The Handicap Principle: A Missing Piece of Darwin's Puzzle

THE HANDICAP PRINCIPLE

This page intentionally left blank

THE HANDICAP PRINCIPLE A Missing Piece of Darwin's Puzzle

AMOTZ AND AVISHA G ZAHAV I with Naama Zahavi-Ely and Melvin Patrick Ely Illustrations by Amir Balaban

OXFORD UNIVERSITY PRES S NEW YOR K OXFOR D

Oxford Universit y Press Oxford Ne w York Athens Aucklan d Bangko k Bogot a Bueno s Aires Calcutt a Cape Town Chenna i Da r es Salaam Delh i Florenc e Hon g Kong Istanbul Karach i Kual a Lumpur Madri d Melbourn e Mexic o City Mumbai Nairob i Pari s Sa o Paolo Singapor e Taipe i Toky o Toront o Warsa w and associate d companies in Berlin Ibada n

Copyright © 199 7 by Amotz and Avisha g Zahavi First published b y Oxford Universit y Press, Inc., 1997 First issued a s an Oxford Universit y Press paperback, 1999 Oxford i s a registered trademar k of Oxford Universit y Press All rights reserved. No par t of this publication may be reproduced , stored i n a retrieval system , or transmitted , in any form o r b y any means, electronic, mechanical, photocopying, recording, or otherwise, without the prio r permission o f Oxford Universit y Press. Library of Congress Cataloging-in-Publication Data Zahavi, Amotz. The handicap principl e : a missing piece of Darwin's puzzle / b y Amotz and Avishag Zahavi; translatio n b y Naama Zahavi-Ely and Melvin Patrick Ely. p. cm . Translated fro m Hebrew . Includes bibliographica l reference s and index. ISBN 0-19-510035- 2 ISBN 0-19-512914- 8 (Pbk. ) 1. Anima l behavior. 2 . Anima l communication I . Zahavi , Avishag. II . Title . QL751.Z44 199 7 96-4237 4 591.59—DC20

1 3 5 7 9 1 0 8 6 42 Printed in the United States of America on acid-fre e pape r Illustrations by Amir Balaban

CONTENTS

Acknowledgments xi I N T R O D U C T I O N TH E GAZELLE, TH E WOLF, AN D TH E PEACOCK'S TAIL xiii

P A R T I PARTNER S I N COMMUNICATION C H A P T E R 1 PREY-PREDATO R INTERACTION S 3

Alarm calls : a message to friend s or t o foes ? Stottin g a s communication with predators. Call s by prey during pursuit: the merlin and the skylark. Warning color s (aposemati c coloration). Signalin g to prey . Cooperatio n between pre y an d predator s withou t communication . Preconditions fo r prey-predator communication. CHAPTER2

COMMUNICATIO N BETWEE N RIVAL S 1

5

Threats a s a substitut e fo r aggression : ar e human s wors e tha n beasts ? Handicaps as keys for reliable threats. Threats by approaching. Threats by stretching—deceit or exposure to danger? Threats by vocalization. Other

CONTENTS

VI

threats. Socia l hierarchies an d duel s betwee n equals . Ca n threa t signal s evolve for the good of the group? The drawbacks of "group selection." C H A P T E R 3 MAT E SELECTIO N 2

5

The conflic t inheren t i n courtship . Courtshi p handicap s an d th e infor mation the y convey : feeding ability , superterritories , courtshi p vocaliza tion, colors , scent s (pheromones) , artifact s an d constructions , com binations of signals, long tails, movements an d dances. Leks: congregating for display . Polymorphic specie s and male s tha t mimi c females. Fisher' s model o f th e Runawa y Process compare d wit h th e Handica p Principle . Utilitarian selectio n an d signa l selection . P A R T I I METHOD S O F COMMUNICATIO N CHAPTER4

TH E FALLAC Y O F SPECIES-SPECIFI C SIGNAL S 4

3

Did decoratio n evolv e to identify species, gender, an d age? The evolutio n of common markings through competition betwee n individuals . Markings and th e trait s the y advertise : line s an d stripes , patche s an d frame s that bring ou t bod y parts . Implications o f the Handica p Principle : th e us e of markings to focu s o n features. Symmetry. "Eye" patterns . Th e evolutio n of markings ; polymorphis m an d convergence . Facia l marking s an d th e direction o f gaze. Status badges o r handicaps? Ar e there signal s without handicaps? Ar e there an y signals that ar e conventions? The definitio n of signals; inflation as a test of the theor y of signal selection. C H A P T E R 5 MOVEMENT S A ND RITUALIZATIO N 6

1

Difficult movements . Ritual fighting. Ritualization: does i t reduce th e information conveyed ? How a ritualized signa l evolves. CHAPTER6

VOCALIZATION S 6

9

The correlatio n o f voice with postur e an d tension . Th e information con veyed b y vocal signals . Animal vocabularies: th e connectio n betwee n th e message and its vocal pattern. Rhythm. Vocal patterns used ove r distance. Why shout ? Th e duratio n o f vocalization: request s an d commands . Dialogues an d their significance . Mimicry. Do animal s have verbal language? CHAPTER7

BOD Y PARTS THA T SERV E A S SIGNAL S 8

1

Long tails : are they signals ? Bristling hair o r feathers : an illusio n o f size, or a handicap? Mane s an d crests . Handicap s tha t interfer e with vision . Body parts that emphasize the direction o f gaze. Body parts that handica p

Contents

VII

fighting. Can body parts evolve to reduce the cost of signals? The evolution of horns an d antlers . Signal selection an d th e evolutio n o f feathers. C H A P T E R 8 TH E USE OF COLO R FOR SHOWING OFF 9

3

Black in the desert. Black and white in open spaces . Colors in forests and on cora l reefs. The us e of two colors. Gloss y colors an d movement . Exceptions t o the rules . C H A P T E R 9 CHEMICA L COMMUNICATION S 1O

1

Pheromones i n butterflies and moths: chemical handicaps. Yeast sex pheromones an d propheromones : th e rol e o f glycoproteins . Chemica l com munication within the multicellular body. P A R T II I TH E HANDICA P PRINCIPLE I N SOCIA L SYSTEM S C H A P T E R 1 0 TESTIN G T H E BOND 1 1

1

Testing b y imposition . Aggressio n i n courtship . Hide-and-seek : gentle r testing in courtship. Clumping, and preening others (allopreening). Group dances and simila r rituals. C H A P T E R 1 1 PARENT S A N D OFFSPRIN G 1 1

9

Threats o f self-injury: th e weapon o f the weaker partner. Other methods of blackmail. Exploitation o f offspring b y parents. C H A P T E R 1 2 BABBLERS , COMPETITIO N F O R PRESTIGE, A N D T H E 1 2 EVOLUTION O F ALTRUIS M

5

Territories, groups , and nonterritorial individuals . Rank, avoidance of incest, an d th e lif e strateg y of male s an d o f females . Th e compositio n o f groups; coalition s o f males and o f females. Struggles without aggressio n (well, almost). Theories tha t explain altruism : group-selection theor y and its failings. The theor y of reciprocal altruis m and th e problem of enforcement. Competition over altruistic acts in babblers: sentinel activities, feeding of nestlings, feedin g of other adults (allofeeding), mobbing. Altruism as a substitute fo r threats. Rank and prestige. "Shyness " ove r copulatio n as a test of male prestige. Reasons for and consequences of living in groups. Prestige an d the evolutio n o f altruism: altruism as a handicap. C H A P T E R 1 3 TH E SOCIAL INSECTS : WH Y HEL P TH E QUEEN? 15

1

The evolution of social structures in the social insects: conditions that favor collaboration: foo d storage and helpless offspring; the haplodiploid mech-

viii C O N T E N T

S

anism of gender determination; selection through queens only, or selection through worker s too ? Ho w insec t colonie s form . Wh y d o th e worker s work for the colony? altruism and prestige; queen pheromones an d prestige; the handicap in queen pheromone. Kin selection theory and its drawbacks: parasitis m amon g kin, o r Haldane's othe r brothers; ar e offsprin g "kin"? Partnership s amon g kin : wh y i t make s sens e t o joi n th e famil y business. The kin effect . C H A P T E R 1 4 T H E PARENTING COUPL E 1 6

9

Paternity an d mate-guarding . Taking car e of the youn g to gai n prestige . Other mean s of showin g off to one' s mate . Dominance betwee n mates . Conditions for female dominance. The parental couple as a partnership. C H A P T E R 1 5 SOCIA L AMEBA S (CELLULA R SLIM E MOLDS ) 17

7

The life cycle of cellular slime molds. Forming the stalk: altruistic suicide? The individua l selection hypothesis . DIF a s a poison. The differenc e be tween prestalk and prespore amebas . Some remaining questions. When is a chemical a signal? C H A P T E R 1 6 PARASIT E AN D HOS T 18

5

An arms race or a state of equilibrium? European cuckoo s and reed warblers. Grea t spotte d cuckoo s an d crows . The prestige model. Th e Mafi a model. Acceptin g a parasit e t o minimiz e damage . Neuterin g th e host . From parasit e to collaborator . Th e less virulent parasite as a collaborato r against it s virulent variant. The implication s o f assuming a state of equilibrium. C H A P T E R 1 7 INFORMATIO N CENTER S 1 9

7

Food sources and social organization: the white wagtail. Communal roosts as information centers . Insurance agains t evi l days : winter gathering s o f rooks. Flock s an d loners : the communa l roos t o f kites i n Coto Donana. Bright adult s an d dul l youngsters: handicaps in food squabbles. How information repositorie s work . Huma n gatherings . Communa l display s at gatherings: promoting the roost or mutual testing? P A R T I V HUMAN S CHAPTER 1

8 HUMAN

S 2O

9

Innate behavio r i n humans . The huma n body an d it s decorations : hair ; eyes, eyebrows, and eyelashes ; nose and facia l wrinkles ; chi n an d beard ;

Contents

IX

red cheek s an d lips; menstruation; breasts and body fat; clothing. Testin g the huma n socia l bond: the huma n sexua l act as a test o f the bond, self endangerment i n humans : suicid e a s a cr y fo r help . Huma n language : communication withou t reliability . Decoration , esthetics , an d th e evolu tion of art. Altruism and mora l behavior. Epilogue 229 Notes 231 Bibliography 245 List of Figures 261 Index 267

This page intentionally left blank

ACKNOWLEDGMENTS

O

ur daughter s Naam a and Tirza took par t fro m th e very beginning i n the discussions fro m whic h ou r idea s emerged , an d thei r suggestion s hav e helped us present those ideas here. Naama's presence in these discussions, and tha t o f her husban d Melvi n Patric k El y in late r years , equipped th e tw o of them, workin g fro m ou r origina l Hebrew text , t o writ e the versio n o f this book that now appears in English and other languages. We also thank Amir Balaban for the skil l and artistr y he brought t o the task of illustrating the book . Azaria Alo n encourage d an d helpe d u s throughout , bu t especiall y wit h th e difficult earl y steps . W e wis h t o than k Daniel a Atzmony , Helen a Cronin , Pau l Eckman, Michal Gil, Jehoshua Kugler , Arnon Lotem, Jonathan Wright , Mina Yarom, Yoram Yom-Tov, and Zohar Zuk-Rimon for reading and commenting on the whole manuscrip t o r on som e of its chapters . The first drafts of this book were written well over ten years ago. It is impossible for u s to acknowledg e al l our friends , students , an d colleague s who rea d some of these drafts an d made important comment s and suggestions for improvements, or those wh o provide d photograph s o r videocassettes , o n which Ami r Balaban has based som e of his lively drawings . W e d o hope that these friends wil l forgive u s for no t mentionin g al l of them b y name.

xii A C K N O W L E D G M E N T

S

We thank the staf f o f Oxford University Press, especially our edito r Kirk Jensen, ou r productio n edito r Kimberl y Torre-Tasso, an d ou r devote d copyeditor , Nora Cavin. We gratefull y acknowledg e the effort s o f our literar y agent, Richard Balkin. Special thanks ar e due to the man y volunteers and students wh o helped with our field observations of the babblers , gettin g to know these cooperativel y living birds individually and recording data about their behavior. But above all we thank the Societ y for th e Protection o f Nature in Israel (SPNI) for allowing all of us t o live and work in its Field Stud y Center a t Hatzeva. The income from th e Hebrew editio n of this book is dedicated to a fund which will b e use d t o continu e ou r stud y of th e babbler s an d t o maintai n the Sheza f Nature Reserve , where these bird s ca n continue t o live , protected fro m th e up heavals created by modern life. We hope that the English version and other translations of this book will widen th e circl e of friends of the Sheza f Nature Reserve and its babblers. Amotz and Avishag Zahavi

I N T R O D U C T I O N

THE GAZELLE , TH E WOLF, AND TH E PEACOCK'S TAIL

W

e start with a scene of a gazelle resting or grazin g in the desert . I t is nearly invisible; the colo r o f its coat blends well with the deser t landscape. A wol f appears . On e woul d expec t th e gazell e t o freez e o r crouch an d do its utmost to avoid being seen . But no: it rises, barks, and thumps the groun d wit h it s forefeet, al l the while watchin g th e wolf. The thump s o f th e gazelle's hoove s carr y through th e deser t groun d ove r long distances ; its curve d horns and th e dark-and-ligh t pattern o n its face clearl y reveal that the gazelle is in fact lookin g a t its enemy. If the wolf come s nearer, one would expec t the gazell e to flee as fast a s it can. But n o again : often th e gazell e jumps high on al l four leg s several times and only then begins to run, wagging its short black tail against its conspicuous white rump, which has a black border. These high jumps are very clearly linked to the approach of the wolf. Yet a gazelle escaping immediate, urgent danger—such as hunters in a jeep—flees in a n entirely different manner : it run s away silently a t great speed , making good us e of the topography to concea l its escape. Why does the gazelle reveal itself to a predator tha t migh t not otherwis e spo t it? Why does it waste time and energ y jumping up an d dow n (stotting ) instea d of running away as fast a s it can ? The gazell e is signaling to th e predato r tha t it has

xiv I N T R O D U C T I O

N

seen it ; by "wasting " time and b y jumping high i n th e ai r rather tha n boundin g away, it demonstrates i n a reliable way that it is able to outrun th e wolf. The wolf , upon learnin g tha t i t has lost it s chanc e to surpris e its prey, and tha t thi s gazelle is i n tip-to p physica l shape , ma y decid e t o mov e o n t o anothe r area ; o r i t may decide to look for more promisin g prey. Even partie s i n the mos t adversaria l relationships, suc h a s prey an d predator , may communicate , provide d tha t the y have a common interest : in this case , both want to avoi d a pointless chase . The gazelle tries to convince the wolf that it is not the eas y prey the wolf is looking for , and that the wolf would be wasting time and energy by chasin g it. Even i f the gazell e is sure that i t ca n outrun th e wolf , i t to o would prefe r to avoid an exhausting chase. But in order t o convince the wolf no t to give chase, the gazelle has to expend preciou s tim e and energy that i t will need should th e wol f ignore its signals and decid e t o chas e it anyway . The encounter betwee n th e gazelle and the wolf dramatize s the basic theme of this book: in order to be effective, signal s have to be reliable; in order to be reliable , signals have to be costly . The hig h cos t tha t animal signaling often involve s is clearly seen in the cas e of the peacock . Mos t peopl e hav e seen an d admire d a peacock, spreadin g an d quivering his enormous tail—a fan of glistening feathers, adorned with blue and green "eyes." But to be abl e t o put o n such shows , peacocks hav e to drag massive tails around mos t o f the year . By managing to fin d foo d an d avoi d predator s despit e such a burden, a peacock prove s tha t he is the high-qualit y mate that th e peahe n is seeking to fathe r her futur e chicks . This i s another basi c them e o f our book: tha t there i s a logica l relationshi p betwee n th e signa l and th e messag e i t conveys . Th e gazell e display s its confidence in its ability to outrun th e predato r by drawin g attentio n t o itsel f an d b y expendin g precious tim e an d energ y that i t will need shoul d its signal not be heeded. The peacock proves his strength and agility by carrying a heavy load, as does a stag carryin g heav y antlers . Eac h signa l is closely related to its message. A perso n ca n signal courage by courting danger , for example, bu t courtin g dange r doe s no t attes t t o wealth—whic h ma y be demon strated b y wasting money. The investmen t that animal s make in signal s is similar to the "handicaps " imposed o n th e stronge r contestant s in a game or a sporting event : for example, th e removal of the superior player's queen in a chess match, the extra weight the swifte r race horse mus t carry , or the scor e o f several strokes tha t the more accomplishe d golfer start s with. A handica p prove s beyon d a doubt tha t th e victor's wi n is due to mastery , no t chance . Th e peacock' s tai l an d th e stag' s antler s ar e no t mer e disabilities; rather , the y ar e handicap s i n thi s ver y specia l sense : the y allo w a n individual animal to demonstrat e it s quality.

Introduction x

v

Are animal signals always reliable? We believe that most signals are: before an individual acts on information it receives through signal s from anothe r individual , it first needs to chec k the reliability of that information. We suggest a very simple principle: if a given signal requires the signaler t o invest more in the signal than it would gai n by conveyin g phony information, then fakin g i s unprofitable an d th e signal is therefore credible. I f a gazelle that is slow or weak sends the wolf a phony signal about its speed and strength by stotting, it wastes what little strength it has on puny jumps that will only convince th e wolf that it is easy prey. Such a gazelle would d o better t o fle e fo r its life an d hope fo r the best . T o gauge the reliabilit y of a signal , then , on e ha s t o conside r th e investmen t i t entails . Th e cost—th e handicap tha t th e signale r take s on—guarantees that th e signal is reliable. When w e first suggested this Handica p Princi ple i n 1975 l i t wa s almos t unanimousl y rejected. Many papers were published using formal, explicit mathematical models "proving " tha t the Handica p Principle doe s not work,2 or that it might apply only under ver y special conditions. 3 Thi s trend change d in 199 0 whe n Ala n Grafe n publishe d two papers 4 using differen t mathematica l models t o sho w tha t the Handicap Principle i s generally applicable, an d that it is a sound principle that can ensure reliability in communicatio n betwee n competin g organisms . Since then , th e Handica p Principl e ha s becom e widely accepted. 3 Throughout al l these years , while our colleague s were debating th e validity of the principle , w e continued t o observ e an d explor e th e livin g world aroun d us . The Handicap Principl e reveale d to us an endless arra y of new ways to understand such phenomena as the extreme expenditures often involve d i n sexual advertise ment, th e evolutionar y enigma of anima l altruism , an d th e working s o f collabo rative systems in the animal kingdom, which could not easily be explained in terms of straightforward , utilitarian natura l selection . Our investigation s into the ramifications of the Handicap Principl e coincide d with ou r studie s o f the babblers, group-livin g songbirds, at Hatzeva in the Arava Valley of southern Israel . W e have been studyin g these small desert birds fo r th e last twenty-fiv e years. The y ar e used to ou r presence ; we often ge t within a few feet or even inches of them. As far as they are concerned, we are not much differen t from the herds of goats and camels with which they share the desert. We can watch them very closely, observe the fine details o f their behavior, an d hear the soft an d subde vocalizations wit h which they communicate with one another. We can identify eac h individual bird b y the colore d le g bands w e put o n it s legs when it was a nestling. All members o f a babbler grou p participat e i n th e defens e of thei r territor y and i n the car e of nestlings, eve n when the nestling s ar e not thei r ow n offspring .

