened to what is perceived. Here is the third principle unique to the ecological approach. What is perceived first and foremost are the affordancesof events, places, and things. An affordancefor perceptioncanbe thought of as the utility of someenvironmental resourcefor acting by an animal possessingthe appropriate action system. Texts have affordancestoo, divers ones, for an organism possessingthe active skills required for reading. The affordancesbranch out to all aspects of our lives in the caseof reading. A text may afford finding out how to bake a Genevoisecake, or the telephone number to call a broker in San Francisco , or words and music for singing a hymn. The act of reading will be different, depending on the kind of information available, and part of learning to read is learning how to obtain most economically and use the kind of information in a text. I did not try to .comparereading with direct perception when I was engagedin reading research.But the connectionswere becoming apparent when Harry Levin and I wrote our book on reading. One chaptercaptures the aooroach more than the others, the one called "Learning from Read. L . ing." It emphasizesthe functional aspectsof the reading task- the extraction of relevant information, use of the information, and what kinds of information are extracted. It is a far cry from someof the models described

Readingin Retrospect : Perception , Cognition, or Both?

495

in the book's final chapter, ones that depend on ludicrously complex diagrams with flow charts and boxes. The emphasisin the leaming-fromreading chapter is on the acquisition of knowledge- what reading, like perception, is all about- a very cognitive process.

VI Perceptual Development from the Ecological Approach (1972 to the present)

Introduction to Part VI

My years of researchon the reading process culminated in the book Psychologyof Reading(a title deliberately copied from Huey's book, published a half-century earlier). I always thought of this work as being a diversion from my concernwith perceptualdevelopment, and in a sensea battle had been won. The stagnation in experimentalresearchon reading was overcome and the field looked healthy and flourishing. I had another reasonfor moving on. I had finally acquiredmy own laboratory. I could at last take any direction

of research I wished .

Since 1891, Cornell's laboratory of psychology had been situated in Morrill Hall, the oldest building on the Cornell Campus. The two top floors of this building had housed Titchener's laboratory, and they were still in use. The shop, the animals, the graduate students, and most of the lab space were on the top floor , very hot in summer and without even an elevator to haul up equipment and animals. For many years I had been

obliged to work in someoneelse's laboratory, at first for lack of rank and support, and later for lack of space. A lot of the reading researchwas conductedin schools, but even so the situation was desperate.I was always quite successful in getting grants , but what to do with the equipment , the

data to be stored, and personnelbecamemore and more perplexing. I was not the only one with a spaceproblem. My husbandneededmore laboratory space and so did others. One chairman after another fought the administration on the spaceproblem. Finally, in 1967, the departmentwas given temporary researchspaceat the airport in a building formerly used for researchby General Electric. Along with this compromise came a promise of a new building for psychology. The spaceat the airport laboratory was good- lots of it, air conditioning, plenty of parking- but of course, it was inconveniently far from campus. In the end, only the department shop and the Gibsons and their graduate students moved out there. We had a number of graduate stu-

500

PartVI

dents and many visitors came from overseasand other universities, so we weathered the move without too much disruption and even enjoyed our rather exclusive club. It was a great day, however, when the new building was finally completed and everyone moved back together on campus, in 1972. The great advantage for me was that I was given the opportunity to plan my own laboratory. I chose to plan a laboratory for research on perception in infants, and I began switching my research priorities. Fortunately, a number of excellent graduatestudentswere interestedin working in this area. There was a richness of problems to work on. No one had as yet examined any aspect of the new ecological approach in researchwith infants. I had become very~ interested in that approach as JamesGibson worked it out and formulated the conceptsneededto embracea radically new way of thinking about perception. The ideasworked especiallywell, it seemedto me, for a behaviorally and developmentally oriented domain, such as perception in preverbal infants. The chance to pursue the new ideasand establisha new laboratory camenone too soon, becauseI had to retire (formally, that is) in 1979. JamesGibson died that sameyear, leaving me to extend his ideasto perception in infancy without his help (although I had the assistanceof somevery bright young colleagues). He did help plan one of the earliestexperimentson perception of affordancesin infants. This was an experiment on looming (Schiff, 1965), already studied in infants (Bower, Broughton, and Moore 1970, Ball and Tronick 1971), but we adaptedit to ask a more specificquestion about perception of affordances for collision: will an infant detect the difference between approach of a closed contour that affords passing through (an aperture) and one that affords an abrupt stop (an obstruction or obstacle). One of the following papersdescribesthe information provided in the experiment for differentiating the two cases . The topics that I set out to study in the new infant laboratory included ones that were generating new researchwith adults, but they were presented so as to be appropriate for infants. They included: the perceptual pickup of invariants in an array of information; the perception of multimodally specifiedevents; perception of affordances , and development of mobility . Affordances begged for attention, becausethat was the last important concept launchedby my husband. There was very little research availableas yet, and the conceptwas misunderstoodand oversimplifiedby many people. In the courseof researchon infants' perception of the affordancesof surfaces , I becameinterestedin the conceptof mobility in general and how it progresses. That also is a natural problem for the ecological approach, which stressesthe inseparabilityof perceptionand action, and so our researchwas later extendedto it (Gibson and Schmuckler1989).

Perceptual Development fromtheEcological Approach(1972to thepresent )

501

Researchover theseyearsand attempts with others to refine conceptsof the ecologicalapproachand carry it further have gradually had the effect of directing my theoretical concernstoward more epistemologicalquestions. As a young scientist such questions concerned me very little . I am a natural behaviorist. But concentrating on what infants perceive and how they come to know the world forcesone to confront the big epistemological questions(as Piaget had long been telling us). I was not content with Piaget's answers, so I have been attempting to formulate my own. Several of the papers that follow introduce these concerns. Perception as the foundation for knowledge of the world and the role of exploratory behavior as an active searchfor knowledge emerge as major themes in an ecological approach to development. Learning is again implicated, and a new path to perceptualleaming unfolds.

27 TheSenses asInformation -SeekingSystems Eleanor Gibson , James J. Gibson

Thisfinal sectionbeginswith two short newspaper essayswritten by invitation for the London Times Literary Supplement. Togethertheypresenta confronta tion of two radicallyopposed contemporary viewsof perception : thedirectpercep tion view of the ecological approachand the indirectperceptionview of Richard Gregory , spokesman for a traditional, concept -mediatedapproachto perception . Gregory , a prominentBritish psychologist , and a friend of many yearsstanding , proposedthe debate . Thebrief sparringputs the controversyin a nutshell , some timescloseto caricature , but theissuesaredrawnand the locusof disagreement is clear. Two issuesstand out in theseessays : where the action is, and where the informationis. For the Gibsons , perceptionis itselfan active, information-seeking process of searching ambientarraysof energyfor informationaboutthesurrounding environment . For Gregory , who alsoclaimsto bean activist, the actionis in a kind of thinkingor problem-solvingprocess following afterperception , whichfor him is a passiveintakeof whateverenergyhappensto fall on receptors . For the Gibsons , theambientarray is rich in informationthat specifies layout, objects , and eventsin theworld. For Gregory , only dregsor hints of informationareavailable to perception , which must be supplemented by inference , piecing togetherthe evidencewith the aid of past experience , working like a detectiveor a solverof puzzles . How thepast experience , which mustitself havebegunas impoverished , becomes meaningfulis a problemand may leavethe theoristwith the disquieting beliefthat everydayperceptions are "ima~inativeconstructions " or "fictions ." ~ Gregorycomparesthe Gibsons ' position with B. F. Skinner's psychology , implying that it is an elementaristic , one-to-oneS-R kind of position. What the ecologicalapproachactually doesis stressthe interleavingand inseparabilityof perceptionand action. A perceptualsystem incorporates actions(e.g., headand eyemovements , accomodation , etc. in the visualsystem ), perception guidesaction, (London ) TimesLiterarySupplement , 1972, June23, 711- 712.

504 and as

E. J. Gibson & J. J. Gibson

action an

informs

inseparable

mental

and

guided

like

the

the

all

we

with

the

know

problem

ability

to

deal

ment

that

we

individuals

are , like

. It

behave

is our

we

must

thrnuQ ~. ~. _ ~- ' h . and

first

what

. The

are

exception

. Man for

They Hence

also is

adaptation The

answer

systems

-

the

attending

to

organism

on for

their

agree in he

question

and

are that

that

an

animal

that

. What

innately

has

given

produced

or of from

of

the

controversial

organs first

what

life

of

clear

all

, a mate must

with

of the

, escape

information

exciting or

. None

his

fresh

, and

sleeping

of would

so

, and

he .

us

believes certainly

environment

information

information answers

sensory means

, shelter

Gregory

getting

.

.

world

. Richard

.

evolution

. The

have

, grazing sorts

the

its

surrounding

, sexually flying

to de -

perceive

the

world

. He

sense

them

they

provide

food

of question

.

seems -

is the

equipment

course

the

interacting , and

nature

his

, unfriendly

events

constantly

the

of

find

walking

comes

something

the

hazards

about

is

in

activities

must

the the

.

of

around

from

from

it . He

is

this

organ

survival

with

now

related

for

world

is

is

course

function

, friendly

safe

; and

, doing

done

in

the

information

the

information

edible

layout

in of

obtain approach

environment

learn

species

environmental

are

the to

the

their

survive from

about

information has

question

getting

that

spatial

this

individual

organs

places

Information that

the

and

objects

; about

an

must

predators

about

of to

of

, and

goes move

an

and

this

order

continuously

understand

genetically

ability

in

as

what

we

-

and

perceive

. How

on

environ

species

a developmental

to

that

from

the

for

question

endowed

a

the

about we

evolution

obvious

as

must

as

that head veridical

of

information

seeking

same

use we

people

comes

events

environment

the

stimuli

adaptation

and

that

from

the

with

both

in

seems

of

the

the

claiming

remarkably

and

form

question .

in

perception

that

behavior

is

by

-

computer

approach

it

it

action

Develop

everyday

think

and

be

(see

perception

actually

with

must

with

are

reception endowed

of

our

cope

is the

asks

animals

we

, up - to - date

-

asks

. But

environment

serves

, it

direct

, objects

, there

relations beginning ~

it

for

to

the this

than

the

view

and

, stresses

, dismissing

substances

biologist

other

there

and

a

problem

a current

, our . very

the

. Indeed

the

are

in

perception

, obliged

. For

it

of the

hand

rather that

perception

research

other

illusions

account

of of

the

ecological

with

all

psychologist

and

signed

not

function

species

ask

or

obtain

question

no

to and

relations

maintain

adaptive

attempt

at

O J ' -' . - - --Jfrom -information

The

surfaces it

successfully

on , and

The

an the

.

on

study focus

, on

implication

up

illusions

in with

the

popular

studying

problems

given

are

a

is an

"

years

. 6 ) . Gregory of

" inside have

recent

become

, No

. There

who

can

has

desirability

tough

Gregory

. In

, 1989

locomotion

neglects

to

system

Psychology

programs like

perception

is

getting

, taking from ? This

what is

the

Sensesas Information-SeekingSystems

505

There is a contradiction at the very heart of the existing theories of

senseperception which can be expressedby the following two assertions: the

senses

cannot

be

trusted

and

the

senses

can

be

trusted

. From

the

biological point of view it would seemthat the sensesmustbe trusted since they are all we have. How else is an observer to keep in touch with the environment? But everything we know about the physiology of the senses seemsto show that the inputs of the sensory nerves are inadequatefor keeping in touch with the environment, sincethey are nothing but signals touched off by stimuli . How to resolve this contradiction

is the chief

problem for understandingsenseperception. Existing theories try to resolve this contradiction by assuming either that the inputs of the sensory nerves are corrected by the brain , or that the

corresponding senseimpressionsare interpreted by the mind. There are many such theories of correction , inference, interpretation , compensation ,

equilibration, organization and the like, all taking it for granted that the process of perception is some kind of operation on the deliverances of

sense.But there is another way of resolving the contradiction that seemsto us more promising . It is to assume that the inputs of the sensory nerves are

merely incidentalto the processof perceptionand that the useful sensesare actually perceptual systems. These are systems which adjust the sense organs instead of just receiving stimuli ; systems that have output as well as

input; systems that explore the light and sound and pressureof the environment

for the information

contained . This is information

about

the

sources of the stimulus energy , not merely signals. Professor Gregory believes that there is "a cognitive element in percep-

tion." He is saying that one has to know something about the environment in advancebefore he can perceiveit properly. But there is a dilemma here. Surely one cannot know anything about the environment exceptas he perceivesit, or has perceivedit . The trouble comesfrom the long-standing assumptionthat the sensesat birth can deliver nothing but meaninglesssignalsover the sensorynerves, signals that have to be interpreted in the slow course of learning by association. On that assumption it follows that knowing must precede perceiving. But perhapsthe long-standing assumptionis wrong. We might assumeinstead that the senses , even at birth, are perceptual systemsthat pick up facts about the world. On the first assumptionlearning is a matter of supplementingthe infant's bare sensationswith memories. (But memories of what? Sensations ?) On the secondassumptionlearning is a matter of distinguishing the information, of discriminating not of associating. Bare sensations have no meaning until they have been enriched in some unknown way by past experience. But primitive perceptions have primitive meanings from the very outset of life .

506

E. J. Gibson& J. J. Gibson

Psychologistsand physiologists who are of ProfessorGregory's persuasion take it for granted that lights, sounds, chemicals , and contactsare mere stimuli for the receptors in the eyes, ears, nose, mouth , and skin, and they suppose

that

stimuli

as such carry

no information

about

their

sources

in the

world . But we reject this doctrine and try to show that an array of light, for example, specifiesthe surfacesfrom which the light is reflected. At least it does so in the world outside the laboratory . Similarly , a natural flow of

sound specifiesthe vibratory event from which it comes, a taste or an odor specifiesthe substancefrom which it emanates , and the sequenceof pressures obtained by feeling an object specifiesthe object. The notion of information in stimulation is novel and unfamiliar , but it is the basis for a new theory of perception . In life , the sea of stimulus

energy

in which

an observer

is immersed

is

always an array and always a flow . The stimuli as such, the pin -pricks of

light or sound or touch, do not carry information about their sources. But the invariant properties of the flowing array of stimulation do carry information. They specify the objects of the world and the layout of its surfaces.They are invariants under transformation, non-changeunderlying change. Note that they are not in any sensepictures or imagesof objects and of layouts as so many psychologistshave been tempted to think. Nor are they signalsfrom the objectsand surfacesof the environment like dots and dashesin a code. They are mathematicalrelations in a flowing array, nothing less. The demonstration that a natural array of light specifiesthe surfaces from which the light is reflected, together with the layout and composition of thesesurfaces , dependson a new approachto optics, an approachthat is ecological instead of physical. It begins with ambient light at a point of observationinsteadof radiant light from the source. It studiesthe structure and transformations

of this ambient

array

instead

of rays or waves

or

photon-paths. It treats the eye as an organ instead of a cameraand rejects not only the doctrine that the retinal image is a picture but even the notion

that it is an image, properly speaking. The old theories of sense perception assumed that it consisted of the operations of the mind upon the data of sense. If this notion sounded too

philosophical it could be made to sound more scientific by assertingthat perception consistedof the processingby the brain of the signalsarriving over the sensory nerves. This is the modem formula , but actually it is the same old theory . It still says that the act of perceiving is something that occurs wholly inside the head. The new theory that we advocate says that the act of perceiving occurs in a circular process from the sense organs to

the brain then back to the senseorgans, and so on. It involves exploration by the eyes of the whole array of light and exploration by the handsof the

Sensesas Information-SeekingSystems

507

whole layout of surfacesaround one. Man's delicately mobile postural system, which includes the eyes, head, hands, and body, is beautifully adaptedfor this activity . Perceptiontherefore does not have to be conceivedas the interpreting of messagesor the learning of the so-called I'sensory code." It is the exploring of an array, the enhancing of available information, and the optimizing of its pickup. The eyes, for example, look around, focus their lenseson details of the world, and modulate the intensity of the light when the illumination is too high or too low. For listening, the head turns to equalizeintensity of input to the two earsso as to point the head towards the source of sound .

The assertionthat the information in stimulation specifiesits sourcesin the world does not imply that this information is automatically picked up. It is available, but it mayor

may not be perceived . An observer must

extract the information from the flowing array of stimulation. And he must often learn to do so. What is it that the human observer learns? We suggest

that, beginning as an infant, he learnsthe distinctive featuresof objects, the layout of places in the environment , and the invariant features of events- A

human observer also perceivesrepresentationsof things and places and events, of course, and in that case the information coming from the picture or the television screen is essentially the same as it is when it comes from the environment . Finally , a human observer learns to extract information

from the constituentsof spokenand written language, but this is information of a quite different sort . It is not essentially- the same as that which comesdirectly from the environment. The child's learning about the world from speech, and then from writing , is a much more complex processthan learning about the world from what we call firsthand experience . Here is a brief accountof the developmentof perception of objects. The processbegins in the newborn infant with visual attention to certainsalient stimulusproperties that carry information: motion, brightnesscontrast, and the kind of contrast provided by the edge of a surfacein the world. The infant's attention is "caught" by these properties. The world he peceives, then, is not at all a "blooming, buzzing confusion," as William Jamesput it, for he at least seessurfacesand edges. But this is only the beginning, since objects gradually becomedifferentiated from one another by their distinctive features, that is, by attributes that render each object different from

other objects. For example, babiesdifferentiatehumanfacesfrom non-faces in their environment very early, although it is doubtful that they perceive the relations between the features of a face before they are three months old , or thereabouts . Individual

faces are not differentiated

from one another

until six months have passed. Other properties of an object suchas its size and shapeare differentiated within the first few months of life, before the

508

E. J. Gibson& J. J. Gibson

baby can walk or even reach. There is no substance in the old notion

that such visual attributes must gain their meaning from touching and .

graspIng

.

A human face, of course, has properties that are not constant over time

aswell as propertiesthat are. The movement of the facial musclesproduces different expressionsthat portend different events. Moreover a moving face usually produces sounds. Interestingly enough, an infant at twenty days perceives the voice as coming from the face- he does not seem to have to learn to connect these sensations by associating the sound with the

sight. An object comesto be perceivedas permanenteven when it is partially or entirely hidden by another object . If a screen is drawn in front of an

object so that it is gradually concealedand then gradually revealedagain, an infant soon learns that it has not gone out of existence and expects its

reappearance . There is optical information for its continued existenceand for its only having gone out of sight. This is not the same thing as remembering- the object . Later on, when a child has learned names for familiar objects that he hasdistinguishedfrom one anotherby their distinctive features, he knows things about objects that he can remember and think about, but perceptualdifferentiation is basic to this knowledge. The differentiation

of the features of the environmental layout also

develops without having to be supplementedby knowledge. When a crawling infant is placedon a platform with a visual cliff on one side and a very shallow drop-off on the other side (but actually a glass surface of support on both sides) the infant will crawl to its mother over the shallow

side but not over the deep side. Is this becauseit has "knowledge" that a cliff is dangerous? This seemsunlikely. The baby hasno past experienceof falling and surely does not inherit racial memoriesof falling. What about the perceiving of events? Events occur over time and are of

many degreesof complexity, sincea short episodemay be embeddedin a much longer episode. If perception were really basedon single elementary sensations , each successive

sensation

would

have to be somehow

inte -

grated for the total event to be perceived. Again, it seemsthat learning proceedsby differentiation, not by integration. For example, if an object aovroaches

an observer

on

a collision

course , he

will

blink

or

duck

or

L L dodge so as to

mitigate or avoid the collision. The optical information for this imminent event is the progressivemagnification of a silhouette in the field of view. Experiments consisting of the display of this information have been done with several species of animals and with human infants . The shadow of an object is cast on a translucent screen in front of the observer and its size is increased at an accelerating rate. The adult human

observerperceivesa virtual object approachinghim. Turtles facedwith this

Senses as Information -Seeking Systems

509

display pull their headswithin their shells. Monkeys cry out and rush to the rear of the cage. Human infants, at two weeks, respondconsistently with a backward jerk of the head and by raising the hands. A little later they differentiate between the information for an object on a collision course

and that for an object on a non-collision course. This differencedependson the symmetry of magnification. A perception of this sort can hardly be a matter of successivesensations . It must be that optical motions of different kinds are distinguished from one another as perceptual development proceeds. Different events involving motions are differentiated very early in life. The sameanimal that retreats from an approaching object may follow a retreating object . Baby chicks run away when they are faced with the

optically expanding shadow on the screen, but they move towards an optically contracting shadow on the screen. This responseto the diminishing shadow is related to the imprinting that occurs early in the life of a young bird suchasa chick or duckling. It runs after a retreating mother and thus succeeds in staying with its protector and with its kind . And it demonstrates for us, incidentally , that two contrasting kinds of events are

distinguished. The early developmentof perceptionseemsto us clearly to demonstrate the picking-up of information that is available in stimulation and not the supplementingof sensationsby memoriesof past experience, or by some kind of knowledge. But, the reader may ask, what about symbols like words? They are perceivedtoo. Aren't they at least a clear caseof supplementing auditory sensationswith an associatedmeaning? The analysisof the information in a speechevent tells us that it hasthree quite different kinds of information, all of which must be comprehended. There is the sound itself, the phonetic sequence , to be perceived. There is the syntactic information , the rule system that governs how words are put

together. And there is semanticinformation, the "meaning." How does a child learn to pick up all this information? One thing seemscertain- he does not simply learn by association. How then? The first essentialto this developmentis what the linguists call segmentation of the sounds of the speechsystem. Speechcomes in a physically continuous flow , usually without

the separation we seem to hear. This

streammust be analysed. It is analysedat many levels, the lowest of which is considered to be the phoneme and the higher levels being syllables, words, phrases, etc. But phonemesthemselvesmust be differentiated. They are differentiated from one another by sets of contrastive features. These

distinctive features have a developmental sequenceof their own, as the linguist RomanJakobsonhas taught us. The first differentiation is between the optimal vowel and the optimal consonant, and development goes on from there in a seriesof ordered splittings.

510

E. J. Gibson& J. J. Gibson

It seemsunquestionable that this processmustbe oneof differentiation , not of association. The features cannot be associated with anything , since, as Jakobson said, they indicate mere "otherness." The same twelve pairs of contrastive

features serve to differentiate

, in various

combinations

, all the

phonemesin humanspeech.The phonemesthemselvesare abstracted,by a processof analysis, for one cannotbe heardalone, choppedout of a speech segment. Yet we do all differentiate it and acknowledgeits constancy. We do not learn to perceive phonological featuresof speech, then, by adding something on. The secondessentialin the learning of speechis grammar. Noone has succeeded in accounting for a child 's acquisition of grammar by an associative process. A child 's first sentences are not copies of the sentences of

adults, but they neverthelessfollow rules of grammaticalconstruction in accordancewith the relations expressed , suchas agent-action, agent-object, and action -object . What the child has learned appears to be the result of an

inductive process- the extraction of relations from information presented to him in adult speech. The third essentialin learning speechis meaning. How do words come to have meaning for the child ? By associating a word with a referent , like the word " kitty " with the animal referred to? This is the answer that used

to be given, but it seemsunlikely. Meaning in speechis not conveyed by single words, but always in a relational context. For example, when a child says JJkitty all gone" or JJherekitty " he is referring to an event in the world. The meaning of the event has been perfectly clear to him for some time. What he has succeededin observing is the correspondencebetween the event itself and what

someone

said about it while

it was occurring .

Children begin by making predicationsabout the immediateenvironment. Again , there seems to be an inductive process involved , an extracting of the relation

between

the two kinds of information

, one in the event itself

and the other in the spokenwords. By this brief survey of the development of perception we have tried to show that a child useshis "senses " in an active and adaptive way to extract information that is present in the ongoing flow of events in his environ -

ment. He does not use previous knowledge to interpret his sensations , or to supplementthem. He could not do so, for he must begin by picking up this knowledge from what goes on around him. The pick-up comesfrom differentiating the complex, embedded, relational, dynamic structure of the world

.

28

Perception

as

Thoughts and

Bell

Eleanor

a Foundation

Inspired

by

Papers

for

Knowledge of

:

Fraiberg

ugi ] . Gibson

I include the brief set of remarks that follow because they illustrate the emergence of interest in questions about the origins of knowledge . The occasion was a symposium at Cornell organized as a memorial for Eric Lenneberg and it focussed on language . There were a number of invited speakers arranged on panels, one of which was composed of Selma Fraiberg , Ursula Bellugi , and Hermine Sinclair . I was asked, as one of the host faculty , to comment at the end of the panel on the three papers. Finding a common denominator was not easy, but I came up with one, the problem of reference, which arose often in the conference. Which comes first , the meaningful percept, or the linguistic symbol ? In other words , what is the origin of the meanings in language , the things that we talk about ? These remarks were never published , although the speakers' papers were collected in a book, but my students were impressed with them . One student , Katherine Loveland , who had a major interest in language development , based her dissertation research on them . She liked the suggestion that the blind child 's difficulty in achieving correct usage of the terms "you " and "! " was at least partly owing to lack of visual information for learning about what Piagetians call "coordination of perspectives." " You " and "I" refer to a speaker in a given place, and the words are interchangeable , as the roles are. Developmentally , learning about the stability of the layout whilst station -points are changing isn 't accomplished until about two and a half years by even the sighted child . The blind child , unable to watch perspective transformations and occlusion -disocclusion as he moves around , is badly handicapped , lacking a prime source of information the persistence of the layout along with changing views of it .

for

Loveland 's research (1984) was a longitudinal developmental study of perspective-taking and achievement of correct usage of personal pronouns . She showed that knowledge of perspectives is always reached before consistently correct

Discussionpreparedfor the Lenneberg Symposium , CornellUniversity, Ithaca, N.Y., May 1976.

522

E. J. Gibson

pronounusage , usuallyshortly beforeit. Thisfinding supportsthe notionof precedence of knowledgeof eventsand featuresof the real world in the genesisof linguisticreference . How elsecould it happen , anyway? The deprivationof the blind child and its ratherspecificeffectson languagedevelopment area persuasive argumentfor it, as thefollowing essaycontends . - - - 0

J

'

,

-

-

The Origin of Linguistic Meanings

Where do the meanings expressed in language come from? Like Eric Lenneberg, I am of the opinion that linguistic meaningsare rooted in the perception of meaningful events in the world . And I would add, in the accompanying proprioception of oneself in the world . Meanings are anchored in knowledge obtained from perceiving the real world and thus they derive from the meanings of things, events, and the spatia/ layout. It is

the latter, the spatial layout and the meanings it affords, that I am concerned with in the point I have chosen to elaborate. It is the juncture where

all three papers came together and struck a responsivechord in my own thinking. What Specifies the Self, and Especially , 1? I was impressedby Dr . Fraiberg's evidence that JJ A delay in the acquisition of self-referencepronouns and the concept JI' is a unique problem for

theblind youngchild." Exactlywhat is it that is missing,perceptually , that makesthis meaning so elusive for a child who cannot see? What specifies the self- myself- n I can think of a number of things that specify the self, and someof them seemto be perfectly available, perceptually, to a blind child. What are they? A useful specificationof oneselfmust be invariant, relative to oneself. An early incidence of such an invariant is the infant hearing himself cry , and

at the sametime activating the peculiar breathing and vibratory patterns that are contingent with it . Hearing the cry alone would not suffice- but hearing a cry that is always contingent on unique breathing and activa tion of the vocal cords permits control of that activity ; it affords self-

activation- pretty good information for oneself. A little later, perhaps, as the child's use of arms and legs becomesagile, all the kinesthetic information for lifting a limb can be experiencedcontingently with feeling an object or a surfacethat is touched or struck- it might be, in fact, one's own mouth that a thumb makescontact with . A seeing infant can witness his hand come into view , contingent on activity of the trunk and arm. But feeling the thumb get into the mouth and tasting it is not a bad substitute . Neither is grasping something that results in

Perceptionas Foundation for Knowledge

523

unique pressuresand sounds, like crumpling paper. Preyer, (1890) discussing development of the "feeling of self," (his phrase) said an "important factor is the perceptionof a changeproducedby one's own activity . . . an extremely significant day in the life of the infant is the one in which he first experiencesthe connectionof a movementexecutedby himselfwith a sense impressionfollowing upon it," (p. 191 FE ). Preyer thought this happened between the 45th and 55th week, when children do a lot of experimenting

like shakingkeys, opening and closing things, and pressingswitches. Many theoristshave remarked, in the past, on the supposeddifficulty of

distinguishingone's self from the environment . Lest sucha permanent confusion actually come to pass, nature has been extremely generous (especiallyto the sighted child) with meansfor discovering this distinction. A meansoften remarkedon is the discovery of one's self in the mirror. A seeing child has the opportunity to move his arms, or grimace before the mirror , and to perceive a remarkable contingency in his reflection , informa tion for one of the most interesting kinds of intermodal invariance . Preyer

set the beginning of this discovery at about the 17th month. But he observed that there was a delay, after this, in the clear use of names, and

especiallythe correct use of personal pronouns. "Many head-strong children," he said, "have a strongly marked 'I'-feeling without calling themselvesanything but their names. Theseobservationsmake it clear that the 'I' feeling, according to the facts given above, is present much earlier" (pp . 202 - 203 ff .).

One other invariant that we, as sighted persons, have always with us is the view straight ahead, the visual field. In the first place, even if we are sitting still, there is the frame provided by the nose and the eyebrows, always occluding the samearea of the layout on either side. And if we move , the expansion pattern of the visual field that results has a motionless

center- a beautiful, external, invariant referencepoint that exactly specifies the movement of the self. Sincepeople don't remain fixed, like trees, they need movable invariants, and they have one. But there is one trouble with all of these sources that specify the

meaningof "myself." None of them seemsto correspondperfectly with the use of the word I. I is not a name for me alone, it is a name for everybody . What PerceptualExperienceCorrespondswith the Meaning of the Word 11

Either the hypothesis I began with, that verbal meanings derive from perceptually given experience, is wrong, or there must be something else that specifiesI; and you, too, when you say the word I! To say that I is relative to oneselfis not enough. It is relative to everyone. I think there is just one thing that is specificto this meaning, and that is the station point

524

E. J. Gibson

where

the

about

it

A

speaker

is

spatial

and

layout

light

,

sources

being

.

reflected

in

this

at

~ he

the

room

world

see

we

same

both

of

see

at

and

you

station

is

of

,

experimenters

,

Pick

children

know

a

than

not

spectives

"

different

an

could

.

child

for

these

cerns

stitute

toward

,

an

4

the

yours

child

speaker

,

' s

when

3

learns

years

I

am

;

-

1

an

downward

area

.

that

in

although

which

they

Other

year

pioneer

age

between

.

2

their

evidence

of

,

perceive

of

Since

the

years

/

to

development

.

relations

the

only

they

had

expe

points

.

experiments

olds

-

Their

( Shantz

realize

meaning

too

of

clear

spatial

why

that

echo

-

of

blind

array

an

is

defined

array

II

the

by

child

be

of

hearing

per

speakers

-

at

accomplishment

would

the

should

coordination

but

,

I

I

blind

and

by

meaning

a

layout

obtained

ambient

are

or

-

of

seems

necessarily

elude

a

blind

.

adult

I

of

eventu

convincing

subjects

be

arguments

for

nod

is

the

normal

able

meaning

for

map

at

that

it

Comprehending

time

9

very

possible

objective

in

some

a

cognitive

that

eventually

locations

difficult

the

correct

,

While

being

.

point

.

or

shown

is

station

,

coordination

to

revising

found

toys

perspective

Perceiving

children

to

time

and

annually

)

younger

have

II

,

.

correct

you

the

7

1974

include

)

will

the

when

between

scheme

of

you

(

and

set

been

(

time

invariantly

,

I

, "

.

locating

one

handicapped

and

.

' Ioffset

environment

if

that

speaking

through

are

that

then

was

move

and

,

owns

what

they

perspective

about

Fariello

who

see

extended

or

from

things

moments

if

place

,

put

layout

can

and

an

,

be

Some

successive

,

-

lnhelder

"

hypothesis

s

am

and

fraction

did

the

only

I

achievement

speaking

Piaget

1971

'

are

you

can

may

but

floor

layout

observer

at

always

I

objective

only

speaker

refer

around

Watson

is

eye

.

he

,

objective

an

An

a

this

spatial

the

now

any

.

the

then

anyone

is

it

when

and

an

investigation

,

)

But

and

surfaces

as

and

in

perspective

.

,

at

)

perspectives

walked

more

of

have

,

had

just

rienced

We

we

furniture

of

that

see

disagreement

.

view

visible

or

They

you

a

as

hidden

that

considerable

Kosslyn

If

do

around

observation

one

when

research

If

how

1966

from

you

terms

and

Gibson

point

perspective

mine

.

,

of

propose

to

J

be

layout

s

.

perspectives

are

.

I

coordination

has

same

Piagetian

objectively

and

and

bounces

have

what

of

'

J

not

see

learned

;

There

had

,

walls

-

station

points

I

versa

space

and

adjacent

and

is

vice

II

,

point

and

point

has

say

another

else

arrive

will

surrounding

in

,

(

it

sequence

someone

, "

ally

station

light

would

surfaces

perspectives

room

and

cannot

layout

take

,

the

whose

spatial

to

a

occluded

I

taking

can

the

a

ambient

some

to

.

the

some

I

be

before

can

is

there

as

move

occluded

as

-

point

or

will

can

the

What

everything

,

eye

such

any

standpoint

in

.

Light

by

room

that

located

?

correct

,

conversation

you

someone

,

or

.

he

in

we

an

are

several

things

should

Indication

of

ordinary

conversation

addressing

,

station

indicating

follow

.

points

One

con

should

.

And

you

so

.

We

-

sub

it

point

-

does

.

to

525

Perception as Foundation for Knowledge ourselves

. We

when

the

stood

, or

you

third

Dr

I are

I learned

of

just

recently

."

space

in

order

something side " and

like " more

relatively

little ,

question

here

I - you

to and

contrast

spatial the

same

How

is

Can

one

sense

time

, for

say

Meaning

especially

,

question

equilibrium

tween

them

A that

I

is

to

of

invariant

for

.

The

) suggested

) is

to

a station

to

be

blind

have to to

similar

a

this

to

the

in

the

point

acquired

child

on

she

at

-

later

, of

is

learned

about course

his

may

then the

rest

of

in

,

a

the

would

a

a

kind

of

be

first

pressure

,

" core

assimilated

"

self to

concept

of ,

self

such some

a

of

them

- image child

the

little

,

later in

as

,

as

about

and

it . A

f.

perceives

moves

reflections

and of

, shaking

he

pattern

by be -

you

information

as

expansion

or , superior

meaning

child a

of

resultant

the

new

relation

developed

perceiving

, and

, followed

the a

the

. Sinclair

between

learning when

and

a

patient

confusions to

the

of

disequilibrium in

and

. Dr

in

equilibrium

compensatory

sources

actions form

the

already

feeling

"

an of

agent

occur

,

' s paper

role

early

multimodal

sounds

ahead

. Sinclair

the

,

specifications

establish

conflict

has

I

the

mechanisms

to

. He

and

between

,

should

of

Dr

terms

of

conflict

of all

that

reversal

notion

resultant

, and

meaning with

analogous

basis

" straight

latter

the

the

grasping

hearing

invariance

two

a

regulatory

what

everyone

the of

and

me

apply

for

on

the

language

the

of

or

a name

contingencies thing

( 1948

example

can

my

?

touches

of by

disequilibrium

presumably

mirror

learning

seems we

" by

answer

out

relating

or

mechanisms

understallding

f . Perhaps

saw

compared

( The

, pointed

the

child

how

that

resolved

. Her

ultimate

the

the is

in

correspondence

regulatory in

and

normal

about its

the

that

1973

expect

Learned

anything

emphasized

conflict

a

.

sign

. We

would

, as

signers

this

.)

relative

either

child

to

" signing

in

signed

pronouns of

.

location

this

sentence

deaf

.

space

attempt

of

. Frishberg

a

of

the

positions

that

personal

Brown

. I would

Dr

,

world

. Bellugi ( as

being

when

,

one

use often

a

deaf

.

discovering is

Dr

surfaces

Verbal

? This

has

from contrast

child

the

of

come there in

the a

will

in

he

the

orthogonal

, establishing

signs

not

an

intricate

specific

between

in

by

this

where for

boundaries

to

we

he

do

indicating

signer

the

back

should

up

, a

even

Language

very

setting

within

acquiring

grows

of

blind

,

story

ago

by

Sign , and

of

refer

minutes ."

then

both

layout

the

self

he

a

he ; we

scene

makes

notion

relations

trouble if

has

The

indicate

expect

child

telling or

then

that a few than one

should

blind

for

to

the

gestures

objects

will

indicating

American

language

when

and

signer

the

obvious

stage

persons The

We

that

left

. In

us , sign

imaginary

different

space

told

person

recently

sitting by

has

an

a third

has

been

indicated

. Bellugi

develop

toward

person

had

and

As

glance

the

Zazzo

' s parents

-

526

E. J. Gibson

have

communicated

name

, but

himself can

real

conflict

an

eventually

and

how

with

might

he

assimilate here

The

child

blind

mirror

image

conflict

, when

for , for

. The

enhances eye

late

slow

in

perceive

about

reason

for

a deaf

child

knowledge

of

is signed

to , that

and

that

reference

But

he

does

, the

doesn

the

spatial the

of

pronominally

is being

raised

much

and eye

. He

sees

his

must than

can ' t

, I think he

less

, knowledge points

in

that

. I is where

your

systems for

it . He with

to

, so himself use

so via

to

knows

, if he point

use

space

of

. Until

push

and

him

referring

self - reference the

a

a station

everybody

as

!) .

developing

articulated

of

to

the

opportunity

layout

occur

do

the

, and

) goes

, learn

could

less

deprivation

applied

concept

he has

resolving

writing

the

of

visual

within

to

pronoun

in

possibilities

with

is a

, ( lacking

example

hieroglyphic

perspectives

one

conflict

with

sooner

I in

with

station

with

excellent

corresponds

or

bilingually

for

, since

so much

spatial

? There

at a concept

slower

" : for

goes

to

self . But

.

start

between

the

has

personal

dealing

in words very

use

? He

persons

disequilibrium way

perspectives

discovering

its

be

of

selves

conflict to

he has

mobility

layout

the

also

relations

of

eye ( the

to

' t hear

language

of the

in

this

will

of arrives

by

only

concept

points

self

themselves

belonged

his

he

station

it , because

and

retardation

when

of

to

name to

perceiving

he

and

a multitude

a concept

development

is ( a good

superior

movable

) ; and

him

. His

resolved

with

instance

space

the

What

be

to

assimilated

11 Is there be

" coordination

objective

space

and

developing

he does

developing

about

been

only

layout

referring pronouns

have

can

may

information

, at first

personal

you

that

spatial

for

by already

objective

rich

him

word

via

1 even

.

he to

a

to

it

sign if he

.1

Note 1. As

it turns

out

spoken -written

( 1989 ), the deaf language

child ' s achievement

is as difficult

of correct

as for the hearing

pronoun

child , regardless

usage

of early

in the signing

.