XVI I N T R O D U C T I O

N

The birds perform many other altruistic acts, such as feeding other adult member s of thei r grou p an d standin g guar d whil e th e res t o f th e grou p i s feeding. Such activities are difficul t fo r researcher s o f evolution t o explain. We discovere d tha t these altruisti c behavior s serv e to advertis e each babbler . I n othe r words , each babbler's investment in altruistic acts demonstrates the validity of its claim to social status—to prestige. We learned from babblers how important it is to animal s t o be able to measure the strength of social bonds; that need explained the babblers ' group "dances," and their clumping together to rest. The Handicap Principl e also explains th e birds ' shynes s abou t se x an d eve n th e detail s o f thei r ver y subtl e decorative markings. We did not start out seeking a unifying principle i n biological communication . All we attempted t o d o a t first, in 1973 , was to explai n the evolutio n o f the peacock's tail to a student an d colleague , Yoa v Sagi , who—fo r goo d reason , a s it turned out—coul d no t see the logic in Fisher's runawa y process, then the current theory. Ou r broa d applicatio n o f the Handica p Principl e develope d slowly : one finding led to another, until we realized we were dealing with a general principle . Many people, includin g those who now accept th e Handica p Principle , have not ye t attempted to appl y i t broadly to biological signaling . W e think tha t th e explanations an d models we present i n this book ar e more plausible, an d fit the known facts better, than those commonly found in textbooks. We expect our ideas to serv e a s a starting point fo r mor e detaile d studie s an d experiments . Som e of our explanations will prove valid as they stand; in other cases, reality will turn out to be even more fascinating and complex than we now imagine. We hope that this book wil l encourag e th e searc h fo r reliabilit y and fo r it s unavoidabl e cos t i n all systems of biological signaling. W e believe that this search wil l change our under standing of the natura l world in a myriad of ways.

P A R T1

PARTNERS I N

COMMUNICATION

This page intentionally left blank

C H A P T E R1

PREY-PREDATOR INTERACTIONS

ALARM CALLS: A MESSAG E TO FRIENDS OR TO FOES ? abblers mov e aroun d their territor y durin g the day, looking for food o n th e ground amon g the deser t trees an d bushes. One o f the grou p ofte n perche s on a treetop, actin g as sentinel. When th e sentinel, or any other babbler, see s a bir d o f prey , or raptor , i n th e distance , i t emit s a loud "bark. " Upon hearin g such barks, babblers ofte n rais e their heads an d sca n the sky . If the rapto r i s far away an d doe s no t pos e an y immediate dange r t o th e feedin g group , the y go on searching for food. But when a n abrupt, frightened bar k indicates imminent danger, th e birds, including the sentinel, immediately jump for cover. Often, however, precisely afte r suc h a fright, th e entir e grou p goe s u p t o th e to p o f th e tre e i n which they have taken shelter an d joins the sentine l in barking and calling loudly. In mos t studie s o f animal behavior suc h vocalizations are classified a s warning calls, issued by animals to others o f their species. That idea has a certain superficia l plausibility. Bu t year s of observation raise d s o many puzzling question s tha t w e finally dared as k ourselves whethe r th e call s were indee d mean t as warnings. The barks rin g out before the rapto r ha s had an y chance to notic e th e grou p o r pos e any dange r t o it. The call s seem t o be a n unjustified ris k for the barkin g birds: a feeding grou p of drab an d well-camouflaged babblers ha s a good chanc e of being overlooked b y a rapto r altogether , ye t th e bark s disclos e thei r locatio n ove r a

B

3

4 PARTNER

S I N COMMUNICATIO N

considerable distance . In fact, when we try to find babblers in their vast territories, which often cove r more than a square kilometer, we have a good chance of missing them whe n the y feed quietl y amon g the bushes; bu t a s soon a s we see a raptor in the distanc e we stop the ca r and listen fo r barking babblers. Th e sharp barks th e birds emi t periodicall y a s long a s a raptor i s around allo w us—and probably th e raptor—to locate th e group. The call s seem to serve the interests o f the predato r rather than the group . One coul d possibl y understan d th e first bark a s a warning to th e group , an d perhaps th e secon d an d th e thir d a s effort s t o mak e al l members o f th e grou p aware of the danger . But wha t is the poin t o f repeating th e call s afte r th e entir e group ha s already taken cover ? And who m ar e they "warning," when th e entir e group joins the sentine l a t the top o f the canopy, barking together ? And if the bar k is a warning to th e res t of the group , why i s i t s o loud ? Babbler s ofte n vocaliz e softl y whil e feeding, bu t th e "warnin g calls " ar e hundred s o f time s louder tha n thes e sound s an d ma y be hear d ove r hal f a mile away. Why do babblers rais e the volume of their calls to a level that ca n alert a predator to their presence pre cisely when a predator i s in th e area—an d befor e i t ha s had a chance to notice th e well-camouflaged group? There is also a theoretical difficult y regardin g the evolution o f warning calls. For a trait—such as the tendency to call out when one spots a predator—to spread through a population b y natural selection, that trait has to improve the chance s tha t th e specifi c individua l wh o possesses it will survive and pass it on to descendants. Bu t the investment in so-called warning calls is made by the callers , while the benefit goes to the listeners. 1 Indeed, caller s an d listener s ar e no t eve n necessaril y of th e sam e species : a t Hatzeva, where we observe, the babblers, shrikes , blackstarts, bulbuls, and wheatears also call out loudly at the appearanc e of raptors. Some of these birds, such as shrikes in winter, ar e solitary and have no collaborato r t o warn, not eve n a mate. The fac t tha t th e barks tell anybod y who happens t o be around , including othe r bird species—an d for that matter, including us—tha t there i s a raptor i n the sky does not mean that they evolved for that purpose. S o we searched for an alternative explanation. Loud call s mak e sens e whe n th e listene r i s distant. 2 Sinc e th e rapto r i s fa r away fro m th e sentinel , i s i t possibl e tha t th e intende d listene r i s actuall y th e raptor? Th e notio n make s eve n mor e sens e i n view of the callers ' behavio r afte r a rapto r lands : th e babbler s ofte n approac h th e predato r an d "mob " it . On e does not hea r an y clear separatio n betwee n th e warnin g call s an d th e mobbin g calls: on e behavio r merge s into th e other . I t i s widely accepte d tha t mobbin g i s directed a t the predator 3; is it possible tha t the "warning " call s ar e also directe d at it ?

Prey-Predator Interactions 5 Communication requires cooperation. There is no point in talking unless someone is listening, and there is no point i n listening unless it might be to the listener's advantage to do so. These two conditions ar e met only if there is a common interest between th e parties . Wha t ca n be th e commo n interes t betwee n a rapto r an d a babbler? A rapto r ca n catc h a babbler onl y b y surprise, o r whe n th e babble r i s far fro m cover . Babblers ar e masters of the thickets. Inside th e vegetation, thei r stron g feet , lon g tails , an d short wines enable the m t6o dodee and outma,T . , neuver an y predator. I t i s not unusua l to se e a group o f babblers hoppin g an d callin g loudl y in a bush o r a tree that a raptor has perched in. Indeed, at every loud noise, suspicious movement, or urgent warning call, babblers hurry into a thicket, where a predator ha s no chance against them; a predator tha t knows i t has been see n b y babblers i s wasting its time if it stay s in the area . And as long as the predator remain s in the vicinity, the babblers have to stay near bushes and track the raptor's movements to make sure it does not catc h them by surprise. If th e babbler s notif y th e rapto r tha t the y hav e see n it , bot h partie s gain . Th e raptor move s o n t o anothe r feedin g ground , t o tr y an d surpris e other prey ; th e babblers ca n resum e thei r feeding . I t make s sens e fo r th e babbler s t o signa l t o the raptor, and for the rapto r to pay attention t o their signal. But on e coul d argu e that onc e th e raptor s hav e learned—or evolved th e ability—to pa y attentio n t o th e "warnin g calls, " babbler s coul d "cheat, " emittin g warning barks periodically, whether o r not they see a raptor. The raptor, however, has a defense agains t such trickery : it put s th e burde n o f proof o n th e babblers , reacting to call s only when it is sure that they are directed a t it individually. The behavior o f the "warning" babble r fits this hypothesis: the babbler climb s to th e to p o f the tre e t o bark, eve n though i t could hav e watched th e rapto r and called fro m withi n th e canopy—a s i t doe s when i t i s really frightened . The bir d barks, disclosin g it s location, rathe r tha n trilling , a s babblers d o under othe r circumstances, and which makes them much more difficult t o locate.4 A babbler who would chea t by going to the to p o f the canop y and barking before it actually saw a predator woul d expose itself to raptors it might not have noticed. That risk helps ensure that if a babbler goe s to the top o f the tree and declares it has seen a raptor, it has indeed see n one. If th e rapto r lands , th e babbler s ofte n approac h i t an d mo b it , erasin g an y doubt th e rapto r ma y have tha t i t i s indee d th e objec t o f thei r calls . Th e call s change fro m bark s t o trills , interrupte d b y tzwicks wheneve r th e rapto r moves , which proves that the babblers ar e in fact watching it continuously. Bulbuls, blackstarts, sunbirds , warblers , an d wheatears al l join th e gathering , so the rapto r ca n see tha t the y to o ar e awar e o f it s presence . B y taking th e calculate d ris k o f ap proaching th e predator, th e birds increas e the reliabilit y o f the message, which is

6

PARTNERS I N COMMUNICATIO N

that the predato r ha s no chanc e of catching them. Indeed, they may even expose the raptor to its own enemies, such as bigger predators or humans, who may notice the commotion and come to investigate . Having los t th e chanc e t o catc h it s prey by surprise an d itsel f expose d t o risk , th e rapto r i s per suaded t o leav e the area . Bu t mobbing expose s th e babblers a s well: SordaP collecte d evidenc e showing that mobbin g bird s ar e occasionall y caugh t b y th e raptors they mob. This ris k is the price birds have to pay to convince the raptor that they are indeed aware of its presence. 6 Unfortunately we cannot collect dat a on whether the behavio r o f th e babbler s encourage s raptor s t o move on. Although the babbler s ar e used t o u s and do no t min d ou r movin g amon g them , th e raptor s certainly d o mind, an d our presence causes them to move away.

STOTTING A S COMMUNICATION WIT H PREDATOR S In th e introductio n w e describe d th e behavio r o f gazelles toward wolves . Thes e behaviors were onc e thought t o be warning signals meant to alert other gazelles. 7 But thi s explanation pose s th e sam e sort o f question s a s the contentio n tha t th e babblers' bark s are warning calls: the thumping , barking, an d slotting—jumpin g straight u p o n al l four legs—d o indeed war n othe r gazelle s of the predator , bu t why haven' t gazelle s evolved a more discree t wa y to aler t eac h other ? Again , we propose that the gazelle's behavior is in fact aimed at the predator. First, the gazelle shows the predator tha t it has seen it, by stamping its feet an d turning the blackand-white "flag" of its behind toward s the predator. Then it shows off its strength and fitness by jumping straight up . Only a gazelle certai n of its ability to outrun a predator dare s squande r it s strengt h i n this way. Stotting provide s a distant ob server with clea r evidence o f the abilit y of the gazelle to jump: a high jump can be assessed fro m an y direction, while long jump s coul d b e properl y evaluate d only when observed from th e side . If the wolf wants to selec t a gazelle it has a chance to overtake , it makes sense for th e predato r t o watc h th e stottin g o f gazelle s befor e i t choose s an d start s pursuing it s prey . Th e chas e is a matter o f lif e an d deat h fo r th e gazelle ; it als o requires a high investmen t fro m th e wolf . Anybod y who ha s see n a wolf afte r a chase, successful or otherwise, can testify to the wolf's exhaustion. If it fails, it may be a while befor e i t ca n gather th e strength for another attempt. Hence both the gazelle an d th e wol f benefi t fro m communicating , as long a s the communicatio n

Prey-Predator Interactions 7 is guaranteed to be reliable. The wolf avoids exhausting itself in pursuit of a gazelle it cannot catc h and saves its energy for a more promising chase; likewise the gazell e will hav e th e strengt h t o fle e fro m a predato r wh o ma y fee l u p t o th e tas k of overtaking it. Since 1977 , whe n w e firs t publishe d th e idea that stotting and "warning" vocalization s are i n fac t directe d a t predators, 8 zoologist s have collected som e evidenc e t o suppor t th e hypothesis that individual prey benefit by signaling t o predators . Fitzgibbo n an d Fan shawe9 collecte d dat a i n th e 1980 s i n th e Serengeti t o test thi s hypothesis. Fro m a hilltop, they watched th e hunting strategies o f spotte d hyena s and o f hunting dogs , wh o catc h thei r pre y by runnin g i t down i n a lengthy chase, rather than hunting by surprise, like cheetah s and lions. Some gazelle s stotte d repeatedl y whe n dog s o r hyena s approached, while other s did not attempt to stot but rathe r fled right away. Both spotted hyena s and hunting dogs wen t afte r gazelle s tha t di d no t stot , o r stotte d onl y a little; the y avoide d chasing those who gav e an impressive display of stotting before their escape. Thus it seems that predators d o indeed observ e stotting in order to estimate an individual's abilit y to flee. Hasson an d hi s colleagues 10 collecte d evidenc e o f simila r behavio r i n zebra tailed lizards . When a lizard i s surprised i n th e ope n i t run s awa y a t to p speed . But when i t is near its hole o r near a bush wher e it is able to hide, it does not flee but instea d stops an d moves its conspicuous tail from sid e to side. The model they used t o explai n th e behavio r o f th e lizar d was th e on e w e use d t o explai n th e behavior o f the babbler s an d the gazelles. 11 We were no t th e firs t t o sugges t tha t potentia l pre y may communicate with predators. Smythe 12 observed the behavior o f maras—large Patagonian rodents— when the y meet humans. He realize d that their behavior wa s analogous to that of stotting gazelles; but maras are solitary, and so he could not interpret that behavior as a warning t o other members o f their species. Smythe concluded that the fleeing maras signal to predators—that the former manipulate the latter, "inviting" pursuit only when they are ready for it. Smythe believe d tha t predator s wer e stimulate d t o pursu e fleeing prey eve n when the y stoo d t o los e b y doin g so . It di d no t occu r t o Smyth e that pre y an d predators may have a common interest . He wa s aware o f the shortcoming s o f his explanation; he suggested that evolution was not yet complete, an d that eventually the predator s migh t evolve the abilit y to ignore such temptations. Bu t any feature whose benefits we do not understan d ca n be dismisse d by saying it is not ye t well adapted an d will change in the future . W e thin k it more likely that most feature s are well adapted and in equilibrium with their environment; experience has shown us that thi s approac h produce s bette r explanations . Eve n when on e has no ready explanation fo r a given phenomenon, i t is best t o assum e that we are still missing

H

partners in communication

a par t o f th e pictur e an d continu e th e search . Suc h a strateg y is more likel y t o generate new findings. The socia l interaction s amon g childre n i n a game o f ta g serv e a s a beautifu l model o f prey-predator interactions . Once a catcher—"it"—is chosen , childre n who kno w themselve s t o b e slowe r tha n th e catche r tr y t o ru n a s fa r awa y a s possible. B y contrast, thos e wh o kno w themselve s t o b e faste r tha n th e catche r stay nearb y or eve n approac h th e catcher , "inviting " pursuit. The catche r ofte n ignores th e stronge r an d pursues th e weak, who see k refug e a t the fa r end o f the playground. Th e handicap—th e ris k o f approachin g th e catcher—take n o n b y those who are confident in their abilit y to outrun him or her convinces the catcher not t o waste any effort tryin g to catc h them.

CALLS B Y PREY DURING PURSUIT : THE MERLIN AN D TH E SKYLARK Rhisiart, an d late r Cresswell, 13 studie d ho w traine d merlins—a species of falcon—select thei r prey among skylarks. Whe n th e falcone r sight s prey , he remove s the merlin's hood an d releases it to the chase. Rhisiart found tha t whe n th e skylar k sing s whil e fleeing , th e merlin is likely to abort th e chase. When th e lark does not sing , the merlin is more likely to continue the pursuit an d is often abl e to catch th e lark . What coul d b e the connectio n betwee n th e song and the chase? If we assume that som e larks are faster tha n the merlin an d som e are slower, it makes sense for the merlin t o tr y to selec t an d chas e individual s it can overtake. It i s also in th e interest o f a skylark that flies faste r tha n the merlin to let the merli n kno w that it cannot be caught. To convince the merlin of its superior abilities, the skylark must do something that a slower lar k would not be able to do. Singing while flyin g is a good indicato r of the lark's abilities, sinc e it displays the bird's capacity to diver t a part o f its respirator y potential whil e still flying a t least a s fast a s the merlin . A skylark tha t need s ever y ounce o f strength i t ha s t o fly cannot sin g a t th e sam e time. We can see the similarity between th e vocalizing of the chased skylark and that of children . A child fleeing during a play-chase ofte n jeer s a t his pursuer. Suc h vocalization provide s reliabl e informatio n abou t th e reserve s o f strengt h a t th e child's disposal . I t makes sense to invest some air in a jeer if it saves one a greater investment i n a prolonged flight. A child running a s fast a s he ca n cannot affor d to jeer at his pursuer; if he does , the breathless quavering of his voice reveals that his energy will soon give out. The difficulty of vocalizing while running thus renders the jeering a reliable signal; i t therefore make s sens e for th e pursue r t o listen t o

Prey-Predator Interactions 9 vocalizations in order t o asses s the abilit y of the runne r t o flee. Both th e pursue r and th e pursued gai n from communicating . Once again we have drawn a parallel between the behavior of animals and that of humans. Some deride suc h comparisons as "anthropomorphisms" an d conside r them "unscientific. " Bu t a model i s a tool; and anthropomorphi c models , i f anything, are closer to the reality of animal behavior than mathematical ones. Needless to say , models ar e not proofs ; they ar e suggestions. Models ca n help on e devis e experiments and plan the collection o f data to test interpretations. We use humans as a model only to improve our understandin g of the behavior of other organisms and th e logi c behin d thei r behavior . Afte r all , mathematica l models—muc h i n vogue among zoologists thes e days—ar e als o just that: models, which have to b e tested, an d which may or may not reflec t reality .

WARNING COLOR S (APOSEMATI C COLORATION ) Many poisonous animal s have bright coloratio n tha t stands out from their surroundings. These bright colors—aposematic coloration—have long been considered a warnin g t o predators , an d indeed , man y predators avoi d these animals . But how did th e coloring evolve ? We sugges t that the evolutio n o f aposematic coloration parallel s the evolution of all other signals tha t serv e th e commo n interes t o f prey an d predator. T o indicate reliably that the signale r is unpalatable, colo r mus t be s o conspicuous that no palatable prey can afford t o have it. In fact , the show y colors d o attract the notice of predators. There is now some evidence tha t naive and inexperienced predator s actuall y learn to avoid these colorful animal s by catching and tastin g them;14 althoug h th e predators ma y wound their gaudy prey while getting this first taste, they find it unpalatable an d let it go. Indeed, man y poisonou s insect s hav e especiall y stron g cuticl e an d regenerativ e ability, which lets them recove r from suc h injuries. Thus, by being highly colore d the prey undertakes an increased risk of being spotte d b y predators, bu t tha t risk is balanced b y the likelihood tha t most predators wil l avoid it. What abou t mimic s who look like poisonous animal s but i n fac t ar e not poi sonous? Som e researchers are rather too prone t o assum e that a certain anima l is mimicking another ; on e o f th e classi c textbook example s o f butterfl y mimicr y turned ou t to be even more toxic than the species it was "mimicking."15 Nevertheless, tru e mimicr y does exist . There i s a strictly limited evolutionar y niche for such cheating: in cases where the receiver of the signal—th e predator— does no t benefi t b y making the effor t neede d t o tel l th e differenc e betwee n th e poisonous species and the mimic. If the poison is very dangerous and the predator' s

1O

PARTNERS I N COMMUNICATIO N

food plentiful , an d i f there ar e few mimics, then th e ris k of its mistakenly eating the poisonous pre y may outweigh th e small benefit it gains by eating a mimic. But the greate r th e numbe r o f mimics and th e scarce r the predator' s food , th e mor e tempting th e cheater s become , an d th e greate r th e ris k t o the m tha t a predato r will learn t o distinguish betwee n the m an d their poisonous model . A t that point, the mimics' prominent colorin g become s a severe drawback.

SIGNALING TO PRE Y Not al l communication betwee n pre y and predato r run s fro m th e forme r to th e latter. The interaction betwee n th e tiger and the bull convinced us that predator s may signa l their pre y as well. Th e tiger , onc e ther e i s no chanc e it ca n catc h th e bull by surprise, moves in an arc around it. The bull endeavors constantl y to fac e the tige r an d point it s horns a t it, to prevent th e tiger fro m pouncin g o n its back. The tiger, indeed, i s looking fo r a n opening t o d o jus t that, while tryin g to avoi d the bull's sharp horns. The tiger's black and white ears, which point forward, make it eas y to tel l exactl y what th e tige r i s looking a t an d s o telegraph it s intentions . Thus th e tige r force s th e bull t o react ; the tiger ca n then bette r asses s the agility of th e bull , its abilit y to defen d itself , an d it s weaknesses. This informatio n help s the tige r attac k whe n an d wher e i t i s sur e t o hav e a n advantag e ove r th e bull, without running int o its horns . The tiger is telling the truth. It makes real preparations for attac k an d stop s i f and whe n i t i s clear tha t th e bul l is read y t o repe l th e attack . Th e tige r want s t o fin d a weakness i n th e bull' s defenses ; the mor e precisel y th e predator display s its location an d its intended moves , the more reliabl e an d informativ e the bull's reactio n is , and the easie r i t i s for th e tige r t o identif y tha t weakness. If the tige r di d no t displa y it s positio n an d movement s clearly, it could not rely on its prey's responses and might then underestimate the bull' s agility , strike, and find itself impaled o n the bull's horns . The interactio n betwee n tige r an d bul l i s simila r t o that between two boxers in a ring. They do not as a rule go directly for a knockout blow bu t firs t attemp t explorator y punches . A majo r blo w tha t misse s it s target would thro w th e attacke r of f balance an d leav e hi m vulnerable . Boxer s usually start with blow s tha t ar e strong enoug h t o forc e th e opponen t t o defen d himself but d o not involve too great a risk. The exploratory punches test the ability of the opponent t o react . A knockou t i s usuall y attempte d onl y afte r th e attacke r ha s assessed his rival. Some se e these explorator y punche s a s efforts t o mislea d th e opponent . Bu t

Prey-Predator Interactions 1

1

misleading i s a less effectiv e strateg y than on e tha t utilize s rea l advantages, since its success depends o n the opponent's stupidity and on chance, rather than on the attacker's rea l ability. Early in a bout, man y o f a boxer's move s ar e made t o determine th e opponent' s agilit y in defense . W e believ e thes e exercise s hel p th e fighter—or th e predator—no t b y misleadin g bu t rathe r b y offerin g precis e information abou t th e attacker' s repertoire . Th e mor e reliabl e th e informatio n provided b y the attacker , the more he finds out; eventually, he can attack what he has learned are his opponent's most vulnerable points. The defender has no choice but t o sho w the attacke r something of his defensive ability. Of course , a defender certain of his strength and prowess may choose to ignore the attacker's maneuvers, showing a cal m an d a lac k o f respons e which themselve s ma y persuade th e ag gressor of the defender' s abilit y to withstand a n attack. Eshel16 also suggested tha t predators ma y communicate with their pre y in order t o identif y th e mos t vulnerabl e individuals . The predato r let s th e prey know of it s presence ; a t tha t point , difference s i n th e behavio r o f th e potentia l pre y help the predato r discer n which are the most vulnerable . Kruuk 17 describe s how spotted hyena s choose prey among a herd o f gnu: one or several hyenas rush into the herd, then stop and survey the fleeing prey. Only when they find an individual in worse shape than the others d o they—and any other hyenas who may be watching fro m th e side—giv e chase . Even afte r suc h selection , onl y 30 t o 4 0 percen t of suc h chase s ar e successful . Th e prohibitiv e odd s agains t the hyenas ' catching a physicall y fit gnu mak e crysta l clear their interes t i n communicatin g wit h thei r prey. The distanc e fro m whic h th e predato r "announces " itsel f mus t no t b e to o great, though : otherwis e al l the pre y animal s will b e abl e t o escape . Eshe l suggested that th e color s o f leopards an d other spotted cat s reflect this strategic fact. The spots on the leopard' s coat merge into a uniform gray-brown , camouflagin g the leopard , a s long a s the anima l i s to o fa r awa y from it s pre y t o pursu e i t effectively. Bu t the sam e spots becom e distinc t a t closer range , making the leopar d stand ou t an d announcin g th e terribl e menac e t o it s prey , whic h the n panic , showing the leopard whic h individua l it ought t o pursue. Lions, which ar e larger and hun t i n groups , see m t o choos e individua l pre y les s selectively . Thei r col oration, which i s equally cryptic at all distances, enables the m to com e very close and tak e prey by surprise.

COOPERATION BETWEE N PRE Y AND PREDATOR S WITHOUT COMMUNICATIO N A commo n interes t betwee n pre y an d predato r ca n evolv e behavior s tha t serv e that interes t withou t communication . A floc k o f starling s fleein g a rapto r ofte n engage i n ver y comple x maneuvers : shar p turns , quic k shift s o f direction , re -

12 P A R T N E R S I N C O M M U N

I CAT IO N

peated change s o f altitude , an d s o on . Thi s curiou s fligh t pat tern ha s commonl y bee n though t t o b e a communa l defens e system: the errati c turning s o f the flock were believe d t o mak e it mor e difficul t fo r th e rapto r t o targe t a particula r starlin g without collidin g wit h another . Bu t Eshel 18 suggeste d a n addi tional explanation: that the maneuvers, which afte r al l are phys.j^ically strenuous , ma y be a strateg y by which th e starling s for the weakes t o f their numbe r t o fal l behin d a s quickly as possible, to becom e eas y prey for th e raptor . Tha t serve s the interes t of al l othe r member s o f th e flock , wh o thereb y avoi d a long , exhausting chase.

PRECONDITIONS FOR PREY-PREDATOR COMMUNICATION Genuine warnin g behaviors d o exist , of course. Sinc e successfu l adaptation s ar e those that increase successful reproduction , parent s ma y benefit b y warning their offspring, eve n when they put themselve s at risk. Also, as we will see in chapter 6, the babbler who emits the so-called "warning calls" may increase its prestige within its group. Bu t agai n and agai n we find that behavior s tha t ma y seem t o serv e as warnings t o othe r member s o f the species , an d that hav e thu s been considere d altruistic, are in fact bette r understood t o be signals sent to predators t o serve the interest o f the individua l sending them. Communication, again, requires collaboration. I n order fo r communication to take place, both the signaler and the receiver must benefit fro m it . Two conditions are necessary for communication to evolve, and they are especially clear in the case of prey-predato r interaction : th e partie s mus t have a commo n interest , an d th e signals used must be ones that cannot be faked. Communication between prey and predator makes sense whenever some of the prey can escape the predator an d the latter has to target one of the laggards. Under these conditions, a mechanism for such communication will probably evolve. This is not necessaril y a tribute t o th e wisdo m o f either predato r o r prey. Rather, it is testimony t o th e powe r o f natural selectio n t o evolv e in pre y behaviors tha t will help the m escape, and in predators th e ability to heed signals that will spare them fruitless chases . Such systems may evolve even in microorganisms—between bacteria and thei r hosts, as we shall see in chapter 16 . All animals, humans included, ca n behave wisely without bein g awar e of th e wisdom of their action s o r the logic behind them . This wisdom an d this logic are expressed i n a behaviora l mechanis m tha t evolve d throug h natura l selection , a process by which those who adop t effectiv e mechanism s survive, while those who

Prey-Predator Interactions

13

do not perish , o r d o not reproduc e t o the sam e degree. Bot h predato r an d prey are probably unaware of the significance of their behaviors, just as they are unaware of the functioning of their brains, their kidneys, or their muscles. There is no reason to admir e these behavior s an y more than we admire the functionin g of other biological mechanisms—o r any less. All are marvels that deserv e ou r appreciation .

This page intentionally left blank

C H A P T E R

2

COMMUNICATION BETWEEN RIVAL S

THREATS A S A SUBSTITUTE FOR AGGRESSION: ARE HUMANS WORSE THAN BEASTS?

R

arely attack each othe r withou t firs t signalin g their intentions . Most of th e time , the y d o no t attack a t all , an d th e conflic t betwee n the m i s resolved b y a n exchang e o f threats . Thes e signal s com e i n man y forms : singing, aerial displays, electric pulses (in the electri c eel), puffs o f noxious chemicals, posturing . Th e son g o f th e nightingal e advertise s th e bird' s readines s t o defend it s territory and deter s its rivals. Red deer walk side by side and roar. Fish swim in parallel to one another, extending their fins. What about humans? There is a common notion, endorsed b y Konrad Lorenz, that th e tendenc y to escalat e conflict s int o comba t an d th e readines s t o kil l opponents i s unique among humans. 1 But in fac t th e typica l study of animals done at the tim e simply did no t las t long enoug h fo r researcher s t o determin e this. I n most long-term studies observers have recorded conflicts that escalate into fights, which have occasionally resulted i n woundings and eve n death. In our own longterm study of the babblers at Hatzeva—over 20,000 person-hours of observation— there have been eyewitness accounts of some 20 fights that resulted in killings; we have found indirect evidence that several more have occurred. Among humans, too, most conflicts are resolved by threats rather than by actual violence. Thi s i s true in particular on the persona l level, but als o on the interna 15

16 P A R T N E R

S I N COMMUNICATIO N

tional level. Wars ar e terrible, an d partl y for that very reason , mos t conflict s ar e resolve d withou t them. Th e reaso n w e regar d humans a s uniquely aggressive i s the degre e o f attentio n w e an d ou r news medi a pa y to case s that d o escalat e into violence. I f we were t o tur n of f the TV , cancel th e newspaper, and rely on our own observations—as does a zoologist studyin g an animal species—the picture would be very different. W e would conclude , rightly , that humans do not as a rul e kil l o r woun d eac h other . Humans , lik e other animals , resolv e mos t conflicts by communicating—which often include s th e exchang e of threats. Mammals, birds, reptiles , fish , insects—i n fact , al l living creature s tha t com municate in an y way—use threats. Resolvin g a conflict b y threats alon e prevent s the loss of time and energy , and the ris k of injury o r death, that attends an actual battle. I t is easy to understand wha t the winner gain s by using threats rather than fighting; but wh y should threat s make the other party back down ? What i s it that convinces one of the opponents t o give up a bit of food, a potential mate, a territory without even trying to fight? Maynard Smith and Parker2 put forwar d the obviou s argument tha t i f on e i s goin g t o los e anyway , i t i s better t o los e withou t bein g defeated i n a fight. Yet how doe s on e kno w on e is going t o lose? Wha t i s it tha t convinces on e o f the tw o rivals that it s defeat is inevitable, or tha t what it woul d gain by winning is not worth th e cos t it would incu r by fighting?

HANDICAPS AS KEY S FO R RELIABLE THREATS In 197 7 we suggested a solution:3 the threats themselves can be reliable indicator s of eac h rival' s chance s i n a fight . Reliable threats , b y definition , ar e signal s that enable on e rival, the one who is likelier to win a fight, to threaten more effectivel y than th e on e who is likelier t o lose. How ca n a signal work thi s way? The threa t itself mus t increas e th e ris k tha t th e threatene r wil l b e attacked , o r wil l b e a t a disadvantage if attacked; an animal that is genuinely willing to fight and confident of it s abilit y wil l fin d suc h a ris k reasonable , bu t on e lackin g strengt h o r moti vation wil l fin d th e ris k excessiv e an d thu s b e unabl e t o threate n t o th e sam e degree. I n othe r words , fo r a threat t o b e reliable , th e signa l mus t increas e th e danger to the threatener—and an escalation of the threat mus t increase that danger eve n further . There has to be a fundamental relationship betwee n th e particular pattern , or form, o f a reliable signa l and the specifi c message the signal conveys—to be more exact, between th e cos t of that pattern an d the message. What is an honest threat? An honest threat communicates reliably one's ability and willingnes s t o fight. A reliable threa t leaves th e signale r ope n t o attack . This

Communication between Rivals

17

increased risk is acceptable to the honest signaler—th e one who thinks the objec tive is worth a fight, and tha t it ca n win agains t that particular opponent. Such a threatener ha s alread y decided t o fight if its opponent doe s not retreat , an d th e increased likelihood that it will actually have to fight does not deter it. A bluffer— one who tries to gain by threats alone but i s not really willing to back up its threats by fighting—woul d fin d th e increase d likelihood o f being attacke d to o risky . A quick look a t known threa t signals illustrates this point .

THREATS BY APPROACHING Threatening b y approachin g a riva l i s ver y commo n indeed . Approachin g i s a reliable threat because by approaching it s rival, the signaler opens itself to attack. If suc h signals were a pure convention, then an y movement would b e a s likely as any othe r t o tak e on th e meanin g of threat. If the mos t clearl y visible movemen t were the aim, then a movement sideways would have been better than a movement forward or backward . But a movemen t sideway s is less risk y and thus , thoug h clearer, is less reliable. Likewise, if signals were arbitrary conventions, a movement backward would have been just as likely to become a threat as a movement forward. We kno w o f n o case , however, i n whic h a movement awa y fro m one' s riva l has become such a signal. The stanc e o f th e defian t human male is well known to us all: a straight back, the chest thrown out, th e shoulder s back , chi n up . Thi s i s a very inefficient an d risk y postur e i n whic h t o ente r hand-to-hand combat . Th e uptilte d chi n i s ex posed t o blows ; th e erec t bod y make s it difficul t for th e threatener to launch a surprise attack or to change position a t all. It is the exac t opposite o f a boxer's o r wrestler' s stanc e in th e ring : ready to attack or avoid an attack, the boxer keeps his chin down, clos e to his chest, and his body coiled lik e a spring ; he balances on th e ball s of his feet , o n th e alert, ready to seiz e the first opportunity. O f course , boxer s canno t resolv e th e matc h b y threats : both hav e already committed themselves to fight. The threatener , b y contrast , uses precisely this vulnerability to strengthe n his threat. B y standing up straigh t he gives up th e benefi t of a good defensiv e stance and th e option of a surprise attack. In the day s before razors, a man's thrown-out chin presente d anothe r risk : it brought the threatener's beard nearer to his rival and mad e i t easie r for th e latte r to gra b it . B y putting his chi n out , a threatene r shows his confidence that his rival will not dar e or will not be able to grab him by the beard or punch hi m on the chin—and that he, the threatener, is still confident

18

PARTNERS I N COMMUNICATIO N

of winning the fight if the othe r does dare. A version of this signal is familiar fro m a hundre d movie s an d cartoons : th e toug h gu y points to hi s outthrust chi n an d taunts, "C'mon, big shot, lay one on me! "

THREATS BY STRETCHING-DECEIT OR EXPOSURE TO DANGER? Many animal s threaten b y stretchin g thei r bodies . I n suc h cases , th e threatener s do no t sho w thei r weapons—teet h o r claws—t o their opponent s but rathe r expose thei r whol e bodie s t o attack . Threatenin g dogs , a familia r example , stan d side by side o n tiptoe , bodie s stretched an d hai r raised. 4 Suc h a posture i s commonly explaine d a s an attemp t b y the threatene r t o presen t itsel f a s bigger tha n it reall y i s an d thu s t o dete r rivals. 3 Th e stretche d bod y doe s indee d provid e precise informatio n abou t th e threatener' s size , bu t stretc h a s it may , no anima l can becom e bigge r tha n i t reall y is , an d th e smalle r o f th e tw o wil l stil l see m smaller tha n th e other . Stretching is therefore not likel y to make a rival misjudge one' s size . I s i t done , then , t o sho w off one's actual size? Apparently not: eve n rival s wh o know eac h othe r well , suc h a s longtime neighbor s who ar e wel l awar e o f eac h other' s tru e size , threaten b y stretching their bodies. Th e only point in signalin g is to affec t th e other' s behavior , an d a fact that is known already is unlikely to change that behavior. Keepin g it s bod y stretche d expose s th e threatener to risk; why take a risk in order t o convey information that the rival has already? The point o f the stretche d bod y is not t o show size but t o conve y confidence. Key t o an y predictio n abou t th e result s o f a figh t i s th e willingnes s o f a give n individual to inves t i n that fight—in other words, it s motivation. One wh o is not sure the objec t of the fight is worth th e risk of injury will hesitate before exposing its bod y t o danger . Motivatio n change s wit h circumstances—unlik e bod y size , which is easily determined an d doe s not chang e from da y to day. The appearance of anothe r riva l o n th e scene , th e approac h o f a predator , th e soun d o f one' s offspring callin g for help—all can instantly change one's willingness to take risks. Thus, stretchin g one' s bod y i n th e presenc e o f a rival i s a reliable, moment-by moment indicator of one's current willingness to engage in the prospective conflict. Such stretching displays often las t for a good while before coming to a resolution: Glutton-Brock and his colleagues describ e red dee r walking in parallel along their

Communication Between Rivals 1

9

territorial line , bodie s stretched , fo r hal f a n hou r befor e on e o f them decide s t o attack or to withdraw. 6 Like stretching, bristling hair among mammals and the stretched fins of threatening fish are commonly explained a s attempts to appear bigger. Such explanations assume that the rival cannot detec t th e truth. Indeed, fish with stretched fins look much bigger tha n fish that hold their fins closely folded; but sinc e the strength of a fish in a fight depends o n its muscles, not it s fins, this deceit would onl y work if the rival mistakes the fins for muscles. Yet the coloring of the fins makes it unlikely that an y such decei t i s intended: th e patter n an d th e colo r o f the fi n ar e almos t always differen t fro m thos e o f th e body , an d th e stretche d fin s thu s loo k ver y different fro m th e fish's muscular body.' A threatenin g fis h stretchin g it s fin s i s bot h revealin g th e precis e siz e o f it s muscles and exposing its body to attack. By contrast, a fish that is having the worst of a battle bu t i s still willing t o figh t find s refug e i n a corner o f the aquariu m o r in a narrow gap between rocks , fins held tight and read y for instant motion, teet h exposed. I t i s ready to bite it s attacker at the first opportunity an d i s minimizing its rival' s abilit y t o ge t a t it . Th e fis h tha t i s losin g th e battl e canno t affor d t o expend much on showing off; it conveys only its grim willingness to go on fighting. Its positio n i s quit e unlik e th e threatenin g stanc e o f a fish confident tha t i t ca n win, whic h show s of f its self-assuranc e wit h a stretched postur e tha t prevent s i t from springin g into action an d opens i t up t o its rival's attack.