References Brown , R . A first 1973 .

language . The early stages . Cambridge

, Mass .: Harvard

Gibson , J. J. The senses considered as perceptual systems . Boston : Houghton Kosslyn

, S. M ., Pick , H . L ., and Fariello , G . R . Cognitive

maps

University -Mifflin

in children

Press ,

, 1966 .

and men . Child

Development , 1974 , 45 , 707 - 716 . Preyer , W . The mind of the child . Part II . The development and Co ., 1890 . Shantz , C . U ., and Watson Child

, J. S. Spatial

abilities

of intellect . New

and spatial

egocentrism

Development , 1971 , 42 , 171 - 182 .

Zazzo , R . Image

du corps

et conscience

de soi . Enfance , 1948 , 1, 29 - 43 .

York : D . Appleton in the young

child .

29 Perceptionof Invariantsby Five-Month-Old Infants: Differentiation of Two Types of Motion EleanorJ. Gibson , C. J. Owsley , J. Johnston - '" Thiswasthefirst studyundertaken aftermy returnto research on infantpercep tion. Thehighest prioritytopicat thattimeseemed to beinfonnation: howdoes dynamic information in anambient opticarrayspecify criticalproperties of objects in theworld, andhowearlyareinfantscapable of pickingit up differentially ?I choserigidity-nonrigidityas an importantpropertyfor investigation , because rigidityor elasticity of substance is a propertywith obviousutilityfor infants . It distinguishes surfaces thataffordcomfortable posture(rest ) or locomotion , chew ablethingsin the mouth , hardobjectsthat affordgripping , and so on. The distinction is specified visuallyby a contrast in motioninformation overtime(see paper15 on rigid motion ) and it is alsospecified multimodally , sincehaptic exploration yieldsinformationin patternsof mechanical resistance to pressure . Furthermore , thepropertyisa persisting oneandit isspecified by information that is invariantovershapes , changes of source of pressure , observer position , etc., an example of perceptual constancy . Wechose habituation to a visualdisplayasourmethod , combining it with a so-called"infantcontrol " method(Horowitzet al. 1972 ). Thelattermethodhas theadvantage thattheinfantmaylearntoshutoff(control ) thedisplaybyturning awayitsgaze . Timesfor habituation arerendered lessrandomandmorereliable with thisprocedure . Thequestion of howto determine whetheran infantdetects aninvariantproperty ofsubstance wassolvedby usingageneralization procedure . Thesameobjectwasput throughthreedifferent typesof rigidmotionfor habitua tion. Followinghabituation , theinfantswereshowneithera fourthtypeof rigid motionor a deforming elasticmotionof thesameobject . If an infantgeneralized habituation to thenewrigid motion , but not to thedeforming motion , hemust detect an invariantof information overtherigidmotions . Infantsof five monthsgavereliableevidence of differentiating rigid from nonrigidmotionsover variationsof presentation , so we concluded that they

Developmental Psychology ,1978 ,14 ,407 -415 .Copyright 1978 the American Psycho logical Association .Reprinted by permission of the publisher . by

528

E. J. Gibson, C. J. Owsley, & J. Johnston

were detecting invariance ofa common property. Later experiments (notreprinted here) showed thatinvariance ofsubstance wasstillperceived overa shape change,

andthatinfantsas youngas threemonths detected theinvariant information (Gibson et al. 1979). A finalexperiment in theseriesreversed theorderof

presentations, showing thathabituation tovaried deforming motions generalized andwasperceived ascontrasting witha rigidmotion (Walker etal.1980), sothe contrast canbedifferentiated eitherway.Thesurfaces andsubstances ofthingsin theworldare boundto havevariedaffordances thatmustbedifferentiated for

successful action, andinformational invariants mustspecify them. Properties of

substance aredetected veryearlyviaat leastoneof theseinvariants, optical informationfor rigidity.

Five-month-oldinfantswere habituatedto three types of visuallypresented

rigidmotion, withduration of fixation as thedependent measure. After reaching acriterion ofhabituation, afourth rigidmotion (nothabituated) and a deformation werepresented. Dishabituation wassignificantly greaterto the

deformation thanto the fourthrigidmotion.Comparison witha no-change controlconditionshoweda significant difference for deformation whenit

followed thefourthrigidmotion. Itwasinferred thattheinfants detected the invariant property ofrigidity anddifferentiated it fromthedeformation.

Thisexperiment isthefirstofa seriesdesigned to investigate theinfants perception ofinvariants giventhrough stimulus information. 1 It haslong been known that human adults perceive objects in their environment as

constantin sizeandshape,despitea changingopticarrayproduced by

movementof the observerfromone observationpoint to anotheror by movementof the objectitselfin the surrounding spatiallayout.As the observermoves,or as objectsmovein thislayout,aspectsof targetobjects

areprogressively occluded andrevealed withconsequent changes in the stimulus arraythatreachesthe observersreceptors; yet the information

generally remains adequate forthedistalobjectto be apprehended for whatit is.Thisstoryholdsformanykindsof information presentedin an

ambient opticarray.It probably alsoholdsforinformation presented inan

ambientacousticarray,althoughthat has been less thoroughlyinvesti-

gated.Werecognize voicesdespite noisesenveloping them,distance, and variationsin acousticvariablessuchas intensityand pitchcausedby whispering,head colds,and shouting.

Constancyisonlyoneexample oftheinteresting andhighlyadaptive phenomena ofhuman(andanimal) cognition thatwereferto asperception of invariants. Apprehension of conservation whena row of objectsis lengthened (orshortened) is another,a phenomenon introduced to psy-

Perception of Invariants

529

chology byPiaget (1952); andsoisperceived unifyandpersistence ofan object whenit isoccluded anddisoccluded, a phenomenon introduced by Michotte (Michotte, ThinŁs, &CrabbØ, 1964). Permanence andobjectivity ofthespatial layouttheground andskyaround usandthearrangement

of thingsin themwith respectto one anotherand to ushas more

recently beenrecognized asanimportant kindofinvariant andisbeing studied undersuchrubncs asperspective takingandcognitive maps.

Manyofthesephenomena havereceived someattention fromdevelopmental psychologists instudies witholderchildren, forexample, studies of sizeconstancy in children between5 and 12 yearsof age by Piaget andLambercier (seePiaget, 1961). Investigation ofperceived invariants in

veryearlydevelopment has beenscanty,exceptfor a few studiesof constancy ininfants(e.g.,Bower, 1964,1966;Day&McKenzie, 1973).But

before infancy isover,a childlivesandactsintheworldwithapparent

confidence. It seemsunlikely thatthesurfaces andobjectsof theenvironment appearto shrinkand expandas he movesor is moved,or as the

objects move. Perception ofinvariants surely hasa longdevelopmental

history,andonethatbeginsearly,perhapswithdetectionof someof the permanent properties ofobjects. Itisthisaspectofperceptual development

thatwehavechosen toinvestigate, beginning withsome simple properties ofmotionthatcarryinformation forinvariant properties ofthings.The

subjectsin the experimentto be describedwere5-month-oldinfants.

Specifically, weaskedwhether themotions inthestimulus arraythatare causedbymovement ofa rigidobjectprovideinformation foraninvariant property,even thoughthe particularrigidmotionsdiffer.To answerthis question,we comparedperceptionof such motions with a motion of

deformation thatspecifies elasticity in contrast to ridgidity to theadult

perceiver(Fieandt& Gibson,1959).Foran infantto detectthisinformation

at anearlyagewouldbehighlyadaptive, sincevisualrecognition ofan object asrigid(unyielding tothetouchifbumped intoorfallen upon,for

instance) versuselastic(yielding and potentially soft or evenanimate)

couldprovide foreknowledge oftheconsequences oftouching orcollision.

Knowledge of affordancesgiventhusis oneof the mostusefulachievementsof perceptualdevelopment. Whetherthe affordance mustbe learned by association oftheinvariant givenbymotionwithtactualandkinesthetic

experiences givenbypressure is a separate question. Ourprediction was

simply thatyoung infants would distinguish a setofrigidmotions, specify-

ing the sameinvariant,froma motion of deformation.

Thehabituation paradigm waschosen. Theessential procedure involved arepeated display ofonetypeofmotion forhabituation, followed bya test fordishabituation withtheothertype.Theinfantsfixation wasmonitored,

and lookingtimewasusedto determinehabituationand to test for dis-

habituation whenthenewmotion wasintroduced. Intheclassic paradigm,

530

E. J. Gibson, C. J. Owsley, & J. Johnston

one

infers

that

the

ested

that

in

was

a

subject mere

looking

that

the

motions

.

presented

with

prediction

is

deformation

a

the

fourth

correct ,

rigid

of

since

motion

motion

to

habituation

,

and

is the

same

a

the

a

type

,

were .

greater ,

property

,

rigid

infants

array

-

therefore

deformation

the

invariant

, as

different

of

with

inter concern

,

three

half

in

The

motions

significantly

present

not

modified

to ,

be

.

rigid

but

half

should

were

however

was

sequence

motion

projects

, in

one

indicates

. We

paradigm

information

still

dishabituation

display

perceived

usual

dishabituation

new

for

the

of

not

rigid

,

test

in

was . The

habituated

end

the

change

invariant

was

the

at

change

any

deformation

infant At

time a

of

something with

fourth

in

detected

detection

whether

contrasted so

rise has

If

our

for

the

whereas

the

.

Method Object

and

The

object

was

a

Motions

round

Figure

29

obvious

. 1 ) . It

axis

were

around

motion

squeezing

Object

of

29

in

into

the

10

holes

in

in

the

frontal

object

axis subject

about

the

by

object

for cm

, dappled

so

the

embodied

, 30

in

the

two

diameter

with

types

and

black

spots

to

the

of

applying

to

produce

plane

by of

,

the

an

a rigid

motion

of cm

motion thick

provide

( see some

rigid

. Four

around

motions

were

cyclical . The

with

to

all

and

10

( rotation

around

the

vertical

motions

vertical and

fro

axis on

and

).

then

,

the

. Each

consisted

fingers

axis

rigid

continuous

deformation

or

, whose

the

( displacement

displacement

motion

experimenter

object

rotation

looming

pressure

either

produced

backside

, and

) . The 300

as

were

.1

undergoing

6 .2

.

of

horizontal to

color

manipulated

rotation

perpendicular

information

rubber

variation

be

the

foam

yellow

motions

:

stimulus of

texture

inserted

displayed

rigid

Figure

and

. All

rotation

light

could

deformation

were

provide piece

was

object

fingers

to

disklike

color

The

Displayed

chosen

of releasing

a

Perception of Invariants

531

it,producing a rippling (aspongy appearance) ofitssurface. Thedeformation

was also cyclicaland continuous.

StimulusInformation

In allfourcasesof rigidmotion,the cross-ratios of surface elements remain

invariant overtime(Gibson, 1957; inpress). Thisinvariant relationship isinformationforthepersistence oftheobjects surface structure. More simply stated, asthe

object undergoes rigid motion, anyalignment oftexture elements intheprojected

array remainsaligned (Hay, 1966).

Indeforming motion, thecross-ratios ofthetexture elements change overtime. Moresimply stated, alignments oftexture elements intheprojected arraydonot remain aligned. Theobjectssurface structure, ratherthanpersisting, is contin-

uouslychanging. Thepatternof change, however, neednotbe random. In the presentexperiment, thepatternof changewasquiteregularin thattherewasa cyclical squeezing andreleasing oftheobject.

Thus each typeofmotion canbedescribed byaunique invariant relationship in

theambient opticarray, which canserveasa basisforthemotions differentiation. Subjects

Datafrom 24full-term infants (mean age= 153days; range: 131179days), half

males andhalffemales, arereported below. Datafrom5 otherinfants couldnotbe

usedbecause ofthesubjects fussing. Subjects wereobtained bysending letters to parents whose addresses werelisted inthebirthrecords ofthelocal newspaper. Apparatus

Thesubject wasplaced inaninfant seatlocated ina rectangular curtained booth. 75.8x 148.5cm.Hefaceda plywoodscreenthathada 16 x 16cmwindowthat

couldbe openedandclosedby sliding a blackcardboard screenacrossit. The objectsubtended anangleof28. It wasdisplayed in thewindow in frontofa black curtain. Theexperimenter whoproduced theobjectsmotions stoodbehind

thecurtain, which entirely concealed her.Shepushed herhands through anopen-

inginthecurtain, inserted herfingers intothefinger holesintheobjects back, and moved the object in a specified and carefully practiced routine for each of the five motions.

Anobserver watched thesubjects eyesthrough a peephole below thewindow

(diameter 1.9cm)anddepressed a button whenthesubject fixated theobject. Shewasblindwithrespect to theparticular routine beingpresented andto

theexperimental condition. Thesubjects fixation wasrecorded on a Harvard Apparatus recorder, located ina neighboring roomandmonitored bya second

experimenter.

Experimental Designand Procedure

Theexperiment hada 2 x 2 X2 x 2 factorial design: 2 between-group factors,

order ofpresentation ofmotions intheposttest andsex,and2 within-group factors, Posttest I versus Posttest 2andtypeofmotion presented intheposttest trials.

532

E. J. Gibson , C. J. Owsley , & J. Johnston

A habituationparadigmwas used. The generalprocedurehad three parts: a pretest, a habituation series, and a posttest. All 24 subjectswere presentedwith three rigid motions in a consecutive fashion in the habituation series. The three rigid motions that a single subject observedwere counterbalancedacrosssubjects.

Thenext motionin the sequence waspresented whenan infantlookedawayfrom the current display for 2 sec (Horowitz , Paden, Bhana, & Self, 1972). This routine was continued

until

a criterion

of habituation

was reached for each motion

sepa -

rately . The criterion for any subject was defined as one half of his first or second

habituation exposure for that motion, whichever was longer. If the criterion for habituation was met for a certain motion , but was not yet reached for one or both

of the other two motions, that motion was not dropped from the series. In other

words, all threemotionswerepresented until all hadmet their respectivecriteria. When the criteria for habituation had been met, a posttest was presentedthat included a fourth rigid motion (the one that had not been seenin the habituation series) and the deformation. One group of 12 subjectssaw the new rigid motion first , whereas the other 12 subjects saw the deformation first . The next motion in

the posttest waspresentedwhena subjectlookedawayfrom the currentdisplay for 2 sec. The posttest was repeated, maintaining order of presentation. Males and females were divided evenly between the two groups . A pretest was given before the habituation

series. It was identical to the

posttest, except that it was not repeated. In other words, it consisted of a single presentationof the fourth rigid motion followed by the deformation, or the reverse order, depending upon what group the subject was in. The purpose of the pretest was to determine whether one of the motions was intrinsically more interesting than the other .

Results

Subjectswere generally attentive to the moving displays, but looking time varied considerablyfrom one infant to another. Mean total looking time across trials to criterion was 153 .70 sec (SO = 110 . 37 ). Mean duration of the first fixation was 18 .98 sec (SO = 14 .17 ). Mean duration of the last look before criterion was 6 .05 sec (SO = 10 .31 ). Mean number of trials to the final criterion was 11 .13 (SO = 3 .18 ).

The comparison of interest was between two difference scores: fixation time on the deformation

minus the fixation

time on the last trial of habitua -

tion compared to the fixation time on the fourth rigid motion minus the fixation

time

on the last trial

of habituation

. The means of these

differencesare presentedin Figure29.2, the two orderspresentedsepa rately and the two testswithin eachorder presentedseparately. Figure 29.2 indicates

that for each order and for each test , differences

in fixation

on

deformation were larger than differencesin fixation on the fourth rigid motion

.

Since the distributions of looking time on the criterion trial and on the

posttest trials were postively skewedand sincecell meanswere nonhomo-

533

Perceptionof Invariants

Figure29.2 Meandifferences in fixationtime betweentrialsfollowingcriterionandthe trial just preceding . (Blackcolumns indicate differences for thefourthrigidmotion . Whitecolumns indicatedifferences for the deformation . Cross -hatched columns aredifferences in a nochange condition .)

geneous

, log

times

( see

Figures

29

Cohen

transformations

Winer

,

. 3 and

29

numbers at

criterion

trial

ences

in

at

of

in

made

no

posttest

effect

- 400

&

, since

they

are

,

1975

trials

, but

number

of

subjects

decreases

to

criterion

. At

trials of

the

points

plotted

were

) . The

from as

are

, owing

trials

. All

left

to

individual the

are

subjects

the

of

in

following

to

half

fixation

graphed

habituation

reads

one

all

plotted

criterion

one

least

on

data

curves

Gelber

habituation

performed

) . Transformed

habituation

Cohen

of

I ) , were

of each

data

Increment

motions

. 397

=

are

from

the differ

subjects

are

.

Analyses

analysis

tests

end

the

transformed

it

the

number

represented

Primary

, pp

( y

. 4 . Backward ( see

negative

represented

An

1971

' s procedure

given

, log

and of

sex was

2

, the

was factors

performed being

on

Order

Fixation

. There which

type

of

motion

main

effect

of

posttest

was revealed

. The

main

significant

no

the

was

same

effect ,

F (I ,

of

no

x

difference

. There I

variance

main

pattern differential

20

) =

the Sex

difference x

effect

of

of

x

order

was

. Nor

increment <

. 005

in ,

with

of

that

first that

was fixation a

the

Differential

, indicating

presented

, indicating results

9 . 69 , P

scores

Posttest

in

both there to

greater

the Post

a the

-

main two incre

-

534

E. J. Gibson, C. J. Owsley, & J. Johnston

1 .5

0

R

.

.

~

+

>-

<. 9 0

(L + A) 90 -' 3l1\Jl.l 9NI)tOO-'

~ w

:2: 1 . 0 ~

<.9 Z ~ 0 0

R2

- 1

0 .5

- 11

- 10

- 9

- 8

- 7

TRIALS PRE

-TEST

- 6

FROM

- 5

-4

-3

- 2

- 1

+1

CR \ TERION

+2

+3

+4

POST -TEST

HABITUATION

Figure 29 .3

Backwardhabituation curve and pre- and posttests for subjectsreceiving deformation first in the pre - and posttests .

1.

1 .0

0 .5

- 11

- 10

- 9

- 8

TRIALS PRE

- TEST

- 7

- 6

FROM HABITUATION

- 5

- 4

CRITERION

- 3

- 2

- 1

+ 1

+ 2

+ 3

+ 4

POST-TEST

Figure 29.4 Backward habituation curve and pre- and posttests for subjectsreceiving rigid motion first in the pre- and posttests.

Perception of Invariants

535

men!: to deformation. Nineteen of 24 subjects showed thispattern. The other5 lookedlongerat thefourthrigidmotion.

Theonlysignificant interaction wasbetween thesex,posttest, and

differential increment factors,F(1,20)= 4.94,p < .04;malesshoweda

greaterincreaseto rigidmotionin Posttest1 than in Posttest2, while

femalesshoweda greaterincreaseto rigidmotionin Posttest2 thanin Posttest1. Malesshoweda greaterincreaseto deformationin Posttest2

thaninPosttest 1,while females showed thesamemagnitude ofincrease in

both posttests.

Ananalysis ofvariance wasalsoperformed ontherawscores, despite theskewness ofthedistributions andthenonhomogeneity ofvariance. A

similarpatternof resultswas obtained.The onlymaineffectwas the differential increment factor,F(1,20)= 5.16,p < .034.Therewereno interactions.

It waspossibleto calculate thecomparison of interestin anothermanner

thatincreased thegenerality of theresults. Weaveraged theduration (using transformed data)ofthelastthreelooksinhabituation ofanysubject (thusincluding allthreerigidmotionsto whichhe hadbeenhabituated) andsubtracted thatmeanfromthefixation timeonthenewrigidmotion or onthedeformation. A second analysis ofvariance wasperformed on thesescores. Theresults weresimilar to thefirstanalysis, withtheonly significant effectbeingthedifferential increment factor,F(I,20)= 9.87,

p < .005.

To investigatewhetherthe significant differential incrementin fixation

mightbedueto theeffects ofonlyoneortwooftherigidmotions, two

analyses ofvariance wereperformed. Thefactors wereparticular rigid

motion anddifferential increment inposttest fixation. Inthefirstanalysis, the particular rigidmotionthat eachsubjectreceivedin his lastlookof

habituation provided thelevelsof themotionfactor.In the second,the

particular rigidmotion thatwasviewed intheposttest provided thelevels ofthemotion factor. Thetwoanalyses yielded similar results. Asignificant maineffectof differential increment infixation againrevealed a greater

increment to deformation, F(1,20)= 6.89,p < .016,in thefirstanalysis, andF(I,20)= 8.66,p < .008,inthesecond analysis. Therewasnomain

effectof particular typeof motion,norwastherean interaction of it with thedifferential increment in fixation. Together, theseresultsindicate that

thedifferential increment infixation timewasnotduetotheoverpowering effects ofoneortworigidmotions; rather, theycontributed equally.

Wasdeformation intrinsically moreinteresting thanrigidmotion, thus accounting forlongerfixation onit intheposttest? Ananalysis ofvariance

wasperformed onthepretestfixation times, factors beingtypeofmotion

(rigid vs.deformation) andorder ofpresentation. There werenosignificant effects, indicating thatthetwotypesofmotion wereaboutequally effec-

536

E. 1. Gibson, C. J. Owsley, &tI. Johnston

tive in elicitingfixations beforehabituation. Thusthe greaterfixation incrementto deformationin the posttest was due to the habituationexperi-

enceandsubsequent detectionofnewinformation in thedisplay. Habituation

Several questions mightbeaskedconcerning thetimerequired fora subject to habituate,sometimes referredto as rate of habituation.Sinceallsub-

jectsin thisexperiment wererunto an established criterion (onehalf oftheiroriginal looking time),wenecessarily defined rateintermsofthat criterion. Rateis accordingly definedas totalfixationsummedovertrials untilcriterionis met.Findingswereas follows.First,whena singleratewas calculatedacrossthe three-motionseries,a t test indicatedthat there were no sex differences.Second,there was no correlationbetween age (range:

131179 days)andrate.Third,whenrateswerecompared fororderin the

three-motionsseries,t tests showedno differencesamongthe rates, for

example, thethirdmotion introduced wasnothabituated to morequickly than the firstor second.Fourth,when a one-wayanalysisof variancewas

usedto compare theratesofhabituation to eachofthefourrigidmotions, there was a significanteffectof particularrigidmotion.F(3,68)= 2.87,

p < .05.Inorderfromfastestto slowest rates,themotions werelooming, rotation around the verticalaxis,rotation in the frontalplane,and rotation around the horizontalaxis. Multiplecomparisonswere carriedout using

theTukey(b)test.Theonlysignificant difference wasbetweenlooming and rotation around the horizontal axis, p < .05, clf= 68.

Comparison to No-ChangeControlGroup

Theexperiment wasdesigned so as to addressthe questionof whether dishabituationto deformationwouldbe relativelygreater than dishabitua-

tionto the newrigidmotion.Ourresultsprovidedan affirmative answer. As an afterthought, we askedwhetherthe amountof dishabituation to deformation andto thenewrigidmotionwasgreaterthantheincreasethat wouldbe expected if no changeof motionwasintroduced. Althougha randomassignment of subjectsto groupsat thisdatewasimpossible, 12 subjects(meanage = 152days;range:135161 days)wererun in a no-

changecontrolcondition. Uponreaching criterion, insteadof introducing the newmotionsin the posttest,thesesubjectswerepresentedwiththe

nexttworigidmotions ofthehabituation sequence, justasifsubjects had not met criterionat all. It was expectedthat an increasein fixationon deformationwouldbe significantly greaterthan an increasein fixationin

theno-change posttest.Theexpectation regarding thefourthrigidmotion

was not so clear-cut.Significantdishabituation might occurif subjects

detectedthefourthrigidmotionasbeingdifferent fromtheprevious three.

Perceptionof Invariants

537

Continued habituation might occur if subjectsperceivedan invariant property as remaining in the array and if they were not sensitiveto the change to a fourth rigid motion. In evaluating the results, it should be noted that only Posttest I scores in the original experiment were relevant for comparisonsbetween the no-change control posttest trials and the posttest trials in the original experiment . This is because only two test trials followed criterion in the

no-changecontrol condition. Two difference scores were computed for each subject: The first differ -

encescorewas the old rigid motion that was presentedfirst in the posttest minus the last look in habituation. The mean differencescore is displayed in Figure 2 as the first bar in the no-changecondition. This differencescore is properly compared to difference scores involving rigid motion and deformation

. which also occurred first in Posttest

1. Note that the relevant

deformation difference score (11.45 sec, SD = 35.05) is larger than the no-changedifferencescore (5.67 sec, SD = 11.08), whereasthe rigid motion difference score (2.90 sec, SD = 8.71) is smaller than the no-change differencescore. Mann-Whitney U testswere computed, sincedistributions were not normal and varianceswere nonhomogeneous. Neither comparison was statistically significant, despitethe apparentdifferences. The seconddifferencescorecomputed for the no-changecondition (the last bar in Figure 2) was the old rigid motion that was presentedsecondin the posttest minus the last look in habituation . This difference score is

properly compared to the difference scores involving the fourth rigid motion

and deformation

that occurred second in Posttest

1. Note that the

deformation differencescore(3.67, SD = 5.05) is larger than the no-change difference score (.33, SD = 3.56). A Mann-Whitney U test indicated that this difference was significant, U = 114.5, P < .05, two-tailed. The rigid motion differencescore (.67, SD = 11.27) is slightly larger, but this difference is not statistically significant. Although the difference between the control group and the deformation was significant only when deformation was presented second in the posttest. the results are in the expected direction

in both comparisons .2

Discussion

The results of the presentexperimentimply to us that 5-month-old infants are capableof perceiving rigidity of an object as an invariant property of that object. In contrast to a fourth rigid motion, a motion of deformation was perceivedasdifferent, presumablybecauseit offered information about a new property of the object, that object being otherwise unchanged. Had the experiment merely found that the infants discriminated a single rigid

538

E. J. Gibson , C. J. Owsley , & J. Johnston

motion from a motion of deformation (as an earlier pilot experiment did), there would be no grounds for concluding that an invariant property was detected. The four rigid motions all provide information for unchanging structure of the distal obiect. On the other hand, deformation provides information for changing structure, and a reversible, cyclic deformation of the kind displayed implies that the distal object is elastic. The experiment doesnot tell us whether elasticity, as such, would be perceivedas invariant if other instancesof it were provided, but it does imply that four motions specifying rigidity were perceivedto have an invariant property. Whether that property can only be describedin terms of proximal stimulus information, or can be referred to in distal terms as rigidity is a moot question. Might there be other ways of interpreting the results? Severalhave been suggestedto us. One interpretation, deriving from a constructivist view of oerceotual development in which inference is considered an important ~ ~ ~ process, might argue that the infants categorizethe three rigid motions in the courseof habituation and recognizethe fourth one as a member of the newly acquired category (Bruner, 1973; Bryant, 1974). The flaw in this argumentis a logical one; the infant must perceivesomecommon property in order to form the category. If he is able to detect whatever specifiesthe property of rigidity in all four motions, he is perceiving an invariant property. Assuming a further processof categorizationwould be adding an unparsimonioustheoretical superstructure. Could one, under a kind of mediation stimulus- responseview, assume that a responsemadeto eachof the three habituatedmotions "generalizes " to the fourth? The term generalizationsoundsplausibleand innocuous; but there is again the problem of acquiring a mediating responsethat is the samefor all four motions. What property underliesthe generalization, if it is not the invariant that describesrigidity ? Could one, under a Piagetianumbrella, invoke a common scheme, constructed during viewing of the first three motions, to which the fourth is assimilated ? This view seemsto us the leastlikely of all, since(a) perception is assumedto be "figurative" at this age, a sequenceof arrested"pictures" (Piaget, 1954, p. 4 ff; Piaget, 1971, pp. 248 ff ), whereasthe information for the invariant and its contrasting property are carried by motion; and (b) a schemeis defined as sensorimotor, and the motor aspect, especially one common to the four motions, would be extremely hard to identify. Piaget statesthat "groups of displacement," as thesemotions might be called, are "surely not perceivedas such by the child," the child he referred to being lessthan 9 or 10 months of age (Piaget, 1954, pp. 97 ff.). It will require further researchto settle on any of theseinterpretationsor to generate a better one. A related question, the role of motion as an object-defining property, also deservesinvestigation. Do unique motions

Perception of Invariants

539

enjoy the sameprivilegefor objectidentification , as do staticdistinctive features(onespresumedto be usedfor distinguishingfaces , for instance )? Are they perhapsusedprior to staticfeaturesfor identifyingpropertiesof objectsthat havepermanentaffordances , if not for identifyingthe object itself?It is not at all unlikelythat invariantpropertiesof objectslike rigidity andelasticityaredetectedwell beforeobjectsthemselves areidentifiedon the basisof staticdistinctivefeatures . We hope that experimentsin progresswill clarifythesequestions . Finally, we return to a questionposedearlier. Does detectionof the invariantgiven by visually perceivedrigid motions, as contrastedwith deformation , leadthe infantto expectthat the objectwill feelrigid to the touchwhenpressed , in the first case , andfeelyieldingin the second ? Must this expectationbe acquiredby association of the two experiences ? Earlier in the century, it wasgenerallyacceptedby Titchenerians andfunctional istsalikethat the latter wasthe case . But in recentyears, the doctrinethat "the handteachesthe eye" haslost ground. Infantsdo not employactive touchto explorethe feelof objectsbefore5 monthsafterbirth (seeGibson, 1969, pp. 359- 362), yet they are capableof differentiatingtwo typesof visually presentedmotion that can afford informationfor rigidity and elasticity. To someof us, the visualinformationseemsjust as valid in its specification of rigidity as that given by feeling. Furthermore , multimodal redundantinformationthat specifiesthe samepropertyor the sameobject is frequentlyavailableto a perceiver . An intermodalexperiment , performed assoonasan infantexploressubstances hapticallyby proddingor squeez ing, may tell us whetherthe infantexpectsthe visuallyrigid substance to feelrigid aswell. Notes I. Thisresearch wassupported in partby Training GrantSTOIHDO038I -OSfromthe National Institute of ChildHealth andHuman Development ; theSusan LinnSage Fund ofCornell University ; andNational Science Foundation GrantBNS76 -I4942to thefirst author . Thanks aredueto DaleKlopfer for assistance in monitoring equipment anddata analysis . A briefer version ofthispaper waspresented atameeting of theSociety forResearch in ChildDevelopment , NewOrleans , March1977 . 2. Thecontrol - experimental comparisons werealsocomputed usingthelog-transformed data . Since thelog conversions rendered thedistributions normal andthevariances homogeneous , two-tailedt testswerecarried out. Therigidmotion difference scores did notdifferfromtheircorresponding no-change difference scores , bothyielding f values at p > .5. Whendeformation waspresented first,thedeformation difference score didnot significantly differfromtheno-change difference score , t(22) = 1.03. P< .32, whereas whendeformation waspresented second , thedeformation - nochange comparison approached significance , t(22) = 1.80,P< .086.

540

E. J. Gibson, C. J. Owsley, & J. Johnston

References Bower. T. G. R. Discrimination of depth in pre-motor infants. Psychonomic Science , 1964. 1, 368. Bower, T . G. R. Slant perception and shape constancy in infants. Science , 1966, 151, 832834. Bruner, J. S. Beyondthe informationgiven. New York : Norton . 1973. Bryant. P. E. Perception and understandingin youngchildren. London: Methuen, 1974. Cohen. L. B.. & Gelber. E. R. Infant visual memory. In L. B. Cohen & P. Salapatek(Eds.), Infant perception : Fromsensationto cognition,(Vol . 1). New York: Academic Press, 1975. Day, R. H., & McKenzie. B. E. Perceptualshapeconstancy in early infancy. Perception , 1973, 2, 315- 320. Fieandt, K., &: Gibson. J. J. The sensitivity of the eye to two kinds of continuous transformation of a shadow pattern. Journalof ExperimentalPsychology , 1959, 57, 344- 347. Gibson, E. J. Principlesof perceptuallearningand development . New York: Appleton-CenturyCrofts, 1969. Gibson, J. J. Optical motions and transformations as stimuli for visual perception. Psycho logicalReview, 1957, 64, 288- 295. Gibson, J. J. The ecologicalapproachto visual perception . Boston: Houghton-Mifflin , 1979. Hay, J. C. Optical motions and space perception: An extension of Gibson's analysis. Psychological Review, 1966, 73, 550- 565. Horowitz , F. D ., Paden, L., Bhana, K., & Self, P. An infant-control procedure for studying fixations. Developmental Psychology , 1972, 7, 90. Michotte , A ., Thines, G" & Crabbe, G. Les complements amodaux des structures perceptives. In A . Michotte & J. Nuttin (Eds.), Studia Psychologica . Louvain, France: Publications Universitaires de Louvain, 1964. Piaget, J. Thechild's conceptionof number. London: Routledge & Kegan Paul, 1952. Piaget, J. Theconstructionof reality in the child. New York : BasicBooks, 1954. Piaget, J. Lesmecanismes perceptifs . Paris: PressesUniversitaires de France, 1961. Piaget, J. Perception and knowledge . Edinburgh, Scotland: University of Edinburgh Press, 1971. Winer, B. J. Statisticalprinciplesin experimentaldesign.New York : McGraw-Hill , 1971.

30

Developmentof Knowledgeof Visual-Tactual Affordances

of Substance

EleanorJ. Gibson, Arlene S. Walker

The previouspaper showedthat rigidiiy-nonrigidity is distinguishedby very younginfants by meansof opticalinformationgiven in motionover time. Two questionsarise. Do infants detecta meaning {affordance} for this property, or are

they only discriminatinga differencein stimulation, not a propertyof an object existing in the world outside them? Second , would they distinguish and recognize

hapticinformationfor thesameproperty, that is, feel thedifference , and recognize it as the sameone, characterizingan externalobject? The two questions , one havingto do with perceivingan affordance for action, theotherhavingto do with multimodal information specifying the sameproperty, are of great importancefor the ecological approach. Information is not the same concept as stimulation striking a receptorsurface: it is information for something, about somethingin the world . As such, it should be abstract , not tied to one senSOni svstem . We would

exceptto find generalization betweenbimodal presentations , even at an early age,

denotingpick up of a propertyof an externalobject , not mereregisteringof input to a singlereceptormodality. I havearguedin otherpapers(Gibson 1983, on development of intermodal unity, and Walker-Andrewsand Gibson1986, on development of bimodalperception) that infants do not differentiate modalities as such at an early age. Information is amodal, rarely modality specific, and information for an external

eventmay bespecifiedby morethan oneperceptualsystem . Thesubstance of an object , for example , may be specifiedvisually, haptically, and evenauditorily. Think of a steel bar, for instance- it looks hard and rigid, feels hard and cold, and has a metallic ring when struck. Information is conveyedover time and can have amodal, abstract propertiestogetherspecifying the sameaffordance.

As adultswe distinguishwhetherwe recognize a substance by odor, sight, or feel, but it may be that young infants perceive the affordance of a substance

without distinguishingthe sensoryquality (smell, vision, whatever ) of the inChild Development , 1984, 55, 453- 460.

542

E. J. Gibson& A. S. Walker

formation

that

qualities

, beginning

any

specified

case , the

specified

paper

year

At

the

same

If

the

visually

time , we

detects

after

it ? It

something

external

in haptic

felt

exploration

of or

visual

, and

12 months

preference

test

objects

moving

in a pattern differently

with

and of this

able

. A

deforming characteristic young

affordance

of the

laboratory

Previous

research

in this

not

visually

differentiate

only

detect of

optical ~

rigid

mental elasticity motions Megaw 1980 move

motion

. ! The

evidence

-

of

substance

( Gibson - Nyce ) . Adults so as to

for

know provide

; Walker that

. The

substance

has

objects patterns

shown

over

of

the

2

preferentially

yielded

. A

compar

allowed

-

them

to

familiarization

moving

, together

rigidly in

or

a manner

, suggest some

that from

that

quite

substances

3 - 5 month deforming

is invariant

proposition

over

is also detect a

, 1978 , Owsley feel

1

handled

moving

object

of

and

familiarized

haptic

specifying

3 months

that

for

a , a

and

.

that

, Gibson

infants

of either

object

looked

objects

the

results

motions

Johnston

.

of 2 films

substance

- old

for

object ,

manipulation

. Infants

in the light

real

at

an

an object sec of

and of

.

substance

to the problem

of a rigid

substance

invariants

invariant , &

type

with

longer

of

in

to the substance

with

object

1 - month

a soft

rigidity

infants as

, Owsley

, 1979

or

symmetrical

that

the

with

rigid

information

infant

of

presentation

familiarization

substance

the

an

a meaning

later

60

manner

to

with

intermodal

in

experimentally

characteristic

appropriate looks

detect

infants

the

detects

return

dark

of an elastic

looked

of the novel

. We

simultaneous

preferences

infants

property by

an affordance

a property

be investigated

with

whether

appropriately

in the

infants

perceive

acted

. Following

a hard

looking

. These

has

is enough

experiment

on

the baby

familiarized

an

visually

is , recognizes

) substance

first

of either

that

us that

it can

experiment

tested

-

in a pattern

in

third

objects then

infant

given

1986 ) . In

is detecting

is perceiving

characteristic

longer

replication

and

was

light

substance

tell

that

, 1 moving

substances

mouth

were

sensory

Gibson

a substantial

specified

some

infant

how

(spongy

identical

throw

bimodally

the

particular and

whether

when

a nonrigid

of it , perhaps

an elastic

results

to

the

-Andrews

at one year .

does that

that

to itself . If

more

and

and

it -

to differentiate

investigates

substance

of.T Iperceiving ~ affordances "

hard

hoped

indicates

come ( Walker

as the same

a rigid

the

having

certainly

Infants

here

at one month

between

baby

later

months

is perceived

of life -

distinguishing

only

six

reprinted

haptically

the first

it , and

around

hard

transformation

class

of

; Gibson , Megaw and

rigid

infants but

different

supported optical

- old ones

by

also forms

experi

information

different , Owsley

deforming , Walker

- Nyce

, &

when

manipulated

( perspective

for

, &

Bahrick

transforma

,

-

Knowledge of Visual-Tactual Affordances of Substance

543

tions) that specify rigidity visually and that objects that feel spongy and elastic when manipulated can move in deforming patterns that specify elasticity visually. Visual recognition of theseproperties of substancesis of great utility for development. For example, an elasticsurfacebelow affords jumping upon, but a hard one may afford breaking a leg. Chewablethings canbe seento move in deforming patterns. Hard onesthat might injure the mouth

do not . Thus , information

about

the affordance

of a substance is

available to both haptic and visual pickup systems. When adults perceive the utility of this information for appropriateactions, we say that they have detected 1979 ).

the affordance

of the substance (E. J. Gibson , 1982 ; J. J. Gibson ,

How early can infants obtain this knowledge about a useful intennodal invariant? Preliminary experimentsextending over severalyears were instructive and infonned us that manual exploration of a hidden object is extremely difficult to obtain before 10- 12 months of age. For this reasons,

the first two experiments to be reported were performed with subjects about 1 year old. The third experiment was performed with very young subjects (1 month ), who explored haptically by mouth . Experiment 1

This experiment employed a method adapted from Gottfried and Rose (1980 ) and Soroka , Corter , and Abramovitch

(1979 ). The infants (12 months

old) were presentedwith the objects to be manipulated in total darkness with only the object 's location specified by a dot of luminescent paint . The

infant's behavior was videotaped in infrared light and observed by an auxiliary experimenter on a TV screen in another room . This procedure allowed us to ask an important question for investigating the perception of affordances: When an infant makes an unexpected contact by touch with an

unknown object, does exploratory behavior ensuethat is appropriate for

distinguishingobjectsof differentsubstances , in one caserigid (hard), in the other elastic (yielding- )?

Following manipulation of the object, we presentedthe infant, without its being moved or taken into the light, with motion picture films of two objects, side by side, one being moved in a pattern characteristicof a rigid object and one deforming in a pattern characteristicof an elastic object. Thus, optical information about the two substanceswas availablel given solely by motion , since the two objects were otherwise identical .

This procedurewas designedto answerthe question, Does the infant show, by preferential looking at one or the other of these films, that intermodal correspondenceof substancescan be recognized following exploratory handling of a substancein darkness ?

544

E. J. Gibson& A. S. Walker

Method Subjects The subjectswere approximately 1 year old, ranging in age from 356 to 392 days. There were 18 boys and 14 girls. Numbers of boys and girls familiarizedwith the hard object were equal. For the soft object, there were 10 boys and six girls. Apparatus The infant sat on its mother's lap at a table 91 x 61 cm, facing a wooden frame the width of the table. A translucentscreen(71 x 41 cm) was mounted in the frame 41 cm above the table. Motion picturescould be back-projected on this screen, which was at a comfortable eye height for the infants, as they looked up at it slightly. Mounted in the wood under the screenand 28 cm above the table's surfacewas the lens of a TV camera (GeneralElectricinfrared video cameraCCTV). The camerawas positioned so as to include the movements of the babies' hands and arms as well as their head and eyes. Two infrared lamps (Kodak Satellitefilter no. 11) were positioned on opposite sides of the table below the screen. Behind the screenwere mounted two 8 mm Bolex motion picture projectors arranged so as to project two films side by side on the screen. The objects filmed each cast an image approximately 17 cm in diameter on the screen. An experimenter sat at one side of the mother's chair, arranged so that she could place and remove objects in front of the infant and pressthe switch that started the two projectors. Small observation holes were cut in the wooden frameholding the screenat a comfortableheight for observing the infant's eye movements. Button boxes were placed behind the screenfor recording. When one was pressed, a pen on an event recorder in the next room was activated. A TV monitor in the next room was available for a second experimenter, who watched the infant on the screen and timed its hand movementswith another button box. The times were accumulated on a digital clock. An intercom system permitted communicationwith the experimenterin the room with the infant. Objects The objects that the infant handled were in one caserigid (hard), madeof wood with a thin coating of spongerubber; in the other they were elastic (soft), cut from sponge rubber. Both objects were circular, 61 cm in diameter and 2! cm thick. In the center of each was a single dot of luminescentpaint. A sufficient supply of theseobjects was constructedfor two of either kind to be availableto the infant throughout the first part of the experiment. Motion picturedisplays Two motion pictures were made. One was of a circular spongerubber object (18.5 cm in diameter) being moved in a series

Knowledge

of

projective

of

same

a

black

were

an

so

adult

the

typical

other

.

objects

of

The

The

pick

darkness

two

identical

table

,

for

the

in

a

sit

table

the

baby

pulated

infant

the

event

right

recorder

or

left

to

checked

. 90

eye

s

The

and

,

complete

for

the

to

a

failure

the

.

elastic

to

one

in

with

the

Half

the

;

was

checked

objects

subjects

.

whose

,

two

to

two

data

because

tape

the

observer

data

to

to

buttons

eyes

the

.

and

The

did

had

achieved

not

been

agreements

the

objects

and

were

infant

fussiness

16

run

discarded

of

depicting

The

,

so

.

infants

were

were

transmit

photograph

hard

Eight

film

observers

and

continued

with

32

The

-

with

films

looking

s

If

mani

room

The

two

'

had

.

was

of

.

experimenter

left

to

on

attention

the

the

baby

-

objects

experimenter

the

the

one

the

.

the

the

.

infant

the

discover

the

baby

,

box

pushing

on

sec

button

experiments

also

,

had

the

projected

familiarized

of

a

on

slightly

the

on

-

before

in

it

,

to

' s

projectors

had

watching

previous

infant

60

com

hard

around

infants

half

using

by

the

it

the

ones

up

experimenter

on

right

pick

When

of

it

put

requested

not

objects

total

)

itself

been

.

soft

to

handling

-

which

received

left

catch

the

,

to

infants

did

one

experi

it

then

(

the

the

a

lighted

before

dimmed

having

turned

.

dimly

spongy

infant

and

screen

or

the

look

,

table

was

whether

film

be

were

total

were

experimenter

Half

infant

the

,

camera

experiment

handle

sec

screen

in

a

,

the

video

infants

ones

for

and

the

could

Sixteen

object

indicated

of

movements

)

gloved

identical

gradually

while

movements

reliability

.

1978

They

motions

not

The

might

signaled

30

eye

side

for

above

. ,

was

The

the

or

.

object

objects

behind

which

was

al

.

spongy

the

ones

identical

on

,

was

know

an

of

side

experimenter

.

in

moved

other

'

et

them

a

on

mother

If

another

total

baby

)

The

the

.

threw

the

lap

.

spots

monitor

a

,

.

possible

substance

watched

the

up

s

toy

visible

experimenter

the

familiarized

other

characteristic

Gibson

were

were

( hard

luminescent

or

for

the

the

removed

projected

the

texture

and

removed

ones

them

as

TV

see

visible

,

lights

was

rigid

substituted

or

'

a

The

toy

soft

the

the

(

some

noted

with

spots

pick

or

;

moving

case

be

mother

,

either

quietly

,

one

the

equipment

received

that

it

watching

failure

half

dropped

retrieved

soft

luminescent

minutes

so

its

.

,

the

as

it

hand

one

will

minute

the

objects

few

the

and

and

to

on

a

manipulate

,

objects

hand

sat

and

and

it

object

545

pattern

was

in

,

Substance

.

about

up

with

objects

object

rigid

provide

The

object

filmed

infant

room

could

rigid

a

to

to

.

moving

handled

Procedure

plete

a

objects

actually

mental

the

of

deforming

applied

background

only

a

is

spotted

black

that

Affordances

in

pressure

lightly

a

,

Tactual

of

moved

when

against

in

being

object

objects

filmed

-

characteristic

object

spongy

The

Visual

transformations

the

of

of

-

,

and

with

who

five

one

the

did

not

because

of

because

of

.

546 Table

E. J. Gibson& A. S. Walker 30 .1

Frequencyof Occurrence of Exploratory Manipulations, Experiment 1 ~

~

&

-

Throws

Hard Soft ~P <

-

/

Presses

Drops -

Touches

Strikes

Mouths

50

26

31

40 ~

21

95 ~

61

60

9

13

.05 .

Results

Exploratoryhandling Videotapes were examined for evidence of haptic exploratory behavior. A coding schemewas designed to include all the behaviors that we could distinguish. Two independent coders (not the authors), who had previously practiced on tapes of other experiments, viewed the tapes. The tapes were reviewed in casesof disagreement , and an agreement was reached. Some of the coded behaviors occurred very rarely or not at all (e.g., hands to parent or experimenter , passes from hand

to hand, or holds both) and were dropped from the analysis. Other distinctions proved to be too fine. The categorieswere collapsed to yield five overall . Total frequencies for these five cate .gories are entered in Table 30 .1.

"Presses " refers to gripping an object with adjustmentof the fingers so as to apply pressure. Movement of the fingers against the object, with taut muscles, was the criterion. The grip was sometimesreleasedand tightened in a cyclic fashion. "Touches," in contrast, refers to touching lightly or relaxed holding without gripping or changing pressure. "Strikes" refers to production of an impact on the table or on a secondobject. To test differences in type of manipulation of the two substances , Mann-Whitney U tests were performed on each of the five categories. The differenceswere significant as follows: mouths, U(16,16) = 99, N .S.; throws/ drops U(16,16) = 81, N .S.; touches, U(16,16) = 79, P < .10; strikes , U (16 ,16 ) = 71 , P < .05 ; presses , U (16 ,16 ) = 65 , P < .02 (all two -

tailed). Frequency of striking was associatedsignificantly with the hard object ; pressing (squeezing) was associated with the soft one. Looking preference Twenty -three of the infants looked first to the film

representing the familiarized substance , a significant preferenceby a binomial

test , p < .01 . Mean duration

of the first look was 2 .009 sec when

it was familiar and .844 sec when it was novel, as portrayed in the bar graphs of Figure 30.1. An analysisof variancewas run with group (soft or hard) x side (right or left) X preference, with duration of first look (familiar vs. novel) as the dependentmeasure. The main effect of duration was

Knowledge of Visual -Tactual Affordances of Substance 547 Experiment 1- dark

Experiment 2 - light

.90

~ .80 - . 70 't 0 c .60

.584 *

551 .449

., 8 .50 5

.416

.40

Q.

e

.30

Q.. .20 . 10 familiar

novel

familiar

novel

Figure 30 .1 Mean length of first looks and proportion of total looking time for the familiar versus the novel object in Experiment 1 (familiarization in the dark ) and Experiment 2 (familiarization in the light ).

significant , F ( I ,24 ) = were significant . When

total

sidered , 20

4 .65 , P <

looking

of

the

time

32

familiarized

substance

. The

substance

was

difference

did

not

for

infants

familiarized This

.0413 . No other

the

whole

showed

mean .551

reach

effects

30 - sec test

a preference

proportion (SO =

or interactions

period

for

of time , t (31 ) =

assigned

spent

con at

looking

the

at the

in Figure

1 .54 , P <

for

was

looking

.186 ), as portrayed

significance

were no significant effects of substance infant , or side of presentation .

main

30 .1 .

. 134 . There

familiarization

, sex of

Discussion Both

questions

mative

asked

. Infants

given

in the introduction an opportunity

have to

substance in darkness did exhibit diversified object was more often pressed or squeezed on

a resisting

consequences a surface

surface . Pressure than

leads

Thus , detection actions

to

pressure

ensued . Then , when

acteristic substances

of the two were

affordances objects

substances

distinguished

were by more

object

than was

moving

first

leads a rigid

striking

made

possible

for

glances

visual

unknown

. A spongy often struck to

different

object

against

a spongy , and

in differential

presented

in the affir of

exploration one more

object ; striking

consequences

answered an object

manual , a hard

a spongy

on a rigid

different

of different

on

been

explore

one .

different

patterns

char -

comparison

and longer

first

, the

glances

548

E. J. Gibson & A . S. Walker

at whichever substancehad been familiarized by manual exploration, as if confirming what had been touched. Why do these infants selectthe familiarized substancefor the first look and look longer at it when it is the first look? A few babiesdid look at the novel substancesfirst, as is almost bound to happen when films are suddenly flashedon in a dark room. If the head is turned to one side, the first glance is apt to fall there. But in thesecases , the look was shorter, and the babiesturned to the other object. Unimodal familiarization and preference tests (all visual, for example) have commonly shown a novelty preference on the preferencetest. If the baby recognizesthe object as the sameone, it should presumablyselectsomething novel to look at, the argument goes. 'Thequestionariseswhether infants familiarizedhaptically with an object in the light, so that they could seeas well as feel it, would look preferentially at the film displaying a novel property. Experiment 2 was designed to answerthis questionby providing visual familiarization as well as a visual test. Experiment2

Subjects The subjectswere approximately 1 year old, ranging in age from 355 to 384 days. There were 17 boys and 15 girls. Numbers of boys and girls familiarized with the hard objects were equal. For the soft object, there were nine boys and sevengirls. An additional eight infants' data were not included- five becauseof equipment failure, two becauseof crying, and one becausethe baby had Down's Syndrome. Procedure This experiment replicated the first as closely as possible, except that the infants were familiarizedhaptically with an object in a lighted room, so that they could seethe object aswell as handleit . Otherwise, the procedurewas nearly identical. Thirty -two infants were given either two hard or two soft objectsto manipulate, but in this casethe lights were dimmed only slightly. Again, 16 infants were familiarized with the hard objects and 16 with the soft. Following 60 sec of manipulation by the infant (again timed by an experimenterwatching the TV screen), the lights were dimmed further and the films were presentedfor 30 sec. Looking time was monitored asbefore. Results

Exploratory handling The infants ' exploratory handling of the objects was coded by two observers. The useful categories of Experiment 1 were

Knowledge of Visual-Tactual Affordances of Substance included

and

the

some

lighted

at

the

a

,

than

babies

seems

frequent

something

of

puted

,

.S

76

holds

,

N

,

P

<

.;

. 06

; ,

Looking

preference

looking

time

in

, 24

ward

)

the

substance

.

P

<

)

( 16

=

as

before

. 85

,

The

the

of

mean

. 584

92

.;

.S

first

30

.;

the

vastly

more

struck

against

more

holds

,

U

U

were

( 16

looks

/

frequent

tests

,

and

throws

mouths

looks

and

( 16

,

, 16

)

com

, 16

,

drops

. 1

.

=

of

1

the

When

. 18

)

=

-

119

,

U

( 16

, 16

)

=

U

( 16

, 16

)

=

104

,

=

proportion

of

) ,

time

a

significant

;

.

spent

of .

direction

infants

N

.S

.;

,

looks

When

mean

total

preference

looking

=

familiar

. 683

sec

directed

looking

for

proportion

are

length a

were

total

These

and showed

novel

first

a

proportion examined

the the

sec

32

substance

showed

the were

,

mean

Twenty

. 174

was

U

rotates

substances

infants

=

.S

;

.;

considered

,

32

( SO

.S

. 0001

N

'

familiarized

the

,

Figure

familiar

N

N

frequent

manipulatory

considerably

112

<

familiar

were

.

,

P

the

again

- Whitney

The

.

infants

in

. 038

,

= . S

=

96

24

N

and

:

<

=

)

Mann

) .

because

)

was

the

this

more

,

( squeezing

was

.

it

object

show

doing

Perhaps

held

one

action

, 16

)

=

) ,

the

looks

P

of

23

, 16

111

graphs

4

was . 01

U )

, 16

, 16

( 16

they

hard

to

and

.

of

the

turning

while

shows

pressing

the

because

observing

as

varied

.2

as

others

U

( I6

novel

first

,

,

substance

the

bar

film

considered

, 16

U

( 16

,

( 16

'

=

U

Both

infants

preference

( I

strikes U

in

,

,

,

to

portrayed

F

looks

But

the

than

30

object

and

although

touches

presses

both

the

:

and

.S

,

babies

follows

the

,

and

occurred

Table

.

object

often

the

as

.,

holds

of

soft

more

some

at

pertinent

such

more

as

informative

the

rotating

considerably

looking

now

ones

sometimes

1 ,

less

with

were

included

was

busy

that

unexpected

Experiment

so

behavior

These

( manipulation

in

were

added

some

activity

overall

100

and

parent

manipulatory

N

.

time

to

in

were

environment

same

object

others

549

time

the

at

;

-

was

familiarized

the

,

to

familiarized

t ( 31

)

=

2

. 723

,

.

Discussion Familiarization in both Experiments 1 and 2 resulted in a preferencefor looking at the familiarizedsubstance , when familiarizationwas haptic alone and when it was both haptic and visual. However one explains the direction of the preference,the evidencethat there is one is convincing. Because it is about equally evident whether familiarizationtakesplacein the dark or the light, we can infer that there is intermodal knowledge of somecommon properties of substancesby 1 year of age. From these data, it is unclear whether infants can extract invariant information or whether they have associatedhaptic experiencewith visual experiencein similar situations when both haptic and visual information were available. Finding that infants have intermodal knowledge of substancebefore they have gained control of opportunities for simultaneousmanipulation and looking would

r~~]L'J~~[I:-:-!(J~J~]~~=I'j~[lJ~ .~rJI ]~Uiil ~!t[lJ J[-= ;[~ ~tj~ '~~1[~~J~ 'JlllJ ~:~=J~ J['~~~.JL 'J~JIJJrll ~_J[tJ~ J~ 'J~~LU ~~flli ~~J~J=1 r~"~ (t~~~l.1~1[-.;1"

550 E. J. Gibson& A. S. Walker

Knowledge ofVisual-Tactual Affordances ofSubstance 551 arguefor the viewthat theremaybe directdetectionof intermodalin-

variants without associative mediation. Experiment 3 addresses thispoint. Experiment3

Thisexperiment wasdesigned toprovide asclose a parallel aspossible withthefirstone,butwithmuch younger infants (about I month ofage). Since manual exploration wasoutofthequestion, oralexploration pro-

videdthehapticexperience forfamiliarization. Wereasoned thatinfantsof thisage,whilehavingexperience ofmouthing varioussubstances, havenot

hadtheopportunity ofsimultaneously watching thepatterns ofmovement produced bytheirmouthing. Theexperiment wasinmostcases performed in thehomesof thesubjects, ratherthanthelaboratory, sinceparents wereoftenreluctant tobringoutsuchyounginfants. Thischange necessitatedportable equipment andratherdrastic differences inthedisplay for testingpreferential lookingforexample, a display ofrealobjects rather

thanfilmsforthe preference test. Subjects

Thesubjects were32infants ranging from26to45days(median age,30 days). Overall, 56infants participated, 24ofwhom didnotcomplete the

experimentbecauseof sleepingor crying,or had to be eliminatedbecause

of equipment failureor experimenter error.Twenty-four of the infants werebrought to thelaboratory bytheirparents. Theprocedure wasthe samein thesecases,exceptthat the screenswerebuiltin and a Harvard

Apparatus event recorder was used.

Apparatusand Procedure for Familiarization

Theinfantreclined at a comfortable anglein an infantseat,withfoam

rubber padsateachsideofitsheadtokeepit frominclining to oneside. Theseatwasplaced onthefloorina three-sided enclosure formed by portable blackpanels thatscreened offtheroom.Thebabysmother was askedto inserttheobjectforfamiliarization inthebabysmouthandhold

it gentlyat oneenduntilaskedto withdraw it. Therewasno needto dim

thelight,sincethemothershandconcealed the objectfromtheinfants

view. Themother kneeled behind theinfant andwasnotvisible. Therigid (hard) object wasacylinder oflucite 1cmindiameter and7.5cmlong.The elastic (soft) object wasa stickofporous plastic ofapproximately, the

samedimensions cutfroma cellulose dishsponge(boiling rendered this

objectquiteflexible as wellas sterile). Halfthe babiesreceivedthe hard

object, andhalfreceived thesoftone.Theinfants mouthing wasobserved through a holeinoneoftheblackpanels andwastimedbyanobserver.

552

E. J. Gibson& A. s. Walker

When the infant had mouthed the object for 60 sec, the parent was asked to withdraw it . Apparatusand Procedure for VisualPreference Test The objects for display following familiarization were two cylinders of the same length as the familiarized object and 3 cm in diameter. One was constructed of foam rubber and was very pliable; the other was made of wood covered with a thin layer of the samefoam rubber so that the two looked identical except when manipulated. The screendirectly in front of the infant had two slots through which an experimenter's hands could be thrust to provide the display. They were centered, side by side, 22 cm apart. The experimenterwore black gloves and was trained to manipulatethe objects simultaneouslyso that the speed of movement was approximately equal. Both were gripped at either end of the long axis. One (the hard one) was rotated around the vertical and horizontal axesin arcsof about 15 degreesto provide a seriesof perspective transformations; the other was squeezed and releasedrhythmically. The moving display was approximately 27 cm from the infant's eyes, slightly above eye level. The test was divided in halves. For each infant, one half the time the familiarizedobject was displayedto the infant's right, the other half on its left. Each of these displays lasted 30 sec. Half the infants had the familiarizedobject presentedfirst on the right, and the other half had it first on the left. The dual test was necessaryto control for side bias, sinceat 4 weeksinfants do not always have good head control. Looking time to one side or the other was monitored by an experimenter concealedbehind the screen, who watched the infant's eyes through a peephole in the screen. She pressedone of two button boxes while the infant's gazewas on one side or the other. Looking times were recordedon a portable Rustrakevent recorder. In most cases , three experimenterswere present; when only two were present, the observer was kept ignorant of either the substanceof the familiarizedobject or the position of eachobject during the preferencetest. Results The results of interest in this experiment were the infants' looking preferences. When total looking time for the combined test periods was considered, the proportion of looking time for the novel substancewas greater for 23 of the 32 infants (significantat p < 0.1 by a binomial test). The mean proportion of time spent looking at the novel substancewas .605; at the familiar it was .395. This differenceis significant, t(31) = 2.915, P < .0066. An analysisof variancewith substance , side (right or left), test order (first or second), and sex as factors showed none of these factors to be signifi-

Knowledgeof Visual-TactualAffordancesof Substance 553 cant. Means for the groups familiarized on the soft or the hard substance show the sametrend as the overall mean. For infants familiarized on the soft substance , meanproportion of time looking to the novel was .571; to the familiar it was .429. For infants familiarized to the hard substance , the mean proportion of looking to the novel was .639; to the familiar it was .361. Proportion of looking time to the novel substancewas .553 on test 1; it was .666 on test 2. The location and duration of first looks, although a very informative measurefor older subjects, was not considereduseful at this age becauseof possiblesidebias resulting from poor headcontrol and long reaction times. Discussion These data support the conclusion that by 1 month infants may detect information for substance that is accessible intermodally , as indicated by the preference for looking at the novel display following oral haptic fami liarization . It will be noted that the displays were identical in appearance except for type of motion , which was typical of the movement of a rigid object in one case and of a deformable one in the other . We were not able to obtain qualitative data on mouthing the two substances in this experiment , as we would have liked . There is evidence that babies do show some differentiation in oral exploration (Alegria & Noirot , 1982; Allen2 ), although it has not as yet been related to differences in substance. General Discussion All three experiments converge to show that intermodal information about substances, carried by differentiating motions , is picked up by infants , very likely by 1 month of age. But why did the older infants look preferentially at the familiar substance whereas the younger ones showed a preference for looking at the novel one? Many factors may constrain the preference (see Meltzoff , 1981). The explanation may rest in some developmental change. It has been suggested that very young infants show a familiar preference (Meltz off & Borton , 1979), or that preference is related to duration of familiarization and that younger babies need more time to achieve familiarity (Rose, Gottfried , Melloy -Carminar , & Bridger , 1982). Neither of these explanations fits the present case. Perhaps older babies have a stronger tendency to unify multi modal experiences; Bahrick (in press) found that babies detected an intermodal invariant specifying substance (heard and seen) and looked preferentially at a film depicting the event just heard. Her subjects were 4 months old . An alternative explanation should be considered . It has been suggested that similarity of the preference situation to the familiarizing one may affect

554

E. J. Gibson& A. S. Walker

the choice of familiar or novel (Rolfe & Day ; 1981; Ruff, 1981). In Experi -

ments 1 and 2, movies rather than real objects were presented for the visual preferencetest. The objects and their substanceswere not actually the same; they were pictured in a film raised slightly above the subject's eyes, whereasthe felt object was a solid "thing" resting on the table. Both films were novel , but one offered the same invariant

information

for identi -

fying a substancethat feeling it could provide, and a tendencyto unify two impressions, gained visually and haptically, was apparently strong. To put it another way , the infant detected an intermodal invariant , a very impor -

tant accomplishmentfor economy of perceiving and eventual comprehension of an external world with permanent distinguishing properties of things. In Experiment 3, the visual display consistedof real solid objects of the same length and only slightly greater in diameter than the ones mouthed . There were no distinctive visual features except the motions ,

which were also relevant for the haptic experienceand served to differentiate

the substances

in both

cases . Motion

is in any

case very

attractive

to young infants. It seemsreasonableto hypothesizethat the novel motion caught their attention. However the preferencefor a novel display versus a familiar display is to be explained, our experiments converge toward the conclusion that detection of intermodal invariant information for substancesthat are rigid or elastic is possiblein young infants. It is likely that this information can be detected without previous association of unimodal experiences.

A final pair of questionsarisethat invite speculation: What is perceived? What is the nature of the correspondencewhen an intermodal invariant is detected? Gibson's (1979) ecological approach to perception hypothesizes that what is perceived, even very early in life, is the affordanceof something, although exactly what affordancesare perceived undoubtedly changesdevelopmentally. Affordance refersto actions- what canbe done with something. Is it, perhaps, the affordanceof the substancesthat is perceived in unimodal presentations, either haptic or optical, that accountsfor the correspondencewe take for granted in normal multimodal perceiving? Notes 1. We wish to thank Lorraine Bahrick and Ryan Bliss for their very active and creative

participation in this research. Description of procedures that were tried and found unworkable are available in a previous report to the Spencer Foundation. Weare also indebted to Janet Weinstein and Patty Phillips for coding of tapes and assistancein running subjects. This researchwas supported by a grant from the SpencerFoundation. 2. Allen, T. W . Oral-oral and oral-visual objectdiscrimination. Paper presented at the third International

Conference

on Infant

Studies , Austin

, Texas , March

1982 .

Knowledge of Visual-Tactual Affordances of Substance

555

References Alegria, J., & Noirot , E. On early development of oriented mouthing in neonates. In J. Mehler, M . Garret, & E. Walker (Eds.), Perspectives on mentalrepresentation . Hillsdale, N .J.: Erlbaum , 1982 .

Bahrick, L. E. Infants' perception of substance and temporal synchrony in multimodal events. Infant Behaviorand Development , in press. Gibson, E. J. The concept of affordancesin development: The renascenceof functionalism. In W . A . Collins (Ed.), The Minnesota symposiaon child development , Vol . 15: The conceptof development . Hillsdale, N .J.: Erlbaum, 1982. Gibson, E. J., Owsley, C. J., & Johnston, J. Perceptionof invariants by five-month-old infants: Differentiation of two types of motion. Developmental Psychology , 1978, 14, 407- 415. Gibson, E. J., Owsley, C. J., Walker, A . S., & Megaw-Nyce, J. Development of the perception of invariants: Substanceand shape. Perception , 1979, 8, 609- 619. Gibson, J. J. The ecologicalapproachto visual perception . Boston: Houghton Mifflin , 1979. Gottfried , A . W ., & Rose, S. A . Tactual recognition memory in infants. Child Development , 1980

, 51 , 69 - 71 .

Meltzoff , A . N . Imitation, intermodal coordination, and representation in early infancy. In G. Butterworth (Ed.), Infancyand epistemology . Brighton: Harvester Press, 1981. Meltzoff , A . N ., & Borton, R. W . Intermodal matching by human neonates. Nature, 1979, 282 , 403 - 404 .

Rolfe , S. A ., & Day , R. H . Effects of similarity and dissimilarity

between familiarization

and

test objects on recognition memory in infants following unimodal and bimodal familiarization . Child Development, 1981 , 52 , 1308 - 1312 .

Rose, S. A ., Gottfried, A . W ., Melloy -Carminar, P., & Bridger, W . H. Familiarity and novelty preferencesin infant recognition memory: Implications for information processing. Developmental Psychology , 1982, 18, 704- 713. Ruff, H. A . Effect of context on infants' responsesto novel objects. Developmental Psychology , 1981

, 17 , 87 - 89 .

Soroka , S. M ., Corter , C . M " & Abramovitch

, R , Infants ' tactual

discrimination

of novel

and

familiar tactual stimuli. Child Development , 1979, 50, 1251- 1253. Walker, A . 5., Gibson, E. J., Owsley, C. J., Megaw-Nyce, J., & Bahrick, L. Detection of elasticity as an invariant property of objects by young infants. Perception , 1980, 9, 713 - 718 .

31

The Conceptof Affordancesin Development: The Renascence

of Functionalism

Eleanor ] . Gibson

This paper is included becauseit marked an important occasion. The symposium

tookplacein thefall of 1980, only a little morethan a yearafter thepublication of JamesGibson's The EcologicalApproach to Visual Perception. That book presentedthefirst lengthy discussionof the conceptof affordances. The symposium was an opportunity to introduce the conceptto developmentalpsychologists, relate

it to its historicalbackground , and show how it might profitably be appliedto developmental researchand theory. Here the paperhas beendrasticallycut by removingthesectionson thehistoryoffunctionalismand thedescriptionof several other experimentsso as to introduce only the conceptitself and a major illustrative experiment .

At the time the paper was written, no researchat all had beenpublished under the tenn "a./fordance , " although earlier work existed that could be cited as rel-

evant: mostparticularly, the work by Walk and myselfon the visualcliff, which examinedperceptionof the affordanceof a surfaceof support by crawling infants and many other animals, and the work by William Schiff on perception of

imminentcollision,or "looming" (Schiff1965). Schiffhadobserved responses to a looming shadow in a number of animals. A few years later, severalexperimenters

observed responses to loomingin humaninfants. Thepaperthat follows describes an experimentperformed by John Carroll and myself the first undertaken exclusively in an effort to study perception of affordances. We sought to determine

whetherthe avoidanceresponse to a loomingeventwas affectedin an adaptive fashionwhentheloomingobjectcontainedan apertureratherthan constitutingan obstruction. The distinction seemedto pose a truly diagnostic question as to

whetherthe infants' actionwasguidedadaptivelyby the eventpresented , indicating perception of the contrasting affordances. In recent years, more experiments have beenperfonned on perceptionof affordances , and the questionsasked have In W . A . Collins (ed.), The Conceptof Development . The Minnesota Symposiumon Child Psychology , Vol . 15. Hillsdale, N .].: L. Erlbaum, 1982, 55- 81.

558

E. J. Gibson

been formulated much more specifically to reflect the concept's reference to an animal -environment fit . There have been experiments on perception of apertures , in fact , asking whether an animal perceives the exact fit of the aperture in relation to its own size (Warren and Whang 1987 , Palmer 1987) . The concept of affor dances has been sharpened and its theoretical value enhanced greatly by this later research, but I think the emphasis here on the functional nature of the concept and its place in psychology was important for its acceptance by many developmental psychologists , v ~ .

It is a matter

of common

ideas are new . Physicists

agreement

among

scientists

find new particles , evolutionary

that

not

biologists

many and

geneticists move toward new models of speciation and evolutionary change (Lewin 1980 ), but the discoveries and new views always have a past . Still , over the decades or at least centuries , science progresses , sometimes in spurts and sometimes slowly , to new understandings . Often , this happens because a way of viewing the phenomena of some scientific discipline , once discarded as worn out and unproductive , appears in a new light ; reclothed and fresh , it points the way out of traps in which our thinking had become stalled and stereotyped , allowing us to take a naive look at what we are trying

to understand , unfettered

by paradigms

place of the original problems . My chapter focuses on a concept

that had slowly

taken the

that is new and fresh , with

great

possibilities for offering insights into development . The concept is that of affordance , taken from a way of thinking about perception and the world to be perceived that was recently described by James Gibson in his last book , The Ecological Approach to Visual Perception. This way of thinking has a past in the early flourishing of a peculiarly American psychology , the psychology of functionalism , as opposed to the structuralism of the British empiricists and many German psychologists , which found its ultimate elaboration and also its nadir in Titchener ' s laboratory . Functionalism

declined

in the finally

fossilized

and unproductive

blind

alley of 5-R theory . But its earlier beginnings tie in remarkably well with modem notions of ecology and the mutual , reciprocal relations of living creatures and their environments . 50 does the concept of affordance . The affordances of the environment

are what it offers the animal , what

it provides or furnishes , either for good or ill . The verb to afford is found in the dictionary , but the noun affordance is not . I have made it up . I mean by it something that refers to both the environment and the animal in a way that no existing term does . It implies the comple mentarity of the animal and the environment . The antecedents of the term and the history of the concept will be treated later ; for the present , let us consider examples of an affordance .

Concept of Affordances in Development

559

If a terrestrial surface is nearly horizontal (instead of slanted), nearly

flat (insteadof convex or concave), and sufficiently extended(relative to the size of the animal), and if its substanceis rigid (relative to the weight of the animal), then the surfaceaffordssupport. It is a surfaceof support, and we call it a substratum, ground, or floor. It is stand-onable, permitting an upright posture for quadrupedsand bipeds. It is therefore

walk -on -able and run -over -able . It is not sink -into -able like

a surfaceof water or a swamp, that is, for heavy terrestrial animals. Support for water bugs is different (J. ]. Gibson 1979, 127). The descriptionjust given soundslike a physical description, and so it is. But notice that it is also relative to the animal. The properties that define the affordance have unity only " relative to the posture and behavior of the

animalbeing considered." The mutuality of the affordancesof the environment and the behavior of the animal is the essentialthing. ]. ]. Gibson (1979 ) continues :

An important fact about the affordances of the environment is that

they are in a senseobjective, real, and physical, unlike values and meanings, which are often supposedto be subjective, phenomenal, and mental . But, actually , an affordance is neither an objective prop erty nor a subjective property ; or it is both if you like . An affordance

cuts across the dichotomy of subjective-objective and helps us to understandits inadequacy. It is equally a fact of the environment and a fact of behavior. It is both physical and psychical, yet neither. An affordance points both ways to the environment and to the observer (p . 129 ).

Notice that the animal's behavior and perception are both involved here. The animal behaves in accordance with the affordances of the environment

,

and this depends on his perceiving them. Weare at once concerned, as psychologists, with what he can do and also with what he can perceive. Acting appropriately implies perception of the affordancesoffered. As a developmentalpsychologist, a rather grand program immediately opensup before me : I need to find out what the environment

affordances-

offers in the way of

how to describe them, what the appropriate behaviors are-

and also whether and when they are perceivedas affordances. Perceiving an affordanceimplies perception that is meaningful, unitary, utilitarian , and continuous

over

time

to the extent

that

environmental

events that pertain to the observer may require . To what extent must young creatures (human or otherwise ) learn to perceive them? And if they must learn , how

is it done ? Affordances

are not invented

or read into

events by the perceiver . They are there to be perceived . A lever affords

facilitation of moving something, even in the caseof a small child who is

560

E. J. Gibson

as yet ignorantof its utility. He simply doesnot perceiveits' affordance . (Materialdeletedhere.)