THREATS BY VOCALIZATIO N Morton8 notice d tha t i n vertebrates, threatenin g vocalizations are usually low in frequency, i n contrast to appeasement calls , which are more high-pitched. Morton also observe d tha t large r individuals usuall y have lower voice s than smalle r one s and suggeste d that a low-frequency cal l is intended t o reflec t th e threatener's size. This raise s a number o f questions: Whe n rival s are looking a t each other, why should they use hearing rather than sight to assess each other's size ? If a low voice can frighte n rivals , ho w i s i t tha t animal s hav e no t simpl y evolve d longe r voca l cords, so as to threaten better? And most of all, during conflict, why does the same individual emi t high-pitche d sound s when frightene d rathe r tha n stickin g to th e lowest-pitched sound s it can make? The pitch of the voic e reliabl y discloses th e tensio n o f the signaler' s body. 9 A tense bod y make s a more high-pitche d soun d tha n a relaxe d one . A frightened individual i s tensed t o tak e flight or t o fight back. Onl y on e who is relaxed, no t poised t o tak e instant action , ca n sound a low-pitched, threatenin g note . Suc h an individual disclose s reliabl y that i t doe s no t fea r it s rival ; i t i s no t coile d lik e a tightly wound sprin g an d thu s has exposed itsel f t o a first strike. This—th e cos t

2O P A R T N E R

S I N COMMUNICATIO N

involved i n makin g such a sound i n a rival's presence—i s th e ver y element tha t makes the messag e reliable. Th e threatenin g call s o f two rival s confronting eac h other are true indicators of each one's willingness to risk attack. In a famou s experiment , Davie s and Halliday 10 foun d a correlatio n between th e pitc h o f mal e toads ' croaks an d thei r abilit y t o fen d of f challengers i n th e contes t ove r a female. A mal e toa d find s a femal e ready t o la y an d grip s (amplexes ) her, waiting for her t o discharge her eggs so that he can fertilize them. Rival toads try to kick him off and take his place. In response , th e firs t croaks . Davie s an d Hallida y foun d tha t th e croak s o f th e amplexing toa d correlate d wit h it s bod y size , an d tha t large r toad s displace d smaller ones . The y suggeste d tha t th e pitc h o f the cal l displaye d th e siz e of th e threatening toad an d thereby its potential t o win the fight. But wh y should other toad s liste n t o th e pitc h o f its croak s to asses s the size of a toad that is right in front o f them? We suggest that the croa k displays not size but rathe r the toad's self-confidence. If a toad riding a female with his rival present fears bein g kicke d off , he grab s he r tightly—renderin g himself unabl e t o emi t a low, relaxe d croak . Only on e certai n o f his abilit y to sta y o n no matte r what, or confident tha t hi s rival will not dar e challenge him, can afford t o rela x enough t o emit a low-pitched croak . Obviously , thi s confidenc e ha s somethin g t o d o wit h body size : smalle r toad s woul d naturall y fear large r ones . Bu t i f pitc h indicate s only body size , each toad's croaks would sta y at the sam e pitch n o matte r whic h rival i t wa s confronting ; if pitc h reflect s confidenc e an d motivatio n rathe r tha n size, though, then the same toad would emit a low, relaxed croak when confronting a small rival and a higher, tenser one when confrontin g one larger than himself. It would b e simpl e and interesting to test thi s hypothesis . Among humans, too, postur e affect s th e abilit y to produc e a persuasive vocal threat. Actor s ca n sound convincingl y threatening eve n thoug h the y have no intention a t all of actually fighting with a n onstage "rival." We were told by Nissan Nativ, actin g teacher and hea d o f a theater school i n Tel Aviv, that a n actor does not have to worry about how to sound menacing: once he or she assumes the body posture o f a confident , aggressive individual, th e correc t tona l qualit y comes of itself. Does this mean that actors would make perfect bluffers i n the real world? No t at all . Th e acto r i s no t actuall y in dange r o n stage . I n rea l life , facin g an actual enemy who migh t attac k at any moment, eve n the bes t acto r woul d find it jus t about impossibl e t o bluf f an d threate n wit h a relaxed , low-pitche d voice . An y tension i n hi s body—any readines s to figh t o r flee instantly—would reflec t itsel f in hi s vocalization. Vocal threat s ar e als o use d fro m a distance , when attac k is not ye t imminent

Communication Between Rivals

21

and tensio n versu s relaxatio n i s irrelevant. In suc h case s the voca l signa l conveys other qualities . On e o f them w e have learned , a s we have s o many things, fro m the babblers. " Th e long-distanc e threa t cal l o f babblers is a sequence o f several loud syllable s clearl y separate d b y precise intervals. I t give s a good indicatio n of the motivation o f the calle r and its willingness t o fight. When a group o f babblers advances toward a border clas h with anothe r group, one can easily distinguish the calls of the front-runner s from th e call s of those lagging behind. Th e one s in th e front emi t rhythmi c and clearl y defined syllables; those i n the rear , who see m less eager to fight, emit softer sounds , with less precise intervals. Once the rival groups engage, th e call s o f those actuall y fightin g becom e metallic and highe r i n pitch , and the intervals between th e syllables are less precise. One can "hear" the tension in their bodies . Why d o th e babbler s use precisely space d syllable s onl y when the y ar e eager to fight? In order t o emit rhythmic, regularly spaced, an d clearly defined syllables, one ha s t o concentrat e o n th e ac t o f calling . An y distraction—suc h a s a glance sideways—distorts bot h th e rhyth m an d th e precisio n o f sound ; a n individua l cannot a t one an d th e sam e time collec t informatio n an d concentrat e o n performance. A cal l compose d o f precise , rhythmi c syllable s testifie s tha t th e calle r is deliberately deprivin g itself of information, which mean s either tha t it is very sure of itself or that i t is very motivated t o attack , or both. 12 A human being in control of a situation, too , tend s t o issu e threat s i n a n ordered, rhythmi c cadence . Th e even bea t increase s the effectivenes s of the threat , sinc e it show s confidenc e and demonstrates th e threatener's abilit y to concentrate on the threat regardless of the fact tha t h e or sh e may thus be deprive d o f crucial information.

OTHER THREATS Staring a t a rival also impairs one's abilit y to collec t in formation; the risk is small for a dominant individual, who is not likely t o be attacke d from behind in its own territory or group, but i t may be too high for one less in control of its surroundings. Hence , starin g is perceived a s a threat. Th e individua l tha t stare s is dangerous eithe r be cause i t i s dominan t withi n it s grou p o r becaus e i t i s so highly motivated t o attack that it does no t see k to gather all the informatio n it can. Contempt ca n als o be use d t o dete r rivals . A Holly wood sherif f enterin g a rustler's hideout with arms folded and pistol s holstere d i s showing hi s confidence : if he were t o ente r gu n in hand, it migh t increase the chance s o f an immediate victory, but b y displaying his confidence, he may get his man without the need for any gunplay. Likewis e one who

22

PARTNERS I N COMMUNICATIO N

turns his back o n a rival shows his contempt—and his confidence ; a dog turning aside t o urinat e i n th e middl e o f a confrontatio n show s i t i s confiden t tha t hi s opponent wil l no t dar e attack him—an d that i f attacked, he will stil l win, eve n when caugh t with one leg in the air. Still another wa y to issu e a threat i s to sho w the degre e of harm one is willing to suffe r i n orde r t o win . When ant s us e formi c acid an d bee s us e chemical s in their fighting , the y creat e a n environmen t hostil e t o bot h signale r an d rival . I t makes sense that one able and willing to endure the noxious chemicals displays in a reliable way its ability and commitmen t to the fight.

SOCIAL HIERARCHIES AND DUEL S BETWEEN EQUALS Most conflicts , i n fact, occu r between rival s that hav e clashed wit h eac h othe r i n the past. In such cases the previous loser tends to submit without fighting. In social systems where th e sam e individuals mee t repeatedly , these historie s o f previou s conflict produc e a socia l hierarchy— a peckin g order—i n whic h eac h indivi dual know s it s rank fro m experienc e and submit s t o higher-rankin g individuals . Often th e mer e hin t o f a threa t b y a dominan t individua l i s enough t o dete r a subordinate.13 When bot h rival s are the same size an d equall y stron g and willing to fight, one rival cannot threaten reliably more than the other, and there is no way to win without fighting. Darling,14 who studied re d deer, mentions that actual physical clashe s occu r onl y betwee n male s o f simila r size ; i n other cases , one of the contestant s withdraws following an exchange o f threats. Barret e an d Vandal 13 observe d 131 4 confrontations betwee n reindeer . Onl y i n six of these di d the exchang e of threats develo p int o a n actual fight; in all six cases , the male s were very evenly matched i n siz e and strength. Bu t eve n betwee n rival s wh o ar e equally stron g and motivated, threats still have their use: any injury to one of the contestant s in the fight changes the balance of strength, opening th e way to another roun d o f exchange d threats , whic h ma y well persuad e on e tha t h e n o longer has a chance of winning and shoul d withdraw.

Communication Between Rivals 2

3

CAN THREA T SIGNALS EVOLVE FOR THE GOOD OF THE GROUP? THE DRAWBACKS OF "GROUP SELECTION" According t o th e theor y of "grou p selection, " groups tha t us e threats instea d of aggression have an advantage over and eventually supplant groups whose members always figh t eac h othe r t o th e bitte r end. 16 There was a major discussio n o f this idea i n th e 1960s , whic h resulte d i n th e practicall y unanimou s conclusio n tha t group selectio n canno t explain the evolution o f traits, except perhap s unde r very special circumstances. 17 Ye t explanation s base d o n group selectio n ar e still common in popular literature . According t o grou p selectio n theory , a population—a grou p o f individuals— that solve s conflict s b y threat s rathe r tha n b y fighting has th e advantag e over a population o f aggressors who wound an d kill each other. The efficien t grou p supplants the less efficient, mor e self-destructive one. According to this theory, selection ca n ac t o n th e grou p a s a unit—thu s th e nam e "grou p selection." 18 Thi s model—which ha s no t bee n supporte d b y observation—suggest s tha t eac h individual submit s t o threat s fro m other grou p member s because fighting would harm the group. By this logic, there is no need for any connection between a threat and th e actua l ability of the threatener . An y gesture can be a threat, a s long a s it is so accepted by the group . The problem is that if such signals are only conventions and bear no connectio n to the threatener' s rea l ability , then a mutant who doe s not obe y the "rules" — who fight s rathe r tha n givin g u p somethin g importan t jus t becaus e i t i s threat ened—will reproduc e bette r tha n on e who submits . Its descendant s wil l inheri t the trai t o f fightin g rathe r tha n obeyin g th e rules , an d the y to o wil l reproduc e better, unti l the y form a significant par t o f the group . Thu s ove r time the grou p will lose the share d advantage of using threats rathe r than fighting. Some argu e that society will punis h those who break th e rules . Thi s onl y expands th e problem , whic h becomes: Wh o i s this "society " that will impose punishment? Ho w wil l th e punishin g work ? Th e exception , o f course , i s when th e individuals metin g out punishmen t receiv e direct compensatio n fo r their efforts , like police and judges in human society. 19 A group-selectionis t migh t retor t tha t a group i n which combativ e dissident s come to prevail will ultimately die out because of excessive aggression. But before that happens, wha t is to stop som e of the rule-breakers from movin g into a "non aggressive" grou p o f their species—wher e they would agai n reproduce more successfully than other grou p members until that group, too, loses the ability to avoid violence by using threats? A trait that harms or deprives the individual who bears it an d onl y benefits "th e group " canno t surviv e natural selection , eve n if threatening rather than fighting is "good for the group." In contrast , our model, th e Handicap Principle , show s how the pattern o f the

24 PARTNER

S I N COMMUNICATIO N

signal itself ca n prevent cheatin g an d ca n show the rea l motivation an d abilit y of the threatener i n a reliable way. The patterns of threat signals—like those of other signals—are neither random no r arbitrary. Th e need for reliability links the signa l directly t o th e messag e i t convey s an d guarantee s tha t th e cos t o f th e signa l i s reasonable fo r a n hones t signale r bu t prohibitiv e fo r a cheater. I n th e cas e of a threat, the cost that the honest signaler i s willing t o undertake, and that a cheater cannot afford , i s the ris k that the signa l entails. Rather than being a mere convention, th e signa l itself provide s th e genuin e informatio n tha t i s needed t o resolv e the conflict. Thus there is an optimal signal fo r every threat-—and indeed for every particular message . The Handica p Principl e no t onl y assert s tha t thi s i s so bu t explains why.

C H A P T E R3

MATE SELECTION

S

exual display s often attai n giganti c dimensions an d tak e bizarre forms. Th peacock's tail grow s int o the well-know n large , colorfu l fan ; peacocks pre senting themselve s t o peahen s spen d tim e an d energ y holdin g thei r tail s spread ope n an d uprigh t an d vibratin g them rapidly . Male s o f many other bir d species gro w long tail feathers for sexual displays. Relative to their size, some male pheasants and birds of paradise have tails almost as large as the peacock's, an d th e males of some specie s of songbird, such a s the viduas o f Africa, have tails that are proportionately eve n longer . Male songbirds sin g to attrac t mates; some o f them inves t most o f the da y in singing. Others, includin g man y larks, sing during strenuou s fligh t displays . Male sage grouse an d mal e ruff s danc e in special arena s for man y days to compet e fo r the favor s o f females. Blackcock s vocaliz e an d displa y thei r tail s a s they dance , broadcasting thei r messag e by means of several modalities a t once . Sexual displa y i s no t uniqu e t o birds . Mammals , reptiles , amphibians , fish, insects—all invest in sexual displays, and in each of these group s there are species in which the investment attains striking, even fantastic dimensions. Crickets , grasshoppers, an d cicadas sing for many hours to attract partners. Fireflies display with 25

26 PARTNER

S I N COMMUNICATIO N

flashes of light. Moths and many mammals emit scents. Even one-called organisms like yeast and alga e emit pheromones t o attrac t mates. Sometimes bodil y growth s ar e mean s o f sexua l display: newts gro w a finlike nuptial cres t alon g thei r backs; the whit e pelican grow s a bump o f flesh between its eyes . The male s of many bird specie s suc h as shrikes, terns, an d gulls , as well as some insects, court their mates with nuptial gift s o f food, leaves, or twigs. Male bowerbirds buil d comple x bower s an d decorat e the m wit h shell s an d flowers, bones, insec t skeletons , an d colorfu l fruits t o attrac t females ; som e male fish and crabs buil d san d castle s o n rock s expose d t o th e waves , castle s that hav e to b e continually rebuilt. The extravagan t dimensions o f sex ual displays make them see m like crazy fashio n shows . Bu t hav e the y ru n be yon d reasonabl e evolutionar y control , a s s i commonly assumed? Or do the dimensions o f th e display s an d th e spe cific form s they take serve a purpose?

THE CONFLICT INHERENT IN COURTSHIP Williams1 emphasized the competitive aspect of courtship: males and females have conflicting interests. Each wants the highest-quality mate it can get—the mate that can best improve its offspring's gene s and, depending o n the gender an d species , best rais e those young . Williams therefor e suggested that durin g courtshi p male s and females ca n be seen as opponents. Th e male, like a good salesman, does whatever he ca n to impress females, while the goa l of the female , like that of a shrewd customer, is to check the merchandise and accept onl y proven quality . Of course , females also advertise themselve s to males , with the sam e ensuing conflict o f interests.2 Still, the possibilities open to them are different. Th e number of a female's offsprin g i s limited mos t b y he r ow n capacit y to produc e egg s o r undergo pregnancies , while a male's breedin g success depends mor e on the number a s well a s the qualit y of the female s he ca n persuade t o breed wit h him. Fo r the sak e of convenience, an d sinc e males as a rule invest more in advertising tha n females, we will discuss the issue mostly in terms of males as presenters and females as choosers . How can males prove themselves to be superior? Williams did not address that question. Bu t in the previous chapter s we have seen how reliable communication develops between enemie s of different specie s and between rival s within a species. The sam e logi c applie s t o sexua l display. 3 Her e to o th e conflic t o f interests — between mal e and female—i s ofte n resolve d b y communication , which depend s on the evolutio n o f reliable signals.

Mate Selection 1

7

Signals ar e reliabl e i f a cheater canno t gai n by using them—if th e investmen t in the signa l is a reasonable on e fo r a truthful suitor t o make, but prohibitiv e o r unprofitable fo r a cheater. The more the suitor stand s to gain, and the bigger th e loss t o on e wh o accept s a fals e suitor , th e mor e th e signale r mus t inves t i n th e signal in order t o reliabl y demonstrate his superiority. We assum e that th e specifi c investmen t a signaler makes is directly linked t o the messag e of the signal . Male rivals are only interested i n the fightin g abilit y of their adversaries ; predators ar e concerned onl y with their prey's ability to escape . Individuals looking for mates, however, are interested i n a wide range of qualities. What d o animal s loo k fo r i n a mate ? Tha t depends . Wher e bot h mal e an d female tak e care of the young , the ideal mate is not jus t o f superior geneti c qual ity—quality tha t wil l b e passe d o n t o th e offspring—bu t als o skille d enoug h t o provide fo r its family and committed to parenting the offspring effectively . I n such species, th e mal e can commi t t o onl y one female at a time—or to a t most a few. A female in such a species may well have to compromise on quality in order t o get a mal e willing to commi t to her . A t the othe r en d o f the spectru m ar e species in which male s have no parenta l involvemen t whatsoever. I n thes e specie s a femal e can concentrate o n finding the most superior sper m donor she can, even if she has to shar e hi s favor s wit h man y other females . In suc h specie s a few outstandin g males get most of the females, while young and low-quality males do not copulat e at all. Courtship signals thu s conve y different message s i n different species . As with all signals, we expect to find a direct relationship between the investment in sexual advertising an d th e specifi c information th e courtshi p signa l convey s t o th e se lecting party.

COURTSHIP HANDICAPS AND THE INFORMATION THEY CONVEY deeding Ability Courtship feeding—a s i n terns , shrikes , an d grea t tits—reliabl y demonstrate s both th e feeder' s abilit y to giv e up a good portio n o f the foo d he collects an d his interest i n the particula r female . Th e femal e gain s both th e actua l food provide d and th e knowledg e tha t th e mal e woul d b e a goo d provide r fo r he r offspring . The more foo d the male brings to the female, the more reliabl e th e message: that he is a good collecto r o f food. Indeed, Nisbet 4 found that male terns who provide more food durin g courtship als o feed their offsprin g better . Courtshi p feedin g is a muc h bette r indicato r o f th e male' s qualit y as a hunte r tha n a mer e sho w of strength woul d be . Th e effor t require d t o fee d th e femal e prevents pretens e b y

28 PARTNER

S I N COMMUNICATIO N

a mal e who ca n barel y fee d himself . It als o prevents male s from courtin g many females simultaneously . White pelicans , bot h mal e an d female , gro w fleshy bumps between their eyes when they are ready to breed. The bump interferes with the pelican's ability to see the area around the tip of its bill. In orde r t o catc h prey, a pelican with a bump ha s to remember where it last saw its prey and project the prey's likely movements. An inexperienced or inept pelican would not be able to do so. A pelican that can fish and maintain itself in spite of the handicap of its bump is reliably demonstrating its expertise in fishing. Later, when th e pelicans hav e t o fee d thei r brood o f fou r o r fiv e demandin g young , the bum p shrinks an d they are able again to hunt mor e efficiently . Singing can also demonstrate the ability to provide. The time invested in singing cannot be use d fo r foraging. A courting male who handicaps himsel f b y singin g continuously provide s evidenc e tha t h e needs les s tim e to forage , eithe r becaus e he is very efficient o r because his territory is very rich. Wilhelm and his colleagues3 studied the effec t o f supplementary feedin g o n the singin g o f yellow-bellied sunbirds. They found that males who were not given insects did not sing, while thos e who receive d insect s an d suga r wate r san g ofte n an d a t length . Tim e spen t i n sentinel activity, or in dancing displays, can also indicate expertise in finding food, especially when the "waste " of time comes early in the mornin g after a long, cold night without food.

Superterritories O'Donald6 ha s suggeste d tha t th e siz e o f a n animal' s territor y ca n serv e a s a n advertisement. Male s an d female s o f man y species protec t a territory an d chas e others away from it. Often the territory is far larger tha n is needed to provide food or shelte r fo r a pair o f mates and thei r offspring. Som e suggest that these bigge r territories preven t overexploitatio n o f food resources , and thu s a richer resourc e remains t o sustai n th e populatio n a s a whole i n th e future. 7 Bu t thi s argumen t depends o n th e mechanis m o f group selection , which, a s we have seen , i s questionable. Whe n som e males protect territorie s tha t ar e larger than the y need, th e population a s a whole might gain from th e preservatio n o f future foo d resources; but th e males who hold smaller territories would gai n the most, since they would enjoy th e resource s preserve d b y earlier resident s withou t squanderin g thei r own efforts protectin g large r territorie s tha n the y need; the y would therefor e be abl e to devot e mor e o f their energ y t o reproduction . Thus, ove r th e generations , th e tendency t o hol d a larger territor y than one needs woul d disappear . Why , then,

Mate Selection 2

9

should a n individua l spen d tim e an d energ y an d eve n figh t t o protec t a bigger territory than it needs? The ric h buil d gran d home s an d mansion s no t becaus e the y nee d the m a s shelter fo r themselves or their childre n bu t i n order t o proclaim their status . The cost o f building an d maintainin g a palatial mansion advertises the owner's wealt h to his competitors an d colleagues . Just s o among many species of animals: a large territory proves the male's superiority and attracts a good mate. Indeed, observations sho w tha t female s settle first in th e large r an d riche r territories ; ver y small territories ma y attract no female s a t all. These superterritorie s may indeed preserve the resources available to the whole populatio n o f th e species ; bu t this is only a by-product and doe s no t play any direct role in the evolution of the trait of establishing superterritories. A good analogy is provided by the vast hunting lands maintained by European nobleme n i n pas t centuries , while throng s o f peasants starved . These territorie s were used fo r spor t huntin g and wer e strictl y guarded agains t poachers. Suc h superterritories serve d to show off thei r owners ' wealth , authority , and powe r t o thei r peers, who m the y invited to th e hunt . Man y of these sam e hunting ground s b y now have become national parks, whic h preserv e fo r peopl e o f ou r ow n tim e th e animal s and th e natural forests o f Europe—bu t thi s i s hardl y th e reaso n th e nobleme n establishe d an d maintained their hunting estates generations ago. In just the same way, males hold large territories because by doing so they deter their rivals and attract good females. Large territories may indeed preserv e the species ' resource s of food against overexploitation—but that is a side effect , no t th e superterritories ' actual purpose.

Courtship Vocalization Many animals vocalize during courtship . Lions , tigers, and dee r roar ; cicadas and crickets chirp ; bird s sing . Courtshi p call s ca n b e dangerous : Ryan an d hi s colleagues8 showe d tha t frog-eatin g bats locat e frog s b y th e amphibians ' courtshi p calls. Onl y a mal e frog tha t ca n successfull y avoi d bats despit e disclosin g its location t o the m ca n affor d t o croa k much . Courtshi p call s ca n als o be ver y demanding: Glutton-Brock and Albon 9 found that red deer are often exhauste d afte r a roaring contest with rivals. Only a strong, well-muscled individua l ca n roar loudly for a long time. The detail s o f a call , it s tempo , an d th e numbe r o f syllable s in a phrase ca n demonstrate th e caller' s quality . The son g o f th e grea t tit i s a series o f precisely spaced syllables . Lambrechts and Dhondt 10 found a positive correlation betwee n

3O P A R T N E R

S I N COMMUNICATIO N

the numbe r o f syllable s i n a phras e an d th e rhythmi c precisio n o f th e las t fe w syllables on the one hand, and the reproductive success of the singer on the other . The abilit y to maintain both th e temp o an d the patter n o f syllables at the en d of a lon g phras e woul d see m to b e a good indicato r o f quality . As was discussed in chapter 2 , th e precis e performance o f a cal l demands concentration . A less abl e male would probabl y fin d i t difficul t t o concentrat e fo r long ; afte r all , th e son g reveals his location, and he must keep a n eye out for approaching rivals and predators. Th e singin g conveys his confidenc e or lac k of it, informatio n important t o a female who must decide whether t o accep t him a s her mate. 11

Colors The adul t males o f many species o f birds ar e far more colorfu l tha n female s an d young males: examples include peacocks, ducks , birds o f paradise, an d sunbirds . Colorful plumag e attracts rivals and predator s an d thu s serves as a reliable signal of quality : only males of high qualit y can risk advertising their location. Conspic uous coloratio n als o emphasixe s the exac t shape , posture , an d movement s o f its bearer. A high-quality individua l "wears " brigh t colorin g well ; o n a low-quality one the sam e coloring accentuate s imperfections. 12

Scents (Pheromones) Scents also serve to attract mates. Studies have found that some female insects can identify a dominant mal e by its scent alone. 13 Many male mammals, too, use scent to attrac t females and dete r rivals . We kno w a great dea l abou t th e chemistr y of pheromones—chemicals that are produced by one individual in order to influence the action s o f others ; bu t ver y few scientist s hav e trie d t o explai n th e adaptiv e significance o f specific chemicals. It woul d b e fascinatin g t o discove r wha t infor mation the pheromone s o f each specie s provide abou t thei r producers . The mai n componen t o f the pheromon e secrete d b y male arcteid an d danai d butterflies i s a derivative of a n alkaloid—a stron g poiso n produce d by plants fo r their ow n protection. 14 Arctei d larva e ca n metaboliz e th e poiso n an d thu s tak e advantage of a food sourc e that is not availabl e to most other animals ; the poiso n that the y take into thei r syste m then help s protec t the m fro m predators . A male arcteid butterfl y secretin g thi s pheromon e testifie s tha t a s a larva he wa s able t o feed o n the poisonous plants ; the concentratio n of poison i n the pheromone demonstrates th e male' s relativ e physiological abilit y to dea l with the poison . Danai d butterflies secrete a similar pheromone, derive d from plants that they eat as adults. Eisner an d Meinwal d furthe r propose d tha t th e pheromon e coul d functio n a s a chemical yardstick by which females gauge the poison load their suitors carry . The alkaloid i n the pheromon e ma y also sho w tha t th e mal e will probably b e abl e to

Mate Selection 3

1

pass on a good quantity of alkaloid to the female during copulation, to protect he r and their offsprin g agains t predators . The monarch butterfly is one species in the danaid family that does not use these alkaloid derivative s in its courtship. It is a migratory species that winters every year in California, Mexico, and Florida. Thi s species instead courts its females by means of aggressiv e displays. Male s bounc e o n passin g female s an d thro w the m t o th e ground, wher e the y copulate. 15 Th e abilit y of the mal e to us e force i n copulatio n may attest to his strength an d stamina, which are necessary for long migrations.

Artifacts and Constructions A cichlid fish of Lake Malawi in Africa gathers a pile of sand and forms a depression in it for the female to lay eggs in. This nest is not used to raise young: both sand and eggs get washed awa y by the waves in the lake. 16 In ou r opinion , i t is precisely the difficulty involve d in gathering a pile of sand again and agai n in a wave-swept area that enables th e male to show off his ability reliably. Crabs who live in the tidal area also build san d castles, which have to be rebuilt a t every low tide.17 As mentioned earlier , courtship display s are especially elaborate in species where the males do not take car e o f offsprin g an d ar e fre e t o concentrat e their effort s o n attractin g female s al l seaso n long . The to p performer s get most o f the females , while most males do not copulat e at all. The most famous of thes e specie s includ e peacocks , ruffs , grouse , manakins, birds of paradise, an d bowerbirds . Male bowerbirds, a family found onl y i n Aus tralia and Papua, spend much of their time building bowers o f twig s o n th e ground . Thes e structure s have no utilitarian purpose; thei r only function is to serve as stages on which males perform courtshi p displays . Th e bowers o f each specie s hav e their ow n structur e and characteristi c decoration. 18 One specie s build s a platform of twigs a meter in diameter; on the platform , tw o parallel rows o f standing twigs, each half a meter long, form a corridor i n which copulation take s place. The male of another specie s builds a mossy wall abou t fou r inche s hig h aroun d a courtyard, a t th e cente r of which, a s a rule, is a sapling; th e bir d cover s a meter o f the sapling' s trun k wit h woven twigs . Still anothe r specie s cover s th e courtyard , makin g a hut wit h onl y one smal l entrance . Th e builde r adorn s hi s bowe r wit h rar e ornaments , suc h as the feather s of birds o f paradise, or with fres h flowers , which mus t be continuall y replaced, showing off his ability to find decorations. In fact , the two highly dec orative feathers on the head of the king-of-Saxony bird of paradise are valued both by bowerbirds an d by tribesmen o f Papua Ne w Guinea. 19 The female visits the bower several times and carefully examines both structure

32 P A R T N E R

S I N COMMUNICATIO N

and builder befor e deciding whether to copulate wit h the hopeful candidate. The female copulates onl y once before egg-laying. She may well visit a bower, examin e it, watch the male perform hi s dances, even take part in the dancing and court the male befor e decidin g t o pas s him up . I t seem s tha t th e decoratio n o f the bowe r plays a large part in the female' s choice o f a father for he r offspring . The numbe r o f decoration s i s importan t t o female s too . The y prefe r male s whose bowers are richly decorated. Male s steal decorations from neighboring bow ers. Borgi a trie d addin g individuall y marked rar e an d sought-afte r ornament s t o one se t o f bowers , the n t o another ; h e foun d tha t n o matte r whic h bower s h e added the ornaments to, they always wound up in the bowers of the most successful males.20 To be attractive to females, a male bowerbird has to spend considerable time an d talen t buildin g hi s bower , collectin g decorations , placin g them , an d guarding them agains t hi s neighbors, not to mention performing o n the stag e he has built. He thus proves that he is stronger and more energetic than his neighbors, that h e ca n fee d himsel f adequatel y an d i s stil l abl e t o buil d an d decorat e hi s elaborate bower , guar d it agains t competitors, an d raid their bowers. 21 When w e visited Borgi a in Australia , we sa w some bower s o f the grea t bow erbird. Th e platfor m a t th e entranc e t o th e avenu e include d fla t stone s fro m a riverbed, bleached bone s (mostl y vertebrae), broken glass—mos t of it green—and colorful bit s o f foi l an d cardboard . Borgia observe d tha t th e arrangemen t o f th e decorations wa s not random . Th e gree n bit s o f glass were consistentl y placed by the norther n entry , whic h i n th e souther n hemispher e face s th e sun . Th e glas s glistens i n th e sunlight , an d th e mal e display s his viole t plume s agains t a shiny color-coordinated backdrop .

Combinations of Signals In mos t cases , severa l signal s of differen t kind s ar e use d i n courtship . I n birds , special feathers , brigh t colors , singin g an d calling , dancing , and gift-giving—th e last thre e of which deman d tim e and energy—al l pla y their part . Eac h modalit y brings out a particular qualit y of the male; the female can then use several criteria to asses s the male . Let's tak e the displa y of the peacock a s an example. Th e mal e holds hi s tail uprigh t and spread out—which demands considerable effort . From tim e to tim e he shake s his tail vigorously; this requires ye t more effor t an d produce s a remarkabl e rattle. Th e "eye " pattern s o n th e peacock' s tail , the glisten of his feathers, the crow n o n his head, all add up to a symphony of shape, color, pattern, movement and sound— a performanc e tha t i s announced wit h periodic roars . Each aspec t of the displa y seems to convey specific, reliable information about a particular featur e o f th e male . Th e lon g tai l feather s ar e grow n ove r a period o f

Mate Selection 3

3

several months, during a time of the year when food is scarce. Unhealthy birds arrest the process , s o a male who display s a set of perfect tai l feathers advertises that h e has been i n good health an d has managed to find food even during molt season . The long , heavy, brightly colored tai l als o attest s to th e owner' s strengt h an d skill, for he has succeeded in avoiding predator s despite such a burden. By holding his tail upright an d shaking it, the peacock proves his stamina, and his roars show he is not afrai d t o disclos e hi s location t o rival s and predators. The perfection of the tail' s pattern testifie s tha t th e peacock' s developmen t whil e the tail grew was perfectly coordinated , a s we shall see in chapter 8 . Each of these criteria seems to be minutel y scrutinized b y females: Petrie an d he r colleague s foun d tha t takin g out eve n five of the hundred an d fifty or so feathers in a peacock's tail reduces his ability to attract females to his dancing arena. 22

Long Tails Moller23 investigated the long outer tai l feathers of barn swallows, small songbirds that catc h thei r foo d i n flight . Bar n swallow s hav e long , forke d tails ; th e oute r feathers ar e longer i n adul t male s than i n females or in young males. When Molle r adde d extra pieces of tail feather to the tails of some males an d shortene d th e tai l feather s o f others, he discovered that thos e with th e longer tails, whether natura l or artificially enhanced , found mate s more easil y than thos e with th e shorter or shortene d tails , and tha t the y got to copulat e wit h additiona l female s a s wel l (extrapair copulation) . Bu t th e inadverten t cheaters—the male s wit h artificiall y length ened tails—pai d a heavy price. The added length apparently impaired thei r ability to fly. They could not hunt large insects, and their physical condition deteriorated : they did not molt well afte r th e breeding season , and none of them returne d from their winter migratio n the next spring , while many of the other male s did retur n to breed i n the same area. Smith and Montgomerie 24 repeated Moller's experiments and found that males with eithe r naturall y long o r artificiall y lengthene d tail s found mates earlie r an d started breeding earlie r tha n othe r swallows . Bu t when the y teste d fo r paternit y among the nestling s by DNA fingerprinting, they found that only half of the nestlings i n the nest s o f the inadverten t "cheaters"—th e males with enhanced tails—were i n fact th e offsprin g o f those males , compared wit h 9 5 percent i n the nest s of male s whose tail s were eithe r naturall y long o r artificiall y shortened. 25 Smit h and Montgomerie suggested tha t the glued-on tails were too much of a burden for males who were not fit to carry them, and that suc h males could not prevent their females fro m consortin g with other males.

34 P A R T N E R

S I N COMMUNICATIO N

The experiments o f Moller and of Smith and Montgomerie with barn swallows, and o f Evans an d Thomas 26 wit h th e scarlet-tufte d malachite sunbird , sho w th e price paid b y a bird wh o carrie s a longer tai l tha n h e ca n handle. Thu s it woul d seem tha t longe r tails , like the bar n swallow's , th e peacock's , an d thos e of many other birds, are reliable indicators o f the agilit y of strong, experience d bird s and are attractive to female s for precisely that reason .

Movements and Dances The displa y flights and dance s of male birds usuall y involve movements tha t ar e not commo n i n their everyday lives. Turtledoves walk when the y forage, bu t the y hop aroun d the female they are courting. Falcons fla p their wings slowly in regular flight, bu t th e male flutters hi s quickly in aerial courtship displays. The tropicbird even flie s backward . In man y species o f songbirds th e mal e takes o n th e burde n of singing while in flight . In som e specie s whos e males take n o par t i n parenting , male s spen d a great deal of effort for many hours a day, many days in a row, on courtship dances. The females visi t th e dancin g arenas , called leks, wher e the y watch the performance s and choos e fathers for their offspring ; ver y few of the dancer s are chosen. A small number o f outstanding males get most of the females, while young and low-quality males do not ge t a chance to copulate . It ca n b e difficul t fo r huma n observer s t o identif y th e criteri a b y which th e females select their mates—after all , humans often do not know enough to appreciate the difficultie s involve d i n the particula r dance. Gibson and his colleagues 27 showed tha t i n sag e grous e th e bes t indicato r o f a successfu l mal e i s a certai n vocalization and a certain pause within it, which accompany a particular movement in th e dance . Evidently, only the mos t superio r male s can achieve this particular combination within th e strenuou s choreography of the dance . As in human gymnastics competitions, in which most o f the move s are part of a specific repertoir e and ar e performed within a highly circumscribed framework, this very standardization i s what enable s competitor s t o best display subtle advantage s and demon strate the abilit y to execute difficult combination s of movements.

LEKS: CONGREGATING FO R DISPLA Y In mos t specie s male s chas e their rival s as far awa y as the y possibly can , but i n most lekkin g specie s male s congregate, eac h i n his ow n miniterritory within th e lek, in order to display. Communal leks are found among insects, fish, amphibians, birds, an d mammals , but onl y among species whos e males d o no t participat e in parenting. Amon g ruffs , man y species o f grouse , an d severa l specie s o f bir d of

Mate Selection 3

5

paradise, males court an d displa y in leks; among ruffs an d sag e grouse, hundred s of birds may congregate in one lek. Lekking males do not have equal success in reproduction. Mos t males ultimately do not copulate a t all; two or three males—who are usually in the center of the lek— may d o ove r 9 0 percent o f the copulating . Why , then , d o th e othe r male s com e together where the y hav e little to gain ? Probably because elsewhere thei r chance s are smaller still. Petrie an d her colleagues note that on days when many females visit the top male, other males may attract some females too.28 Hoglund and his colleague s found tha t the more males in a lek of ruffs, th e more females there are in proportion to males.29 Females, afte r all, find it easier to choose a father for their offspring when they ca n compar e males side by side . Mor e tha n on e femal e ca n affor d t o choos e the same male; they do not mind sharing him with others because they will not have his help in raising their young anyway. Under such circumstances, it is in the interest of female s to b e abl e t o compar e a s many males a s possible, as easily a s possible. Gibson an d Hoglund foun d that young female grouse tend to prefer males selected by older, experience d females : they watc h th e experience d ones , the n choos e the same males. 30 If fo r thes e reasons females ignore small congregations of males an d insist on the greater choice afforded b y larger ones , males are forced to compromis e with other males and congregate . Manakins ar e smal l birds o f th e America n tropic s that displa y in leks.31 Among them there are males who dance in a miniterritory within the lek in groups of two to five. Only th e to p mal e of each group, wh o i s olde r than the others, copulates with the females who visit the courting arena . Th e other s cooperat e wit h hi m i n th e dance without immediate reward; but som e of them can benefit i n the long run. After hi s death, it seems that his top helper , wh o is usually at least six years old, inherits the top male's miniterritory. 32 Female manakins seem to prefer the group performers. No wonder : th e to p mal e in a group aren a demonstrates both his dancing ability and the deference that other males pay him, which makes hi m al l the mor e attractiv e to females . The othe r member s o f the grou p gain a s well: the y ca n both practic e thei r dancin g an d increas e their chance s to inherit a good miniterritory.