TheDevelopmental Study ofAffordances Someaffordances seemto be peceived , by someanimals , withoutthe necessity of learning of anykind. A newlyhatched chickperceives the peckability of abit of grainatonce . Spalding 's (1875 ) swallows perceived theopenvistasforflightpathsdespite havingbeenreared insidea small box. Butasfor human creatures , we knowlittle or nothingaboutthe development of perception of affordances . It hasbeenfashionable for psychologists , in ourrecent pastat least , to assume thatallthemeanings (inthesense of information foraction ) thatI havebeencalling affordances mustbelearned throughassociation withaction , thoughtherehasbeen littleagreement abouthowthatmighthappen . Luckily , thepresent state of ourtechnology forstudying perceptual development in infancy issuffi cientto permitusto pursue serious investigations of howperception of various affordances develops . Thequestion I askis this: Wherethereis invariant information thatspecifies anobject , a place , or anevent , and affords information foraction , istheaffordance detected byaninfant ?If so, when , andhowdoestheperception ofaffordances develop ?It makes sense to beginwithaffordances of events involving objects , surfaces , andpropertieslikesubstance thatseem to bebasicfortheyounghuman creature . Thedirectwayof asking thequestion is to choose a significant eventfor display , in whichinvariant information fortheenvironmental eventcanbe specified , andto seewhether appropriate behavior ensues . Doestheinfant actin accord withthespecified consequences (in thesense thatthefinal trajectory of theballisspecified forthecatcher earlier in thetransforming flightpath )? A less -directwayof askingmy question is to lookfor appreciation of intermodal relations in anevent . Affordances haveinterrnodal consequences thatimplyabstract relations . Knowledge of these , if it canbe demonstrated , implies perceived meaning , notmeresensory processing of proximal stimuli . Evidence of thiskindismorecontroversial , I admit . TheAlfordanceof a Surfaceof Support About 20 yearsago, RichardWalk and I devisedan experimentalsetupthat we dubbed a "visual cliff." We were set on investigating the detection of what we referred to as depth at an edge, and the method had two great advantages. The information for differentiating affordances (though it never occurred to us then to use that word) was easily specified, and the opportunity for behaving appropriately (or inappropriately) was available.

Concept ofAffordances in Development 561

Thecliffoffers,as I wouldnowput it, information for a solidsurfaceof supportfor the animalon the nearside,but not the far.The two surfaces

arespecified byoptical information forgradients oftexture, andbytrans-

formations in the ambientarraywhenthe animalmovesits heador locomotes.Perspective transformations ofthepatterned texturespecify thatthe surfaceis rigid,andthe depthof the droppedsurfaceon the farsideis specified by occlusion of textureundertheedgebetweenthetwoas the perceiver moveshishead.Information supplied bydisparity gradients may also be available. Themainthingis thatthereis information fora surfacethatis walk-on-

able,thatcanbestepped onorcrawled on,ononeside,butfora place

thatdoesnot affordsupportor affordsfallingoffon the other.Precocial

animals likechicks andbabygoatshavebeenshownto perceive these affordances at once,andso do ratsthathavebeenrearedin thedarkforan

appreciable time.Human infants mostly perceive theaffordance bythetime

they can locomote (crawl)on their own.

Whether human infants could perceive thedepthattheedgebefore they

were able to crawlwas a moot questionfor sometime.In an effortto answerit, Camposand his colleagues(Camposet al. 1978)decidedto use

a different measure of thebabysresponsiveness, thechangein itsheart

ratewhenit wasplacedon a Plexiglassurfaceandallowedto viewa surface

affording supportdirectlyunderit,ascompared withone4-1/2feetbelow. Thenotionwasthat if the babyperceived dangerwhentherewasno

underlying surface ofsupport, itsheartrateshould accelerate ascompared

withthesafe side,indicating an emotionof anxiety.Infantsunderfive monthsgaveno indication of an accelerated heartrate,but,in fact,there wasa relative deceleration, a changethatis generally thoughtto accompanyraptattention in infants. Indication of acceleration beganto appear

onlyafterabout ninemonths, wellaftermostinfants respond appropriately

in the cliff situation.

Thisdisjunction of actionanda presumed indicator of emotionunder-

lines anotheraspect of affordances.They are not the attachmentto a

perception offeelings ofpleasantness or unpleasantness. Theyareinforma-

tionforbehavior thatisofsomepotential utilitytotheanimal. Aroadmay afford crossing (ifthelightsaregreen), butit doesnotnecessarily giveone

pleasureto observethis,especially if onedoesnot wishto crossthe road.I

doubtthata mountain goatpeering overa steepcragisafraid orcharged withanykindofemotion; hesimply doesnotstepoff.Manypeople do become afraidofheights at somepoint,butthisfearisprobably learned longaftermotorpatterns forresponding appropriately tosurfaces ofsupporthavedeveloped. Thispointhasbeenclarified recently insomeelegant

research byRaderandhercolleagues (Rader, Bausano, andRichards 1980).

562

E. J. Gibson

Graspability andPalpability : Affordances of Objects Objects and the substancesthey are madeof have innumerableaffordances , but one of the simplest ones to describe and to observe in an infant is graspability. An object of a certain size (not too big and not too small in relation to the infant's hands), not too far away for the length of the baby's arms, and having the property of substantiality- that is, more or less opaque and rigid- provides optical information in the ambient lighted array for graspability. Size, distance, and texture are optically specifiedin well-understood ways, as is opaquenessas contrasted with transparency. Optical information for rigidity is specifiedwhen slight movementsof the object occur. This information must be detectedby an infant for an object's affordance of graspability to be observed. We cannote that it hasbeenwhen an infant reachesappropriately for an object and adjusts the shape of its hands during the reachso as to grasp it . There hasbeen considerableresearchon reaching and grasping in recent years, with some disagreementas to the time of onset of successfulreachingand grasping. But the evidencenow is fairly conclusivethat by three months or so, infants do not reach toward an object that is too far away or not of a graspablesize. Field (1977) tested infants at two and five months with objects of varied sizes at variable distances. At two months, infants distinguished objects that were too far away from ones within reach by appropriate arm adduction (even when the retinal image was held constant). At five months, there was evidenceof adjusted reaching for objects at variable distances. Bruner and Koslowski (1972) varied the size of objects presentedto infants from two and a half to five months. Objects of graspablesize were differentiated by different arm and hand movementsby three months. Objects within reachbut too large to grasp were approachedwith different hand and arm movements than ones of graspablesize. It is even more interesting to considerthe affordanceof a moving object. Infants ---------- are --- notoriously ~ very- interestedin moving things and watch them attentively as neonates. But when are they perceived as something graspable, when the information of velocity and trajectory must be perceived, as well as that for size, substantiality, and so on? There is available some remarkableresearchon the baby as a catcherby von Hofsten and Lindhagen (1979). They presentedseatedinfants beginning about fourteen weeks with an attractive moving object (a fishing lure). It was mounted on the end of a sort of boom that swung round on a pivot so that it cameoccasionally within the baby's reach. Infants successfullyreachedfor and caught this object at about eighteen weeks, the same time that they had mastered reaching for stationary ones. Von Hofsten ( 1980) later analyzed the infants' arm movements during reaching, over the period from eighteen to

Concept of Affordances in Development

563

thirty-six weeksof age. Themovements improvedin skilloverthistime, becoming moreeconomical (fewerseparate smallmovements ) andmore ballistic . Butevenat thebeginning , thebabies put a handoutto catchthe objectnotwhere it wasseen atagiveninstant butwhere it wouldbe,predicting theobject 's trajectory . Lateron, someof theinfantsuseda handto trail theobjectandchase it, speeding upfor thecatch , butI ammostimpressed by theearliest reaction , beingreadyfor thecatchin therightplaceat the righttime. VonHofsten 's (1980 ) resultsimplya basichumancapacity to coordinate one's behavior temporally with external events , ashesays : "to foreseein one's actionsfuturelocationsof movingobjects ." The timing consideration is overwhelming proof , it seems to me , of a perceived affordance . Objectsnotonlyafford(or donotafford ) grasping ; theyaffordactivities havingto do with variableproperties of thesubstance theyaremadeof, properties discoverable by palpating butalsopotentiallyvisiblewhenthe objectis movedor palpated in appropriate ways.Of course , surfaces have substantial properties too, evenwhentheyarenot surfaces of detached objects . Somesubstances affordwalkingon or poundingwith or being pounded ; othersafforddrinkingor bathing ; andothersaffordsqueezing andall kindsof contactcomfort . We referto themby suchtermsas rigid, fluid, elastic , or spongy . Information for perceiving thesesubstances visually , in particular thedifference between rigidandelasticsubstances , is givenonlyby motion.Theseareimportantproperties of thingsat a very earlyage. Theearliestactivitiesof a babyincludea lot of chewingand mouthing and,alittlelater,alot of throwingandbanging . I havebeenvery interested in earlydifferentiation of theaffordance of rigid ascontrasted with deformable elasticsubstances andwith someyoungcolleagues have donea numberof experiments on theirdetection(Gibson , Owsley , and Johnston 1978 ; Gibsonet al. 1979 ; Walkeret al. 1981 ). (Materialdeleted here.) Obstacles andPassages : Affordances of Events

Thereis an experimental situation , the so-calledloomingexperiment , in whichthe perceived affordance is that of an imminentcollisionto be avoided . Theinformation for this eventwasdescribed by James Gibson thirtyyearsagowhenhesuggested thatopticalinformation for thespatial layoutconsisted of spatiotemporal flow fieldsratherthana sequence of staticretinalimages (VisualWorld , 1950 ). In a paperonvisuallycontrolled locomotion (Gibson1958 ), hewrote: Approach to a solidsurface is specified by a centrifugal flow of the textureof the opticarray.Approachto an objectis specified by a

564

E. J. Gibson

magnification of the closedcontour in the array correspondingto the edgesof the object. A uniformrate of approachis accompaniedby an accelerated rate of magnification. At the theoretical point where the eye touchesthe object, the latter will intercept a visual angle of 180 ; the magnificationreachesan explosive rate in the last momentsbefore contact. This acceleratedexpansionin the field of view specifiesim-

minent collision, and it is unquestionably an effectivestimulusfor behaviour in animalswith well-developed visual systems. In man, it produces eye blinking and aversive movements of the head, even when the stimulus is a harmless magnification of a shadow on a translucent

screen (p . 188 ).

In 1965, Schiff published a monograph describing researchon looming with several species of animal (monkeys , fiddler crabs, frogs , chicks, ki ~~ens,

and an occasionalhuman). The loom was produced with a shadow cast on a screen , and the inverse transformation

, a minification

of the shadow , was

used as a control . The subjects all demonstrated some form of avoidance behavior , however

different

their visual and response systems , to the flow

pattern of acceleratedmagnification. Some years later, researchersbegan to study the sensitivity of human infants to optical information for impending collision. Bower, Broughton, and Moore (1970) reported that very young infants respondedwith avoidance behavior that included head withdrawal , eye widening , and raising

the handsbetween the object and the face. Both a real approachingobject and a magnified shadow were displayed. Ball and Tronick (1971) reported similar behavior in one-month-old infants with both types of display and compared the collision condition (symmetrical expansion) with an asym-

metricalopticalexpansionpatternthat providedinformationfor a "miss" course. The infants in this casetracked the display to the side rather than exhibiting avoidanceor withdrawing. But a differential heart rate, showing accelerationto the collision display and not to withdrawal, occurred only with older infants after eight months. Avoidance responses , in this case as well as with the cliff, are apparently not necessarilyaccompaniedby physiological indicators of fear or anxiety. Since that time , a number of experiments using more refined response

measureshave been performed with looming displays, and carefullongi tudinal comparisons have been made (reviewed by Yonas 1981). Yonas'

own experiments on sensitivity to optical information for collision have made particular use of blinking, a responsethat can be very reliably observed, and have been extended to the study of preterm and postterm infants , as well as full -term ones (Petterson , Yonas , and Fisch 1980 ). The

latter experimentsprovided evidencethat maturation plays an important role in the development of the perceivedaffordanceof optical information

Concept of Affordances in Development

565

for impending collision. Defensive blinking to appropriate displays was advanced in postterm six-week-old infants as compared with full-term

infants at six weeks .

Research of my own with John Carroll extends the usefulness of the concept of affordance in the analysis of how appropriate ways of acting

develop in responseto optical information about the layout of the environment . The paragraph I quoted from Gibson (1958) described the informa -

tion for approachof an object as "magnificationof the closedcontour in the array correspondingto the edgesof the object." But this is only part of the story . There are obstacles with which one may collide , but there are also

paths that open up and afford locomotion through a passageway.In both cases , there are contours or edges, but the affordancefor locomotion varies with the presenceor absenceof other information. Quoting Gibson (1958) .

agaIn :

In short , the lay of the land, the jumping -off places, the interspaces, barriers and obstacles, as well as the level stretches, are given by the

geometry of the optic array. Depending on the locomotor capacities of the animal, this terrain provides definite possibilities or impossibili -

ties for crawling, walking, climbing and the like. And if the animal can discriminate

the textural

variables

it can discriminate

among

potential paths for locomotion. A potential path is a stretch of surface extending away from the animalwhich affords the kind of locomotion for which the animal is equipped. A barrier or obstacle is a surface which does not afford locomotion

(p . 192 ).

This description emphasizesthe role of textural variables, but Gibson had not yet combined this variable with information deriving from flow patterns producedas a creaturemoved forward along a path or toward an obstacle, or as an obstacle or a contour around an opening moved toward

him/ her. What is the optical information for approachingan opening that has to be steered through, as opposed to an obstacle with the same contours that will bar locomotion and afford collision? According to Gibson (1979), when an opening is to be entered, movement toward it is

accompaniedby " opening up of a vista," by "magnification or gain of structure

inside the contour

and not loss outside " (p . 234 ). As a barrier is

approached(or an obstacleapproachesone), it occludesmore and more of the backgroundstructure until it is lost entirely at 180 . Our experiment undertook to contrast these two situations with infants

as subjects. Do infants, before they achieveindependentlocomotion of any kind, differentiatebetweena path aheadthat provides an opening vista and might afford passagefrom an obstacle that progressively occludesmore and more of the backgroundstructure, even as its own textured structureis

566

E. J. Gibson

magnified? To put it another way, if a rectangularcontour is approaching or approachedby an animal, when is it perceivedas affording a barrier that makesstopping obligatory and when is it perceivedas a vista, an opening, or aperture that affords passingthrough? What is the information for the affordancein either case, and when is the affordancedetected? We choseto present the two situations in the form of something moving toward the infant, as is done in the looming experiment. An obstacleor an aperture of the samemeasurementswas moved toward the stationary subject. The optical flow pattern of acceleratedexpansionof the contour as it approachedis the sameas if the infant approachedit, and deletion or accretion of background structure would also be the same. The optical expansionpattern of the contour of the obstacleand of the aperturewould be identical to one another. Panelsproviding an obstacle or an aperture condition were moved toward an infant in approachtrials and were moved away in withdrawal trials. The infants' behavior was recorded on videotape, and pressureof the subject's head against a foam rubber headrest was measured. The subjectsin this experiment were approximately three months old. The infant sat at the rear of a booth surrounded by curtains on three sides, all textured with the samepattern. A flat table extended in front of the subject. At the far end of the booth could be placed either of two movable panels, both covered with the sametextured material as the walls of the booth, so that, when flush with the rear wall, they covered it and tended to fade into it . Movement of either panel resulted in optical information for edges at the contours of the obstacleor the opening. The panelswere supportedon invisible carts that were moved toward or away from the subject by an unseen experimenter, at a constant velocity. A panel's approach trip began at the rear wall and ended just short of the infant's face. The movement of the panellastedapproximately six seconds. Eachinfant took part in four conditions of the experiment: (1) ten trials with the obstacleapproaching(Obstacle Approach); (2) which were alternated with ten trials with the obstaclewithdrawing; and (3) ten trials with the apertureapproaching(Aperture Approach); (4) alternatedwith ten trials with the aperturewithdrawing. The order of the main condition (obstacle or aperture) was alternatedfrom one subjectto the next. The results indicated that theseinfants definitely differentiated between the two approachconditions, although no infant cried or becamealarmed as the obstacle approached(this finding agreeswith all the more recent researchon looming, which the obstaclecondition resembledin its arrangements). The results of change in head pressureare especially interesting. We calculateda mean curve of head pressurechangefor every infant in all four conditions, and also a composite group curve for each of the four

Concept of Affordances in Development

567

11 . 2

OBSTACLE ZOOM

OBSTACLE

.... ' "

LOOM

9 .6

" ,

M

5

" ,

8.0

"in E

-

" - ", APERTURE

~

6 4

0)

.

ZOO

"

M . .. ..

- ....... .

.

. . . .

.

. .

. .. . . .

I"i ' " u. .

..

." .

.

. .

:J

....... . . . .

' " -

. .

W

.

.

.

.

.

.

.

.

4 .8

. L - ,

.

.

.

'-

.

.

~

~

-... -

.

. .

.

.. .. ..

.

.

..."' , ...7 "-"",

~ W II : 0-

.. .. . "

.

.

"",,'" . . .

3

2

......"""' APERTURE

- - ", "' - ' "

...

LOOM

- - -- - ... -

.

-

-

-

. . .

0

", '

<X: W

"

", "' - ' "

/'

I

1 .

,, '

0

- 1 .6 0

1

3

4

TIME

( sec )

6

7

Figure 31 .1

conditions. The group curvesmay be seenin the figure. Approach trials in both main conditions are grossly different from withdrawal trials, with positive slopes for approach (increasingpressureof the head against the headrest) and negative slopesfor withdrawal, where the infants' postures relaxed and backward

head pressure was released . However , curves for the

two approach conditions differ from one another . Head pressure increases

very rapidly in the "ObstacleApproach" condition, rising steeply until the cessation of the panel's movement . In the "Aperture Approach " condition , head pressure increases for the first three seconds of the panel's movement

and then shows little change. Typically, infants in this condition seemedto track the top edge of the opening upward at first, but, as it came closer, they turned their headstoward a side of the opening. Regressioncoefficientswere calculatedfor the slopesof eachof the four mean curves for all the individual

infants , and these bear out the group

trends. For example, the coefficient of the positive slope for "Obstacle Approach" is higher than for "Aperture Approach" for twelve of the sixteen infants. In an analysisof variance peformed on these coefficients, the effect of approach condition was significant (p = < .018), but the interaction of condition and order was not , so the two conditions no matter which came first .

differed

568

E. J. Gibson

Differences in head pressure are only part of the story , however . A more

complex pattern of behaviors differentiated the two approachconditions. The videotapes of each infant were analyzed and coded for a number of different behaviors. Behaviors that differed significantly and reliably in frequency of occurrence(more frequent for the obstacle's approach) were reaching toward the panel, excited movement, arms up, head turned to avoid, anticipation, frowning, and head withdrawal. Blinking camecloseto significance. Two behaviors, head turning to track and passive interest,

werenotedsignificantlymoreoften asthe apertureapproached . Reaching and touching were especiallynotable as the obstacleapproached , frequently with pressureexerted on the surfaceof the obstacle. The beginning of the reach more often resembled the "placing" response of a kitten extending its forepaws as it is moved toward a surface, than a defensive response of hands raised before the face .

Altogether, the patterns of behavior and the differencesappearedto be appropriateindicators of the affordancesof the two situations, approachof an impassablesurfaceor obstaclein one caseand a vista or passagewayin the other. Potential paths for and barriers to locomotion may be differentiated early in life, before self-initiated locomotion has begun. (Material deleted here .)

Conclusion

Development is best viewed in the context of an organismin a relationship of mutuality with its environment. Perceptionservesthe function of keeping an animal

in touch

with

the information

in its environment

about

affordancesfor actions, like going placesand making use of surrounding objects in ways that serve its needs. Evolution of the species has done a lot

to provide an individual with the meansfor accomplishingthis, but in the development of an individual , we need to study the way these means are

refinedby ongoingdifferentiationof perceivedaffordances , of actions , and the way the two are related .

A functionalist perspectivekeepsin mind the utility of actions for supporting a creature's needs. But the creature, human in the case that interests

me most, is at once an actor and a perceiver. He cannot act adaptively without perceiving the affordances of his habitat . But on the other hand, he cannot perceive effectively without acting . Weare , even as infants , active

seekersof information, and as we perform actions, we put ourselvesin a position to obtain more information . The flow of perceiving and acting is continuous

and unbroken .

This point brings me to another , very important , facet of a functionalist view . It shuns atomism in any form . It does not seek components or final

particles of any kind, whether stimuli, or sensations , or memory traces, or

569

Concept of Affordances in Development

responses . The flow of information around us that our environs provide is continuous, and the things that are specified, the surfacesand placesand people, are unitary and meaningful. It is the meanings that we need to extract. The complementarityof the animal and its environment is a whole and must be studied as such. The more we try to decomposethis complementarity by looking for elements, the more likely we are to sacrificethe meaningswe are looking for. It is a sort of corollary of these points that we should try to study development in the most lifelike possible surroundings. That does not mean that we should shun the laboratory, but rather that we should make conditions of our experiments as much like real ones as possible. The researchI have been describing seemsto me to do that by attempting to simulatereal situationswithout sacrificingcontrol. I have alreadyalludedto the program of researchthat this perspective encourages . The study of affordancesof the things and placesof the world, how they are perceived and lead to action is the way to proceed. Primitive affordancesin the young infant are being studied, and some appearto be tuned in innately, but perception is bound to action, and most affordancesmust be detected in the course of action. New actions become possible, as development progresses , and thus new information for affordancesis continually becoming available. The old functionalistswere essentiallypragmatists, and at one point this pragmatismled to formulation of the goal of psychology as the prediction and control of behavior. I have never liked this slogan, becauseit was not clear who was going to do the controlling. It is control of one's own behavior that is important. Behavior occursin an environment, and adaptive control requiresperceiving the affordancesof that environment. References Ball,W. A., andTronick , E. Infantresponses to impending collision : Opticalandreal.Science , 1971 , 171,818- 820. Bower , T. G. R., Broughton , M. M., andMoore , M. K. Infantresponses to approaching objects : An indicator of response to distal variables . Perception and Psychophysics , 1970 , 9. 193- 196. Bruner

, J. 5 ., and Koslowski , B . Visually Perception , 1972 , 1 , 3 - 14 .

Campos

, J. J., Hiatt

on the visual York : Plenum Dewey Field

, J. The

reflex

, J. Coordination 103 , 48 - 57 .

, S ., Ramsay

, D ., Henderson

cliff . In M . Lewis , 1978 . arc of

concept vision

prepared

and

, C ., and

L . Rosenblum

in psychology and

constituents

prehension

Svejda

manipulatory

, M . The

emergence

( Eds .) , The development

. Psychological in

of

young

Review infants

, 1896

. Child

action

of

.

fear

of affect . New

, 3 , 357 - 370 .

Development

, 1977

,

Garcia, J. Tilting at the paper windmills of Academe. American Psychologist , 1981, 36, 149- 158.

570

E. J. Gibson

Gibson, E. J., Carroll, J., and Ferwerda, J. Differentiation of an aperture as contrasted with an obstacle under conditions of optical motion. In Final Reportto the SpencerFoundation . August I , 1980. Gibson, E. J., Owsley, C. J., and Johnston, J. Perception of invariants by 5-month-old infants: Differentiation of two types of motion . Developmental Psychology , 1978, 14, 407- 415. Gibson, E. J., Owsley, C. }., Walker, A ., and Megaw-Nyce, J. Development of the perception of invariants: Substanceand shape. Perception , 1979, 8, 609- 619. Gibson, J. J. Theperceptionof the visual world. Boston: Houghton-Mifflin , 1950. Gibson, J. J. Visually controlled locomotion and visual orientation in animals. British Journal of Psychology , 1958, 49, 182- 194. Gibson, J. J. Theecologicalapproachto visual perception . Boston: Houghton-Mifflin , 1979. Heidbreder, E. Sevenpsychologies . New York : Century, 1933. Hofsten, C. von . Predictive reaching for moving objects by human infants. journal of ExperimentalChild Psychology , 1980, 30, 369- 382. Hofsten, C. von. and Lindhagen, K. Observations on the development of reaching for moving objects. Journalof ExperimentalChild Psychology , 1979, 28, 158- 173. Hruska, D ., and Yonas, A . Developmental changes in cardiac responses to the optical stimulus of impending collision. Paperpresented at the meetingof the Societyfor Psycho physiologicalResearch , St. Louis, October 1971. James,W . Principlesof psychology . New York : Holt , 1890. Johansson, G., von Hofts~en, C., and Janssen , G. Event Perception. Annual Reviewsof Psychology , 1980, 31, 27- 63. Koffka, K. K. Principlesof Gestaltpsychology . New York: Harcourt, 1935. Kohler, W . Thementalityof apes.New York : Harcourt, 1925. Lewin, R. Evolutionary theory under fire. Science , 1980, 210, 883- 887. Murchison, C. Psychologies of 1930. Worcester, Mass.: Clark University Press, 1930. Petterson, L., Yon as, A ., and Fisch, R. O . The development of blinking in response ~o impending collision in preterm, full-term, and postterm infants. Infant Behaviorand Development , 1980, 3, 155- 165. Rader, M ., Bausano, M ., and Richards, J. E. On the nature of the visual-cliff avoidance responsein human infants. Child Development , 1980, 51, 61- 68. Ryan, T. A . Dynamic, physiognomic, and other neglected properties of perceived objects: A new approach to comprehending. Americanjournal of Psychology , 1938, 51, 629650. Schiff, W . The perception of impending collision: A study of visually directed avoidant behavior. Psycholofl .ical Monographs, 1965, 79 (Whole No . 604). Spalding, D. A . Instinct and acquisition. Nature, 1875, 12, 507- 508. Spelke, E. Infants' intermodal perception of events. CognitivePsychology , 1976, 8, 533- 560. Spelke, E. Intermodalexplorationby 4-month-old infants: Perceptionand knowledgeof auditoryvisual events . Ph.D . thesis, Cornell University , 1978. Spelke, E. Perceiving bimodally specifiedevents in infancy. DevelopmentalPsychology , 1979, 15, 626- 636. Tolman, E. C. Purposivebehaviorin animalsand men. New York: Century, 1932. Walker, A . S. The perceptionof facial and vocal expressions by human infants. Ph.D . thesis, Cornell University , 1980. Walker, A ., Gibson, E. J., Owsley, C. J., Megaw-Nyce, J., and Bahrick, L. Detection of elasticity as an invariant property of objects by young infants. Perception , 1980, 9, 713- 718. Yonas, A . Infants' responsesto optical information for collision. In R. N . Aslin, J. Alberts, and M . Peterson, (Eds.), Sensoryand perceptualdevelopment : Influencesof geneticand experientialfactors, 1981.

32 Detection

of

Crawling

and

E . J . Gibson

the

T raversability

Walking

of Surfaces

Infants

, G . Riccio , M . A . Schmuckler

T . A . Stoffregen

by

, D . Rosenberg

,

, J . Taormina

This paper records the results of a program of experimentson perception of the affordanceof a surface with regard to locomotion- in other words, its traversability . The method is directly descended from myoId studieswith the visual cliff; this time, we built what we referred to as walkways, with interchangeable surfaces. The particular affordanceinvestigated has to do with substance - what the surface is made of- and so is also related to the previous investigations of detectingrigidity of objectsoptically and haptically. A new question is introduced that has to do with the concept of affordances- that is, the influence of the particular action to be performed in relation to the environmental property that is required to support it . The concept of affordance holds that the utility of an environmental support must be relative to the structure and dynamic propertiesof the organism involved in an interaction with the environment. The shoulderwidth of a walker, for example, would define the minimal width of an aperture so as to afford passing through it; the weight of an animal would limit the surfacesthat might bear it; and its manipulatory action system would determine whether it could pick up berries or hold a spoon. Likewise, the level of developmentof an animal of a given specieswould constrain the affordance of environmental supports for certain actions such as jumping , even when the speciesat maturity is capableof such an interaction. Whether an animal perceives the affordancesfor acting appropriately, in agreementwith both environmental and organismic constraints, is a questionfor research(seeNewell [ 1986] for a useful discussionof constraints on development of perceptionand action). It is also essentialto ask what the information for the affordanceis, and how the animal picks up that information (if it does). We asked thesequestionswith regard to infants in early stages of locomotion, comparing infants who were only capableof crawling with infants who had begun to walk . Journalof Experimental Psychology : HumanPerception andPerformance , 1987, 13, 4, 533- 544. Copyright 1987by the AmericanPsychological Association . Reprintedby permissionof the publisher .

572

E. J. Gibson , et al.

Walking

requires

surface

.

asked

supportable to

crawlers

,

that

,

of

Many

in

whether

not

? Are

they

relation

a

difficult

to

they .

on

a

newly

aware

of

the

a

rigid

explored

Do

, given

the

task

between

properties

do

constraints ,

, and their

crawlers

systems

were

reported

here

its

these

major

,

pliant highly

the

walkers

surfaces

differ

available

from

means

capabilities

supportability

, or

of

four

of

a

studies

and

this

ment

property lA

, crawling

surface

: a that

duration

a

of

lack

surface

coded

of

to

be

from

walkers

,

in

rigid

formation deformable

in

the the

and

the

use

infant

nonrigid

surface

, as

an

. ,

differed affordance

.

a

toy

supportability

. ( 1989

active

) .

spontaneous to

capitalize

to

for

2 ,

of

theory

would

of

and

8,

detecting

and

surface

displacement

in , and

Experi

-

forcing 4A

a and

8

traversability

varied walkers predict

in

locomotion

walkway

exploration

crawlers

, were

deforming

replicated

3A

in

locomotion

more

dual

-

deforming

activity

mode

a

,

Experi

surfaces a

initiate

were

with

. In

and

, the

information

Patterns

both

displacement

, and

traversal

deformable

two

rigid

typical

Experiments haptic

or

with

findings

presented

, or

standard

that

for

rigid

felt

behavior

. and

its

possible

watch

affordance

Latency and

the

. These

surfaces optical

and

and

suggesting

surfaces

presented

presented

nonrigid

was

two of

-

as

with

fashion

either

were

with

,

of

the

looked

exploratory

crawlers

experi

such

Schmuckler

, haptically

exploration

traversability

which

between

both

. Compared , more

perceived 18 ,

investigated of

not

of was

optically

felt

haptic

do

's

adaptive

presented

infants that

evidence

an

,

to

and

perception

either

and

and

latency

but

influences

choice

looked

videotapes

longer

the

walking

the in

to

surface

Gibson

Some

.

surface

surface

visual

elicited

ment

and

it

.

surface

have

the

in

years the

opportunity

indicate

of

offers

specified

standard both

of

use

five of

the

was

and

. The

was

to

experiments

investigated

surface

succession

as

of

Others

briej1y

. environment

we

supporting

." like

so

infants

period features

" holes

reported

these of

. the

affordances ""

In

all

part

a

other

information

are

of

the

with

surface

experiments

on

the

a

over

do

regards

of

on

outcome

exvloration

to

as

sources

performing

of

performed

have

integrity

secondary tool

Some

I on

own

and of

a

The

new

experiments not

opacity

or

that

is

distinguish

?

ments

use

sets

their

traversed

ones about

actions

, which

would

deformable

information

perform

balance

walkers

and

obtain

locomotion

bipedal

whether

surfaces

actively

them

maintaining

We

with in

the

the case

in of

a

.

How does anyone decide whether an unfamiliar surfacestretching ahead over the ground is safefor travel by foot, by car, by ski, or by any other mode of transportation?! Particularly, how does a young member of the human species, newly ready to undertake self-initiated and self-propelled locomotion, detect the properties of a surfacethat make it safeto traverse? Is adult tuition necessary ? Reinforcementby trial and error? Or is in-

T raversability of Surfaces

573

formation for traversable properties detected, and if so, when does this occur? What is the relation between perception and action in the novice percei ver - actor? The questions relate to the concept of affordance (E. J. Gibson , 1982;

E. J. Gibson, 1984; J. J. Gibson, 1979). According to Gibson (1979), what is perceived first and foremost is utility for action- what an event or the layout of the environment or an object offers for doing something, either for good or ill . The concept refers to the unique relation between the potential actions of an animal and some aspectof its environment such as a place(e.g., a cave, which may be enteredand will afford shelter), an object (e.g., a stone, which may be graspedor hurled), or an event (e.g., a wave, which may be "ridden" or may engulf an animal). Objects, places, and events that provide support for adaptive actionsfor someanimalsmay not for others of different size, bodily structure, and organic requirements. The relation is an ecological one, the result of the evolution of an animal in a given niche. The relation is real, not simply "in the eye of the beholder ," and it may not be perceived by the animal. In some instances, perception of an affordance may be innate, in the sense that the animal

detectsinformation for it in an optical or other array and actsappropriately in a way characteristic of its species (e.g ., defensive or avoidance activities

of many animalsin responseto a "looming " object or shadow}; but it may be learned, as would be the affordanceof a tool for prodding something out of reach or for operating as a lever . The information for an affordance

is twofold : There must be information specifying properties of the thing perceived and also information about relevant capacities and structural

constraintsof the perceiver. A good examplethat has been studied experimentally is the affordanceof stair-height for climbing (Warren, 1984). To perceive the optimal stair-height, the perceiver must detect not only information for the height of the riser but also the constraintsimposedby his or her own stature. This apparently is done so well that the relation can be

describedmathematicallyand is predictive for an adult climber of stairs. In the experiments to be described, two classes of subjects were observed :

infants capableof crawling but not walking and infants capableof walking. Bipedal locomotion, as compared with crawling on four limbs, imposes constraints that may changethe affordanceof a surfacefor traversal. We were interested to know whether infants whose motor development had attainedbipedal locomotion acteddifferently than onescapableof crawling only, depending on a surface's affordance for traversability. Will they perceive surfaceproperties differently, in relation to their affordancefor an action of which they are newly capable? The information availablefor perceiving the traversability of surfacesis potentially manifold. There may be optical information for a surface's

574

E. J. Gibson, et al.

rigidity or elasticity and for discontinuities suchas holes or obstacles. The degree of rigidity and weight-bearing properties of a surface's substance may be detectedhaptically by exploring the surfacewith a hand or with a foot or a tool . Events that we witness, such as impact of an object striking a potential surfaceof support, can provide dynamic optical and acoustical information about the traversability of a surface. In a familiar situation most

of us rely heavily on barely attended optical information for guiding locomotion, but in a strangeplaceor under unusualclimatic conditions, we actively obtain information from all the available sources.

In this researchwe investigated perception of traversability in infants making independenttrips in an unfamiliar environment. The overall plan was to put young ambulatory infants in a novel situation where surfaces

varying in properties defining traversability or nontraversability stretched between the infant and a customary objective (a parent ). We observed the infant 's spontaneous behavior before embarking on a trip over the surface,

including latency to leave a starting position, the exploratory behavior (both visual and haptic), and the displacement(evasiveor playful) behavior that occurred, as well as the manner of locomotion

once the trip was

undertaken. Both crawling and walking subjectswere observedwhen presentedwith surfaceshaving different affordancesfor walking as compared with crawling .

The rigidity of a surface- its resistanceto deformation- was chosenas the property to be focused on in the experiments to be reported.2 This property was of particular interest for severalreasons. It can be specified both optically and haptically, permitting an analysis of what information specifying an affordanceis used. Earlier research(Gibson, Owsley, & Johnston, 1978; Gibson , Owsley , Walker , & Megaw -Nyce , 1979; Gibson &

Walker, 1984) showed that even precrawling infants can discriminate differencesin rigidity of objects that can be mouthed or handled. Here we askedwhether the affordanceof rigid and nonrigid surfaceswas detected with respectto locomotion. The following experimentssystematicallyvary the deformability of a surface, the sourcesof information that specify its deformability, and the locomotor development of the infant subjects. We asked whether the infants differentiated these surfaces behaviorally in both

exploratory and executive actions, whether they employed appropriate meansof detecting the relevant information, and whether an infant's locomotor development (crawling or walking) was related to differential behavior on surfacesspecifying rigidity versusnonrigidity . We reasonedthat maintaining upright posture and walking would be difficult on a deforming surface(weight must be supported and balancemaintained on one foot as the other is lifted and moves forward ), whereas it would

be feasible on a

rigid surface. Crawling, with four limbs to support the weight as the infant

T raversability of Surfaces

575

progressed, would be feasibleon both types of surface. A baby capableof walking was expected to detect the surface's affordancefor maintaining balanceand moving aheadbipedally, in contrast to one capableonly of crawling. We were especiallyconcernedto observethe infants' exploratory activities when confronted with these surfaces , in addition to their hesitation or lack of it in traversing them. Exploratory behavior of infants in handling and examining objects has been studied and is relatively - skillful in infants of the ages we observed (Palmer, 1985; Ruff, 1982, 1984), but we know little about it in the caseof a surfaceto be traversed by an ambulatory infant. Experiment lA

Two surfaceswere selectedfor the basicexperiment. One (to be termed the rigid surface) presentedboth optical and haptic information for rigidity ; the second (to be termed the deformingsurface) presented both optical and haptic information for non-rigidity . It was hypothesized that (a) infants would tend to embark upon the surfacepresenting information for good traversability more readily than the one not specifying that affordance, (b) that exploratory behavior would take placeand delay embarkationon a surfacespecifying poor traversability, (c) and that locomotor development would be linked to exploration and detection of traversability of the surface. General Method

SubjectsSubjectsincluded16crawlersand16walkers . Agesrangedfrom 8 to 14 monthsfor crawlers(M age= 10.25 months) and from 10 to 21 months(M age= 14.44 months) for walkers . Two infantsdid not yield usabledata, due to temperamental problems , experimenter error, and/ or because they refusedto cross eithersurface , in which casethey werereplaced . Only babieswho couldwalk at least 10 stepsindependentlywere classifiedas walkers . Subjectswere located throughbirth noticesin thelocalnewspaper , lettersof explanationweresentto the parents , andfinally, telephonecallsweremadeto solicitparticipation . Apparatus Walkwayswere constructedto hold differentexperimentalsurfaces (seeFigure32.1). A walkwaywassturdilybuilt of wood, 222 X 104cm in length andwidth. Framesat 85 cm abovethe floor heldthe experimental surfaces , which could(with one exception ) be removedandreplaced . A superstructure wasbuilt abovethe framesthat heldthe surfaces . The superstructure on the two long sides hadstrong, large-meshednetsstretchedalongthe sidesfor the infants' protection. At oneend(the startingposition), a low woodenfencepaddedwith foamrubber formeda backing.Curtainshungbehindthe fenceto concealanexperimenter . The

Traversability of Surfaces

577

Designand procedure Each infant was presented, once only, with both surfaces , alternating order over subjects. The baby was encouragedto play until it appeared to be comfortable and at ease. The parent then placed him or her on the starting platform in a sitting position, facing the far end. An experimenter remained standing behind the infant, concealedby curtains, while the parent walked to the far end and facedthe infant. If the baby did not look immediately at the parent, the experimenter pointed to direct the baby' 5 gaze. The parent was requested to smile at the baby silently for 30 s. After 30 5, if the infant was still on the platform, the parent was signaledto call the baby and urge him or her to come. After 60 s, the parent was signaled to take down a concealedtoy (a red plastic ring holding a bunch of metal keys), shakethe toy , and continue to urge the baby to come. If the baby still remained on the starting platform after 120s, the trial was terminated. During the procedure on the walkway, the infant was continuously videotaped. Treatment of results

Because our major data consisted of video -tapes of the in -

fants' behavior in a relatively free situation, it was necessaryto develop a procedure for coding behavior from the tapes that was reliable and yet captured the essenceof activity that might inform us about the child's perception of surface properties and detection of their affordances.We were particularly concernedwith the child's activities before leaving the platform- for example, whether the child simply implored help from the parent or whether the child engaged in his or her own information gathering or testing. Performancesof both kinds occurred, but study of the tapesof a pilot study made it clear that the latter kind of activity was more frequent and most likely to prove enlightening for a study of how affordancesare detected. Four major categoriesof activity were adopted: (a) latency to leave the starting platform, beginning when the parent appearedand ending when the child had three limbs off the platform; (b) accumulated duration of visual exploration (scanning over the surface) before leaving the platform; (c) accumulated duration of haptic exploration before leaving the platform (touching the surfacewas counted as haptic exploration only when accompaniedby looking); and (d) displacement , a term borrowed from ethological research, to which the present researchbearsa strong relation methodologically. Displacement designates behavior directed away from the "problem" confronting the infant. The child was occasionally playful (playing with the protective nets and rubber bumpers) but often was evasive (e.g., attempting

to part the rear curtains and I'escape" from

the situation). The category of displacement did not include behavior directed exclusively at the parent, suchas holding out arms toward her or him, pointing at the surfaceand looking at the parent, vocalizing with apparentintent to communicate , and so forth .