POLYMORPHIC SPECIE S AND MALE S THAT MIMIC FEMALE S In leks of ruffs, one finds both dark-collared and light-collared individuals. 33 Darkruffed male s fight for an d acquir e miniterritories withi n th e lek, while males with

3 66 P A R T N E R

S I N COMMUNICATIO N

light ruffs mov e frequently from on e territory to another , sharing i t wit h th e dark-collare d mal e wh o preside s i n each. The light-collare d ruff s displa y subservience towar d the dominant dark-collare d one s by crouching with their beaks touching the ground when they approach, but even so, females copulate with the light-colored ones willingly. In fact , sometime s female s see k ou t th e light-collare d males, even though nearby in the very same arena a darkcollared male is waiting.34 Why do the dark-collared ruff s allow the light-collared one s to sta y in their arenas? Because female s fancy light- a s well a s dark-collared males, i t ma y well be tha t arena s with light-collare d ruff s presen t attrac t female s better tha n arena s withou t any ; the dark-collare d male s therefor e welcome th e light-collared one s an d procee d t o "shar e th e wealth." But why do females seek to copulate with th e subservient light-collared ruffs ? Females wh o copulat e wit h bot h varietie s (morphs ) o f ruf f ar e likel y doin g right by their offspring . Th e difference s betwee n th e tw o morphs see m to be genetic. Sinc e both morph s hav e evolved, it would see m tha t eac h ha s its own advantages. If these advantages could b e combined i n the same individual male, one would assume that natural selection woul d hav e merged the two morphs long ago. We ar e not familia r enoug h with th e life strategy of ruffs outsid e breedin g seaso n to tell what the advantage s of each morph are ; but simila r cases among fishes and crustaceans sugges t tha t eac h morph i s best fitte d fo r a specific ecologica l niche . Both dark-collared an d light-collare d ruff s thu s have a good chanc e in life, an d it is to the female' s advantag e to have some offspring o f each morph; she hedges her bets by having some offspring tha t are best suite d fo r one life strategy, and other s best suite d for the other . Adult male bluegill sunfish may be either small or large. The large ones defend breeding territories an d court females; th e small males join them during courting, and both fertiliz e the eggs laid in the large male's territory. 3' The color an d movements o f th e smal l male ar e simila r t o thos e o f the female . Th e larg e territoria l males do no t chas e the smal l males away as they do other large males. The common explanation for this behavior is that the small males "cheat" both the large territorial males and the females by pretending t o be females themselves, getting access t o a large male's territory and fertilizing the eggs females lay unbeknownst to eithe r the female s or the big male. But human observers do not find it difficul t t o distinguis h th e smal l "mimetic" male s from females ; wh y should i t be difficul t fo r members of the sunfish' s ow n species, who afte r al l have a lot more to lose by their "mistake"? 36 We thin k tha t i n fac t ther e is no cheatin g involved , tha t rathe r thi s i s a reproduction arrangemen t tha t satisfie s al l parties . Th e tw o mal e form s fac e

Mate Selection 3

7

different constraint s o f predation , feeding , an d s o on . Bot h surviv e an d bree d successfully. T o hav e offsprin g o f both morphs , whic h raise s th e odd s tha t he r offspring wil l surviv e an d breed , th e female' s egg s mus t b e fertilize d b y bot h morphs; therefor e sh e ma y wel l prefe r territorie s i n whic h bot h morph s ar e present. Thu s i t i s mor e likel y tha t th e larg e dominan t mal e wil l attrac t a fe male i f h e accept s a smal l mal e a s a partner . Sinc e bot h female s an d smal l males displa y submissiv e behavior , eac h fo r it s ow n reasons , i t i s not surprisin g that the y use simila r signals ; afte r all , the y are sendin g the sam e message . But that doe s no t mea n tha t th e larg e territoria l mal e o r th e femal e mistak e th e small mal e for anothe r female . True, thi s i s speculation; but th e suggestio n that the smal l mal e i s cheatin g i s als o speculative , an d a fa r les s likel y explanatio n in ou r opinion . Similar "cheating" systems have been describe d i n fish , reptiles , birds, an d insects and are probably present i n other form s o f lif e a s well. W e believ e tha t a clos e loo k at th e fact s i n eac h cas e would re veal that each party to the system is actually straightforwardl y servin g its ow n bes t interest s unde r give n conditions, withou t deception . A s we shall see in chapter 16 , there are even socia l system s in whic h tw o species, hos t an d parasite , cooper ate in order t o reproduce. I n the case of some of these interactions, too, one party has been describe d a s cheating the other ; an d the y too ar e far more interesting if one look s a t the m a s forced collaboratio n betwee n tw o partie s wit h conflicting interests, eac h o f whom trie s t o ge t th e bes t i t can . Th e resul t i s a compromis e between th e opposing interest s o f the parties rather than a temporary triumph by one or the other throug h cheating.

FISHER'S MODEL OF THE RUNAWAY PROCESS COMPARED WITH THE HANDICAP PRINCIPLE According to our theory , "waste" in sexual display is evidence of the advertiser' s quality, bu t ver y different explanations hav e been offere d eve r sinc e the day s of Charles Darwi n himself . Th e issu e firs t cam e u p i n Darwin' s O n th e Origin of Species. Darwi n suggeste d i n 1859 37 that the feature s an d qualitie s o f each specie s are formed by the process of natural selection, in which the more efficient survived

38 P A R T N E R

S I N COMMUNICATIO N

and reproduce d whil e other s di d not . Thi s theory , thoug h genera l an d all encompassing, di d not explain th e waste involved in the showing of f that precedes sexual reproduction . Darwi n coul d no t se e ho w a n investmen t i n showin g of f increases a n individual's efficiency . H e suggested , then , tha t ther e ar e two kind s of selection : natura l selection, whic h make s a n animal best suite d t o it s environment, and sexual selection, which assists an animal in competing within its gender for th e chanc e to reproduce . Darwin38 thu s define d sexua l selectio n a s competition wit h member s o f th e same sex. With thi s definitio n Darwin lumpe d togethe r straightforward , efficient fighting between riva l suitors with features that enable an individual to deter rivals of the sam e species and gender by means of threats, and with features that attract potential mates. Darwin di d no t se e a problem i n the evolutio n o f bizarre signals that functio n i n sexual advertisement—he simply turned his observations into an explanation. Th e simpl e fac t tha t bizarr e signal s attrac t mate s an d dete r rival s justified fo r him the investment animals make in these signals. He di d not as k why waste attracts mates and deter s rivals . Rather, he treated these effects a s a given.39 In th e earl y twentieth century , Fisher recognize d th e proble m presente d b y animals' preference for wasteful signals. He stated , rightly , that female preference is an adaptation like an y other produce d b y natural selection; h e then aske d why females prefe r wasteful males.40 The mode l Fisher propose d t o answer this question use d to be the only one available, and many still believe i n it. The model can be found in almost any book o n evolution. 41 Fisher believe d tha t th e mal e who show s of f is no bette r tha n th e mal e who does not sho w off. According t o that premise, th e showing off itself is a drawback , and thus the show-of f males are less well adapted than their fellows. 42 According to Fisher , th e onl y advantage the ostentatiou s male s have is the fac t tha t females consider the m attractive . Since such males pass on the show-off character to thei r offspring, thos e offsprin g wil l show of f and wil l be attractiv e to female s too. According to thi s script, male s gain by investing in showing off because by showing off they attract more females. Females lose by having offspring who waste resources on showin g off , but the y have n o other choice : onl y wastefu l offsprin g wil l b e attractive to other females, wh o in turn will have inherited from thei r mothers the tendency to be attracte d to wastefulness. Fisher's mode l ca n be see n a s a catch-22, in which each individual male in th e population waste s resource s o n showin g of f solel y becaus e thi s i s the accepte d method o f courtship i n his species. Severa l mathematical models support th e ide a that onc e som e of the populatio n conside r a particular random featur e to b e attractive, the n a "runaway " proces s ma y develop b y which th e featur e spread s quickly throughou t th e population . I n suc h a population , a femal e wh o wen t against the trend and selected a male who is more efficient an d shows off less might have mor e efficien t mal e offspring—bu t thes e offsprin g woul d fin d themselve s without mates , since other female s would no t choos e them. Fisher assume d that th e proces s start s when som e females, who selec t a male

Mate Selection 3

9

by som e feature that trul y correlates t o hi s general quality , ge t more an d bette r offspring tha n female s who mat e indiscriminately. Bu t in Fisher's view , once th e daughters of these females inherit the tendency to select males by this feature, then males who exaggerate the feature—irrespective of their real quality—will get more mates. From that point on , according to Fisher, females choose males not by their quality in general but rather by the exaggerated feature, which in his view no longer correlates t o th e male' s quality . In fact , Fishe r sai d tha t th e exaggerate d feature decreases the rea l quality of the males , but tha t th e proces s i s driven forwar d by the preference of the females trapped in the catch-22. In other words, females now prefer male s with the exaggerated feature because other female s prefer them. 43 But ther e i s a major proble m wit h Fisher's model . Th e sam e wasteful characteristics tha t attrac t mate s als o deter rival s of th e sam e sex, an d w e have to ask why. If only females reacted to a feature that did not correlat e to real quality, one might conceivably explain the value of having sons who bear the same feature. But we find that not only do these exaggerated features attract females, they deter rival males too . If the signal indeed has no connection to the real quality of the male as a rival, then a male who is not deterre d by the feature will succeed better tha n those who are. He will produc e offsprin g wh o ar e likewis e unfazed; even tually the arbitrar y feature will di e ou t a s a threatenin g signal in the entire population. Ye t this does not happen in nature ; in the rea l world, rivals in many cases remain intimidated b y th e sam e supposedl y arbitrar y feature s that attract females, or by features similar to them. In fact, Fisher himself noticed thi s weakness in his model. Bu t as Fisher ha d n o wa y of explainin g ho w displa y coul d b e correlated t o prowess, he suggested that wit h time , rival males will stop reactin g to features that amoun t to mere "war-paint." 44 By Fisher's mode l there need not be any correlation betwee n th e male's actual prowess and the female's choice, and indeed, Fisher assumed that the wastefulness of the signa l decreases the male's tru e prowess. Yet many findings suggest that in fact th e most extravagant males are also the fittest. 45 The real question is not whether Fisher's model is internally logical or whether it ca n be expresse d i n mathematical terms, but rathe r how one ca n best explai n ostentatious waste in nature, including showing off toward rivals. Any explanation must sho w why such high-cost signals as the singin g of male songbirds, the large, heavy, branche d antler s o f male deer, an d th e colorfu l plumage an d dancin g of various male birds often deter rivals at least as effectively a s as they attract females. In fact , Andersso n remarks that it is often difficul t t o tell whether a given feature is used more for the former purpose or the latter.46 Fisher's model explains neither the evolutio n o f features used to dete r rivals , nor th e logical connection betwee n signals and their messages.

4O P A R T N E R

S I N COMMUNICATIO N

Unlike Fisher' s model, th e theor y we offer—th e Handica p Principle—doe s explain th e relationshi p betwee n th e specifi c wa y in which a n anima l shows off and th e individual' s qualit y both i n courtship , wher e i t attract s potentia l mates , and in competition with rivals, whom it deters. According to our model, the cost— the "waste"—i s the very element that makes the showing off reliable. The female in thi s mode l i s not a sill y creatur e attracte d t o extravagan t males just becaus e "every female is"; by selecting as a mate a male who ca n afford lavis h displays, she is choosing a good fathe r fo r her offspring .

UTILITARIAN SELECTION AND SIGNA L SELECTION We believe that natural selection encompasse s two different, an d often opposing , processes. On e kin d o f selection favors straightforwar d efficiency, an d it works in all areas except signaling. This selection makes features—other than signals—more effective an d less costly; we suggest calling it "utilitarian selection. " The other kind of selection, b y which signal s evolve, results in costly features an d trait s that loo k like "waste." It is precisely this costliness, th e signaler's investment in the signals, that makes signals reliable. We suggest calling this process "signa l selection. " Darwin included i n sexual selection—the competition fo r mates—both signals on the one hand and features that actually enable an animal to fight more efficientl y with rival s of the sam e species an d gender o n the othe r hand. Ou r definition , by contrast, make s a clea r distinctio n betwee n feature s tha t ca n b e explaine d b y straightforward utilitaria n selection an d thos e that cannot—signals . As we see it, most o f wha t Darwi n define d a s "sexua l selection " i s better understoo d t o b e "signal selection." 47 Signal selection differ s fro m sexual selection i n that it includes all signals—no t jus t those tha t affec t potentia l mate s an d sexua l rivals , but als o signals sen t t o al l other rivals , partners , enemies , o r anybod y else . A t the sam e time, signal selection exclude s features that improve actual fighting ability, which are selected fo r straightforwar d efficiency. The need for reliability explains the multitude o f signal s i n th e natura l world , an d the theory of signal selection thus offers ne w ways o f lookin g a t ever y species o n earth , from microscopi c organism s to humankind itself. How signal s evolve and what their evolution implie s carr y u s throug h th e chap ters that follow .

P A R T I

I

METHODS OF COMMUNICATION

This page intentionally left blank

C H A P T E RR4

THE FALLAC Y O F SPECIES-SPECIFIC SIGNAL S

DID DECORATION EVOLVE TO IDENTIF Y SPECIES, GENDER, AND AGE ?

W

e have seen ho w signal s used b y enemie s o r rival s evolve: the engin e driving this evolutio n i s the commo n interest th e antagonist s have in reliable communication. But are there signal s that evolv e without th e element o f competition or enmity? What abou t those structures and markings that allow on e t o tel l a give n specie s fro m another , o r t o distinguis h th e mal e o f a species fro m th e female , an d th e juvenil e fro m th e adult ? Scientist s cal l thes e species-specific o r set-specific signals, an d the y assum e tha t commo n interest—a n animal's need t o identify other member s of its own species—rather than competition among individuals explains the evolution o f these traits. But we believe tha t specific signals evolv e just as all other signals do—through the competition among individuals to demonstrat e their quality. To mos t people it seem s obvious tha t species-specifi c signals exist s o that we and othe r animal s can tel l on e specie s fro m another . Afte r all , birders i n a new environment ca n identify th e species , an d ofte n th e gende r an d age , of unfamiliar birds by comparing the birds' decorativ e markings with pictures and photographs. We recognize the bulbul by its black head and yellow undertail coverts, the mallard by its green hea d an d the blue on its wing, the blackbird b y its glossy black bod y and orange beak; the markings of the male mallard and blackbird distinguish them 43

44 METHOD

S O F COMMUNICATIO N

as well fro m th e dra b female s o f their ow n species . I n breedin g season , i t is easy to tel l th e specie s o f mal e duck s apar t b y thei r colorfu l plumage . Experience d birders ca n als o distinguis h the m i n other seasons , by subtl e difference s i n th e patterns o f colo r o n thei r wings . I t ca n b e ver y difficul t t o distinguis h betwee n closely relate d species : ther e ar e abou t te n specie s o f stonechat an d wheatea r in Israel, an d the small differences betwee n som e of the females ar e not eas y to spot. Yet experts ca n tell eve n thes e apar t b y subtle, species-specifi c variations i n thei r markings. Lorenz1 was impressed with the multitude of colorful fish in the coral reef. H e suggested that their conspicuous markings help the fish recognize members of their own specie s s o that the y ca n avoi d fightin g wit h member s o f other specie s wh o are no t competin g wit h the m fo r resource s an d mate s an d avoi d courtin g an d mating with members of other species . Wallace, in his argument with Darwin over sexual selection , propose d tha t th e mai n functio n o f male showing of f is specie s recognition;2 Mayr also suggests that most features unique to males developed fo r that reason. 3 This explanation is accepted b y most researchers. After all , zoologists are use d to recognizin g species , gender , and age by animals ' markings; they find it easy to assum e that th e marking s evolved i n order to aid animals in making the same distinctions . Do animal s reall y us e marking s t o identif y member s o f thei r ow n species ? Experiments sa y that the y do . Som e specie s o f gull hav e black head s an d whit e rings aroun d thei r eyes ; other s hav e blac k head s an d n o rings . Whe n Smith 4 painted whit e ring s aroun d th e eye s o f th e latter , othe r gull s treate d the m a s though the y were member s of the specie s with the rings . When he painted blac k over th e whit e ring s of the former , they were treate d b y other gull s a s members of th e ringles s species . Katzir 5 foun d tha t fish in a coral ree f reacte d t o model s painted t o loo k lik e thei r ow n species an d disregarde d model s painte d wit h th e markings o f othe r species . But di d thes e marking s evolv e i n order t o enable animal s t o identif y member s o f thei r species? No t necessarily . We identif y a kangaroo by its special shape and gait. It is very likely that kangaroos themselves do the same. Yet no one suggest s that th e shap e and gai t of kangaroos evolve d i n orde r t o hel p kangaroo s rec ognize eac h othe r a s kangaroos. Th e fac t tha t features are used by animals to identify the species, age , or gende r o f other animal s does no t prove that th e feature s evolved for that purpose. In mos t specie s o f babblers, th e group-livin g desert bir d we study, males an d females look alike. Luckily for us , in the particula r specie s we observe a t Hatzeva there is a minute difference between th e two: females have dark brown irises, while males have lighter, yellowish-tan irises. These birds trea t strange babblers accord-

The Fallacy of Species-Specific Signals 4

5

ing t o thei r gender : male s attac k male s an d cour t females , females attac k strang e females . On e coul d assum e tha t th e difference i n ey e color develope d s o that babblers coul d tel l the gende r of other babblers. But with time we became skeptical: these birds showe d u s that they can tell the gende r o f a strange babbler lon g befor e the y se e the colo r o f it s eyes — perhaps by its voice (females ' voice s ar e higher), or b y characteristics of its flight. If they can tell a stranger's gender fro m a distance , wha t i s th e poin t o f on e smal l gender-specific marker tha t ca n onl y be see n a t clos e quarters ? Would any body sugges t tha t wome n us e eyeline r i n orde r t o le t me n kno w tha t the y ar e women? Me n ca n figure that muc h ou t lon g befor e the y see a woman's face , le t alone her eyes. We were thus led to the conclusion tha t the differen t color s of the male and female babbler's eye s evolved no t to indicate gender but fo r some other reason. And i f the evolutionar y value of species-specific markings is to preven t fight s or t o preven t courtin g amon g member s o f simila r specie s wh o liv e i n th e sam e geographic area, then what is the usefulnes s o f these markings in animals who live at th e to p o f a mountain o r o n a remot e island , fa r fro m simila r species ? Suc h species retai n their specifi c marking s even afte r lon g period s o f isolation. 6 Mutations o f colo r an d marking s are no t unusua l amon g animals in captivity ; there is every reaso n t o assum e that suc h mutations occu r i n th e wil d a s well. W e rarely see them, because they are selected out . Distinctive markings thus have a selective value even i n isolated areas ; otherwise natural selection would not hav e preserved the species-specifi c markings so faithfully . Some of the mos t prominen t an d colorfu l decoration i s found in male birds of species tha t ar e not monogamous , ye t it is precisely these males that ar e especially prone t o mat e with female s o f othe r species . Selander , who remark s on thi s fact , explains it by saying that the polygamous males do not lose much by fathering a few infertile offsprin g amon g their many fertile ones. Bu t fo r a female, eve n in polygamous species, infertile young represent a substantial loss. Raising infertile offsprin g demands as much effort o n the part of the mother as raising fertile young does. Thus a significan t par t o f he r reproductiv e potentia l ultimatel y goe s t o offsprin g wh o cannot themselve s reproduce. 7 I f species-specific decoration an d marking s evolved for th e purpose s o f species identification, why do mistakes happen mor e often precisely in polygamous species, whose males have the mos t distinctive markings? Species-specific marking s raise a muc h mor e basi c question , however : i f th e only purpose o f such markings is to enable animal s to identify their own and other species (o r gender or age within thei r species), it is difficult t o imagine the proces s by which they evolved. Markings can help establish group membership onl y if they are commo n t o al l members o f the group , or a t least t o many . Yet the first individual who bore thos e markings was a minority of one. In orde r for the markings to become "species-specific, " thei r bearer or bearers must have reproduced bette r

46 METHOD

S O F COMMUNICATIO N

than individual s wh o di d no t carr y the sam e markings, unti l with tim e th e trai t spread throughou t th e population. Th e early bearers of the markings had to have some advantag e over individual s wh o di d no t bea r them . Tha t advantag e could not b e tha t i t wa s easier t o identif y the m a s members o f the species , fo r a t that point mos t members of the specie s did no t carry the markings.