All tapes were coded independently by two coders, who practiced on pilot tapes before starting to code. Because the behaviors we were concerned with were

very easy to observe, training offered no problem. A coder watched the tape over and over , stop -watch in hand, and summed the time an infant spent in the various categories to be coded . When detailed activities were to be coded (such as types

of haptic activity , which could be idiosyncratic with respectto a surface), a check list was available. Correlation coefficients were run separately for the four categories to determine interrater reliability . They ranged from .946 to .996.

Traversabilityof Surfaces 579 main effect of surface, F(I , 28) = 10.797, P = .003. The Surfacex Motor Developmentinteraction was near significance,F(I , 28) = 3.175, P = .086. Haptic exploration The walkers spent more than twice as much time in haptic exploration of the waterbed surfaceas of the rigid surface, but there was no differencefor crawlers. Differenceswere not statistically significant, however, for surface, F(l , 28) = 2.308, P = .140, nor was there an interaction with motor development, F(l , 28) = 1.652, P = .209. There were large individual differences, but some walkers felt the surfaceintently and in a distinctive manner, pushing it with both hands and putting considerable weight on it. Watching them on the tapes gives the inescapable impressionthat they were "trying it out." Haptic exploration was codedby two observersfor differential use of hands and feet, and type of action. Comparisonsacrosssurfacessuggestthat there may have been somedifferentiation by type of haptic exploration. Exploration with the feet was rare, but it did occur with half the walkers on the waterbed (not at all with crawlers, however). Feeling or rubbing the surfacesoccurred with more than half of all babies, both walkers and crawlers, and was about the - - -same -----for the two surfaces.Patting the surfacewas fairly frequent, but less so on the waterbed. On the other hand, more pushing (putting weight on the surfacewith both hands) occurredwith the waterbed. Displacement and evasivebehavior The mean time spent in displacement behavior can be comparedin Figure 32.2. The waterbed elicited more displacementand evasivebehaviorin walkers, but less, if anything, in crawlers. Analysis of varianceindicated no main effect of surface, F(l , 28) = 0.727, P = .401, but a very significant interaction between surfacesand motor development, F(l , 28) = 9.107, P = .005. Multivariate analysis In view of the suggesteddisparity betweencrawlers and walkers on the waterbed comparedwith the rigid surfaceon several measures , a multivariate analysis of variance was performed including all four of the measures . For crawlersconsideredseparatelyas a group, the four variablestaken together did not differentiate the two surfaces , F(4, 27) = 1.55, P = .215. But for the walkers, the four variablesdifferentiatedthe two surfacessignificantly, F(4, 27) = 3.59, P = .018. When the two groups (crawlersvs. walkers) were compared, they were differentiatedsignificantly by this analysis, F(4, 25) = 3.06, P = .035. The four dependentmeasuresmade various contributions to the differencesfound in the multivariate analysesof variance. Canonicaldiscriminant analyseswere run to examinethe predictive value of the four measuresin differentiating the rigid surface from the waterbed for walkers. Visual exploration made the largest contribution (structure coefficient = .686),

580

E. J. Gibson, et al.

displacementwas next (structure coefficient = .551), haptic exploration third (structurecoefficient = .490), and latency madethe smallestcontribution (structure coefficient = .324). The predictive value of the four measuresfor differentiating walkers from crawlers on the waterbed/ rigid comparison showed the heaviest weighting for displacement(structure coefficient = .84), visual exploration second (.54), haptic exploration third (.43) and latency fourth (.41). Displacementmay indicate the presenceof conflict when a mother tries to entice her child onto an apparently unsafe surface; the waterbed surface evidently looked uninviting to walkers, but not to crawlers. Discussion With respect to our first question, whether the surfacesare differentiated behaviorally, the results suggest a qualified yes. As we had expected, the riQid as a standard, tended to have a shorter '-' surface. which we adopted latency for initiation of traversal and elicited a moderateamount of visual and haptic exploratory behavior and displacementactivity, as compared with a nonrigid surface (the waterbed) in the case or the walkers. The crawlers, on the other hand, did not differentiate thesetwo surfacesexcept by somewhatlonger visual exploration. This differenceis not surprisingfor an affordancetheory, becausethe waterbed surfaceaffords adequatesupport for crawling but not for walking. Walking infants differentiated this affordance from that of the rigid surface by all four of our measures , with the difference in age groups supported by a significant multivariate analysis. Experiment1B In this experiment we sought to replicate the trends just observed with a different method. The apparatuswas converted to a double-walkway choice A secondwalkway ---- -- - situation. - was built parallel to and joining the A original one, and the starting platform was placed centrally and lengthwise, partially straddling the surfaceof both walkways. A strip of wood (2 in. x 4 in., 5.08 cm x 10.16 cm) continued to the end of the walkways, making a slight (but not impassible) barrier between the two . The baby had to choose one surfaceor the other in leaving the starting platform, thus giving us a new measureof perceived relative traversability. The parent stood in the center, opposite the baby. Becausethere was a supporting post exactly " in the center between the two walkways, the parent was askedto playa kind or peekaboo, alternating the headfrom side to side. Otherwise, the procedure followed that of the first experiment. The rigid textured surfacewas on the baby's right and the waterbed surfaceon its left. The waterbed was gently agitated as before. There were 32 new subjects;

T raversability of Surfaces

581

16 crawlers (M age = 11.28 months; range = 9-13) and 16 walkers (M age = 13.81 months; range = 11- 18). This choice was presentedto each subject only once. Results Only the walkers (13 out of 16) showed a preferencefor the rigid surface over the waterbed (p = .01 by a binomial test). Sevencrawlers chose the rigid surface, 6 chose the waterbed, and 3 did not cross in the 2-min interval. Of 3 walkers who chose the waterbed, none walked. Of the 13 who chose the rigid surface, 7 walked. These results provide converging evidenceto support the trends observedin Experiment 1A. Latenciesfor leaving the platform are consistent with the ratios of choices; that is, for walkers, latency for choosing the rigid side was short, but for the waterbed side was long. For example, the mean latency for all the walkerswho chosethe rigid over the waterbedwas only 9.5 s, whereas it was 31.3 s for the few (3) who chosethe waterbed. For the crawlers, the casewas different; times were no longer (in fact, slightly less) for those who chose the waterbed than for those who chose the rigid surface (55.5 s comparedwith 64.0 s). Times spent in exploration and displacementbehavior tended to be shorter, on the whole, than in earlier experiments, probably due to the availability of an attractive alternative. They showed no particular trend. There was a potential side bias in the W / R (waterbed/ rigid) choice, becausethe waterbed could not be alternated from left to right. For this reason, the W / R condition was repeatedwith a new group of 16 walkers, with the ends of the walkways reversed so that left- right relations were transposed. Again, there was a preferencefor the rigid surface(12 out of 16 subjects, p = .0384). These results support the findings of Experiment 1A and lead us to conclude that there is evidence for a difference in perceived affordancesof a deforming, nonrigid surface, depending on the infants' locomotor development. The walkers perceivedthe waterbed as a surfacethat did not afford walking, but the crawlersdid not. Experiment2

The question of what information the walkers were using to detect the property of nonrigidity was our next focus of investigation. The waterbed presentedboth optical and haptic information for a deforming surface. The babies could have relied on either of these sources alone. It might be suspected(Hatwell, 1985) that they were making little use of haptic information, even though most of them did actively explore the surface manually. We sought to test the hypothesis that only optical information was effective by covering the waterbed with a Plexiglas surface. It could

582

E. J. Gibson, et al.

still

be

agitated

specifying would

tion

for

unseen

firm

to

was

presented

tured

once

rigid

waterbed

was , but

the

surface

the

waterbed

the .

so

that or

The

of

; range not

=

this ,

as a

in

the of

)

allow on

Plexiglas

tex .

rigid

-

The

condi

cm

-

between

deformations

Plexiglas

was

experimental

the

subject

4

for

the

-

infant

and

to

about

-

.

each

standard

space to

) .

( WR

tra

condition

and

subject

placed

experiment 20

of

,

also

surfaces

were

months

) . No

differently

studies

hand

and

test

be informa

the

laid

depth

over

would

the

not

differ

.

.5 -

responded

earlier

control

this

11

fabric

to

optical

condition from

detectable

the

to

Plexiglas

a

surface

walkways

it , with

the

the

information

the

single

same

and

optical

but

designed

waterbed

over

with

reflectance

subjects

months

the

adult

,

alternated

laid

interrnodally

slipperiness

had

an

was

the

that

contradicting

with

order

waterbed

( To

was surface

thus 2

with was

so

available ,

both (R );

Plexiglas of

difference rigid

with

covered

the

touch

conducted

condition

,

remain

the

. Experiment

experiment

was

experimenter

would

be

deformability

The

tion

an

deformation

versed

in

by

to

the

24

walkers

crawlers

(M

age

participated

waterbed

and

=

15

, because

the

rigid

. 69

they

surface

in

the

.

Results Of

the

24

subjects

,

surface

when

Plexiglas

walked

across

the

seem

that

unlike

these

the Analyses

F ( I ,

22

P

. 431

=

, )

=

2 . 156

,

for

Results 32

.3 ,

where

are

.

in

each

child

could

differences sure condition

For

from

for for

the .

,

in

the

not

the

. 266

,

,

for

four

;

F ( I ,

.

measures .

22

haptic

and

,

waterbed

surface

measures

=

would

surfaces

the

rigid

exploration

covered

,

its

on

own

)

=

,

Differ

-

0 . 527

exploration

F ( l ,

Plexiglas

on

for the

22

)

served

.

The

Positive exceeded

.

Data

ease

of

, ;

=

0 . 643

in

Figure

a

Experiment for

,

.

values

The

surface

rigid , the

that the

was

textured surface

for

-

across

represent indicate

that

experi

comparing

comparison

histograms

1 ,

each

rigid

surface as

.

shown in

textured

deforming

control

condition

are conditions

by

score

the

experiment

displacement

scores

thus

its

and

comparable

difference

infant score

experiment

each

visual

,

for

mean

each

experimental

for

P

two

on for

( 7 )

. It

the

walked

four ,

waterbed infants

individuals

patterned

the

1 . 304

exploration

data was

be

of

=

the of

for

infant

the

22

. ,

its

any )

same

separately

and

12

. 156

the

presented

subtracted surface

=

waterbed

experiments

each

P

latency with

the

,

F ( I ,

and number

affordance

no

performed

for ;

displacement

for

along

ment

significant

the

same

,

same

were

the

surface

surface

The

where

were

latency

rigid .

. They

perceived

waterbed

for

the them

surfaces

variance

not

. 476

crossed

experiments

of

were

=

two infants

the

ences P

24

covered

previous

comparing

all

standard

the

so

that mean mea ( rigid

)

Traversability of Surfaces 30 Difference

25

in Seconds between Rand W

20

Surfaces

Latency

15

I

Surfaces

I

5 0

Haptic Explorat ion

Difference inSeconds between Rand W15 Surfaces I

Visual ~ion Difference in Seconds between Rand W

583

0Expt .1Expt .2 20Displacement Difference 15 inSeconds between Rand W10 Surfaces .2

Figure32.3 Comparisonof ExperimentlA - waterbed(W) and rigid (R) surfaces - with Experiment 2- waterbed(Wp andrigid (Rp) surfaces underPlexiglas . (Valuesplotted aremeandifferencescores[W- R] and[Wp- Rp]. All subjectsarewalkers .)

A multivariate analysis of variance was performed on the differences betweenthe standardand comparisonsurfaceswithin the two experiments. When the four variableswere pooled aspredictors in the analysis, the conditions in the two experimentswere significantly differentiated, F(4, 33) = 4.34, P = .006. A canonicaldiscriminant analysisyielded structure coefficients of .75 for displacement, .63 for visual exploration, .33 for latency, and .33 for haptic exploration, the first two having the strongestpredictive value. The comparisonbetween the meansof the two experimentsdoes not support a visual dominancehypothesis, that optical information- in this case, for deformability- will determinethe perceivedaffordanceand override competing haptic information. The infants apparently detected the rigidity of the Plexiglas surfacecovering the waterbed by haptic testing and were capableof using the haptic information. However, it is possible that self-initiated change or nonchangein the optic array (change in the caseof Experiment I and no change in this experiment) is perceived as a consequenceof haptic activity and is attended to as a particularly informative event for discovering the affordanceof a surface. It is also the casethat haptic and optical information are contradictory in this experiment. The next experimentadopted the strategy of varying optical information with out placing it in conflict with haptic. In it we investigated the perception of a surface's affordancewhen optical information was reducedwhile haptic information was still available.

584

E. J. Gibson, et al.

Experiment

In

3A

this

experiment

ability

, optical

was

face

. Black

that

texture

was

for

surface was

previous

a

little

optical

with

the

experiments

infonnation

could

surface

be

provided

textureless The

two

and

16

)

other

;

and

infants

embark

nor

Purdy

,

surfaces their

&

were

there

in ,

, or a

to

each

whereas

non

a

the

the

-

major

1955

),

this

nonrigidity

.

standard

in

touch

case

-

such

any is

Lawrence

hard

rigidity

was

texture

as

sur

Lighting

rigidity used

travers

textureless

were

either

surface

in

was

10 for

or

a

single

Experiment

included ,

.

of

for

presented

age

surface

,

infonnation

months not

any

,

for a

the

, haptic

. The

textured

black

velveteen

.

of

. 63

were

on

for

mean

eye

rigid both

were

14

affordance with

surface

adult

( Gibson

textured

procedure

walkers

months

an

infonnation

optical

surfaces the

surface

not

's

inhomogeneity

obtained

did

surface

walkway

plywood

Because

. Because

good

surface

following

.

specifying

contrasted

a to

reflection

presented

a a

covered

discernible

in

condition

for

presenting

velveteen

homogeneities

It

by

matte

no

information

manipulated

lA

. 25

for

reasons

of

=

at

a

16

-

20

time

,

crawlers

( range

12

=

8 -

months

experimenter

problems

one

were

crawlers

( range

temperamental

,

Subjects

months

walkers for

walkway

.

) .

error

14

Five

, refusal

to

.

Results

Mode 14

of

locomotion

crossed

walking

an

crossed

each

Latency

to

equal the

)

opment

5 . 092

Visual

)

did

. 032

0 . 020

the

, it

. been ,

, but

Fantz

32

to

. 890

,

1958

, and

two

the

surfaces

16

crawlers

.

velveteen a

and

than effect

surface

mean

crawlers

surface

significant

of

for

Both

with

on

of

the

surface

motor

,

devel

-

.

latencies ( in

This

for

fact

, rather

finding ) as

1969

less

)

on visual

that

triggers opposed

- preference ,

embarking

suggests

texture

looking Karmel

histograms

surfaces

black

inhomogeneity

;

of

presents

interaction

more .

.4

showed

longer

no

by

surface

the

) . Fourteen

two

the

perceptible

Attention

;

=

surface

documented 1958

no

the

( visually

exploration

on

, P

rigid

crossed

each

the

variance

elicited

textured

inhomogeneity

( Berlyne

embark

Despite

surface

on

patterned

They

.

on

of

=

=

exploration

velveteen

often

to

, P

28

(6

the .

Figure

. Analysis

, F ( l ,

times

surfaces

locomotion

longer

=

crossed surface

locomotion

surface

F ( l , 28

walkers

of

two

initiate

took

textured

16 velveteen

number

of

to

walkers

black

initiate

latencies

than

All

the

relative

visual to

Analysis

of

black

optical

attention

and

homogeneity

experiments ) .

the

exploration

with variance

has infants indicated

Se Mo Sha "Co Ke -Se 60 5 0 4 30 ~ 20 ~ 10 R Re R Re Cr W Vis 2 Ex I0Cr R Re R Re W Traversability of Surfaces Haptic Exploration

Latency

Figure

32

.4

Responses

of

rigidity

( R )

crawlers

compared

information

no

( RBV

walkers

with

a

between

in

haptic

4

than

no

difference

Displacement

Mean

are

surface

cant

,

These

greater

active

,

P

F

( l

=

,

did

the

. 022

28

results

)

.

=

The

1

. 290

F

( I

suggested

the

optical

and

information

information

with

=

the

but

for

poor

optical

,

)

P

=

1

=

28

)

=

.

The

,

F

0

( I

,

P

=

exploration

.

28

. 441

)

,

,

=

of

. 028

. 512

,

P

surface

the

motor

the

.

Analysis

two

. 896

,

nor

.

on

for

-

more

variance

=

activities

times

nor

histo

no

Analysis

0

,

the

elicited

=

P

. 208

.

surface

displacement

. 4

. 662

. 288

the

black

of

surfaces

development

two

velveteen

,

was

F

variance

( l

,

28

not

)

signifi

=

-

.

the

patterned

28

surface

,

patterned

. 266

,

,

. 72

visual

between

on

( l

1

velveteen

with

,

activity

,

32

interaction

P

F

=

with

in

Figure

for

,

)

patterned

difference

displacement

compared

good

haptic

,

28

surfaces

spent

those

significant

,

black

,

in

exceeded

. 926

offering

good

case

the

times

presented

a

( l

the

that

interaction

indicated

surface

surfaces

F

between

developmental

surfaces

was

show

exploration

any

two

,

As

Figure

indicated

a

offering

interaction

exploration

grams

to

surface

the

developmental

Haptic

5

and

) .

difference

any

585

that

basis

the

of

longer

black

surface

latencies

velveteen

,

but

to

embark

surface

there

was

was

no

and

unattr

indication

-

586

E. J. Gibson, et al.

whatsoeverof a differencebasedon age or motor development. We sought replication of thesefindings in the next experiment. Experiment3B In this experiment, subjectswere presentedwith the dual walkway to force a choice between the black velveteen surface on plywood and the textured surfaceon plywood . The procedure was in every way comparable to Experiment lB. Thirty -two new subjects participated, 16 crawlers (M age = 10.47 months; range = 7.5- 15.5) and 16 walkers (M age = 13.94 months; range = 10- 17). The side on which the black velveteen surfacewas presentedwas alternatedfrom right to left from one subjectto the next. Results Both crawlers and walkers tended to choose the textured rigid surface in preference to the black velveteen one. Crawlers chose the patterned surface in 15 cases (p = .0003 by a binomial test). Walkers chose it in 11 (p = .0592 ), while 4 chose the black velveteen surface. Latencies to leave the platform were rather long but followed the same pattern as the patterns of choice: Crawlers choosing the textured surface had a mean of 53.13 s, while those choosing the black velveteen had a mean of 72.5 s; walkers choosing the textured surface had a mean of 30.25 s, while those choosing the black velveteen had a mean of 67.8 s. Times spent in haptic exploration were not long and did not favor either surface. Of the walkers , 6 of the 11 who chose the patterned surface walked on it . Of the 4 who chose the black velvet surface, I walked . The findings of this experiment support the conclusions drawn in Experi ment 3A. Lack of optical information for traversable properties of a surface both delayed locomotion and led to the choice of an alternative surface. The black velveteen surface was clearly unattractive , compared with the patterned surface. Differences in mean age and in motor development of the subjects made no difference , however . The reason for shunning the black velveteen surface was not , therefore , that it was perceived as not affording walking . In fact, a number of the walkers did walk on it (the same number as on the patterned surface in Experiment 3A). There was no information specifying that it was "unwalkable ." It would seem, rather, that the ambiguity of the optical information led to indecision (witness the increase in displacement activities ). Lack of visually perceptible " surfaciness" apparently leads to wariness in engaging in visually guided locomo tion in any form . The role of haptic information was apparently small in this situation . Why ? In the next experiments , we varied the kind of haptic information available while retaining the black textureless covering .

Traversability of Surfaces

587

Experiment4A The last experimentsleft many questionsabout the information for traversability and how it is used. Our original hypothesis was that infants would explore a surfaceactively, obtaining haptic information as well as optical information about the surface's properties. Experiment2 led us to think that walkers, at least, do so becausea rigid transparentcover over the deforming waterbed changed the pattern of results. Yet the babies explored the black velveteen, texturelesssurfaceonly slightly, and most hesitatedto embarkon it, despite its rigidity . It is the casethat any pressureapplied to the rigid surface, whatever its cover, yields no optical consequences(no event-like visible change). Might optical consequencesbecomeimportant, when static optical information is impoverished (as it was in Experiments 3A and s)? Would optical consequenceslead to enhancedhaptic exploration? Experiment 4A was designed to test this possibility. The waterbed was coveredwith the black velveteenfabric so that optical information was poor unless haptic exploration was engaged in. The waterbed was presentedwithout agitation (no waves) so that optically it resembledthe surface presentedin Experiment 3A, when the black velveteen covered plywood . But if the subject applied pressureto it with haptic testing, the surface would ripple and produce inhomogeneitiesthat were easily visible (there were shadowsand changesin reflectance , revealing the deforming motion). Experiment 4A presentedthis black velveteen surfaceover a waterbed, without agitation, on a single walkway. Each subject was run with the textured surfaceover plywood also so that the design was exactly comparable to Experiment 3A. The subjects were 16 crawlers (M age = 9.78 months; range = 9- 11.5) and 16 walkers (M age = 14.66 months; range = 13- 19 months). Results All of the walkers crossed the rigid surface; 14 of the 16 crossed the waterbed. They differentiated the two surfacesby their mode of locomotion. Ten of the 16 walkers walked to their mothers on the rigid surface; 4 attempted to stand up on the waterbed (but didn't walk). Fifteen of the crawlerscrossedthe rigid surface; 14 crossedthe waterbed. Resultsfor this experiment, accompaniedby similar measuresfor Experiment 3A, are presented in Figure 32.5. The histograms represent mean differencesbetween the texturelessblack velveteen surfaceand the comparison, textured, rigid, control surfacefor each infant, to make possible comparison across experiments. The results are of interest both across experiments (comparisonwith Experiment 3A), and for a comparisonbetween crawlersand walkers in Experiment4A. When Experiments4A and 3A are compared, it may be seen that the underlying waterbed, despite its lack of optical texture, elicited more ex-

T raversability of Surfaces

589

ploratory activity . The differenceis most apparentfor haptic exploration in the caseof the walkers. Visual exploration rose noticeably for both walkers and crawlers. Even a minimal amount of haptic pressureresulted in ripples appearing on the surface, and this optical motion evidently attracted the attention of both groups of infants. The walkers, however, greatly increasedthe haptic activity and showed increasedlatency to initiate locomotion, though not more than in Experiment 3A. The crawlers did not increasetheir latency to embark; in fact, it was less than in 3A. The black textureless surface was apparently less inhibiting to them when some rippling changedthe surface. Displacementactivity occurred moderately, but was considerablylessthan in Experiment3A. This condition resulted in very large individual differences. The great variability may be attributed to the fact that the surface, unlessprodded, provided the same information as the black velveteen over plywood . Haptic activity produced rippling and optical inhomogeneity, but there was a wide range in the time spent in haptic exploration; the range was from 2 s to 115 s in walkers, and from 0.5 s to 44 s in crawlers. Statistical comparisonof this condition with the black velveteenover plywood condition, becauseof this variability, did not yield very significant differences. A multivariate analysis gave nearly significant differences for surface, F(4, 57) = 2.26, P = .07; and developmentalstatus, F(4, 57) = 2.41, P = .06. The interaction was not significant. Nevertheless, the pattern of differenceslooks different for crawlersand walkers and is in line with expectations basedon earlier resultswith the waterbed. These results suggest that the optical consequencesof haptic exploratory activity may constitute important information for detecting properties of a surfacefor traversability. Overall, crawlerslooked at the surface longer in this condition than in Experiment 3A, when no ripples had appeared, but they were not deterred in locomotion by information about the surface's nonrigidity . Walkers, on the other hand, engagedin increased haptic activity, when the underlying surfacewas a waterbed than when it was rigid, looked longer at the consequencesof their activity, and took longer to embarkon the surface. Was it really the optical consequencesof the haptic activity that were informative for them? We tried to push this question further in the next experiment. Experiment4B In this experiment, the waterbed covered with black velveteen was again compared with the rigid textured surface exactly as in Experiment 4A, except that the waterbed was agitated by an unseenexperimenter as in Experiment 1A. Waves were apparent on the otherwise textureless surface upon its presentationto the infant. Subjectswere new groups of 16

590

E. J. Gibson , et al .

crawlers (M age = 10 .09 months ; range = 7- 16 months ) and 16 walkers (M age = 14 .72 months ; range = 9 1/ 2 - 20 months ). Results

Resultsare presentedin Figure 32.5, Experiment 4B. The most interesting comparisonsare for visual and haptic exploration acrossthe three experiments. When the black surfacewas presentedwith waves, visual exploratory activity rose, comparedwith both the other experiments, particularly in the case of walkers. Haptic exploration, though greater than for the texturelessrigid surface, was much reducedfor walkers comparedwith the waterbed presented without waves. The range of time spent in haptic activity in this condition was 0- 24 s, with 9 walkers spending less than 5 s. Patterns of exploration have changed from the "no waves" to the "waves" condition. For the walkers, preponderanceof visual or haptic exploration has traded. When the optical information for deformation was already good, haptic activity that would produce the information was relatively low; when it was poor, haptic activity that would produceit rose. For the crawlers, the trade-off is not apparent. Looking simply increased when there was something to look at. A multivariate comparison of this condition with the black velveteen over plywood yielded a significant difference

for surfaces , F (4, 57 ) = 6 .85 , P = .0001 .

The differencesascribableto developmentalstatus showed up again in this condition, although the interaction overall did not attain significance. Walkers showed greater increases in all the behaviors coded than did crawlers, as earlier waterbed results led us to expect . An analysis of vari -

ancefor latency differencescoresyielded a significant differencefor walkers, F(I , 30) = 5.340, P = .028. The samecomparisonwas not significant for crawlers, F(I , 30) = 0.435, P = .515. Exploratory behavior increasedin both cases , as expected, but overall it increasedmore for walkers. Fourteen of the crawlers crossed the rigid patterned surface, and 12 crossedthe waterbed. Fifteen of the walkers crossedthe rigid patterned surface, but only 9 crossedthe waterbed. Eight of the 16 walkers in this experiment walked to their mothers on the rigid texture control surface. Two tried to stand on the waterbed but fell. One, after 14 s of haptic exploration, stood up, balancedwith armsoutstretchedand feet wide apart for 90 s, and then carefully walked to his mother (amid applause)- the only one of our many subjects to manage this feat. General

Discussion

As infants becomecapableof independentlocomotion, most actively seek information

about an unfamiliar

terrain

ahead of them that must be trav -

ersedto reacha goal. They examinethe terrain both visually and haptically,

Traversability of Surfaces

591

the amount and type of exploration varying with properties of the surface to be traversed, the information specifying theseproperties, and the child's locomotor development. In line with the concept of affordances , we interpret these results as indicating that infants perceive the traversability of surfacesin relation to the mode of locomotion (crawling or walking) characteristic of their developmental stage and that they do so by actively detecting optical and haptic information that specifiestraversability. Are we, in fact, justified in this interpretation? Four separateexperiments confirmed the differencebetween walkers' and crawlers' responsesto the waterbed; unlike crawlers, walkers tended to hesitate longer before crossing it, to choosea rigid, patterned surfaceas an alternative when a choice was offered, to engagein more exploratory activity when confronted with the waterbed, and to divert their attention away from it to a greater extent. In all four experiments, more walked on the rigid surface than on the waterbed, whereasthere was no distinction in mode of locomotion in the black velveteen standard comparisons. But walkers are, in general, older than crawlers. Is this difference in approaching a nonrigid surface due simply to a few months' differencein age? Walking and age are so tightly correlated that we could not, with the number of subjectsavailable to us, separatethe two groups while holding age constant, though there was occasionaloverlapping. Furthermore, mode of locomotion is a dichotomous variable. But age as suchis not a causalvariable- only biological and cognitive processesand constraints that develop with age could be. The variables that we coded might reflect such constraints. Latency in many tasks decreaseswith age (although in the case of the waterbed, latency increased ). Active exploration might increasewith age, and diversion of attention away from the IIcentral" task also might, under some circumstances. Becauseall of these variablesare graded ones, it was possible to correlatethem with age. U sing the data from Experiment I A, we ran correlations between age and all four of the coded measureson both the rigid and the waterbed surfaces , for crawlers alone, for walkers alone, and for the two groups pooled, making 24 correlations in all. Of these, only 5 were significant, all of them occurring for the waterbed surface. Two of these were for the walkers alone: with latency, r = .36, P < .05; with visual exploration, r = .54, P < .001. Two were for the pooled groups: visual exploration, r = 34, P < .01; displacement, r = .26, P < .01 . One was for crawlers alone, with haptic exploration, r = .31, P < .05. Among 24 correlations, a few low but significant ones might be expected to occur at random. The only impressiveone of these is the correlation for the walkers with visual exploration, which seemsto make good sensein terms of the affordance hypothesis; the older walkers may have learned to engagein more extensive visual examinationof a surfacein motion as they gained more experi-

592

E. J. Gibson, et al.

encein balancingon two legs. The lack of any correlation betweenage and exploratory activity on the other surfacesuggests(a) that development of exploratory activity overall has reached basic competenceby the ages tested in this experiment, as we might expect from earlier studiesof object exploration, and (b) that differencesdepend on other factors, such as the task, physical characteristicsof the subject, and the situation encountered (in this casea nonrigid surfaceencounteredby subjectscharacterizedby two different

modes of locomotion

).

During the period of growth between crawling and walking, human infants are presumablydeveloping increasingcommunicativecompetence. It was possible, in our experiments, that some infants would watch the parent and solicit reactions or IIcues" (Sorce, Emde, Campos, &: Klinnert ,

1985), despite the fact that the parent was instructed to smile and to urge the infant in a positive way for all the surfaces presented. For this reason, a

coding system was devised for what we termed communicativebehavior . There were three major categories: negative behavior , positive behavior , and commenting . Negative behavior included vocalizations (e.g., saying

no), whining or crying while looking at the parent, shaking the head, and stretching out arms toward the parent, as if requesting to be picked up. Only behavior while the infant was on the platform was included (very little occurred during traversal of a surface). Positive behavior included positive vocalizations, suchas playful sounds, giggling, and babbling, waving armsand legs, rocking, and smiling. Commenting was any vocalization or pointing that drew the parent's attention to the surface while observing

the parent' s reaction. Theselatter gestureswere infrequent but unmistakeable. The infants did not appear to be seeking the parent' s advice (the parent was, after all, urging them to come), but calling attention to some feature of the surface .

These behaviors were coded for all subjects and all surfaces in Experi -

ments 1A and 3A. The differencesbetween the rigid textured surfaceand the other two (the textured waterbed and the black velveteen over plywood) were small and insignificant, indicating the relative noninformativeness of this kind of behavior compared with behavior directed toward exploring the surfaceor with displacement. Frequencyof communicative behavior did not differentiate the surfaces , not did the quality of it, that is, whether it was positive or negative in affective tone. It appearedfrom these data that communicative behavior in this situation is more likely attributable

to characteristics

of the individual

infant . Differences

between

crawlersand walkers were apparentonly in a few instancesof words (e.g., saying no) in walkers. They were not related to a particular surface. Could the difference between crawlers and walkers be interpreted in terms of familiarity and novelty? Recentresearchon infants has generally capitalized on a preferencefor attending to a novel display rather than a

Traversability of Surfaces

593

highly familiar alternative or for attending again (dishabituating) to a novel display after repeatedpresenationhasled to habituation of another. Infants' reactions to places also show preferential exploration of new territory (Rheingold & Eckerman , 1969; Ross, Rheingold, & Eckerman , 1972; Ross, 1974). We designed the present researchto capitalize on the predilection to attend to something novel by constructing special walkways and by selecting all the surfacespresented to be ones with which our subjects would be unfamiliar (unlike any in their homes or the usualplacesvisited), so as to elicit attention

to them . The fact that order effects were consis -

tently found, even though not significant in anyone experiment, supports the notion that this happened. But unfamiliarity as such seemsinsufficient to explain the differencebetween crawlers and walkers on the waterbed. Parents were asked whether

their child was familiar

with

a waterbed , and

when this was the case, the infant was run in a different condition. Rigid and nonrigid surfacesare in a sensefamiliar (e.g., a hard piece of furniture as opposed to human flesh or a pillow ), but a new one must be identified by available information- whether optical, haptic, or other- and this process may require exploration and observation of its informative consequences . However

unfamiliar , there were few indications

that the waterbed

elicited fear in either group of infants ; rather , the visual attention indicate

interest

in the wavy

appearance

seemed to

.

It might be argued that cognitive ability increasesbetween 10 and 14 months of age and somehowhasproducedan attitude of greater caution or wariness in the older infants . This suggestion cannot easily be refuted and

can be made more than plausibleby consideringhow cognitive processes might interact with changescausedby maturation of a new action system. That achievementof locomotion plays a pervasiverole in developmenthas been suggestedby other authors (Acredolo, 1978; Campos & Bertenthal, 1984; Gustafson, 1984). But as Gustafson pointed out, it may not be considereda causeapart from the experiencesit produces. Bipedallocomotion, especiallyin its early stage, producesunique experienceshaving to do with maintaining equilibrium (an active process, even in adults). Every stagger or waver produces optical flow patterns that control activities essentialto maintaining balancewhile engaged in locomotion, especially on a "compliant" surface (Lee & Aronson, 1974; Lee & Lishman, 1975; Stoffregen, Schmuckler , & Gibson, in press). These experiencesundoubtedly have two further consequences: (a) to enhance visual attention to the

stability of the surfaceof support and (b) to makethe infant awareof his or her own physical structure and action system. Both consequencescan be seen in our videotapes where a walker, having observed the waterbed in the choice experiment , monitors his or her steps and the floor with great

care in advancing even on the firm side of the double walkway. This behavior, along with manual exploration of a surface, contributes to the

594

E. J. Gibson, et al.

process of learning an affordance not hitherto appreciated, or even relevant for the crawler . Infants learn about the capacities of their own action system in relation to environmental supports . Another way of saying this is that the crawlers may be quite able to discriminate the difference between a rigid and a defonning surface (as other research has indeed suggested) but do not relate the distinction to the action system that involves locomotion . The onset of walking and the constraints imposed by bipedal locomotion lead to discovery of the relevance of rigidity of surfaces for controlling stance and locomotion . In the case of the black velveteen surface on a rigid base, both crawlers and walkers showed longer latencies and greater displacement activity than on the patterned rigid surface. Should this result be interpreted as perception of an affordance7 Might the results be interpreted as a case of misperceiving the surface as specifying depth and thus engaging in avoidant behavior , as in the case of the visual cliff (Gibson & Walk , 1960; Walk & Gibson , 1961)7 There are some distinct differences between the two cases. In the case of the cliff , there is a transparent surface through which the floor , at a depth of (about ) 4 ft (122 cm) below , can be clearly seen. There is ample optical infonnation for depth at an edge, a " falling -off " place. Infants in this situation engage in a large amount of manual exploratory behavior , as we observed in preliminary experiments with a transparent surface on the walkway , although the optical information proved compelling . But the black velveteen surface has no information to specify a falling -off place. There is information for some kind of a surface, because it is continuous and bounded on all four sides, but the absence of texture and inhomogeneities impoverishes optical information for properties of the surface. The result seems best interpretable as wariness in embarking on a surface where optical specification of properties of the surface is weak, whatever the mode of locomotion . The issue of visual dominance over haptic information that comes up in these experiments is a classic and lively one. A seeing adult does appear to depend more heavily on visual information for guidance than on haptic ; the hand is used more often as an "executive organ " than as a perceptual one (Hatwell , 1985). Adults are nevertheless capable of using it for obtaining information , and so are infants by the second half of the first year . The hand may be used in the service of the eye to produce optical consequences. It may be that (when vision is available) haptic exploration waits on visual , requiring inhomogeneity in the optic array to be elicited , and that the haptic activity that ensues is particularly useful if it produces secondary optical consequences, such as visible deformation . But complex patterns of exploration evolve in the course of development , often involving every kind of information available . Consider the patterns of exploratory behavior in a 10-month -old infant handed a bell . The baby looks at the object ,

Traversability of Surfaces

595

reaches for it , brings it to his mouth for haptic exploration , pulls it out and looks at it again, shakes it and listens to the sound, and so forth . This pattern is remarkably well coordinated and maximally infonnative about the properties of the object . It is instructive to compare this highly effective pattern of exploring objects with patterns of exploring a surface. A surface cannot be picked up, put in the mouth , and transferred from hand to hand. But as we have seen, there is, as in object exploration , a kind of canonical pattern , integrated and multimodal . It is not properly described as stimula tion of two receptive systems; it is, rather , an active , coordinated pattern , and it varies depending on the kind and amount of infonnation available in the scene. As in object exploration , the pattern is flexible and is varied to explore potential affordances, tending to optimize the infonnation avail able. Haptic exploration especially varies from delicate brushing to a strong two -handed push. A third point stands out : The exploratory pattern is obviously in the service of action , as the walkers' differentiation of rigid and deformable surfaces tells us, and it results in a kind of perceptual learning . There is no overall role of dominance for haptic or optical information . Both are actively obtained , but circumstances alter the role of the two , both for obtaining the infonnation and in its control over action . Experiments 4A and B showed an exchange in the frequency of haptic and visual exploration ; haptic was more prominent when no " waves" were simply presented to the infant , but visual exploration was more prominent when they were . When the black velveteen surface was rigid (Experiment 3A), exploration of both kinds was at a minimum . There was little to look at, and haptic exploration had no optical consequences. Control of action , however much dependent on results of infonnation getting , is a different problem , and it is closely linked to perceived affordances for executive action , such as going somewhere either on all fours or else on two legs. In the case of a visual cliff with no visible surface at the perceiver' s feet, the affordance for crossing is negative . (This is true even for most adults, however knowledgeable they may be about the presence of Plexiglas .) In the case of an optically impoverished , opaque, rigid surface, hesitation is the usual case, but eventually locomotion ensues. In the case of a defonn ing surface, both optical and haptic information is sought , but action is controlled in relation to motor development and affordance for the type of action . Rather than engaging in speculation over dominance , it seems more profitable to study just how exploratory activity develops and how its consequences are used under varied circumstances for controlling action . The picture this research suggests of the maturing infant confronted with a novel surface is one of a creature prepared (perhaps through evolu tionary history ) to attend to a surface that must be traversed; to explore it actively , both visually and haptically , depending on the information avail able; to observe the consequences of self-initiated exploration ; and to

E. J. Gibson , et al.

596

detect

in

this

suggests

way

that

development

.

these

its

affordances

perception

,

The

capacities

for

exploration

theory

and

of

thus

locomotion

,

in

and

action

affordances

a

are

highlights

provides

a

particular

.

intertwined

the

valuable

mode

closely

interconnection

perspective

It

in

of

on

development

.

Notes

I

.

This

research

and

02

to

Cornell

,

advice

.

.

Other

The

University

Nancy

,

by

National

Health

and

Rader

part

Child

and

the

first

Elizabeth

Science

Human

author

Spelke

,

.

and

Foundation

Grant

Development

The

Grant

authors

James

are

Cutting

,

greatly

all

BNS

-

17207

indebted

of

8209856

lROlHD

whom

to

gave

-

Ulric

invaluable

properties

,

such

as

extension

and

continuity

,

are

investigated

in

further

experi

-

.

number

been

was

in

of

.

ments

3

supported

Institute

Neisser

2

was

National

of

placed

no

on

handhold

walkers

on

the

the

platform

to

pull

rigid

seated

up

on

surface

might

( whether

when

rising

have

or

not

been

they

greater

wished

if

to

stand

infants

had

) ,

because

not

all

there

.

References Acredolo, L. P. (1978). Development of spatial orientation in infancy. Developmental Psychol ogy, 14, 224- 234. Berlyne, D . E. (1958). The influence of the albedo and complexity of stimuli on visual fixation in the human infant. British Journalof Psychology , 56, 315- 318. Campos, J. J., & Bertenthal, B. I. (1984). The importance of self-produced locomotion in infancy. Infant Mental HealthJournal5, 160- 171. Fantz, R. L. (1958). Pattern vision in young infants. Psychological Record8, 43- 47. Gibson, E. J. (1982). The concept of affordancesin development: The renascenceof functionalism. In W . A . Collins (Ed.), The conceptof development(Vol . 15, pp. 55- 81). (The Minnesota Symposium in Child Psychology). Hillsdale, NJ: Erlbaurn. Gibson, E. J. (1984). Perceptualdevelopment from the ecological approach. In M . E. Lamb, A . L. Brown, & B. Rogoff (Eds.), Advancesin developmental psychology(Vol . 3, pp. 243286). Hillsdale, NJ: Erlbaum. Gibson, E. J., Owsley , C. J., & Johnston, J (1978). Perceptionof invariants by nve-month-old infants: Differentiation of two types of motion . Developmental Psychology , 14, 407- 415. Gibson, E. J., Owsley, C. J., Walker, A . S., & Megaw-Nyce, J. S. (1979). Development of the perception of invariants: Substanceand shape. Perception8, 609- 619. Gibson, E. J., & Walk, R. D. (1960). The "visual cliff" ScientificAmerican, 202(4), 64- 71. Gibson, E. J., & Walker, A . S. (1984). Development of knowledge of visual- tactual affordancesof substance. Child Development , 55, 453- 460. Gibson, J. J. (1979). Theecologicalapproachto visual perception . Boston: Houghton-Mifflin . Gibson, J. J., Purdy, J., & Lawrence, L. (1955). A method of controlling stimulation for the study of space perception. The optical tunnel. Journal of ExperimentalPsychology , 50, 1- 14. Gustafson, G. E. (1984). Effectsof the ability to locomote on infants' social and exploratory behaviors: An experimental study. Developmental Psychology20, 397- 405. Hatwell, Y. (1985, July). Motor and cognitivefunction of the hand. Paper presented at the meeting of the International Society for the Study of Behavioral Development, Tours, France.

Traversability of Surfaces

597

Karrnel , B. Z. (1969). The effectof age, complexity, and amountof contouron pattern preferences in humaninfants.Journalof Experimental ChildPsychology , 7, 339- 354. Lee, D. H., & Aronson, E. (1974). Visual proprioceptivecontrol of standingin infants. Perception & Psychophysics , 15, 529- 532. Lee, D. H., & Lishman , J. (1975). Visualproprioceptivecontrolof stance . Journalof Human Movement Studies , 1, 87- 95. Palmer , C. (1985). Infants ' exploration of objects : Relations between perceiving andacting . Unpublisheddissertation , Institute for Child Development , University of Minnesota , Minneapolis , MN. Rheingold , H. L., & Eckerman , C. O. (1969). Theinfant's freeentry into a newenvironment . Journalof Experimental ChildPsychology , 8, 271- 283. Ross, H. S., Rheingold , H. L., & Eckerman , C. O. (1972). Approachand explorationof a novelalternativeby 12-month-old infants.Journalof Experimental ChildPsychology , 13, 85- 93. Ross, H. S. (1974). The influenceof novelty and complexityon exploratorybehaviorin 12-month-old infants.Journalof Experimental ChildPsychology , 17, 436- 451. Ruff, H. (1982). The role of manipulationin infants' responses to invariantpropertiesof objects.Developmental Psychology , 18, 682- 691. Ruff, H. (1984). Infants' manipulativeexplorationof objects: Effectsof age and object characteristics . Development Psychology , 20, 9- 20. Sorce , J . F., Emde , R. N., Campos , J., & Klinnert, M. D. (1985). Maternalemotionalsignaling -: Its effect on the visual cliff behaviorof 1-year-olds. Developmental Psychology , 21, 195- 200. Stoffregen , T A., Schmuckler , M. A., & Gibson,E. J. (in press ). Useof centralandperipheral opticalflow in stanceandlocomotionin youngchildren.Perception . Walk, R. D., & Gibson, E. J. (1961). A comparativeand analyticalstudy of visualdepth perception . Psychological Monographs , 75(No. 15). Warren, W. H. (1984). Perceivingaffordances : Visualguidanceof stair climbing. Journalof Experimental Psychology : HumanPerception andPerformance , 10, 683- 703.

33

Exploratory P ercei

Behavior ving

, Acting

in ,

and

the the

Development Acquiring

of of

Knowledge

Eleanor

] .

Gibson

The last paper that I include has to do with the biggest issue, the origins of knowledge. The terms perceptionand learning were introduced many centuries ago as ways of talking about this issue. How do we know about the world ? How doesthe knowledgewe presumeto have of it get into our heads? Philosophersfirst asked thesequestions (and still do). Developmental psychologistswould seem to have a special obligation to formulate researchablehypotheses , not just broad theories, to addressthe questionsas scientists. There have always beentwo camps among psychologists, those who favor a rationalist approach, believing human creatures to be endowed with some form of knowledge or principles about the world they are going to encounter, and those who favor a more empirical approach. The former group has an obligation to provide a theory of how innately given constraints could comeabout and operate, and the latter has an obligation to show how learning about the world could occur. Most psychologists today would insist that we now know better than to draw a sharp line betweennativists and empiricists becauseboth have something to say. Nevertheless , the differencein emphasisis still very strong, the nativists prompted often by the Cartesian arguments of Chomsky, and the empiricists by fresh evidencethat the developmentof the nervous system is strongly affectedby early encounterswith the world. I favor the empiricist camp in that I want to show how infants can obtain knowledge about the world, use it in interactions with the environment, and thereby enrich their knowledgefurther . But to do this, I have to be a nativist too. I have to accept what I think are facts: that an infant comesequippedwith the systems needed to find out what goes on in the world; that all infants are spontaneouslymotivated to do so; and that ontogeneticdevelopmentis a spiral of maturing systems(perceptual, motor, neural), and give and take with the world. All the ingredients are essential, so the total system we have to deal with is an Excerptsfrom AnnualReviewof Psychology , vol. 39, 1988, 1- 41. Reproduced , with permission, from the AnnualReviewof Psychology , Vol. 39, @ 1988by AnnualReviewsInc.

E. J. Gibson

600 animal

- environment

behind

this

" theory

"

that

an

, of

operation

the

world

animal that

that

are

required

run

their

course

few

the

to

years

I of

acting

think

edge

by

infants

that support

of

formation

will

by

some

Why

is

of

upon elements

. Static

the

light

view

of

retina

, no

sensory

one

the

turn

eyes applied

to

long

Woodworth

often creature

,

active

especially is

a

even

static

about

was

by

in

discovering

. early

These life optimal

. It

that

-

of

-

knowl

of

manual

surfaces

al . 1987

) , or of

).

the is

will

have .

be

fact from

motor

an

optic

means

focus

for

, etc

or

the nature

adjustment

e . g .,

,

) . We

don

't

as

a

turns

,

spectacles

may

looking

. This

as

Dewey

in

his

and experi

system

because But

well head

perceptual

.

the

active

, the

near

a

on is

as

such -

. Such

shutter

, and

far

Gibson

in of

no

array

fell

unrelated

to , interpreted

system

psychologists

exploratory

of

( J . J . Gibson

a

of

essential

stimuli

, perceiving

world

by

a sequence

an

images is

a in -

a perceptual

"

added

in

exploratory

, in

other

to

detail

both

in

aspect

. There

position

also

that

composed to

must

time

, but

termed

" input

had

adjustments ,

information

of

sense

acts

perception

in

head

in

-

and

1990

assume

. Furthermore

functional

developed

touch

random

accommodates

by

develop

( Eppler

et

over

of

with it

information

adjusted

old I

occur

image

image

lens

in

( Gibson

image

system

, the

of

perceiving

properties

.

in

The seek

must

activity

early

is

emergence

the

was

die

visual

the

objects

- gathering

information

we

fixate

, but

of

that

, in

, associated

emphasized . It

on

.

be

, this

hard

as

look

. When

and

a point

dies

and

Review

acquisition

theory

meaningless

experiences

thing

we

not

implicit

a

for

embodied

layout

action

perception

momentary

obtaining

see ,

of

. Functions

about ,

sequence

a knowledge

creating

past

such of

simply

ments

,

perception

a process

be

view

and

of

will

perceptual

being

Cognition

and

, and

,

testing

larger

, and

Annual

and

a

theory

exploratory

activity

about to

of

lies

that

.

of

world

functional

. A

behavior

old

retina

is

factor

exploratory

the

, and

and

aspect

it ? The

the

environment

outcome

the

example

systems

common

for

knowledge

about

essay

article

search

leads

, Action

world

hypotheses : for

kind

provided

the

-

is

, sentient are

species

accept

the

with

the

the

not

size , powers

development

perceptual

incorporates

the

, a motor

part

in

that

some

have

aspect

knowledge

this

the

niche

an

testable

locomotion

Perception

about

related

a to

do on

. Exploratory

acquiring

bipedal -

in

about

speech

to

,

a given

there

I

biomechanical

chose

learning

of

What

Constraints

prepare . I

found

without

of

a in

to

one -

leads

view

sense

as

chose

leads

yet

in

for

point

modem

is

as

, resulting

The

into

of

Implications

be

.

however

thrive

of

to

development .

interchange

I

view

activity

emergence

- made

,

invited

origins are

. This

exploratory

exist

the spiral

ready

evolved

topic

rolled

action

-

similar

and

I was

any the

literally

a

do has

survive

cognition

seeking

Evolution

a developmental

ago on

because ment

form

species

in

Psychology

. of

comes

will

systems

A

system

the

the they

are

young may

be

Developmentof Perceiving , Acting, andAcquiringof Knowledge 601 exploratory even in a skilled observer, becausethey are used to seek information. We live in interaction with a world of happenings, places, and objects. We can know it only through perceptualsystemsequippedto pick up information in an array of energy, such as the optical array. Furthermore, time is required for the adjustmentof the perceptualsystem, for the monitoring of the information being acquired, and for the scanning required by most perceptualsystemsto pick up information (perceiving an object by touching, for example, or locating a sound sourcethrough hearing). Information, accordingly, is picked up over time. Thus if a stable world is to be discovered, there must be temporal invariants of somekind that make constancyof perception possible. I take for granted that perceptual acts extend over time. Perceiving and acting go on in a cycle, each leading to the other. Perceptionoccursover time and is active. Action participatesin perception. Active adjustmentsin the sensory systemsare essential. But action itself may be informative, too. Information about things and events exists in ambient arrays of energy. Actions have consequencesthat turn up new information about the environment. They also provide information about the actor- about where he is, where he is going, what he is doing. All actionshave this property; but it is useful to distinguish executive action from action that is information-gathering. We tend to think of some perceptual systems and the activities that go on within them as primarily information-gathering. The visual and auditory systems, in particular, seem to have little or no executivefunction. (There are exceptions . The eves ~ - -. for example, are used socially in an executive fashion to signal approbation, displeasure , surpriseand so on.) Somesystems, on the other hand, have on the surfacea primary executivefunction, such as the haptic systemsof the mouth and of the hand. The mouth is used for tasting and testing for substantialproperties aswell as for sucking, eating, and speaking. The hand is usedfor examining textures, substantialproperties of objects, shape, and location, as well as for holding, carrying, and lifting . Becauseexecutive functions like lifting can be informative, the distinction between exploratory and executiveactionshassometimesbeenquestioned. But it is a useful distinction for a developmental approach. The possibilities of executive action are minimal in very young infants, but researchin recent years has made it clear that exploratory activities are available and are used in functional ways even in the newborn. Executive actions, such as reaching, grasping, and locomotion, have their own role in perceptual and cognitive development becausethey changethe affordancesof things and places, providing new occasionsfor information-gathering and for acquiring knowledge about what Tolman referredto asthe "causaltexture" of the environment. Cognition, I suggest, restson a foundation of knowledge acquiredasa result of early exploration

602

E. J. Gibson

of events, people, and things. As the baby's perceptual systemsdevelop, exploratory activities are usedto greaterand greateradvantageto discover the affordancesthat are pertinent to each phaseof development. As new action systemsmature, new affordancesopen up and new IIexperimentson the world" can be undertaken, with consequences to be observed. The active obtaining of information that results from the spontaneous actions of the infant is a kind of learning. To say that learning occursonly when actions are repeatedly "reinforced" is to blind ourselvesto the most important kind of learning that underlies our accumulationof knowledge about the world and ourselves. Spontaneousself-initiated actions have consequences , and observation of these is supremely educational. Affor dancesof things generally have to be learned, with the aid of the perceptual systemsand exploratory behavior. Externalreinforcementplays a small role, if any. Intellectualdevelopmentis built on information-gathering, and this is what young creatures(not only human ones) are predestinedto do. They have structures, action patterns, and perceptual systems that are either ready to start doing this at birth or grow into it in a highly adaptive sequenceduring the first year (in human infants). These activities continue as play through the preschoolyears and as deliberatelearning in later life, but the seriousrole they fill is most obvious as they are coming into being. Cognition begins as spontaneousexploratory activity in infancy. Piaget said this long ago. But now researchputs a new face on the story. TheCourseof ExploratoryDevelopment : An Overall Perspective A baby is provided by nature with some very helpful equipment to start its long courseof learning about and interacting with the world. A baby is provided with an urge to use its perceptualsystemsto explore the world; and it is impelled to direct attention outward toward events, objects, and their properties, and the layout of the environment. A baby is also provided with a few ready-to-go exploratory systems, but these changeand develop as sensory processesmature and as new action systemsemerge. There is an order in this development that has interesting implications for cognitive growth . As new actions become possible, new affordancesare brought about; both the information available and the mechanismsfor detecting it increase. Exploratory development during the first year of life occurs as a sequence of phases that build the infant's knowledge of the permanent featuresof the world, of the predictablerelationsbetween events, and of its own capacitiesfor acting on objects and intervening in events. The three phasesthat I am about to suggest are not stages, in a Piagetian sense. They overlap, changeis not "acrossthe board," and absolutetiming varies tremendously from one infant to another. They depend heavily in at least

Development of Perceiving, Acting , and Acquiring of Knowledge

603

one case on growth in anatomical structure. Nevertheless, an order is apparentthat gives direction to developmentand makesclearhow perceptual and action systemscooperatein their developmentto promote cognitive growth .

The first phaseextends from birth through about four months. During this phasethe neonatefocusesattention on events in the immediatevisual surround, within the layout commandedby its limited range of moving gaze. Sensorycapacitiesand exploratory motor abilities are geared to this task, and some serendipitouspossibilities for preliminary learning about features of the grosser layout exist . Visual attention to objects is minimal ,

but discovery of somebasicpropertiesof objectsis madepossibleby visual attention to motion and by the active haptic system of mouthing. Sounds accompanying events are attended to . It is most impressive that these early

exploratory systems, rudimentary as they seem, appearwell coordinated. The second phase, beginning around the fifth month, is a phase of attention to objects. Development of the manualexploratory systemmakes reaching and grasping possible. By the same time visual acuity has increased , and stereoscopicinformation for depth is available. Objects, though presented in a static array, can be explored and their affordancesand distinctive

features learned .

The third phase, beginning around the eighth or ninth month, expands attention to the larger layout, which can only be explored as the baby becomesambulatory. Spontaneous , self-initiated locomotion makes possible discovery of properties of the extendedenvironment around comers, behind obstacles, and behind oneself. Affordances of places for hiding, escaping, and playing are open for investigation. Watching a two-year-old on a playground is a revelation of attention to affordances of things like swings, ladders, bridges, and ropes. After the first year, other phasesmight be identified- e.g., exploring devicesthat have complicatedaffordanceslike mirrors, and tools that must be carried to other objects as well as manipulated. Researchis still scanty in this area. There is also the whole domain of speechdevelopment, in which exploratory activity plays an extensive role during the first year. This domain I reluctantly leave to the experts .

(Here follow sectionson the three phasesof exploration, with presentation of the relevant availableresearch.They are omitted from this excerpt.) Explorationin the Serviceof AcquiringKnowledge The Grounding of Knowledge

In the final accounting, what is the significanceof exploratory activity and its perceptual consequences ? May it not be the essentialingredient for

604

E. J. Gibson

building a foundation of knowledge about the world? Or does intelligence emergeas a separateforce that pulls action- even exploratory activity alonQ '-'. behind it? I have not discussedthe latter idea at all, but the notion that intelligence develops and action somehow follows along has been fairly prevalent during the so-calledcognitive revolution. Beliefsabout and representationsof the world and the self presumablycomefirst, and actions follow after them. This notion is clearly opposedto the points I have been trying to make. Perhapsknowledge eventually becomesa systent of representationsand beliefs about the world (and oneself as an inhabitant of it), but it seemsto me that representationsand beliefsmust be groundedby detection of the surfaces , events, and objects of the layout- the "stuff" of knowledge must somehow be obtained from the world. Furthermore, as living beings we act in the world and necessarilyinteract with the events and furnishings of the layout surrounding us. Our knowledge cannot consist of general abstract properties alone but must relate to the affordancesfor action that the world provides, not only in general beliefs but also in intimate everyday situations whose ever-changing circumstancesdemandgreat flexibility . I have beentrying to show that the young organism, as it grows, has the capability to discoverwhat the world affords and what to do about it. The foundations of the organism's knowledge evolve in an orderly fashion, with something new around the comer in eachphasein a kind of spiralling evolution. What kind of knowledge could result, other than flexible meansof interaction? Predications About the World The knowledge that results from learning affordancesfor action through exploratory activity and observation of its consequencesis, in the beginning, probably entirely utilitarian. Meanings may be confined to situations where interactions are occurring and then can reoccur. It seemsto me that this utilitarian, early, simple knowledge constitutesthe beginning of ability to make predications about the world. For example, objects rest on a ground (but can be lifted from it, if they are the right size and substance ). Ground is always underneath them. Some things are in front of other things. Things can be bumped into. Things can move in the surrounding layout. Somethings makesounds. Someof thesethings are responsive(can eventually be categorized as animate). One can oneself control these responsesby one's own actions(cooing, smiling). Theseare simple examples, but with expanding exploratory and action systems they may become much more elaborateas meansavailable through grasping, manipulation, and later locomotion open up new possibilitiesof learning affordances. Controlled manipulation accompaniedby increasingly mature capabilities of visual observation provides a mechanismfor differentiating affordancesand qualitative properties of things and thus furnishesthe material

Development of Perceiving, Acting, and Acquiring of Knowledge

605

for categorizing, yielding more refined and more general predications. Locomotion with ensuing exploration of places and territories firms up incompleteknowledge and makespossiblepredicationsabout the objectivity and permanenceof the layout and the moveability - of oneself and others . Events, both external and self-perpetrated , present the opportunity

for learning about consequencesof movement, impact, and applying pressures, and thus provide the foundation for discovering causalrelations. I am not suggestingthat predicationsof the kind I have illustrated have been formulated as anything like verbal propositions. Rather, knowledge has been attained that can function as a basis for further categorization and

inference. Learning a vocabulary and a syntax for verbal representationof

predicationsand eventsis an achievementthat presupposes knowledge (something to talk about). It may well have rules of its own, but I doubt that these rules determine

or even select what the infant first attends to and

discoversabout its early environment. I seelittle profit for the scientist in argumentsabout the mental representationof knowledge that cannot be talked about, but I think it must be concededthat such knowledge exists, even in adul~s, and cer~ainly in ~he preverbal child .

Other questions- e.g., how knowledge is organized- are well worth asking and have a good chanceof being answered. An important one has to do with the generalizability of knowledge, sometimesreferred to as IIdomain specificity." After an infant has discoveredan affordancepertaining to one action system, will it transfer appropriately acrossaction systems? Is the affordanceof a substancedetected by mouthing detected as the samewhen the handsbecomeactive in exploring it? Is the differentiation of an aperture and an obstacle by a three-month-old in a looming situation generalizedimmediately to guiding locomotion by a crawler? I doubt that such transfer is automatic in early life, because new action

systemsbring new affordances , and some exploratory practice with them seems essential. But the role of practice would diminish as maturation

winds down. Proliferation of tasks, however, increasesas possibilities of action increase, bringing new opportunities for generalization. So do tasks proliferate with social expectationsof caretakers , and these may engender a new kind of domain specificity as I'training" by society begins. Still more affordancesmust be learnedand the questionof flexibility of generalization over domains can reappear on a new level .

Ontogenesisof PerceptuallyBasedKnowledge

The courseof development of perceptually basedknowledge (knowledge basedon exploratory perceptualsystems) is an orderly one, as I have tried to show. As the phasesof development evolve in the individual, with a focus in eachphase, there is a progressive fanning out. New exploratory systemsdevelop and new action systemsemerge, making new tasks (e.g.,

606

E. J. Gibson

carrying something somewhere) possible. Still, one seesevidenceof earlier phasesforeshadowingthe later ones, as in the caseof the aperture-obstacle distinction. The processdoes not look like disconnectedshoots growing out in different directions, but rather like a spiralling course, an echoing of earlier abilities of affordancedetection plus strengthenedopportunities for discovering new meanings. Perhapsa systemof meaningsbegins its evolution thus. Differentiation is the key processin the kind of developmentI have been describing- differentiation of organs of both perception and action, and differentiation of perceived affordances. But the processis always related to the environment- its resourcesand its constraints. In the caseof the looming experiment in the three-month-old, the information in the optical array, an expanding occluding contour increasing at an acceleratedrate, has the affordanceof imminent collision, calling for such avoidance behavior as the child can muster (head retraction, raising of hands). When a crawler's own locomotion producesan expansionpattern of an object in its path, the information has the affordanceof potential imminent collision, but not in the sameway, becausethe crawler can stop or detour. Furthermore, the information, while similar, is not the same. In the caseof the approaching object, the expansion pattern characterizesonly a part of the total array; but in the caseof the infant's advanceby way of its own locomotion, the expansion encompassesthe total array. The casescall for differentiation, and yet they are closely related; the consequencesof failing to perceive the affordanceare the samebecauseimportant environmental conditions are the same. The system that must be referred to for understandingthe organization of perceivedaffordancesis not the child's own organism alone, despite its manifold relations between perceptual and action systems, but an organism-environment system. Understanding behavioral and cognitive development requires consideration of both as reciprocal entities, a requirement for both the developing child and the psychologist.

Epilogue: Prospectsfor a New Approach to Perceptual Learning

Where has this wandering journey led, if anywhere? One might ask this questionabout the field, of experimentalpsychology as a whole, as I knew it in the thirties and forties. Its face has changedremarkably to the extent that many who would once have been the rising generation of first -rate experimental psychologists are on the verge of casting off the term , leaving

"psychologist" for the clinicians, and rebaptizing themselves "cognitive scientist ." They seek the company of other scholars and see their field as

interdisciplinary, but the favored companionsare now the computer scientists, philosophers, and linguists. The cognitive revolution really turned things around. My own subfieldsof perception and learning are scarcely recognizable. Only a few hardy perception psychologists still study perception of the real world rather than small displays on computer screens. As for learning, only acquisitionof expertiseby adults studying something like math seems to be of concern .

I go along with this trend to only a small extent. I'm happy enough to dissociate myself from the clinicians , and I want companions in other

disciplines. But I want to associatewith the biologists, the people concerned with evolution, the ecologists, and the people studying development. It is the focus on development that deservesthe center of the stage, in my opinion, and with it the realization that we are animalsliving in an environment that we evolved in. We behavein it more or less adaptively to survive

. We occupy

a niche , like other

animals

. Descartes

was wrong

; we

are not a sublime species set apart by reason to rule the world and the

lower orders (although I admit that humanshave done a lot to spoil things for the other species).

Where do I see a path ahead, then? The path I see has been foreshadowedin many ways, as developmentalmilestonesin behavior are, too. That is the way evolution works. The path started with the functional psychology that was so typically American, and flourishedfor awhile in the

608

E. J. Gibson

learning theories of the thirties and forties. Behaviorismwas an important part of the climate for many years, but not Watsonianbehaviorism, because it made too many extravagant claims and rested on too few facts. The behaviorist theoriesof Hull, Tolman, and Skinnerwere the onesthat gained acceptanceand followers. But, except for the hard-core Skinnerians , most of that sturdy movement collapsedwith the cognitive revolution of the sixties, leaving only a few legacies. Not all of the foundation of functionalism collapsed, nor all interest in learning. At the sametime that many psychologistswere being impressed with artificial intelligence and reading Chomsky's review of Skinner's Verbal Behavior , a few other more biologically oriented psychologists were uncovering revolutionary facts about animal learning. Unwelcome as the news was to journal reviewers (Garcia, 1981 ), it turned out that under some circumstancesan animal was IIconditioned" in one trial; that the "conditioned stimulus" could precedethe "reinforcement" by many hours; and that what can be learnedis specificto the species, its habitat and way of life (Garcia, McGowan, and Green 1972). A new generationof learning theorists and experimentershasgrown up, giving rise to an ecologically oriented learning theory (Bolles and Beecher 1988; Johnston and Pietrewicz 1985). These people study what animals learn naturallv - foraging for food, staying - under real-life circumstances clear of predators, developing the means of communication. The new approach is rooted in ecology and evolution, and it does not seek to classify behavior as either innate or learned. It studies learning during development, instead- not in the adult, but in the growing animal as it learnsto survive in its appropriateniche (Johnston1985). This approach to learning has not yet been applied to learning in humans, but it is time to do so. The ecological approach to perception has been bearing fruit in studies of development of perception in infants. I believe renewed study of learning in infants, in ecologically natural and appropriate circumstances , is called for and can be profitably combined with study of early perceptualdevelopment. I now seemy earlier theory of perceptuallearning, publishedin 1969, as flawed by its concentration on learning by an adult or a child who is already a competent speakerof a language. Learning begins long before any languageis available, and before much of any action is. Here is the real role for perceptuallearnmg, and what a lot must go on in just the months before a child can navigate on its own or speaka single word! Experimentswith infants already exist that can be assimilatedto this view, showing that infants learn even before perforrnatory actions are possible, making optimal use of their perceptualsystems. I describea few casesof early learning that seemecologically relevant to a baby's develop-

Epilogue

609

ment in a human environment. The exampleschosenby no meansexhaust the field, but they serve to illustrate its promise. I have selectedfour cases: (1) learning to recognize faces; (2) learning about bimodal specificationof an event; (3) learning about propertiesof an environmentalsupport; and (4) learning about control of an environmental event by one's own action. Others would be ecologically equally relevant, such as phonological perception or turn-taking in communicativeinteraction with a caretaker, but as yet there are no studies with infants that tell us much about the kind of learning that goes on. Actually, very few studies tell us a lot about it, but the ones that follow are suggestive. How early recognition of a human face begins has been a question of interest for a long time, and it inspired a number of studies before very good methods of observing infant perception and learning were achieved. In my 1969 book, I devoted most of a chapter on development of object perception to perception of facesbecausethere was considerableresearch on it (and nothing much else except perception of checkerboards ). I summed it up as follows: Development takes the course of first responding discriminatively with crude compulsory fixation on high contrast edgesor spots in the field. Then follows gradual extraction of distinctive features of the oft-presented face object, differentiated as individual features only. Later comes noticing of invariant relations between featuressuch as the two eyes in a given orientation in the head; later still, the array of featurescharacterizing - a real face as distinct from a dummv's " or from schematicdrawings is distinguished, with no one feature any longer dominant. Eventually, the unique featurepattern characterizinga particular faceis selectivelyrespondedto. I now think that this is all wrong. There have been twenty years of researchon infants' perception of faces since 1969. Most of it has been done with photographs, drawings, or schematicfaces. Take a brief look at it . Following the lead of ethologists, a number of psychologists thought babiesmight have a primitive native attention to a humanfacepattern as a kind of "releaser." In experimentsby Fantz (1963) and others, babieswere shown schematicdrawings of faces to see what they attended to most. Babiespreferred to look at the displays with more information, or higher contrast, or certain spatialfrequencycharacteristicsuntil four months or so, when a preference was found for a "face pattern" (a preference for a properly arranged face vs. a scrambledone). Fagan (1979) summarized many studies by himself and others. Certain features seemedto be more important than others (such as hairline and eyes, contrasted with mouth and nose). At Bye months, speciBcphotographs of subjectsof different sex and ages were discriminated(e.g., a man's face paired with a woman's or

610

E. J. Gibson

either of thesepaired with a baby's). The photographsneededto be upright for this discrimination. Only at seven months did infants recognize a familiarizedfaceand distinguish it from an unfamiliar one when testedwith different poses, suchas a profile or three-quarter view, following a full face. At sevenmonths, infants could also discriminatephotographs on the basis of sex when different photographs were used in the test, providing that several same-sex ones were used in familiarization before the test. Two quite similar facescould be discriminated, provided that severalviews of the sameface were presentedduring familiarization. Faganconcludedthat facial recognition was present at seven months. He thought that this ability was not exclusively for facesbut extended to other abstract patterns. This conclusion is correct, I think, for the material employedtwo-dimensional, static drawings or photographs. But I suspectthat these findings apply only to pictorial perceptionand give us no basisfor a theory of perceptuallearning in infants. It is a notable fact that pictorial perception (that is, getting information aboutsomething from a picture) has been found to mature at about seven months by other psychologistsfor other domains. Yonas (e.g., Yonas et al. 1986) found that information for distance or depth in pictures was not extracted until about seven months; Schmidt and Spelke (1984) have a similar finding for Gestalt laws of organization in patterns. Of those socalled laws, only common fate, involving motion, functions to insure perception of unity in earlier infancy. The others may operate in pictorial perception, but not before sevenmonths or later. Learning to recognize features of faces portrayed in photographs or drawings does not tell us what happensin the kind of perceptuallearning involved in looking at real faces. Quite soon after birth (48 hours), infants prefer to look at the live facesof their own mothers (Bushnellet al. 1989). Furthermore, soon after birth they can imitate a facial gesture of another person, particularly when the mouth is implicated. What arebabiesperceiving then? Perhapswhat they perceiveis an affordance, having to do with a facein a context that is part of an act of communication. Facesare not just visual patterns; they appear as part of an event, one including touch and odors and soundsas well, and the event implies some interaction with another person. At three months, babies are distressedby their mother's immobile, unreactive faces (Tronick et al. 1978). If babies are presented with movies or videotapes of a person behaving in a happy fashion or a distressedfashion, they can identify the appropriate event with a sound track that correspondsto it well before sevenmonths (Walker 1982). Does the baby form an associationbetween modal or other aspectsof an event when it is only a few hours old? I doubt it . All the aspectsare already together in a natural context for the baby to observe, as it usesthe limited exploratory and communicative skills it possesses . I think all the

Epilogue

611

aspectsare simply perceivedas one event to begin with, without analysis into featuresor modally specificexperiences . We started with the wrong end of the stick . Feature analysis and separa-

tion by modes comes later, perhaps with pictorial perception. What is perceivedfirst is the unanalyzedaffordanceof an event in which a caretaker and the baby are actors. This event involves a kind of interpersonalrelation sometimes referred to as I'protocommunication " (Trevarthen 1979). The mother 's voice and facial gestures are prominent in this event and are

clearly responded to by the baby with coos and its own facial gestures. Babiesdistinguish their mothers' voices from others at birth (DeCasperand Fifer 1980 ). This differentiation

may enhance attention

to the facial ges -

tures uniquely synchronized with the vocal information, a unified multimodal event. Recognizingfacesbegins here with multimodal specification of a person having a unique communicativeaffordancefor an infant. Learning about multimodal specificationof an event is my secondcase. What kind of learning, if any, is involved? A number of experimentshave been performed showing that infants between four and sevenmonths may perceive bimodal information specifying an event as belonging together, united in the same event. A method for demonstrating this competence was devisedby Spelke(1976). Infants (aged four months) are shown moving pictures of two events. Both events are filmed with sound tracks

providing the appropriate context. For example, one event in a Spelke experiment was a woman shown playing peek-a-boo, making appropriate exclamations

as she smiled and covered and uncovered

her eyes . The other

was a hand tapping out a rhythm on a little percussion instrument . Babies were shown the films of these two events side by side, but only one

sound track was played, centered between the two films. The babies' eye-movements were monitored for a looking preference. In a number of experimentsbabies have been found to show a preferencefor looking at the event with bimodal specification, perceiving, it would seem, a unified event

.

Careful examination of all the researchof this kind by Marion Eppler (1990) has shown , however , that not all the experiments agree in this

finding. Positive results often favor slightly older babies (five to seven months ); and positive results with younger infants (about four months ) have been found most consistently when a person's face, animated and

conversing, was portrayed. Do infants learn about bimodal specification? If so, do they detect bimodal invariance, abstract information specifying the event? Or are the two modes of information associatedby being experienced contiguously? It is reasonablethat bimodal specificationof persons should come earliest, for the first events that infants are engaged in are social, whereasthey are engagedwith events involving objects somewhat later, increasingly as manual exploration develops enough for objects to -

.