THE EVOLUTION OF COMMON MARKINGS THROUGH COMPETITION BETWEEN INDIVIDUALS We sugges t a very different explanation : these marking s evolve through th e competition that members of the species engag e in to determine their relative quality.8 The marking s that we see as uniform ar e the ver y ones that show most clearly the fine differences amon g individuals regarding the attributes most important to them. The closer the competition—the more evenly matched individuals are with regard to some desirable attribute—th e more helpful unifor m marking s can be in exposing these fine distinctions. Anybody who has ever had to judge an athletic, musical, or beauty contest knows how crucial it is for the athletes to compete under highly regulated an d calibrate d conditions , fo r the musician s to play under simila r conditions, o r fo r th e beaut y contestant s t o appea r i n simila r clothing , precisel y in order t o tell the fine differences betwee n the m and selec t th e best o f the best. 9 Tourists i n th e ol d cit y o f Jerusalem, especiall y those equippe d wit h a goo d guidebook, can identify members of many different group s by their clothing: nonobservant Jews; ultraorthodox Jews, and even members of particular Hassidic communities; Ara b villager s fro m differen t regions , who dres s differentl y fro m eac h other and from Ara b city dwellers; Bedouins, whose clothing is different ye t again. The tourist might think that people wear specifi c clothin g to avoi d th e ris k that, say, a n ultraorthodox Jewish gir l might unwittingly marry a Bedouin or, God forbid, a nonobservant Jew. But any local can easily tell which group a person belongs to, regardless of that person's clothing: by his or her gestures, speech, movements. If a Bedouin wer e t o put o n ultraorthodox Jewish garb , a n ultraorthodox Jewish girl would immediatel y spot him a s a pretender. Why, then , d o peopl e o f a given group mak e the effor t t o dres s "properly"? Why doe s a n ultraorthodox Jew dres s just lik e all the othe r member s of his synagogue? Or , fo r that matter, why do corporat e executives , or members of a biker gang, dres s jus t lik e othe r corporat e executive s o r member s o f th e gang ? Th e clothing withi n eac h grou p i s not reall y identical, o f course . Smal l details—th e quality of materials or workmanship, the fit, the car e taken in dressing, the way a person carries his or her clothing—vary from person to person, and the difference s tell a great dea l abou t eac h wearer' s means , abilities, an d personality , a s well as accentuating posture and behavior. It is precisely the similarity of the clothing that

The Fallacy of Species-Specific Signals 4

7

makes it easier for group members to asses s the difference s amon g them. If each person withi n th e grou p wor e wildl y differen t clothin g i t woul d b e difficul t t o compare them with on e anothe r i n a meaningful way. How exactl y can uniform decoratio n sho w th e difference betwee n individuals ? Let's star t wit h a simple example : let' s sa y we are going t o th e marketplace to find a handmade round plate. The plates are not identical . Some were made by skilled crafts men and are perfectly round; others were less skillfully mad e and ar e slightly elliptical, or their edges are a bit malformed . A small circle in the middle of a plate would mak e it easier for us to tell th e most perfectly round plates from th e less perfect ones . A line around the edge of a plate would mak e it eas y to spo t imperfec t edges. If we wish to selec t perfectl y roun d plates with neat , perfect edges , it would mak e sense for us to selec t fro m amon g plates decorated wit h a circle in the middle and a line around the edge, even if we have t o pa y a somewha t highe r pric e fo r them . Sinc e othe r shopper s ar e als o looking for well-crafted round plates , it would mak e sense in turn for the skille d artisan to paint a circle in the middl e o f each plate and a line around its edge. This decoration wil l clearly benefit the makers of perfect plates, but why should the maker s of less than perfect plates decorate them with a pattern that will bring out thei r imperfections ? They should d o s o because they are competing with yet other craftspeople, whose plates are even worse. Buyers who cannot find or cannot afford th e best plates would thu s be reliably drawn to the better ones amon g the ranks o f the imperfect . Eve n th e maker s of the absolut e wors t o f the plate s ha d better invest in such patterns, since buyers will reject ou t of hand plates that don't bear the pattern, and since there might appear an incompetent artisan making even worse plates than theirs. The proces s is driven by the preferenc e of the buyers. This issu e of choic e i s crucial. An essentia l condition fo r th e developmen t o f any signal is that the receiver of the signal must have at least one other alternative— otherwise ther e i s n o chanc e o f affectin g th e receiver' s action s an d n o poin t i n signaling to it . A gazelle can tel l a wolf reliabl y that it is an excellent runne r an d that th e wol f ha s littl e chanc e o f catchin g it , bu t i f th e wol f doesn' t hav e th e alternative o f findin g othe r prey , i t ha s t o tr y t o catc h tha t gazelle , even thoug h the gazelle has just proven tha t the chas e will be a hard one. If, a s we believe, animals ' markings evolved in order to bring out smal l differ ences betwee n competin g individuals within a species, then th e marking s cannot be arbitrary. Not every decorative pattern would enable us to judge a plate's roundness equally well; many patterns would actuall y disguise imperfections and make it more difficult fo r us to find the better-made plates. Likewise, any markings could serve to identif y species , gender, or age . But i f species-specific markings show off certain importan t trait s i n member s o f th e grou p wh o compet e t o prov e thei r quality, th e markings that evolve have to be thos e that best brin g ou t difference s

48 METHOD

S O F COMMUNICATIO N

regarding these traits . And indeed, onc e one starts looking a t animals with this in mind, examples abound .

MARKINGS AND TH E TRAITS THEY ADVERTISE Lines and Stripes Barlow10 note d tha t fish with unusuall y shaped head s have decorativ e marking s that confor m to the shape of their heads. We likewise find that long fish generally have lengthwise stripes, and high-bodied fish generally have vertical stripes. Barlow assumed tha t th e stripe s camouflag e th e fish. Our ow n experience i s that stripe d fish are easier to spot. Lengthwis e stripe s bring out the length o f the fish, or to be more exact , make it easie r t o spo t sligh t difference s i n length amon g fish of th e same species. Vertical stripes bring ou t the height of the body, and a center strip e brings out symmetr y of form.11 The pattern clearly draw s th e eye of the beholder to whatever is the objec t of the decoration . Species o f coral fish that ar e short , high-bodied, an d thi n ar e decorate d wit h vertical stripes . Moreover , th e vertica l stripe s emphasiz e specifi c proportions : sometimes th e lin e start s a t th e fron t corne r o f th e dorsa l fin and goe s straigh t down, sometime s i t runs from the bac k o f the dorsa l fin to the fron t o f the ana l fin, and sometime s it runs acros s the base of the tail , or acros s the head , through the eye. Since the lines run throug h suc h important bod y parts, they cannot lie: a line tha t wa s longe r becaus e i t ha d a differen t slan t woul d b e obvious , sinc e it would not then g o through th e same features. The paragon o f stripes is, of course, the zebra . The zebra' s stripe s ca n b e see n clearl y onl y b y nearby observers : fro m farthe r awa y the y merge into a discree t gray. 12 I f th e stripe s were mean t purely for camouflage , the y could have been random, lik e a leopard's spots . Bu t a zebra's stripe s emphasize specific body parts: lips and hooves are black, crosswise lines show off the thickness of legs and neck, and other lines show off th e shape of the rump .

Patches and Frames that Bring Out Body Parts Many species of hoofed animal s have markings tha t outline their rumps: patche s of color , as in most antelopes , or stripes like the zebra's . We realize d the value of such markings when we traveled with a student researcher at the Shushlui park in South Africa . A t the time he was studying waterbucks, larg e antelopes with whit e

The Fallacy of Species-Specific Signals 4

9

rumps. Many of the waterbucks were infested with ticks and were thin and weak. Their white rumps made it easy to tell from behin d whic h antelopes were in good shape an d which were thinner than the y should be . O n thi n antelopes , th e white patch had the shape of a pointed ellips e because of the atrophied hin d leg muscles; on healthy animals, the patch looke d nic e and round. Th e hind leg muscles are of immense importance t o th e antelope—the y ar e its "engine, " its driving force. A predator looking for eas y prey, a rival evaluating his chances in a contest, a female looking fo r th e bes t fathe r for he r offspring , al l can benefi t from evaluating th e muscles of the rump . The wing s of many butterflies are outlined with a narrow fram e o f color . I f the win g were uniform in color , i t would b e difficul t t o spo t smal l defects in the edge. The colored outline , on the other hand, is visibl y broke n b y imperfection s i n th e wing' s shape. Suc h breaks can result fro m developmenta l defects, encounter s wit h predator s wh o tak e bite s ou t o f th e wing , o r collision s with har d objects . They are more likely to occur i n older individuals . The outlin e can enable females to spot broken edges and so avoid mating with males who have birth defects , ar e clumsy, or have reached the ag e where they have spent most of their sperm. Similar frames an d colore d edge s o n feathers enable birds to tell th e condition o f those feather s at a glance. Beaks, nails , hooves , fins , spines , tails , an d horn s ar e ofte n a differen t colo r than th e body. The differenc e brings ou t th e shape , size , an d movement o f these parts, just as among humans nail polish show s off the exact shape and movements of nails and fingers .

IMPLICATIONS OF THE HANDICAP PRINCIPLE: THE USE OF MARKINGS TO FOCU S ON FEATURE S Once we realized that there is a relationship between a n animal's decorative markings an d th e informatio n thos e marking s convey , we starte d lookin g a t animal s differently. Til l then, we had bee n lookin g a t animals and tryin g to identif y their species. From the n on , we have been tryin g to see what a n animal's markings can tell us about it s features an d adaptations. Which body parts are emphasized by an animal's decoration, an d what is their importance to that animal? Again, we found that thi s ne w approac h reveale d t o u s many features that w e had neve r notice d before. The yello w dot nea r the bas e o f a surgeonfish's tai l emphasizes the spik e that is locate d there . Th e strip e abov e th e eye s of a trunkfish accentuate s th e specia l shape o f it s head . Th e colore d dot s a t th e tips o f mallards ' an d pelicans ' bill s

5O M E T H O D

S O F COMMUNICATIO N

decorate hook s a t the en d o f their beaks : we never eve n noticed th e hooks unti l we started looking a t markings as signals. Th e re d patc h nea r th e ti p o f th e bea k o n th e lower mandibl e o f th e herrin g gul l show s of f th e exac t dimensions of a thickening there, which is typical in larger gull species . Befor e w e starte d lookin g a t marking s a s meaningful i n this way, we never noticed th e thickening , or remarked that there is no such thickening in the beaks of smalle r gulls. We are not the first to observe the dot on gulls' beaks. Tinbergen13 note d tha t herrin g gull chick s peck a t the colore d dot s o n their parents' beaks , an d tha t th e peck s mov e th e paren t t o regurgitat e th e foo d i t ha s brought it s young. Many ethologists sa w this as the reason the colored do t evolved . But Tinbergen's argumen t doe s not explai n th e exact location o f the dot; and the young of smaller gull species, which don' t hav e a colored do t o n their beaks, also peck a t th e tip s o f thei r parents ' beaks . W e thin k rathe r tha t th e colore d do t evolved t o show off the thickening a t the tip o f the beaks, and to emphasize small differences betwee n individual s regardin g th e siz e of thi s thickening ; it s us e b y chicks is secondary. The fligh t feather s on a stork' s win g an d th e larg e feather s that cove r the m (coverts) ar e black ; th e smal l coverts, which mak e up th e res t o f the wing' s surface, ar e white. Th e wing's aerodynamic properties depend , amon g other factors, on th e thicknes s o f the variou s part s o f the wing . If th e win g wer e colore d uni formly, i t would b e difficul t t o se e the dimension s o f its parts i n relatio n t o on e another. Th e tw o color s sho w of f no t onl y th e genera l shap e o f th e win g bu t also th e relativ e siz e o f it s mai n part s an d change s i n thei r shap e durin g flight , and thu s th e qualit y o f flight . Bustards, stone curlews, and many butterflies are very well-camouflaged on the ground; in flight they display the conspicuous patterns on their wings. The patterns help onlooker s asses s the spee d an d directio n o f the individual' s flight, the manuevers performed, and s o on. Many birds ' neck s ar e a differen t colo r tha n thei r bodies . Th e positio n an d movement of the neck tell a great deal about the intentions o f the bird. Is it calm? Is it on the verg e of flight? Is it tense or frightened? The contrastin g color o f the bird's neck show s off both its intentions an d th e length o f its neck.14 Fish's fins often hav e eyelike patterns o n them. The patter n bring s out movements of the fin: when the fin is held down, the round "eye" change s into an ellipse or eve n int o a line. A fish ready for quic k movemen t fold s it s fin s down ; whe n threatening, a fish spreads its fins, and th e "eye " show s up a s a perfect circle . Color may also be related to health. The rooster's comb , the turkey's featherless head, humans' lips and cheeks, the small patch of bare skin on the forehead of the chick o f th e grea t creste d grebe , al l vary i n colo r dependin g o n th e amoun t of blood they receive: brighter i n healthy individuals, these area s turn pal e an d even

The Fallacy of Species-Specific Signals 5

1

bluish whe n th e bloo d vessel s narro w i n reactio n t o cold , il l health , o r othe r stresses. Hamilton an d Zuk 15 suggested tha t bright color s tel l femal e birds tha t a male is healthy an d i s not sufferin g fro m parasites . Accordin g to them , i t pay s off fo r females t o choos e a brightl y colore d male , sinc e th e brigh t color s sho w of f it s genetic tendenc y t o withstan d parasites— a tendenc y i t woul d pas s o n t o it s offspring. Hamilto n an d Zuk' s articl e led t o man y research projects , som e o f the m showing a reverse correlatio n betwee n brigh t color s i n birds an d th e numbe r o f parasites th e bird s carry. In fact , th e poin t ca n be mad e much more sweepingly : stress create d b y parasite s i s no differen t fro m stres s cause d b y col d o r hunger . The brightl y colore d plumag e o f a mal e i n goo d shap e develop s properly ; th e plumage o f a sickly one cannot. Indeed, Hamilto n an d Zuk refer to the effect the y describe a s a special case of the Handica p Principle . Markings ca n also indicat e how healthy a bird was when i t gre w it s feathers . Bird s mol t a t leas t onc e a year; individuals i n ill health do not molt, or molt only partially. Pattern s typica l o f ne w feathers—fo r in stance, narro w rim s o f colo r tha t wea r of f i n time — give proof tha t th e bird' s plumag e i s new. The plumage of a bird who was in ill health durin g th e mol t i s often partl y made u p o f ol d feathers ; decorativ e pat terns o n ne w feather s mak e thi s readil y apparent . Many bird s hav e feather s wit h a lin e dow n th e mid dle; missin g feather s ar e eas y to detect , becaus e the y distur b th e patter n o f th e combined lines. 16

SYMMETRY The correlatio n o f symmetry to quality , both in animals and i n humans, has bee n studied recentl y by several researchers; the term used in current literature is "fluctuating asymmetry. " Moller 17 foun d tha t th e mal e swallow s preferre d by females had bot h longer an d mor e symmetrica l tails (tha t is, the lef t an d righ t part s were evenly matched) than males who took longer to find mates. Research confirms that humans conside r symmetrica l face s t o b e mor e beautifu l tha n somewha t asym metrical faces. 18 Thornhill 19 foun d tha t mal e scorpio n flie s wh o wo n fight s wit h rivals, and who brought mor e presents to their mates, had more symmetrical wings than the male s they defeated. It i s not clea r whether femal e scorpion flies actually chose male s fo r thei r symmetry ; even whe n th e female s selecte d male s b y smel l alone, without seein g them , the one s the y chose were more symmetrica l than th e ones they passed up. What, then, is the connection between symmetry and quality? Stress or geneti c flaws often caus e asymmetrical development o f body parts. 20

52 METHOD

S O F COMMUNICATIO N

If the proble m wer e simpl y lack of nutrients or of energy, one would expec t that deficiency t o affec t bot h side s o f the bod y evenl y and cause , say, a shorter symmetrical tail rather than a longer, asymmetrical one. W e believe tha t symmetrical growth requires reliable communication within the body.21 For synchronized, symmetrical growth, the center that regulate s growth must send the same messages to both side s of the body, get feedback on the results, and pace further developmen t according to the data received . As we shall see in chapte r 9 , signaling between cell s in the multicellula r bod y requires reliability no less than communication between organism s does. It seems that only a cell in good shap e can afford t o manufacture a quality signal. In other words, reliabl e communicatio n withi n th e bod y demand s a n investmen t o n th e part of the organism, and it may be that under conditions of stress the body cannot spend enoug h o n suc h communicatio n to ensur e symmetry . Thus a symmetrical shape indicates bette r overall healt h durin g development tha n a less symmetrical shape. An awareness and understanding of "esthetic" features such as symmetry, good color, an d patterns that suit what they decorate thus bring real and concrete benefits t o thos e wh o posses s them . Indeed , in the las t chapte r o f this book we will suggest tha t thi s connectio n betwee n "beauty " an d qualit y i s th e basi s fo r th e evolution of the estheti c sense in humans. The righ t decoratio n ca n revea l th e degre e o f symmetry. 22 A lin e dow n th e middle of the body, or down the middle of a body structure, is an obvious example. A circle in the middle of an area that ideally is symmetrical is another. In fact , any symmetrical decoratio n o n bod y part s helps observer s asses s symmetry of shap e and structure; and since the decoration cannot be more symmetrical than the body parts themselves, the informatio n conveyed is reliable.