612

E. J. Gibson

afford actions on them. Some kind of perceptuallearning appearsto take place. What kind? Lorraine Bahrick ( 1988) has addressed this question directly in an experiment with three-month -old infants . The infants were given an oppor -

tunity for perceptualleaming by familiarizing them with two visible and audible events, one at a time , and then assessing the conditions of familiari -

zation that actually resulted in learning. Four groups of infants were given two minutes of familiarization train -

ing with the two event films. One film depicted a hand shaking a clear plastic bottle containing one very large marble. The other depicteda hand shaking a similar bottle containing a large number of very small marbles. The four familiarization conditions varied in their pairings of film and

sound track, as to whether the track was synchronouswith the film or not. Only onecondition (one group of infants) was familiarizedwith films paired with their appropriate, synchronized sound track. A fifth (control) group was familiarized with films of irrelevant events . Following familiarization ,

an intermodal preferencetest was given all groups, with the two films presentedside by side, and a single, centeredsound track played. Resultsof this experiment showed that learning did occur as a result of familiarization, resulting in a preferencefor watching the film specifiedby its appropriate sound track, but learningwas confinedto only onegroup, the one given familiarization with the appropriate , synchronous pairing of

sight and sound. Equal opportunity for associationwith an inappropriate sound track did not lead to a preference for that combination in the intermodal

test .

This ingenious experiment tells us that experience with events may be

important for knowledge of bimodal invarianceof specification; somekind of learning does go on, sincethe control groups of three-month-old infants showedno bimodal preference.It also tells us that the kind of learning that happens is not based on sheer association, since groups receiving inappropriate pairing showed no evidenceof learning. What is happeningthen, when this kind of perceptuallearning occurs? It may be differentiation of the two modes of specificationand detection of an abstractinvariant specifying the sameaffordance. Evidently, opportunities to look and listen to varied kinds of events are important for this kind

of learning to occur. That it is specific to any particular event seems unlikely , for Bahrick found that older infants (over six months ) showed

preferences in her experiment that indicated knowledge of intermodal specificationwithout prior familiarization with that particular event. There is other evidencethat differentiation of modal specificationbegins around six months (Walker -Andrews and Gibson 1986). It may be boosted by

onset of manual exploration and the consequentopportunities for self-

Epilogue

613

initiated events implicating objects and providing multi modal information while they are manipulated. What Bahrick's experiment tells us explicitly, I think, is that association of intermodal experiencesis not a mechanismthat results in knowledge of intermodal specificationof an event. The mechanism , if that word is to be used, must be one of differentiation of multimodal information and detection of an abstractinvariant, probably one specifying an affordance, suchas creating a particular kind of noise. I have mademany referencesnow to affordances , that is, the relationship betweena creaturelike a humaninfant and a way of acting that is appropriate to both the environmental opportunities offered and the potential (organismic, dynamic, etc.) of the actor. An affordancemust be perceived for appropriate action to ensue. Affordances changeas an infant develops and new potentials suchas emerging action systemsarise. Here is a prime casefor perceptuallearning. The infant must detect the information specifying both the required environmental supports and its own capabilities and, above all, the relation between them, a kind of abstract relational invariant. My third caseof learning is of this kind, learning about properties of an environmental support neededfor realization of some action, in this case locomotion. Emergenceof locomotor systemsoccursnaturally during development and requires learning about new capacities and their limits as well as detecting what the environment provides for putting them to use. My example is researchon the affordanceof a surfaceof support for the actions of crawling and walking. Babiescapableof self-produced locomotion observe surfacesthat mayor may not afford support and tend to avoid those that look or feel as though they do not. Richard Walk and I (Walk and Gibson 1961) found, for example, that few crawling infants will move out on a transparentsurfaceeven though it is actually firm and solid to the touch. They expect that a surfaceof support should provide good optical information for solidity. We did not at that time investigate how they cameto expect this, but in recent years, my colleaguesand I (Gibson et ale1987) have investigated properties of surfacesperceivedas necessary for walking by very young walkers(seechapter32). In contrast to crawlers, they explored a surfacefor its affordancefor maintainingan upright posture and supporting bipedal locomotion. I think that emergenceof a new action system, such as walking, instigates exploratory activity, resulting in learning about properties of environmental supportsimplicatedin the activity and also about capacitiesof the actor's body, its competenceor "effectivity." Spontaneousexploration has the immediateresult of producing consequences . Observation of these consequences is the key to perceptualleaming in development.

614

E. j. Gibson

learning that involves self-controlled exploration with observationof its consequences has been shown to occur very early in infants . learning about

consequences

of one ' s own

action

leads

to control

of events

in the

surrounding environment. The baby detects information that specifiesan action performed by itself, and something happening in the world contingent with it . Such cases were described 100 years ago by Preyer (1881),

later by Piaget (1952), and in recent times by followers of Skinner, who refer to them as operant conditioning. Severalexampleshave been wellresearchedin recent years. Non-nutritive sucking on an empty nipple, attached to a pressure transducer, so as to trigger some environmental

event when appropriate pressureis applied, is by now a classictechnique for studying phonetic discrimination by young infants (Eimaset al. 1971). Mouthing is an exploratory activity engagedin by infants soon after birth. If a tool (the nipple in this case) is provided with appropriate attachments, babieslearn to operate it to produce environmental contingenciessuch as a tape recorder playing bits of human speech, even single syllables(BA or PA ). The baby discovers the consequenceof its action, its predictability, and its own control over the event. Siquelandand Delucia (1969) found that babies quickly learn to produce another kind of contingent event, pictures displayed by a slide projector. If the pictures were varied, babies continued to act to obtain them for long periods. One of the most interesting of these experiments was done by Kalnins

and Bruner(1973). A movie was projectedbefore the infant subject, and the nipple and transducerarrangedso that sucking would control the focus of the film. If the babiesengagedin high amplitude sucking at a regular rate, the film was kept in focus, otherwise it was blurred. The babiesprovided with this arrangementquickly learnedto useit, aslong asblur was avoided, and stopped when the focussing mechanismwas detached. The babies learned, in short, to usea tool to optimize their exploratory activity . When suckingresulted in blur, the rate dropped almost at once. A string tied to a baby's wrist or ankle has been used frequently to demonstratehow infants learn to act to produce an environmental consequence. Rovee and Rovee ( 1969) tied a ribbon to a baby 's ankle, with the other end attached to a mobile overhanging the crib in which it lay .

Babiesquickly learnedto kick the leg to make the mobile turn. Babiesin a control group were given the samedisplay, but it was not contingent on their own kicks. They fussed, looked away , and cried after a short time ,

while babies allowed to control the moving display cooed and smiled (Rovee-Collier and Capatides1979). Results similar to these have been obtained

(lewis , Sullivan and Brooks -

Gunn 1985) with a string tied to the baby 's wrist and the other end hitched

to a switch that operateda slide projector and a tape recorder that played a song from SesameStreet. Babiesin both an experimentaland a control

Epilogue

615

group were videotaped and their facial expressionscoded afterward in relation to their activity . The babies in the experimental group exhibited intenseinterest as they discoveredthe consequences of their arm pulls, and acceleratedtheir arm pulling. Babiesin the control group fussed. What is learned in these cases ? The babies are observing the consequencesof an action of their own and learning a causalrelationship in an event sequence . They are discovering what a given action in a specific context affords. The emotional accompanimentsof this learning suggest that there is great satisfactionin discovering predictability, and in control of an event by one's own action. Can we dismissthesecasesof learning to control an event as mere casesof operant conditioning? I think not. The essenceof learning is the exploratory activity , instigated entirely by the infant, and used appropriately for searchingfor information. It is not mere repetition followed by reinforcement . An infant 's behavior is modulated to

test out effectsof action or nonaction. For example, cessationof an interesting consequence (by intervention of an experimenter , for instance) does not result in a neat curve of extinction . It is more likely to result in a burst of action , not unlike what an adult driver does when his car suddenly refuses

to start. Bower (1989) presentsa detaileddescription of an infant's exploratory activity as the experimenterintroduces changesin the consequences of the baby 's actions. The aptness of the spontaneous behavior that ensues

for revealing the underlying IIcausaltexture" of the events going on and the baby 's own relation to them is remarkable.

The old mistakewas to start with static displays in formulating a theory of perceptuallearning. Letters, numbers, and even photographs of human facesare not the natural occasionsfor study of human perceptuallearning in its beginnings. The baby necessarilylearnsby perceiving, and what he has to perceiveare complex, multimodal events that often involve his own activity . This activity is limited to start with - looking, mouthing, kicking. Big changesoccur when a baby begins to grasp and manipulate objects, and still more take place with the advent of mobility . When we study learning in relation to this developmentit takeson a new look. A human infant seeks information

about its environment , actively

searchinga surrounding ambient array of energy, often instigated by the advent of a new action system- a naturally developing one, or in someof the examplesabove, one rigged for the infant by an experimenter. Exploratory activity, spontaneousand self-initiated rather than fortuitously supplied, has the effect of testing the reliability of consequences following the activity . Observation of the relationshipbetween one's own action and the consequentevents is a powerful meansfor perceptuallearning. This is not association

, in the old

sense . Events

are associated

in the causal

texture

of

the environment, and a human infant learns by observing the predictable relations .

616

E. J. Gibson

Are we going to find the answer to learning in IIconnectionism "? I doubt it , because it appears to neglect properties of behavior that stand out even in infants - its directedness, its flexibility with changing context , in short its functional character and adaptiveness. In any case, neural underpinnings of perceptual learning in infants can wait their turn , while we turn our attention again to what it is that preverbal infants learn, using the biologists ' methods of good naturalistic observation and experimental methods that are now available to discover what an infant perceives. Whatever name we choose to give it now I psychology still has a domain in its own right . And so has perception - we perceive to learn, as well as learn to perceive .

References

Anderson , G. J., 1989. Perceptionof three-dimensionalstructurefrom optic ~ flow without locally smoothvelocity. Journalof Experimental Psychology : Perception andPerformance , 15:363- 371. Amoult, M. D., 1953. Transferof predifferentiation trainingin simpleand multipleshape discrimination . Journalof Experimental Psychology , 45:40I - 409. Bahrick , L. E., 1988. lntermodallearningin infancy: Learningon the basisof two kindsof invariantrelationsin audibleandvisibleevents. ChildDevelopment , 59: 197- 209. Ball, W. A., andTronick, E., 1971. Infantresponses to impendingcollision: Opticalandreal. Science , 171: 818- 820. Bass , M. J., andHull C. L., 1934.Theirradiationof a tactileconditionedreflexin man.Journal of Comparative Psychology , 17:47- 65. Bishop,C. H., 1964. Transfereffectsof word andletter trainingin reading.Journalof Verbal Learning andVerbalBehavior , 3: 215- 221. Bishop , C. H., 1976. Orthographicstructure , word recognition , andreading.Ph.D. disserta tion, CornellUniversity. Blake , R. R., and Ramsay , G. V., 1951. Perception : An approach to personality . New York: RonaldPress . Bloomfield , L., 1942. Linguisticsandreading.Elementary EnglishReview , 19: 128- 130. Bolles , R. C., and Beecher , M. D., 1988. Evolutionand learning . Hillsdale , NJ: Erlbaum , Associates . Boring, E. G., 1933. ThePhysical Dimensions of Consciousness . New York: Century. Bower, T. G. R., 1989. TheRationalInfant: Learning in Infancy . New York: W. H. Freeman & Co. Bower, T. G. R., Broughton , M. M., andMoore, M. K., 1970. Infantresponses to approach ing objects:An indicatorof responseto distalvariables . Perception andPsychophysics , 9:193- 196. Bushnell , I. W. R., Sai, F., andMullin, ]. T., 1989. Neonatalrecognitionof the mother's face. BritishJournalof Developmental Psychology , 7: 3- 15. Campos , J. J., Langer , A. and Krowitz, A., 1970. Cardiacresponseon the visual cliff in prelocomotorinfants. Science , 170: 196- 197. Changeux , ]. P., Heidmann , T., andPatte, P., 1984. Learningby selection . In TheBiologyof Learning ed. P. MarlerandH. S. Terrace , 115- 133. Berlin: Springer -Verlag. Condry, S. M., McMahon-Rideout, M., and Levy, A. A., 1979. A developmental investigationof selectiveattentionto graphic,phonetic , andsemanticinformationin words. Perception andPsychophysics , 25: 88- 94.

618

References

DeCasper, A . J., and Fifer, W . P., 1980. Of human bonding: Newborns prefer their mothers' voices . Science, 208 : 1174 - 1176 .

Dewsbury, D . A ., 1989. Comparative psychology, ethology, and animal behavior. Annual Review of Psychology, 40 : 581 - 602 .

Eimas, P. D ., Siqueland, E. R., Jusczyk, P. and Vigorito , J., 1971. Speechperception in infants. Science, 171 : 303 - 306 .

Eppler, M ., 1990. Perceptionand action in infancy: Object manipulation skills and detection of auditory-visual correspondence.Ph.D. dissertation, Emory University , Atlanta. Epstein, W ., 1982. Percept-percept couplings. Perception , 11: 75- 83. Fagan, J. F. III, 1979. The origins of facial pattern recognition. In (Eds.), Psychological Developmentfrom Infancy: Image to Intention ed. Bomstein, M . H. and Kessen, W . 83 - 113 . Hillsdale , NJ: Erlbaum Associates .

Fantz, R. L., 1965. Ontogeny of perception. In Behaviorof Non-human Primates , ed. A . M . Schrier , H . F . Harlow

, and F . Stollnitz

, 365 - 403 . New

York : Academic

Press .

Fran<;es, R., 1962. Le developpement perceptif . Paris: PressesUniversitaires de France. Gallistel , C. R., 1989 . Animal

cognition : The representation

of space, time and number .

Annual Reviewof Psychology , 40 : 155- 189. Garcia, J., 1981. Tilting at the paper mills of academe. AmericanPsychologist , 36 : 149- 158. Garcia, J., McGowan, B. K. and Green, K. F., 1972. Biological constraints on conditioning . In M . E. P. Seligmanand J. L. Hager (Eds.), Biologicalboundariesof learning, New York : App leton-Century -Crofts. Gamer, W . R., 1979. Letter discrimination and identification. In Perception and its development , ed . A . D . Pick , 111 - 144 . Hillsdale

, NJ : Erlbaum

Associates

.

Gibson, E. J., 1953. Improvement in perceptual judgments as a function of controlled practice or training. Psychological Bulletin, 50 : 401- 431. Gibson, E. J., 1959. A re-examination of generalization. Psychological Review, 66: 340- 342. Gibson, E. J., 1963. Perceptual development. In Child Psychology , ed. H. W . Stevenson, J. Kagan, and C. Spiker, 144- 195. 62nd Yearbook National Society for Study of Education, Chicago, Ill .: Univ . of Chicago Press. Gibson, E. J., 1969. Principlesof PerceptualLearningand Development . New York : Appleton Century -Crofts .

Gibson, E. J., 1983. Development of knowledge about intermodal unity : Two views. In Pia.v.et and the foundations of knowledge, ed. L. S. Liben , 19 - 41 . Hillsdale , NJ : Erlbaum Associates

.

Gibson , E. J., 1987 . What does infant perception tell us about theories of perception ? Journal

of ExperimentalPsychology : Human Perception and Performance , 13: 515- 523. Gibson, E. J., and Gibson, J. J., 1950. The identifying response: A study of a neglected form of learning. Abstract in AmericanPsychologist , 7: 276. Gibson , E. J., and Guinet , L ., 1971 . The perception of inflections in brief visual presentations

of words. Journalof VerbalLearningand VerbalBehavior , 10: 182- 189. Gibson, E. J., and Levin, H., 1975. ThePsychologyof Reading . Cambridge, MA : MIT Press. Gibson, E. J., and alum , V ., 1960. Experimentalmethods of studying perception in children. In Handbookof Research in Child Development , ed. P. H. Mussen 311- 373. New York : Wiley . Gibson, E. J., Owsley, C. J., Walker, A . S., and Megaw-Nyce, J., 1979. Development of the perception of invariants: Substanceand shape. Perception , 8 : 609- 619. Gibson, E. J., Riccio, G., Schmuckler, M . A ., Stoffregen, T. A ., Rosenberg, D ., and Taormina, J., 1987. Detection of the traversability of surfacesby crawling and walking infants. Journalof ExperimentalPsychology : Perception and Performance , 13: 533- 544. Gibson, E. J., and Schmuckler, M . A 'I 1989. Going somewhere: An ecological and experimental approach to development of mobility . EcologicalPsychology , 1: 3- 25.

References

619

Gibson, E. J., Shurcliff, A ., and Yonas, A ., 1970. Utilization of spelling patterns by deaf and hearing subjects. In BasicStudieson Reading , ed. H. Levin and J. P. Williams, 57- 73. New York: BasicBooks. Gibson, E. J., and Walk, R. D ., 1960. The "visual cliff." ScientificAmerican, 202 : 64- 71. Gibson, E. J., Walk, R. D ., Pick, H. L., and Tighe, T. J., 1958. The effect of prolonged exposure to visual patterns on learning to discriminate similar and different patterns. Journalof Comparativeand PhysiologicalPsychology , 51 : 584- 587. Gibson, E. J., Walk, R. D ., and Tighe, T. J., 1959. Enhancementand deprivation of visual stimulation during rearing as factors in visual discrimination learning. Journalof Comparativeand PhysiologicalPsychology , 52 : 74- 81. Gibson, E. J., and Yonas, A ., 1966. A developmental study of the effects of visual and auditory interferenceon a visual scanning task. Psychonomic Science , 5 : 163- 164. Gibson, J. J., 1929. The reporduction of visually perceived forms. Journal of Experimental Psychology , 12: 1- 39. Gibson, J. J. ed., 1947. Motion picture testing and research. Report No . 7 in Army Air Forces Aviation Psychology Program ResearchReports. U.S. Government Printing Office, Washington, D.C. Gibson, J. J., 1950. Perception of the Visual World. Boston: Houghton-Mifflin . Gibson, J. J., 1957. Optical motions and transformations as stimuli for visual perception. Psychological Review, 64 : 288- 295. Gibson, J. J., 1966. TheSenses Considered as PerceptualSystems . Boston: Houghton-Mifflin . Gibson, J. J., 1979. The EcologicalApproachto Visual Perception . Boston: Houghton-Mifflin . Reprint, Hillsdale, NJ: Erlbaum Associates, 1986. Hamad, S., 1987. CategoricalPerception . New York: Cambridge University Press. Hatfield, G., and Epstein, W ., 1985. The status of the minimum principle in the theoretical analysis of visual perception. Psychological Bulletin, 97 : 155- 186. Hebb, D . O ., 1937. The innate organization of visual activity : I. Perception of figures by rats reared in total darkness.Journalof GeneticPsychology , 51 : 101- 126. Hebb, D . 0 ., 1949. The Organizationof Behavior.New York: Wiley . Hebb, D . 0 ., 1980. In A History of Psychologyin Autobiography, Vol. VII, ed. R. C. Atkinson, J. Freedman, G. Lindzey, 273- 303. SanFrancisco: W . H. Freeman. - and R. F. Thompson, & Hilgard, E. R., and Marquis, D. G., 1940. Conditioningand Learning. New York: AppletonCentury. Horowitz , F. D ., Paden, L., Bhana, K., and Self, P., 1972. An infant-control procedure for studying fixations. DevelopmentPsychology , 7: 90. Hubel, D . H., and Wiesel, T. N ., 1959. Receptive fields of single neurons in the cat's striate cortex. Journalof Physiology , 148: 574- 591. Huey, E. B., 1908. Psychologyand Peda ~oKYof Readin ~. New York: Macmillan. Hull, C. L., 1929. A functional interpretation of the conditioned reflex. Psychological Review, 36 :498- 511. Hull, C. L., 1935. The conflicting psychologies of leaming- a way out. Psychological Review, 42 :491- 516. Hymovitch , B., 1952. The effects of experiential variations on problem solving in the rat. Journalof Comparativeand PhysiologicalPsychology , 45 : 313- 321. Itard, J., 1894. Rapportset memoiressur Ie Sauvagede L'aveyron. Paris. Trans. by G. & M . Humphrey as The Wild Boy of Aveyron. New York: Appleton -Century-Crofts, 1962. James, W ., 1890. ThePrinciplesof Psychology . New York: Henry Holt . Johansson , G., 1950. Configurationsin EventPerception . Uppsala: Alrnkvist and Wiksell. Johnston, T. D., 1985. Conceptual issuesin the ecological study of learning. In Issuesin the ecologicalstudy of learning, ed. Johnston, T. D., and Pietrewicz, A . T., 1- 24. Hillsdale, NJ: Erlbaum Associates.

620

References

Johnston , T. D., and Peitrewicz , A. T., 1985. Issuesin the Ecological Studyof Learning . Hillsdale,NJ: ErlbaumAssociates . Kalnins , I. V., and Bruner , J. S., 1973. The coordinationof visualobservationand instrumentalbehaviorin earlyinfancy. Perception , 2 :307- 314. Kohler, W., 1925. TheMentalityof Apes . New York. N.Y.: HarcourtBrace& Co. Kosslyn,S. M., 1989. Components of high-levelvision:A cognitive neuroscience analysis . TechnicalReport, Air ForceOfficeof ScientificResearch , BollingAir ForceBase , Washington , D.C. Lashley , K. S. and Russell , J. T., 1934. The mechanism of vision. XI. A preliminarytest of innateorganization . Journalof Genetic Psychology , 45: 136- 144. Lavine, L. a ., 1977. Differentiationof letterlikeformsin prereadingchildren. Development Psychology , 13:89- 94. Lewis, M., Sullivan , M. W., and Brooks-Gunn, J., 1985. Emotionalbehaviorduring the learningof a contingencyin earlyinfancy. BritishJournalof Developmental Psychology , 3: 307- 316. Logvinenko , A. D., 1988. Russiantranslationof J. J. Gibson, TheEcological Approach to Visual Perception . Moscow. Loveland , K., 1984. Learningaboutpointsof view: Spatialperspectiveand the acquisition of I/you. Journalof ChildLanguage , 11:535- 556. Marr, D., 1982. Vision : A Computational Investigation into the HumanRepresentation and Processing of VisualInformation . SanFrancisco : W. H. Freeman & Co. Melzack , R., andScott, T. H., 1957. The effectsof earlyexperience on the response to pain. Journalof Comparative andPhysiological Psychology , 50: 155- 161. Miller, N. E. andDollard, J., 1941. SocialLearning andImitation . New Haven:YaleUniversity Press . Mowrer, O. H., 1947. On the dualnatureof learning:A re-interpretationof "conditioning " and"problem-solving." HarvardEducational Review , 17: 102- 148. Newell, K. M" 1986. Constraintson thedevelopment of coordination . In Motordevelopment in children : Aspects of coordination andcontrol , ed. M. G. WadeandH. T. A. Whiting, 341- 360. Dordrecht:MartinusNijhoff. Palmer , C., 1987. The specificityof infantlocomotionthroughaperturesvarying in width. Paperpresentedat the InternationalConferenceon Event Perceptionand Action, Trieste, Italy, August1987. Piaget,J., 1952. TheOriginsof Intelligence in Children . New York: InternationalUniversities Press . Pick, A. D., 1965. Improvementof visualandtactualform discrimination . Journalof Experi mentalPsychology , 69:331- 339. Preyer,W., 1881. Themindof the child. Pt. II. Thedevelopment of the intellect.New York: Appleton& Co. Riesen , A. H., 1947. The developmentof visualperceptionin manandchimpanzee . Science , 106:107- 108. Robinson , E. S., 1932). Association TheoryToday . New York: Appleton-Century. Romanes , G. J., 1889. MentalEvolutionin Man: Originof HumanFaculty . London: Kegan Paul, French . Rosinski , R. R., 1977. Picture-word interference is semantically based . ChildDevelopment , 48: 643- 647. Rosinski , R. R., Golinkoff, R. M ., andKukish, K. S., 1975. Automaticsemanticprocessing in a picture-word interference task. ChildDevelopment , 46: 247- 253. Rovee, C. K., and Rovee, D. T., 1969. Conjugatereinforcement of infant exploratorybehavior. Journalof Experimental ChildPsychology , 8:33- 39.

References

621

Rovee-Collier, C. K., andCapatides , J. B., 1979. Positivebehavioralcontrastin 3-month-old infantson multipleconjugatereinforcement schedules . Journalof theExperimental Analysisof Behavior . Vol. 32, 1: 15- 27. Rumelhart , D. E., andMcClelland , J. L. (Eds.), (1986). ParallelDistributed Processing : Explora tionsin theMicrostructure of Cognition . Vol. 1, Foundations . Cambridge , MA : MIT Press . Schiff, W., 1965. The perceptionof impendingcollision: A study of visually directed avoidantbehavior.Psychological Monographs , 79:604. Schmidt , H., andSpelke , E. S., 1984. Gestaltrelationsandobjectperceptionin infancy.Paper presented at themeetingof theInternationalConference on InfantStudies , New York. Senden , M. Yon, 1932. RaumundGestaltauffassung beiOperierten Blindgeborenen . Kiel, Germany. Trans. by P. HeathasSpace andSight.London: Methuen& Co., 1960. Siqueland , E. R., andDelucia, C. A., 1969. Visualreinforcement of nonnutritivesuckingin humaninfants. Science , 165: 1144- 1146. Societyfor Researchin Child Development(1970). CognitiveDevelopment in Children . Chicago , IL: Universityof Chicago . - Press Spelke , E. 5., 1976. Infants' intermodalperceptionof events. CognitivePsychology , 8: 553560. Sternberg , S., 1967. Two operationsin character recognition : Someevidencefrom reaction timemeasurements . Perception andPsychophysics , 2: 45- 53. Stevens ' Handbook of Experimental Psychology 1988. 2nd ed. eds. R. Atkinson, R. Hermstein , G. Lindzey,andD. Luce. New York: Wiley. Thompson , W. R., andHeron, W., 1954. The effectsof restrictingearlyexperience on the problem-solvingcapacityof dogs. Canadian Journalof Psychology , 8: 17- 31. Tighe, T. J., and Shepp , B. E., 1983. Perception , Cognition , and Development : Interactional Analyses . Hillsdale , NJ: ErlbaurnAssociates . Trevarthen , C., 1979. Communication and cooperationin early infancy: A descriptionof primary intersubjectivity . In BeforeSpeech , ed. M. Bullowa, 321- 341. New York: CambridgeUniversityPress . Tronick, E. Z., Als, H., Adamson , L., Wise, S., and Brazelton , T. B., 1978. The infant's responseto entrapmentbetweencontradictorymessages in face-to-faceinteraction . Journalof theAmerican Academy of ChildPsychiatory , 17: 1- 13. Ullman, S., 1979. TheInterpretation of VisualMotion. Cambridge , MA: MIT Press . Walk, R. D. and Gibson, E. J., 1961. A comparativeand analyticalstudy of visualdepth perception . Psychological Monographs , No. 519, Vol. 75, No. 15. Walk, R. D., Gibson, E. J., Pick, H. L., and Tighe, T. J., 1958. Furtherexperimentson prolongedexposureto visualforms: The effectof singlestimuli and prior reinforce ment. Journalof Comparative andPhysiological Psychology , 51: 483- 487. Walker, A. 5., 1982. Intermodalperceptionof expressivebehaviorsby humaninfants. Journalof Experimental ChildPsychology , 33: 514- 535. Walker, A., Owsley . L. E... 1980. Detection - --------- , C. J., Megaw ~ -Nvce ~ , -I., Gibson,E. .J.,. andBahrick of elasticityasan invariantpropertyof objectsby young infants. Perception , 9: 713718. Walker-Andrews, A. S., and Gibson, E. J., 1986. What developsin bimodal.. perception .. ? In Advances in infancyresearch , Vol. 4 ed. Lipsitt, L. P. and Rovee-Collier, C., 171- 181. Norwood, NJ: Ablex PublishingCo. Wallach , H., andT. O'Connell, D. N., 1953. Thekineticdeptheffect.Journalof Experimental Psychology , 45:205- 217. Wertheimer , M., 1923. Untersuchungen Zur Lehrevan der Gestalt, II. Psychologische Forschung , 4:301- 350. Warren, W. H., Jr., and Whang, S., 1987. Visualguidanceof walking throughapertures . Journalof Experimental Psychology : HumanPerception andPerformance , 13:371- 383.

622

References

Washburn , M. F., 1908. TheAnimalMind. New York: Macmillan . Weiss, P. A., 1970. Whitherlife science . American Scientist , 58: 156- 163. Wickens , D. D., 1939a ). A studyof voluntaryandinvoluntaryfingerconditioning . Journalof Experimental Psychology , 25: 127- 140. Wickens , D. D., 1939b.Thesimultaneous transferof conditionedexcitationandconditioned inhibition. Journalof Experimental Psychology , 25: 332- 338. Woodworth, R. 5., 1938. Experimental Psychology . New York: Holt. Yonas,A. W., andGibson, E. J., 1967. A developmental studyof feature -processing strategies in letter discrimination . Paperpresentedat EasternPsychologicalAssociation , Boston,MA. Yonas, A., Granrud , C. E., Arterberry, M. E., and Hanson , B. L., 1986. Infants' distance perceptionfrom linearperspective and texturegradients . InfantBehavior andDevelop ment , 9: 247- 256. Zinchenko , V. P., 1966. Perceptionasaction. In Proceedings of the28thInternational Congress of Psychology 64- 73, Symposium 30. Moscow.

Author

Index

Aarons , L ., 331

Beecher , M . D ., 97 , 608

Abramovitch

Beilin , H ., 482

, R ., 543

Abrams , M ., 431 - 432 Acredolo Albert

, L . P ., 593

, R . S., 337

Bellugi, V ., 521, 525 Bergman, R., 202, 223 Berkeley , G ., 356

Alegria, ]., 553

Berko , J., 446

Alekin

Berlyne , D . E., 172 , 361 , 584

, R . 0 ., 335

Allen , T . W ., 553

Bernstein

Ames , A ., 260 , 294

Bertenthal

Ames , E. W ., 374

Bevan , W ., 324

Amidon

Biandra , D ., 431

, A ., 448

, L ., 343 , B . I ., 593

Anderson , G . J., 264

Biederman

Anderson

Biemiller

Anisfeld

, N . S., 462 , M ., 460

Anrep, G. V., 33, 37 Archer , E. ] ., 337 Arnett

, ] ., 342

Amoult

, M . D ., 41

, G . B ., 483 , A ., 460

Binder , A ., 325

Bishop, C., 391, 394, 402, 404, 409, 437, 444 , 462 , 485 Bitterman

, M . E ., 132

Blake , R . R ., 292

Aronson

, E., 593

Blancheteau

Atkinson

, R . C ., 339 , 393

Blankenship, A . B., 57

, F ., 324

Bloom , L ., 455

Attneave Axelrod

, S., 330

Bloomfield

, M ., 341

, L ., 394 , 404

Ayres , ] . J., 344

Bolles , R . C ., 97 , 608

Baddeley, A . D ., 462- 463

Borton , R . W ., 553

Boring, E. G., 4, 6, 26, 283, 311, 314 Bahrick , L . E., 542 , 553 , 612 - 613

Bourdon

Baker , C . H ., 342

Bowden , J. H ., 444

Baker , R . A ., 128

Bower , T . G . R ., 189 , 366 , 385 , 500 , 529 ,

Baldwin

, A ., 173 - 174 , 389

, B ., 233 , 267

564 , 615

Ball , W . A ., 500 , 564

Brault , H ., 326

Barlow , H . B ., 329

Bridger, W . H., 553

Barron

Brill , S. L ., 475

, R . W ., 465 , 479 , 485

Bartlett , F. C ., 294

Britt , S. H ., 57

Bass , M . J., 26 , 28

Broadbent

Bausano , M ., 561

Brogden, W . J., 109

Beatty , F. S., 342

Brooks , V ., 330

, D . E., 328 , 339 , 375 , 406

624

Author Index

Brooks

- Gunn

,

Broughton

,

Brown

,

C

Brown

,

L .

.

Brown

,

R

R

.,

D

Bruce

,

R

Bruner

. .

,

369

J .

,

-

,

Bunch

,

, ,

.

,

Darwin

525

326

Day , R. H ., 529 , 554

467

DeCasper, A . J., 611 Delucia

,

538

292 ,

294

,

562 ,

53

294 ,

311

,

324

-

325

,

356

,

, C . A ., 614

Dember , W . N ., 324

DeMontpelier, G., 324

614 ,

313

,

356

,

DeRivera , J., 336

456

,

DeShazo , D ., 328

62

Dewey , J., 3, 600 Dewsbury , D . A ., 197

610

L .,

362

E .,

M

, L . E., 342 , C ., 4 , 177 - 178

Dennis , W ., 202

I .,

P .

Bryden

,

174

. E .,

W

Bryant

,

444

,

Bryan

564

53

,

M

Bushnell

175

. J .,

E .,

L .,

,

Dale , H . C . A ., 463

463

.,

370 ,

Brush

A

.,

S .,

Brunswik

500

Dameron

173

W

.,

340 338

R

Curti, M., 6

614

M

.,

, ,

.

.,

T

Brownstein Bruce

J .,

M

.

Dodge, R., 389, 405 Doerhring, D . G., 480

538

P .,

344

Dollard , J., 71, 334 , 357 Caldwell

,

Campos

W

,

Cantril

,

I .

H

Carr

,

W H

,

Cattell

105

,

Dominowski 561

,

592

I .

Drever , J. D ., 322 , 324 - 325 , 327 , 343 Duncker

557

,

McK

.,

, C ., 169

Dyk, R., 168

565 362

,

389

,

409

,

423

,

439

,

Eckerman , C . 0 ., 593 ,

I . A

Chananie

,

I .

Changeux

,

Chase

,

W

Chomsky

,

N

Chorover

,

S .

Chistovitch

,

Clark

,

M

S .

Clement Cofer

,

L .

.

N

.,

.,

466

Cohen

,

L .

B .,

Cohen

,

M

.,

,

Conrad

,

R

Cooper

,

F .

,

C ,

,

H

,

Ettlinger, E., 332 Ettlinger, G., 333

.,

R

.

A

472

Evans , J. E ., 36

L ., A

.,

Estes , W . K ., 120

443

.,

, E .

334

Fagan, J. F., III , 609 - 610

543

Fantz , R . L ., 331 , 584 , 609

529

.

Crowder

Eppler, M ., 600, 611 Epstein, W ., 265, 290, 328

Essman , W . B ., 341 -

.,

M

,

362

Ervin , S., 174

462

.,

Ellis , H . C ., 337

Escalona , S., 171

5 .,

. G

Cronbach

-

533

M

.,

608

Engen , T ., 344

467

333

S .

,

Emde , R . N ., 592

333

Condry

Crabbe

599

333

361

G

Corter

L .,

, P . D ., 614

Elam , C . B ., 132

335

,

. E .,

D

.,

.,

467

,

M

A

486

Cohen

,

Eimas

Ekstrand , B . R ., 482

.,

, C

Egeth, H. E., 423, 431

289

289

D

E .,

Edwards , B . J., 330

332

476

.,

.

,

431

P ., .,

,

184

.,

I . G

Christy

.,

D

.

Culler

, R . L ., 482

, D . C ., 431

Dreis , T . A ., 16

614

356

Caviness

Cross

Donderi

593

Downing , J., 393

B .,

444

Cole

-

268

I .,

,

I .,

I .

., .,

Carroll

335

294

, ,

E .,

.,

Capatides Carel

.

Farber , J., 448

393

Fariello , G . R ., 524

188 G

.,

109

Faterson , H ., 168

462 ,

120

Feldman , S. E ., 325

Author

Felzen , E., 460 Fieandt , K., 529 Field , J., 562 Fifer,W. P., 611 Fisch , R. 0 ., 564

Index

Gregory, R. L., 330, 332, 503- 504 Grice , G . R ., 128 Guinet , L ., 446 , 454 , 459 , 461 , 467

Gullicksen , H ., 48 Gunter , R ., 189

Fleishman , E. A., 344 Flock, H., 264 Fodor, J. A., 475 Fowler, W., 332 Fraiberg , 5., 521- 522 Fraisse , P., 325, 341 Frances , R., 321, 323, 328 Frankel , G. W., 486- 487 Freedman , s. J., 322 Freeman , R. B., 189 Fries,C. C., 484 Frishberg , 7, 525 Fruchter , B., 344

Harris , F ., 169

Gagne, R., 201- 202, 291

Hatfield

, G ., 290

Galanter

, E. H ., 324 - 325

Hatwell

, Y ., 581 , 594

Gallistel

, C . R ., 197 - 198

Hauty , G . T ., 342 - 343

Gustafson Guthrie

, G . E., 593

, E. R ., 109 , 120

Haber , R . N ., 462 Halle , M ., 159 , 321 , 326 , 358 Hammond

, M ., 407 , 445 , 458

Hanes , R . M ., 344 Hansen , D ., 484 Harcum , E. R ., 344 Harlow

, H . F ., 138 , 172 , 380

Hamad , S., 321 Harris , C . S., 483 - 484 Harter , N ., 362

Ganz , L ., 331

Hawkins

Garcia , J., 97, 608 Garrett , M . F ., 475

Hay , ] . C., 531 Hay , ] . M ., 265 , 341 , 353 Hayes , W . N ., 185, 338

Garvin , E. A ., 337 - 338

Hebb , D ., 103 , 125 - 127 , 131 , 149 , 294 ,

Gamer , W . R ., 328 , 355 , 361 - 362 , 421

Gelber , E. R ., 533

, W . F ., 340

310 - 311 , 322

Gesell , A ., 4

- 324 , 343 , 357

Hecht , K . E., 267

Ghent , L ., 343

Heidemann

Gibson , J. J., 6, 15, 36 , 71, 157, 201 , 203 ,

Hein , A . V ., 288 , 343 , 353

, T ., 289

237 - 238 , 245 , 247 - 248 , 264 - 265 ,

Held , R ., 288 , 322 , 333 - 334 , 343 , 353

268 - 270 , 283 , 287 - 292 , 305 - 315 , 324 ,

Heller , D . P ., 189 , 190

332 , 334 - 336 , 353 , 355 - 356 , 366 , 373 ,

Helmholtz

375 , 377 - 378 , 397 , 437 , 439 , 488 , 494 , 500 , 513 , 524 , 529 , 531 , 543 , 554 , 557 559 , 563 , 565 , 573 , 584 , 600

-

, H . Yon , 263 , 266 , 269 , 280 ,

293 - 294 , 356 , 472 , 513

Henle , M ., 338 Hermelin

, B ., 332

Gippenreiter, J. B., 353 Gilinsky, A . S., 238

Heron , W ., 125

Glickman

Hess , E. H ., 172

, S. E., 188

Goldiamond

, I ., 340

Goldstein

, M . J., 340

Golinkoff

, R . M ., 390 , 480

Gollin , E. 5 ., 340 , 380

Herter , K ., 189 Hicks , L . H ., 192

Hilgard, E. R., 9, 294 Hill , F . A ., 467 Hoates , D . L ., 361

Goodenough, D ., 168- 169

Hobhouse

Gottfried

Hochberg, J., 268, 329- 330, 353

, A . W ., 543 , 553

Graf , S. A ., 463

Hockett

Graham , C . H ., 267 , 269 , 308 - 310

Hoffman

Green , K. F., 97, 608 Greene , J. E., 342 Gregg , L. W., 339

Hofsten Holton

, L . T ., 4 , 197

, C ., 164 , 175 , J., 480 , C . Yon , 562 - 563 , G ., 488

Honkavaara

, S., 379

625

626

Author Index

Horn , R ., 482 Horowitz

, F . D ., 527 , 532

House , B . J., 339 , 380 Hubel , D . H ., 125 , 355 Hudson , W ., 330

Huey , E. B., 390 , 499 Hull , C ., 4 - 5 , 7, 25 , 28 , 41 - 42 , 79 , 97 ,

Kolers, P . A ., 406 Koslowski , B., 562 Kosslyn,S. M., 321, 524 Kossov,B. B., 327, 339 Krowitz, A., 105 Krueger , L. E., 464- 465 Kukish, K. S., 390

178 , 289 , 354 , 608

Humbert

, C ., 340

LaBerge , D ., 483

Hunt , J. McV ., 324

Lambercier

Hunton

Lane , H . L ., 335

, V . D ., 192

Hyde, T. S., 463 Hyrnovitch , B., 125

, M ., 372

Langdon, J., 250 Langer , A ., 105

Lashley, K., 4, 125, 127, 142, 179 Inhelder , B ., 524

Lavine , L . 0 ., 193 , 437 , 440 , 477 - 478

Irwin , J. M ., 91

Lawrence

, D . H ., 128

Itard , J., 125

Lawrence

, L ., 184

Ittleson

Lee , D . H ., 593

, W . H ., 328

Leeper , R. A ., 294 Jacobsen, C., 101 Jakobson , R., 159 , 162 , 175, 321 , 326 , 358 ,

Leibnitz

, 7, 197

Lemmon

, V . W ., 423

Lenneberg, E., 338, 521- 522

478

James, W ., 3, 47 , 202 , 334 , 357 , 374 , 507

Leontiev

Jeffrey, W . E., 172, 174, 444

Lepley, W . M ., 25, 62

, A . N ., 353 , 357

Jenkins, J. J., 463 Johansson G ., 247 Johnson , S. C ., 193, 432 , 442 Johnson , N . F., 471 , 481 Johnston , J., 542 , 563 , 574 Johnston , T . D ., 97 , 608

Levin , H ., 172 , 174 , 389 , 391 , 404 , 461 , 478 - 480

Levy , A . A ., 472 Lewin

, K ., 296

Lewin

, R ., 558

Lewis

, M ., 614

Liberman

, A . M ., 149 , 335 , 443

Kagan , J., 172 - 173

Liddell , H ., 102 , 119 - 120 , 358

Kalnins , I . V ., 614

Lindhagen, K., 562

Kanfer , F . H ., 340

Karmel , B . Z ., 584

Ling , B. C ., 192 Lipman , E. A ., 109 Lippsitt , L. P., 380 Lishman , J., 593

Kam

Lissina

Kaplan, G., 413, 464 Karman

, M ., 481

, H . W ., 339

, M . I ., 353

Karp , S., 168 - 169

Lloyd, V . V ., 267

Kellogg, W . N ., 119- 120

Locke , J., 293

Kessen

Logvinenko, A . D ., 353

, W ., 174

Kilpatrick, F. P., 328

Lamay

Kimura , D ., 344

London

Klass , Yu . A ., 335

Loveland

Klinnert

Lowenstein

, M . 0 ., 592

, B . F ., 332

, I . D ., 329 , K ., 521 , A . M ., 449

Kluver , H ., 189

Kobrick , J. L., 231 , 233

McBee , G ., 138

Koch , S., 324 , 328

McClelland

Koffka , K ., 6 , 47 , 274 , 296 , 355

Maccoby, E. E., 375- 376

Kohler , l ., 324 , 333 - 334

McCraven

Kohler , W ., 4 , 46 , 149 , 189

McGee , R . K ., 331 - 332

, J. L ., 97 , 483 , V . G ., 62

Author Index McGeoch

, J., 5 , 46 , 59 - 60 , 62 , 83 , 90

McGowan

, B . K ., 97 , 608

Nissen

,

N

Noirot

,

E .,

.

W

.,

627

101

553

McKinney , F., 62 , 90 McMahon

- Rideout

, M ., 472

O ' Connell

, D . N ., 247

McNamara , H . J., 342 McQueen -Irwin , J., 59

O ' Connor

, N ., 332

McVay , S., 181 Madigan, S. A ., 463 Managan, G. L., 342 Marquis, D. G., 9

O 'Neill , B. J., 463

Marr

, D ., 263 , 283

Marx

, M . H ., 326

Maturana

, H . R ., 355

alum

, P ., 147 , 265 , 268 , 323

Dry , N . E., 463 Osser , H ., 345 , 391 , 394 , 397 , 399 , 407 - 408 , 440 , 445 , 458 , 461 , 480

Owsley, C. J., 542, 563, 574 Paden , L ., 532

Mayzner , M . S., 482 - 483

Paivio , A ., 463

Meier , G . W ., 331 - 332

Palmer , C ., 558 , 575

Megaw-Nyce, J., 542, 574 Melloy -Carminar, P., 553

Pastore , N ., 149 , 189 , 329 , 331

Melton

Pavlov , I ., 33 , 37 , 44 , 50 , 360

, A ., 5 , 59 , 91 , 201

Meltzoff

, A . N ., 553

Melzack

, R ., 125

Patte , P ., 289 Pearson , R . G ., 342 - 343 Petterson

, L ., 564

Meyer , D . R., 180 Meyer , P. M ., 188

Phaup, M . R., 335

Michotte

Piaget , J., 147, 162, 190, 288 , 323 - 324 ,

, A ., 529

Mill , J., 311 Mill , J. 5 ., 311 Miller

, G . A ., 174 , 362 , 445

Miller

, M . B ., 361

Miller

, N . E., 71 , 334 , 357

Minsky , M ., 325, 327 Moon

, L . E ., 380

Mooney , C. M ., 343

Pfafflin , S. M ., 336 - 337

343 , 356 , 369 , 372 , 374 , 500 , 524 , 529 , 538 , 614

Pick , A . D ., 391 , 394 , 397 , 401 , 407 - 408 , 445 , 458 , 461 , 480 , 486 - 487

Pick , H . L ., 126 , 135 , 174 , 287 , 353 , 357 , 524

Pietrewicz

, A . T ., 97 , 608

Pilzecker , A ., 46

Moore

, D . J., 335

Pittenger, J. B., 485

Moore

, M . K ., 500 , 564

Postman , L ., 292 , 294 , 316 - 320 , 356

Morton

, J., 462 , 485

Moscovici Mowrer Muller

, 5 ., 340 , O . H ., 107 , 109 , 120

Preyer , W ., 523 , 614

Purdy, J., 202, 584

, G . E., 45

Murdock

, B . B ., 462 - 463

Murphy , G., 323, 327- 328, 339, 341 Murphy , L., 171 Murphy , W . W ., 326 Murrell

Prentice , W . C . H ., 322

, G . A ., 485

Rader , M ., 561 Raffel , G . A ., 92

Ramsay , G . V ., 292 Rasmussen , E. A ., 337

Ratliff , J., 340 Razran , G . H . S., 48

Nealey, S. M ., 330

Reicher , G . M ., 465

Neisser , U ., 325 , 414 , 418 , 464

Reisen , A . H ., 103 , 120 , 322

Newbigging , P. L., 340- 341

Restorff , H . Yon , 47 , 55

Newell

Rheingold, H. L., 593

Newman

, K . M ., 571 , E., 406

Nicholson , M., 6 Nickerson , R. 5., 43I Nishisato , 5., 43I

Rhodes , M . V ., 344 Richards , J. E., 561 Riesen

, A . H ., 331

Riess , B . F ., 372

628

Author Index

Robinson

, E. 5 ., 5 , 15 - 17 , 47

Silfen , C . K ., 374

Robinson

, J. S., 338

Simon , H . A ., 476

Rock , I ., 328

Sinclair , H ., 521 , 525

Rogers, T. 5., 485

Siqueland , E. R., 614

Rolfe , S. A ., 554

Sisson , E. D ., 66

Romanes , G . J., 4 , 177 - 178 , 197

Skinner , B . F., 503 , 512 , 608 , 614

Rose , S. A ., 543 , 553

Smith , E . E., 484

Rosenblatt

, F ., 265 , 325

Smith , H ., 481

Rosenblum

, L . A ., 188

Smith , J., 205 , 207 , 399 , 409 , 440 , 462 , 485

Rosinski , R . R ., 390 - 391 , 461 , 480

Smith , O . W ., 268 , 353

Ross , H . 5 ., 593

Smith , P . C ., 268 , 353

Routtenberg, A ., 188

Smith , W . A ., 249

Rovee , D . T ., 614

Smock , C . D ., 340

Rovee - Collier , C . K ., 614

Solley, C. M ., 323, 327- 328, 339,

Rubenstein

, H ., 340

341 - 342

Rudel , R ., 332

Solomon

Rudin , S. A ., 170

Sorce, J. F., 592

Ruff , H ., 575

Sorenson , R . T ., 430

Rumelhart

Soroka , S. M ., 543

, D . E., 97

, R ., 291

Saiff , E. I ., 185

Spalding, D . A ., 560 Spelke, E. S., 381, 477, 610- 611 Spence, K. W ., 131, 132, 310 Spitz, H. H., 361 Spoehr, K. T., 484

Salten , C ., 380

Stachnik , T ., 482

Samuels, S. J., 444 Santos, J. F., 327

Stagner, R., 170 Steger, J. A ., 48I Sternberg, S., 354

Saravo , A ., 380

Stevenson

Scaggs, E. B., 16 Schapiro, F., 193, 442, 455, 464

Stoffregen, T. A ., 593 Stroop, J. R., 460 Studdert-Kennedy, M ., 443

Rush , G . P ., 379 Russell , B ., 519

Russell, J. T ., 125, 127, 142

Sanderson , W . A ., 338

Schiff , W ., 184 - 186 , 340 , 399 , 409 , 440 , 462 , 485 , 500 , 557 , 564

, H . W ., 138

Sullivan , M . W ., 614

Schlosberg, H., 109, 119- 120 Tallarico

, R . B ., 331

Schmidt , H ., 381 , 610

Teichner

, W . K ., 231 , 233

Schmuckler

Tenney , Y . J., 465 - 466

Schmeidler

, G . R ., 6 , M . A ., 500 , 572 , 593

Schneider , G . E ., 184

Terrace , H . S., 344

Scott , T . H ., 125

Teuber , H . L ., 322

Self , P ., 532

Thines , G ., 529

Sekuler , R . W ., 431 - 432

Thompson, W . R., 125

Selfridge, O . G., 325

Thorndike

Senden , M . Yon , 125 , 329

Tighe, T. J., 126, 135, 141, 287, 330- 331,

Serunian , S. A ., 380 Shankweiler

, D . P ., 443

Shantz , C . U ., 524

Shapiro, F., 421

, E. L ., 4 , 46 , 83 , 97

365

Tikofsky, R. S., 431 Titchener

, E. B ., 3 , 294 , 305 , 307 , 309 , 334 ,

356 , 558

Sheffield , F. D ., 109

Tither , N ., 444

Shepela, S., 376, 448 Shepp, B. E., 365

Tolman

Shurcliff

Torgerson, W ., 359, 442

, A ., 413 , 445 , 458 , 464 , 483

, E. C ., 5 , 17 , 294 , 313 - 314 , 601 ,

608

Author Index Tresselt , M . E., 482 - 483 Trevarthen Tronick

, C ., 183 - 184 , 611

, E. Z ., 500 , 564 , 610

T scherrnak-Seysenegg, A ., 267 T ulving , E., 463

629

Winston , 7, S3 Wise,S. 5., 431 Witkin, H. A., 168 Wittlinger , R. P., 467 Wohlwill

] . F ., 323 , 369 - 370 , 393

Turner , E. A ., 461

Wolf , I . S., 119 - 120

Tyler, R., 393

Woodworth

, R . S., 229 , 294 , 389 - 390 ,

423 , 480 , 600 Uhr , L ., 326

Uleman , J., 326

Yarczower

Ullman , S., 264

Yerkes , R ., 4 , 7

, M ., 338

Yonas , A ., 193 , 373 , 377 , 391 , 413 , 442 ,

Vanderplas, J. M ., 337- 338 Vanderplas, J. N ., 338

Yonas , P . M ., 367 , 437 , 439 , 445

Vellutino

Yum , K . S., 48 , 51 , 74 , 78

, F . R ., 481

Veneger, L. A ., 353 V enezky, R. L., 457 Vernon

, M . D ., 272 , 324 - 325 , 356

Vigotsky , L. S., 353 Vurpillot , E., 326, 353, 377- 378, 479 Vossler , C ., 326 Wallace , J. G ., 330 Wallach

, M . A ., 247 , 249 , 257 , 259 ,

324 - 325

Walk , R . D ., 103 , 126 , 135 - 136 , 141 , 148 , 150 , 156 , 186 - 188 , 287 , 330 - 331 , 339 , 353 , 557 , 560 , 594 , 613

Walker , A . S., 528 , 542 , 563 , 574 , 610 Walker -Andrews Warren

, A . S., 541 , 542 , 612

, J. M ., 138

Warren , W . H ., Jr., 558 , 573 Washburn

, M ., 197

Waters , R . H ., 47 Watson

, J. B ., 10

Watson , J. S., 524 Weber , R . M ., 460

Wehrkamp, R. F., 231, 233 Weinstein

, B ., 138

Weiss , P ., 290 Weisstein

, N ., 483 - 484

Wertheimer

, M ., 247 , 249 - 250 , 273 ,

379 - 381

Whang, S., 558 Wheeler

, D . D ., 465

Wheeler

, K . E., 445 , 461 , 480

White , B ., 185

Whitely , P. L., 57 Wickens

, D . D ., 9 , 463 , 467

Wiesel , T . N ., 125 , 355 Williams

, J., 431

455 , 458 , 464 , 479 , 483 , 564 , 610

Zaporozhets , A. V., 324, 353 Zaslow, M., 465 Zazzo, R., 525 Zeaman , D., 339- 340 Zinchenko , V. P., 332, 353- 354, 374- 375

Subject Index

Abstraction , 182 , 289,318 , 456 , 475 education of, 289 , 339 Accretion anddeletion atanedge , 264 mechanisms of, 289 , 312 , 456 Acquired distinctiveness , 71, 149 , 321 , optimization of, 366 , 373-377 , 485,488 334-335 , 357 asselection , 360-361 Action . See alsoPerception andaction voluntary andinvoluntary , 374 andaffordances , 573 , 594 Avoidance ofshock , 9- 14, 108 - 123 andbelief , 604 executive vs. information gathering , 601 Behaviorism , 3, 293 , 503 , 519 , 608 feedback from , 343 Blindness , 329 , 521-522 Activesearch inperception , 494 , 503-510 , 599-608 Category (categorizing ) Activity , roleofinperception , 360-361 , andinvariance , 538 366 theories ofperceptualleaming , 325 Affordance (s), 9, 105 , 108 , 385,471 Causal texture ofenvironment , 313,456 , indevelopment , 557-569 614-615 andemotion , 561 Cellassembly , 306 offaces , 610-611 Cluster analysis ofgraphic pairs , 432ofgrasp ability , 562-563 434 learning of, 560 Cognition andmaturation ofaction systems , 601animal , 197 602 , 613 andconcepts , 514 ofreading , 390 andperception , 493 oftraversability , 571- 596 andperception oforder , 380 Ambient opticarray , 269 , 283 , 503,506 , Cognitive 527 , 528 development , 600 Aperture psychology , 97 affordance of, 557-558 , 563-568 revolution , 97, 608 information for, 565 science , 471 , 607 Artificial intelligence , 41, 321,608 , 513strategies , 517 514 , 519 Common fate , Wertheimer 'slawof, 247, Association , 41. See alsoPerceptualleaming249 , 273-274 vs. differentiation , 288,509-510 , 612 Computation , 197 , 263,283 lawsof, 3I 7 Computer models ofcognitive processes , theory ofverbal learning , 47- 48 473 Attention , 339-340 , 414-418 Concept formation , 51 indevelopment , 375-376 Conditioning (conditioned response )

632

SubjectIndex

Conditioning(cont.) adaptiveness of, 178 ~

Differentiation

, 26 - 27 , 71 , 148 , 157 , 165 ,

288 , 291 - 303 , 334 - 338

classical , 107

of distances , 218 , 221 - 222

as deductive principle, 26- 27, 41 42

of intermodal

specification

operant , 615

in perceptual

development

in rote learning, 25

in verbal

with

shock , 10 - 14 , 102 , 107 - 123

transfer

of , 107 (see also Transfer , bilateral

)

Confusion acoustic

errors

, 413

of , 44 , 46 , 64

Direct perception , 494 , 503 - 510 Discrimination , 42 opposed

to identification

in verbal

learning

verbal

learning

Displacement

of list items , 26

Distance

, 193 , 359 , 392 , 399 , 421 - 435 ,

442 - 443

as measureof generalization, 74- 78 Connectionism

, 97 , 616

as information

, 614 - 615

for affordances

, 589

Constancy of perception , 311 , 324 , 379 , 528 - 529 , 601

and association theory , 311

of layout, 237- 238, 243- 245 of shape, 247, 250, 260, 528 of size , 188 - 190 , 205 , 237 , 250 , 528 of substance , 247 , 527

Control

on absolute

estimation

effect

on absolute

and relative

of training

of behavior

, 569

of events , 614 control

method

, 527 , 532

learning about control, 614- 615 Correspondence in perception , 290

in perceptual learning, 295- 296, 305306 , 309 , 312 - 313

Cues for distance(depth), 150, 205, 208, 218 , 231 - 232 , 265 - 266 , 270 , 279

models , 25 - 27 , 41 - 69

Deforming vs. rigid motion. SeeRigid(ity ) Depth. SeealsoCues for distance (depth) at an edge, 126, 150, 152- 154, 594 (see also Visual

, 221 - 235

cliff )

kinetic depth effect, 247 motion in depth and slant, 249- 261 objects in depth, 138, 156 separationin depth, 264- 281

, 237 - 246

and motion paralax , 277 Distinctive features , 289 , 321 , 326 - 327 , 358 - 360 , 374 , 377 - 378 learning to detect , 401 - 402 of letters , 162 - 163 , 173 , 192 - 194 , 394 , 399 - 401 , 414 , 418 , 422 , 440 - 442 , 478

Ecological

, 159 - 162 , 509 - 510

view , 97 , 178 , 197 , 607 , 608

of perception 541 , 554

, 292 , 366 , 391 , 493 , 500 ,

of perceptual

development

Economy

principle

increasing

in development

environments

in theory of perceptual 291 - 304 , 305 - 309

Ethology

of

pickup , 377 - 381 , 479 - 485

in perceptualleaming Enrichment enriched

, 499 - 503

, 354 , 368

economy

information

Environment Deductive

,

and ground surface , 281 (see also Chapters 12 , 13 , 14 )

of learning

of action , 595

infant

effect of training 205 - 219

of phonemes

Constructionist view . SeePerceptual learning

, in

)

, 577 , 585

by fractionation

of action , and control

, 324

(see Differentiation

judgment

judgment

Consequences

, 507 - 510 , 606

learning , 41 - 69

graphic, 413- 418 matrix

, 612 - 613

, properties

, 466 , 467 , 472 (see Rearing ) learning

,

of , 181

, 178 , 197 , 437 , 438 , 453

Event perception , 177 , 247 , 508 - 509 Evolution , 97 , 177 , 179 Experimental

neurosis , 102 - 103 , 108 - 109

Exploratory activity , 144 , 182 , 197 , 357 , 360 , 366 , 368 , 384 - 385 , 507 , 548 - 549 , 572 , 600 - 601 in active

touch , 182

in development

of perceiving

, 599 - 608

SubjectIndex andknowledge , 501, 599- 606 in relationto age, 592 andself-control, 614- 615 andtraversability , 575, 578- 579, 584- 585, 589- 590, 594- 595, 613 Extinction in infantcontrolleaming,615 in retroactiveinhibition, 91- 92 Eyeheightasinformationfor egocentric distancejudgments , 206, 237

and

similarity

stimulus

Fear , 105 , 108 Feature ( s). See also Distinctive

features

in

, 325 , 473

262

, 269

435

, 517

laws

of

, 41

, 294

. See

, 41

, 48 , 71 -

- 69

) , 4 , 46 , 296

-

, 312

organization

, 379

also

47

, 67

, 328

, 259

, 366

,

, 432

,

, 381

Generalization

, gradients

of density

over

, 206

ground

of

motion

of

texture

, 231

, 224

- 208

-

463

-

164

- 233

, 268

, 275

, 231

correspondence

, 173

,

, 409

features

of

words

, 457

- 458

,

464

Graphic

symbols

See

, 231

- 266

, 220

- phoneme

163

, 227

, 264

, 207

Grapheme

features )

, 221

, 220 velocity

Graphic

of letters (see Distinctive lists , 321 , 327

, 27

( psychology

Gradients

facial , 609 - 610 of layout , 508

, 24

verballeaming

Gestalt

analyzers , 518 detection

, 26 , 37 generalization

93

of

Faces , perception and recognition of , 191 - 192 , 504 - 505 , 609 - 610

633

also

, 147

, 157

Distinctive

differentiation

-

163

, 192

-

194

.

, 585

,

features

of , 396

-

400

of objects , 157 of tasks , 15 - 17 , 22

Habituation

testing model , 430 - 431 of words , 455 - 467

rate

field of velocity

patterns

change , 264 , 267 - 268

perspective

, 268 , 275 - 277 ,

and species - specific learning , 97 tradition of , 3 - 4 , 101 of conditioned

oral

, 590

, 527

- 532

, 536

haptic

, 546

, 548

, 594

- 595

of

- order 379

in

judgments learning

of

morphology

in

reading

- 543

, 303

, 362

,

" same -

,"

431

449

, 405

- 406

, 409

, 444

-

446

,

485

variables

of , 3 19 in

also

Perceptual

Identification

Generalization as confusion, 74- 78 developmental decreasein, 372 gradients of, 26, 28, 31- 32, 37, 48 as indistinguishability, 73 of invariant, 538 of knowledge, 605 procedure in habituation, 527 sensory generalization with voluntary reactions, 25- 39

, 542 , 289

, 446

Hypotheses

of reading , 494 of tasks , 15 , 17

, 583

, 551

, 448

, 394 -

, 579

, 455

sets

in

549

substance

structure

- 380

in

479

-

exploration

specification

response , 9

view of perception , 177 , 504 , 557 - 558 , 568 , 600 , 606 , 616 view view

587

, 593

of stimulus energy , 506 Functional (ist )

view

exploration

Higher

and equilibrium

habituation

Haptic

Filter , 182 , 289 , 328 , 377 , 456 Flow

and motion 280

- dishabituation

of

perception

, 518

learning

308

, 324

431

-

Identity

in

Illusions

519

same

, 291

- different

, 297

, 512

, 102

decisions

-

, 515 103

, 131

, 191

Inference cognitive

, 272

and

for

cues

in

infants

in

perceiving

. See

, 300

432

, 504

Imprinting

-

learning

depth

, 265

, 385 , 474

, 503

, 511

- 519

,

, 302

,

634

Subject Index

Inference(cont.) perception to, 366, 369- 370 unconscious, 280, 293, 358, 472, 513 Information in stimulation, 150- 152, 181, 203, 288- 290, 295, 355, 488, 504- 507 for depth (distance), 203, 223, 231- 232, 263- 281 ignoring irrelevant, 376- 377 and knowledge, 293, 383, 493 for layout, objects and events, 474 as multimodal, 541 obtaining of, 600- 601 in reading, 392 for rigidity , 531 selective pickup of, 375- 376, 461- 447 in text, 453- 468, 493- 494 Information pickup, 494 economy of, 360, 377- 381, 472, 479- 485 Information processing, 97, 471, 473- 474, 488 and researchon reading, 384, 390, 392, 413, 423, 454 Information theory, 319, 321, 328, 355 Inhibition of delay, 25 disinhibition, 66 in list learning, 67, 91- 92 proactive, 56 spreadof, 33 voluntary, 34- 35 Interference and degree of learning, 58 inter-list, 56 and interval between tasks, 60- 62, 65 intra-list, 54 measuredby recall, 56 theory of forgetting , 26 theory of retroactive inhibition , 42 Intermodal relations as amodal information, 541 in crossmodaltransfer, 321 in multimodal specification of events, 500, 611- 615 of object properties, 539, 541 of substance, 541- 554 and unity , 541 Intrinsic motivation , 289, 450, 456 Invariance (invariant), 183, 354, 360- 361, 367 in events, 455

information

for

perception

in

distance

of

in

perceptual

,

learning

search

for

,

377

,

-

specification

self

,

,

depth

,

527

-

539

289

523

transformation

Kinetic

206

500

379

of

over

,

infants

effect

,

203

,

247

,

246

-

249

,

247

,

494

501

,

521

257

Knowledge

grounding

of

and

,

603

perception

599

of

-

,

-

604

295

,

494

results

,

342

-

343

( see

Reinforcement

"

you

"

and

"

,

,

also

development

here

495

)

Language

"

-

608

"

and

,

there

' 1

"

"

,

5

,

21

521

-

522

525

-

5

26

Law

of

acquaintance

of

assimilation

of

common

,

247

of

identifiability

1

48

organization

( Wertheimer

of

,

,

47

47

fate

,

249

,

274

' s

laws

) ,

379

,

381

Layout

,

spatial

508

layout

Learning

.

as

See

ecological

,

,

identification

in

S

-

reading

,

space

453

perception

theory

theory

of

transfer

,

crawlers

self

in

,

42

,

119

,

572

,

266

383

infants

,

in

,

97

,

walkers

,

Looming

-

-

,

108

,

131

109

,

-

153

of

110

,

,

248

121

,

support

568

,

613

-

414

596

,

593

343

572

185

collision

,

475

,

500

,

vs

Nativism

,

142

behavior

Maturation

280

)

( imminent

Maternal

102

of

,

571

,

,

473

development

- guided

184

259

,

surface

of

291

Transfer

detecting

role

,

467

theory

illusions

526

280

,

,

( see

- factor

and

in

-

-

of

5

in

Locomotion

in

277

learning

R

two

,

Perceptualleaming

608

Helmholtzian

in

objective

also

.

learning

- empiricism

101

) ,

,

557

-

,

,

108

563

,

-

568

103

366

-

367

.

See

also

SubjectIndex Meaning, 49, 148, 297, 314, 318- 319, 455 and affordance, 542 context theory of, 294, 305 in language, 521- 525 loss of, 459 in reading, 409 in verbal learning, 63- 66 in word recognition, 485- 487 of "you" and "I," 521- 526 Mechanismsof perceptuallearning, 309- 312 Mediation additive theories of, 356- 357 mechanismsof, 310- 311 Memory for forms, 71 and kinetic depth effect, 257 and perceptualleaming, 303 Misperception, 303, 515- 517 Mobility in blind child, 526 development of, 500 Morphology , 446- 447 Motion elastic (non-rigid), 249, 528- 531 information for layout, 263 parallax, 150, 263- 281 perspective, 268 rigid (seeRigid(ity )) role in perception, 247, 249 Motivation , intrinsic, 289, 456 Movement, of observer, 525. Seealso Locomotion Multimodal . Seeintermodal

Nativism-empiricism , 142, 148- 149, 197, 292- 293, 329, 599 New look in perception , 292, 473 Objects affordances of, 562- 563 identificationof, 190- 192 perceptionof, 516 perceptionof infantsof, 507- 508 Occlusion , 521, 524 Ontogeny of knowledgeacquisition , 605- 606 of reading,437- 450 Orthographicstructure , 457, 465, 484485. SeealsoSpelling , patterns

635

Parallel processing, 414, 418, 454 Pattern recognition by machine, 325 Perception as adaptive process, 177- 195 as construction, 511- 519 definition of, 181, 289 exploratory activity in, 600- 601 perceptual systems, 504- 505 theories of, 511- 512 Perception and action, reciprocity of, 105, 366, 503- 504, 559, 568, 571 Perception and production issue, 367- 368 Perceptualdevelopment, 147- 175, 321, 323. Seealso Perceptuallearning as adaptive process, 177- 195 in infancy, 288 trends in, 365- 385, 470- 485 Perceptuallearning, 71, 126, 163, 192, 202- 203, 238, 280 associativetheories, 290- 294, 305- 315, 356 as categorizing, 325, 327 as construction process, 293, 325, 471, 473 as differentiation, 287- 385 (seealso Specificity, theory of perceptual learning) of distancejudgments (seeChapters 12, 13. 14) enrichmenttheoriesof, 293- 295, 305, 334 as feature detection, 325 in infancy, 608- 615 inference theories of, 265 as judgmental, 324- 325 of multimodal specification, 612- 613 new approach to, 607- 616 observation of consequencesas key to, 613- 614 and reinforcement, 287, 327 as responselearning, 327 review of, 322- 351 as template matching, 325, 435 what is learned in, 316- 320 Perceptualskills and increasedsensitivity, 344 in reading 344- 345 Perspectives coordination of, 521, 524- 525 and station points, 523- 525, 526 Perspectivetransformations, 203, 247- 262, 274, 521, 542

636

SubjectIndex

Phylogeny , 177, 179 Pictorialperception , 610- 611 Predifferentiation , 41, 63, 321, 336- 338 Preference methodin infantresearch , 542- 554 Pronounceability , 462, 465, 482. Seealso Grapheme -phonemecorrespondence Proprioception , 266 Prototypes , matchingto, 321- 326, 394. SeealsoSchema Psychophysics , 3- 5, 203, 199- 203, 283- 284, 308, 317 absoluteestimation , 206- 219, 222- 227 conceptual scale , 218, 227 constanterror, 206, 212- 217, 224, 229, 230, 240- 244 developmental reductionin error, 372 fractionation , 218, 223, 237- 246 andperceptualleaming , 302- 303 perceptualscale , 218 andpsychologytoday, 282- 283 psychometric functionsfor distance , 229 relativejudgments , 227- 234 same -differentjudgments , 421- 435 tachistoscopic thresholds , 340 variableerror, 214 Weber's law, 231 Punishment , 108. SeealsoReinforcement , shockas Rationalism, 293, 599- 600 Reaction time and confusability of stimuli, 423 developmental reduction in, 372 for discrimination, 36 to generalizedresponses, 33- 36 Reading, 344, 389- 495 decoding in, 393- 394, 402- 404 definitions of, 396 and information processing, 390 for meaning, 391 as tool for learning, 453 units in, 405- 406 and writing systems, 390 Rearing in the dark, 103- 104, 142- 145, 155, 331 deprivation in (blind-born), 125, 329- 330 deprivation of pictures in, 330 in enriched environments, 103, 125- 126, 156, 331

experiments in , 125 - 133, 135 - 139, 321 , 328 , 330 - 332

Reciprocity . Seealso Affordance (s)

of animal and environment (seeEcological view )

of perception and action (seePerception and action )

Reduction of uncertainty . SeeReinforcement Reinforcement

, 49 , 50 , 107 , 237 - 238

as connnnation differential as drive effect

of expectancy , 119

, 43 reduction

, 178 , 354

of correction

as fear reduction observation

as, 245

, 109 , 119

of consequences

, 602

in perceptualleaming, 321, 341- 343 as reduction of uncertainty , 183, 290 , 354 , 361 - 362 , 450 , 457

shock

as, 108 - 123

by visual presentation, 189 Response bias, 340 - 341

Retention and interpolated tasks, 15- 23. See also Retroactive Retroactive

inhibition

inhibition , 15 - 17 , 41 , 44

and degree of learning, 59- 60 as function of generalization, 71- 95 and interval

between

tasks , 60 , 62

and similarity, 57 transfer theory of, 42 Rigid(ity ) information

for , 529 , 541

vs. non-rigid motion, 249 and non-rigidity of substance, 527- 539, 541 - 554

of objects, 260, 269 optical information for, 527, 529 and perspective transformations , 248 250

of surfaces , 260 , 273

and traversability, 571- 596 Rote learning, 25 Rules, use of in reading, 444- 449, 472488

Same-different judgments, latencies, 430 - 431 . See also Confusion Scale of distance

conceptual, 218 perceptual, 218, 324 training with , 221 - 235

, matrix

Subject Index Schema in Hebb's theory, 125, 126, 310, 356 in perception , 325- 327 Scribbling , 439- 440 Search for invariance , 377- 378, 450, 456 strategiesof, 374- 375 visual, andscanningbehavior , 413- 418 Self learningaboutcapacities of, 613- 614 self-producedmovement , 334 specification of, 522- 523, 525- 526 Self-regulatingprocesses , 457 Separation in depth. SeeMotion, parallax Sequential processing of letter features , 423- 435 of word features , 459, 463 Shock . SeeReinforcement Sign(s) languagein deaf, 525- 526 perceptionof, 313, 318- 319 andsignificate , 313 Similarity asgeneralization , 48, 51 andindistinguishability , 73 of tasks , 15- 17, 22 Slant motionlessvs. moving, 256 andrigid motion, 253- 254, 261 andvelocityratios, 264- 265, 270- 271, 280 Species -speci6clearning , 608 Speci6city -

specificity

477

-

295

-

theory

ground

perceptual

297

,

perception

,

394

patterns

,

spelling

-

,

,

357

,

509

,

,

361

-

,

510

289

472

.

,

,

See

290

474

,

-

489

separation

406

-

409

,

437

-

recovery

371

,

602

-

,

296

( see

296

50

)

,

55

,

60

,

360

)

in

,

,

-

(importance

support

by

motion

for

locomotion

, 613

of , 571 - 596

Symbols

, perception

Systems

approach

of , 314 , 318

- environment

- 319 , 516

. 600 . 606

senses as information 503 - 510

Template

, 248

, 105 , 188 , 559 , 561

traversability

animal

of ) , 281

, 572 - 574

- matching

- seeking

model

systems

,

, 430 . See also

Prototype Transactional

theory

, 328

Transfer bilateral cross

, of

conditioned

- modal

inter

- list , 52

intra

- list , 51

of

letter

from

- sound

objects

in reading theory

correspondence to

pictures

words

of

response

, 9 - 14

, 332 - 333

, 402 - 405

, 156

, 449 - 450

retroactive

inhibition

, 16 ,

42 - 43 of

training

for

distance

judgments

,

205 - 207 , 217 Transformation by

disruption

of features ,

in variance pattern

perspective by

prisms

surfaces

, 274 , 357 over

(see Invariance

)

, 268 , 274 (see Perspectives

)

, 321 , 333 - 334 , 274 . See also by

infants

Affordance

(s)

, 571 - 596

,

94

,

Uncertainty, reduction of, 290. Seealso Reinforcement Units embedding of, 365 in reading, 444- 445 (seealso Grapheme-phoneme correspondence )

297

Generalization

, 512 , 558 Meaning

465

603

( stimulation

generalization

44

development

,

information

,

correspondence

perceptual

-

438

( see

- phoneme

of

of

Traversability

correspondence

Spontaneous

of

support

of

sound

surface

topological

also

395

Grapheme

Stimulus

371

development

Spelling

370

,

learning

305

Language

Stages

development

theories

vs . deformable

of

of

of

Speech

in

479

- response

test , 460 . See also

rigid

-

increasing

proximal

Stimulus Stroop Surface

637

)

Verballeaming , 25 Virtual object, 251, 260

638

Subject Index

Visual cliff, 103- 105, 141- 145, 150- 157, 186- 187, 560- 561, 594 Visual dominance over haptic information, 594- 595 Voluntary responses, 25- 39 Whole or part problem, 368- 369 Word (s). SeealsoFeature(s) featuresof, 486- 487 information in, 453- 468, 509- 510 perception, theory of, 453- 468 Writing , concept of, 437, 440