"EYE" PATTERN S The peacock' s lon g tai l feather s take severa l months to grow. The score s of tail feathers are arranged so that eyes form arches ; the arche s ar e ver y orderly, wit h eac h eye exactly halfwa y betwee n tw o eye s i n th e arc h abov e it . Missing eye s are very conspicuous—there is no nee d t o count, an d a peacock canno t hid e th e los s o r imperfec t development o f eve n on e tai l feather . An d a s Petri e found,23 th e respons e of peahens is measurably different i f even only five eyes are missing in a peacock's tail , out o f a hundred fifty or more . For a round ey e pattern to emerge on a lengthening feather , the entire growth and developmen t o f that feather have to be in perfect synchronization. Any irregularity in development would ruin the perfection o f the circle. A circular pattern— and eve n more, a circle within a circle—is thus ideal fo r displayin g regularity in

The Fallacy of Species-Specific Signals 5

3

the proces s of growth. I f the peacock's tail were decorated with lines, it would b e difficult t o b e sur e of deviations : a line o f on e lengt h i s not mor e perfec t tha n a longer o r shorter line. But an imperfectly developed circl e exposes any distortion, in an y direction. This valu e of eyelike patterns can explain their widespread occurrence, a s on the wings of butterflies, mantids, and other insects . Th e standar d explanation—tha t th e eyes frighten off predators, wh o mistake them for the eyes of larg e animals—i s a n insul t t o th e predators ' percep tion; nor doe s it explain patterns composed o f many eyes, or o f serie s o f four o r five concentric circles . W e doub t that an y predator s ar e frightene d of f b y th e eye s o n a peacock's tail. 24

THE EVOLUTION OF MARKINGS; POLYMORPHISM AND CONVERGENC E If indeed marking s are not arbitrary , and certai n markings show off specific qualities bette r tha n others , the n tha t explain s ho w a se t o f marking s ca n sprea d throughout a population. First , member s o f a population star t focusing on a particular feature—the siz e of a beak, say , or certai n movements—because they find that the feature in question convey s essential information either abou t th e overall quality of other individuals, o r about whether they possess certain qualities . Onc e that happens , the n individual s decorate d wit h marking s that brin g ou t thes e features, other thing s being equal , are more likely to be selected a s mates, or avoide d as rivals. They have a selective advantage over others; the y will have more descendants than individual s wh o do no t bear thos e markings. Thus, th e marking s will spread i n th e populatio n an d eventuall y become "species-specific. " When members of a species must compete to show off their superiority in qualities important to the lif e o f that species, individual s who ca n demonstrate suc h superiority have an advantage . The interes t i n th e siz e o r shap e o f a beak, rump , o r tai l leads t o decorative markings that bring out reliably the siz e and shape of that beak, rump, or tail. The marking s of closel y relate d specie s an d eve n o f subspecie s ar e often dissimilar. A new species or subspecies usually evolves when a population o f a certain species adapts to an environmental niche that is somewhat different fro m th e original nich e o f tha t species . Thi s ca n requir e physiological , morphological , o r be havioral adaptations . The ne w nich e ca n chang e th e relativ e importanc e o f certai n features—like , for example , beak size. 25 When thi s happens, it makes sense that the features most important t o members of the "new " specie s ar e different fro m thos e most important to the original species . The decoration bes t suited t o bring out these feature s

54 METHOD

S O F COMMUNICATIO N

will therefor e als o b e different . Individual s wit h decorativ e marking s tha t bes t emphasize these new characteristics will have an advantage ove r individual s wit h the old decorative markings, and the new markings will spread in the population . First come s the evolutio n o f new adaptations ; next, individuals start focusing on these adaptations; and finally, because of that attention, new decoration emerges — the ne w species-specific markings. But the story is not alway s this simple. Sometimes members of the same species have differen t markings . One suc h polymorphic species is the ruf f mentione d i n the previous chapter . Some male ruffs hav e dark ruffs, other s have light ruffs. Th e courtship behavior o f each typ e i s also different . Th e dark-ruffe d males ten d t o hold a territory within the lek and dance in it, while the light-ruffed male s tend to fly fro m on e territory to another, dancing with several of the dark-ruffe d male s in turn. A s we discus s i n detai l i n chapte r 8 , blac k an d othe r dar k color s bette r emphasize siz e and shape , while whit e i s better fo r showin g of f movement. Th e darker colo r o f the territory-holder s better advertise s their stance , while the fac t that th e light-colore d ruff s ar e s o ofte n i n motio n explain s wh y the y advertis e themselves with lighter colors. We don't think that the difference i n colors evolved so tha t wha t w e no w cal l th e dar k an d th e ligh t form s coul d b e distinguished ; rather, we believe it evolved to enable each light-colored ruff to try to show himself as the bes t amon g the light-colore d ruffs , an d t o allo w each dark-colore d ruf f t o compete as successfully a s possible against othe r dark-colored ruffs . The relationship between decoratio n an d its message explains wh y it s o ofte n happen s tha t unrelate d specie s who occupy a similar ecological niche look alike. The decoration of the deser t agamida e of the Middle East is very similar to that of desert iguanas in the Americas, as is the decoration o f rattlesnakes in Americ a an d tha t o f vipers in the Ol d World . Lack remarked that bird s that liv e in similar condition s ten d t o hav e simila r decoration. 26 I n the Argentinean plains, we encountered what we thought was a creste d lark—onl y t o lear n tha t i t wa s i n fac t a larklike brushrunner (Coryphistem alaudina). Species that live i n simila r environments tend t o evolv e similar traits to best dea l with these environments; they then advertis e these traits with the markings tha t best show them off — which tend t o be simila r patterns and colors.

FACIAL MARKINGS AND TH E DIRECTION OF GAZE Most animal s have facia l markings . Often thes e markings decorat e th e eye s an d help onlooker s detec t ey e movement an d determin e the directio n o f the animal's

The Fallacy of Species-Specific Signals

55

gaze. Bird s ofte n hav e eye rings, ey e stripes, or colore d lines runnin g from bea k t o eye . Bird specie s with binocular vision, such as shrikes, stonechats, an d great tits, often hav e ey e stripes convergin g toward th e beak . Whe n both ey e stripes look equa l t o th e observer , th e beak — and th e bird' s gaze—ar e aime d straigh t a t hi m o r her . Any chang e i n th e directio n o f th e bird' s glanc e wil l change th e symmetr y o f th e lines . Th e cream-colore d courser, whose unusual field of vision extends backward, has decorative lines goin g from it s eyes to th e nape of its neck. Bulbuls , partridges , gulls , babblers , an d th e lik e have monocula r vision , tha t is , eac h ey e has a separat e field o f vision ; thes e specie s ofte n hav e ey e rings . Th e bulbul has a black head an d a black eye with a prominent white ey e ring; the whit e rin g enables a distant onlooke r to tell th e directio n o f the bulbul's gaze. Lines, eye brows, and colore d ey e rings make it clear even from a distance what direction a n anima l i s lookin g i n an d demonstrat e hi s o r he r abilit y to stare — which, a s we saw in chapte r 2, is one indication o f his or her leve l o f confidence. This information is reliable: the same decorative pattern that emphasizes confident behavior brings ou t th e hesitan t movement s o f individuals who have not ye t decided wha t t o d o an d ar e still assessin g their options. Th e directio n o f a courting or threatening individual's gaze indicates its interest i n what it is looking at . The ey e decoration o f youngsters, and ofte n o f females, i s less prominent an d revealing than that o f adult males ; for example, young bulbuls lac k the white eye ring tha t characterize s the adul t o f th e species . Whil e youngster s d o benefi t b y showing i n a general wa y what the y ar e intereste d in , the y ar e bette r of f i f they avoid committing themselves—as we shall see in more detail in chapte r 18 .

STATUS BADGES OR HANDICAPS? We hav e see n ho w th e deer' s large , heav y antlers an d th e peacock' s heav y tail enable the m t o advertis e thei r qualit y reliably . No t surprisingly , suc h tail s an d antlers affec t how muc h recognition—prestige—other s accor d an individual. Although the y ar e les s conspicuous , decorativ e marking s ca n als o advertis e socia l status. The grea t tit' s black band, th e white patch o n a magpie's wing, the black patch unde r the beak o f a Harris's sparrow , a n eyelike spot on a fish's fin, and the bulbul's prominen t whit e ey e rin g al l serv e t o proclai m thei r owners ' prestig e (social status) and i n fact ar e termed "badge s o f status." 27 For example , grea t tit s wit h a wider blac k ban d o n thei r underpart s defen d their territorie s an d bree d mor e successfull y tha n thos e wit h a narrower band ;

56 METHOD

S O F COMMUNICATIO N

great tits with narrower band s subordinate themselve s to ones tha t bea r wide r bands. 28 Experiments i n other ani mals confirm the value of such badges o f status. But how can a purely decorativ e markin g be a reliable indicatio n of anything? Evolutionarily speaking, what is there to prevent a young or low-quality bird from developin g a wider band an d gainin g an advantage over its more capabl e rivals? Several researcher s hav e ru n experiment s i n whic h they decorated younger individuals with badges of higher status. I n som e o f these experiments , th e statu s o f such individual s ros e accord ingly, but thi s had costs . Rohwer an d his colleagues 29 painted low-statu s Harris's sparrows with patches of color indicating higher status. They found that the "pre tenders" no t onl y di d no t gai n bu t actuall y fared wors e tha n young , unaltere d control birds : the y were attacke d mor e ofte n b y high-status birds, whic h usually don't bother t o attack youngsters. When researcher s gave the painted bird s injec tions of testosterone—which increased their aggressive behavior and reduced their hesitation—the bird s di d gai n i n status , but the y may have los t mor e tha n the y gained; their bodie s seeme d t o suffe r fro m th e artificiall y highe r level s o f testosterone.30 Th e researcher s asserte d tha t th e increase d aggressio n towar d "impostors" keeps the markings honest an d prevents individuals from puttin g on status badges they have no valid claim to. But if status is determined b y an arbitrary tag, how do the birds recogniz e an "impostor" who bears "fake " badge s tha t d o not fit its "actual"-status? And why should that specific marking hav e evolved as a status badge i n the first place? Status badges, lik e other markings , amplify difference s i n features that directly indicate a n individual's quality . The statu s badge emphasize s the excellenc e o f an individual of higher quality ; the sam e badge bring s ou t th e inferiorit y of a lesser individual. Eye stripes emphasize head movement—an d thus accentuate both the steady gaze of a confident individual an d th e hesitanc y of an insecure individual . The confident animal gains, by drawing attention t o its gaze, but th e insecure one loses. Th e blac k ban d o n a great tit' s breast make s its breast look narrowe r tha n it is ; the wide r th e band , th e narrowe r th e breas t appears. 31 Onl y a high-quality great tit ca n affor d th e narrowin g effec t o f a wide band . Decoration tha t seems at first glance merely to tag its owners as being of higher or lower status is really a standard measuring tool that enables individuals to show off thei r quality—no t as compared wit h those who lack the status badge, who are clearly subordinate, bu t rathe r with individual s who bear the same marking. Such patterns are not physica l handicaps, yet they are handicaps all the same, since they impair th e abilit y o f lower-qualit y individual s t o falsel y presen t themselve s i n a way that would assis t them in social conflicts. In other words, like all signals, these handicaps are more of a burden t o lower-quality individuals than to higher-quality ones.

The Fallacy of Species-Specific Signals

57

In sum , we believe tha t thos e markings that bes t dis play difference s i n qualit y o r motivatio n amon g similar individuals evolv e t o become commo n t o the entir e species, o r t o a n ag e and gende r grou p withi n th e species , for precisel y tha t reason . Sinc e differen t specie s hav e adapted differently , th e decoratio n appropriat e t o thos e species would likewis e be different. Tha t others use these differences t o identif y th e grou p t o whic h a n individua l belongs i s a side effect o f the competitio n amon g individuals within th e group .

ARE THERE SIGNAL S WITHOUT HANDICAPS? Hasson claim s tha t som e signals , which h e call s "amplifiers"—mostl y decorativ e markings—are not costl y in themselves but simpl y show off the qualit y of the signaler without imposing an y handicap o n it. For example , lines along a fish's body or a feather's edge show clearly and reliabl y the length o f the fish or the conditio n of the feather without any significant material cost to the fish or the bird.32 We d o not believ e tha t th e cos t of decorative marking s can be separate d fro m the dimension s o f the structur e the y advertise, or fro m othe r message s they may carry. The line along a fish's body shows how long the fish is; that is its "message." Each individua l fish' s lin e bears a differen t quantitativ e message . The lin e let s a fish show of f clearly how much longer i t is than shorte r fish—bu t onc e i t has th e line, i t als o can' t avoi d showin g of f clearl y ho w muc h shorte r i t i s tha n longe r individuals. Thu s th e cos t o f th e decorativ e markin g is differentia l rathe r tha n uniform an d i s greater for inferior individuals than for superior ones . The material investment in the line may be the same, and may be insignificant, but th e message and it s cos t ar e differen t fo r differen t individuals . Th e sam e markings that sho w off clearl y their bearer' s advantag e over inferiors also impairs that animal's ability to "fudge" an d pretend to be as good a s individuals superior t o it. The higher the quality o f th e individual , th e mor e likel y tha t thos e i t i s compare d wit h wil l b e inferior t o it , an d th e mor e i t benefit s b y bearin g th e markings ; th e lowe r th e individual's quality , the mor e likely it is to be compare d unfavorabl y with better quality individuals, and the more costl y the markings.

ARE THERE AN Y SIGNAL S THAT ARE CONVENTIONS? Are ther e conventiona l signal s tha t d o no t hav e an y competitive aspects ? May nard Smith 33 assume d tha t specifi c signals, such a s those proclaimin g tha t a n in-

58 M E T H O D

S O F COMMUNICATIO N

dividual belong s t o a certai n specie s an d i s o f a particula r gender , ar e signal s that conve y noncompetitive information ; comparing them t o railroa d timetables , he terme d the m "notices, " an d asserte d tha t i n th e cas e o f suc h signal s ther e is no reaso n t o cheat , an d thu s n o reaso n t o prov e thei r reliabilit y an d n o investment require d o f th e signaler . Grafen , wh o develope d a mathematica l mode l that support s th e Handica p Principle , als o assume d tha t som e noncompetitiv e signals exist ; for example , relative s may communicate b y means o f agreed-upon , conventional signals , an d i t wa s hi s contentio n tha t i n suc h a case , guarantee d reliability an d th e necessar y investment woul d no t b e issues. 34 Man y who toda y accept th e principl e tha t th e cos t o f a signal guarantee s its reliabilit y d o no t ye t accept ou r clai m tha t al l signals have a cost—they impose a handicap—and that this i s what guarantee s tha t the y are reliable . "Railroad timetable " signal s o r notice s ar e indee d purely arbitrar y and have no connection t o their message or to cost : it is enough fo r the signa l to be agree d upon , clear, an d efficient . Bu t w e canno t fin d suc h signal s i n nature. The example s in this chapte r al l show a clear relationship betwee n eve n small , unobtrusiv e signal s an d the characteristic s they advertise; all these signals amplif y the abilit y o f the observe r t o spo t superiorit y o r defects in th e animal s that carr y them. Th e mor e flawed the individual , th e costlie r th e pattern. Thi s uneven cos t i s characteristic of signals that aris e through signa l selection, al l of which involve investment in reliability. As we shall see when we discuss chemical communication, even in the cas e of communication withi n th e multicellular body, where ther e is no conflic t betwee n communicating cells, reliability is needed—in this case to prevent errors—and can be achieve d onl y by means of handicaps. A s far a s we know, ther e ar e onl y two biological system s having "agreed-upon," arbitrary signals that do not involv e investment i n reliability : human language—o n whic h we will say more i n chapter s 6 and 18—and , according to the present consensus , the genetic code . But in fact , some believe tha t the genetic cod e i s not arbitrar y either.35 Yet our assertio n tha t there ar e no "ordinary, " agreed-upon signal s in the anima l world i s still contrary to the opinio n o f most researchers .

THE DEFINITION OF SIGNALS; INFLATION AS A TES T OF THE THEORY OF SIGNAL SELECTION But what, after all , are signals? We defin e signals as traits whose value to the signaler is tha t the y conve y information t o thos e wh o receiv e them . Ther e i s a differenc e between traits that evolved fo r other reasons , such as body size or a kangaroo's gait , that ca n an d d o conve y information; an d signals , whic h evolve d solel y t o conve y

The Fallacy of Species-Specific Signals 5

9

information. For example, one can judge the direction o f another's gaze by watching its eyes—ye t ey e movement i s not primaril y a signal. But ey e rings and smal l tuft s of hai r or feather s whose onl y function is to sho w a n observe r th e directio n o f an individual's gaze more clearly, or from a greater distance, are signals. The abilit y to observ e an d understan d signal s is an adaptatio n lik e any other that evolves through utilitarian selection. An individual who pays attention t o unreliable signals will be less successful—will have fewer descendants—than one who insists o n payin g attentio n t o reliabl e signal s only . A s w e hav e seen , reliabilit y demands investmen t b y its very nature—the signaler's investment i n the signa l is the guarante e of its reliability. It is not always easy to judge whether a specific trait is purely a signal or whether it has some other function, but i t is still important to make the distinction. What , then, i s th e fundamenta l differenc e between signal s an d othe r traits ? I t i s th e relationship between th e signa l and it s cost. Ever y trait, whether a signal or not , demands an investment of some sort, and ever y trait is constrained by other traits whose requirement s conflic t wit h it . I n thi s signal s ar e simila r to al l other traits . Changes in the environmen t ma y change the constraint s on the evolution of traits and lesse n o r increas e the cos t involved . For example , th e siz e of a n anima l may be limited by its need to stay small and quick to avoid predators, but if its predators are remove d i t ca n affor d t o increas e in siz e and thu s increas e it s ability to stor e energy, withstan d cold , o r defea t rival s of it s ow n species . Th e specie s the n be comes larger because th e cos t of being big has gone down significantly . If, o n th e othe r hand , th e cos t o f a signal is reduced t o th e exten t tha t every individual can use it equall y well, the n th e signa l can no longer revea l difference s in the qualit y or motivation of individuals. In suc h a case, the signa l loses its value. Because the signal's cost has gone down significantly , the signal is no longer usefu l and wil l disappear. 36 Th e evolutio n o f signals—signa l selection—i s thu s funda mentally different fro m th e evolutio n o f all other adaptations . In trait s othe r tha n signals , the cos t of the trai t i s an unavoidabl e sid e effect . In signals , cost is of the ver y essence; it is necessary to the existenc e of the signal. If ther e i s no cost , nothin g prevent s cheater s from usin g a signal to thei r benefit and t o the detrimen t o f the receivers , and that signal will lose its value as a signal. This ha s happene d no t infrequentl y in huma n history . When mone y is easier to get, i t loses value through inflation. Ornaments that were prized as tokens of wealth when rar e became worthless when easy to obtain . When lac e was made by hand by expert workers, the amount of skilled labor needed to produce it made it very expensive; lace cost—and was worth—its weight in gold to th e ric h an d powerful , wh o wor e i t t o displa y thei r wealth. The developmen t o f machines that could cheaply manufacture lac e indistinguishabl e fro m th e handmad e product37 put a n end t o the use of lace as an indicator of wealth; today , in fact , lac e i s not muc h use d a t all. 38 By

6O M E T H O D

S O F COMMUNICATIO N

contrast, commodities lik e bread o r iron that ar e used to meet direct needs rather than to sen d signal s do not los e their usefulnes s i f their pric e goes down ; rather, they are used all the more.39 Although human cultural and economic developmen t are differen t fro m biologica l evolutionar y systems, the tw o have a very important common element: both emerg e from th e need competitors have to cooperate an d communicate. The significanc e of cos t i s unique t o ou r theor y of signa l selection. N o othe r theory of signal evolution—and certainl y no theor y that assumes the existenc e of agreed-upon, conventional signals—woul d predict that if its cost becomes low, a signal wil l los e value. It shoul d b e possibl e t o tes t thi s hypothesis : eve n thoug h experiments with evolutionar y processe s require many individual subjects an d a long period o f time, they are feasible with unicellular organisms or organisms that multiply rapidly. Meanwhile, we can test our theory against observations of animals in the field. As thi s boo k wa s goin g to press , we learne d o f som e finding s tha t suppor t ou r view of the impact of inflation o n th e effectivenes s o f a signal. In Australia , male satin bowerbirds especiall y favor blu e objects; in their usua l habitat suc h objects consist o f blue feathers , whic h ar e rare , an d blu e flowers that need constan t re placement. Borgia 40 found on averag e five blue feather s pe r bower . Nea r human habitats or picnic grounds, however, blue artifacts—mostly blue plastic—are fairly common, and sati n bowerbirds ca n collect as many as a hundred pe r bower . In mos t areas , male sati n bowerbird s compet e bot h by stealin g blue object s from on e anothe r an d b y destroyin g each other' s bowers . Hunte r an d Dwyer 41 recently found , however, tha t wher e blu e objects ar e abundant, th e males direc t less effort int o stealin g such objects from on e another, an d more into destructio n of their competitors ' bowers, tha n the y d o i n areas wher e such objects ar e rare . Blue objects, where abundant, are no longer significant marks of quality. The birds still collect them , but the y seem to place no more value on them than humans do on chea p lace.

C H A P T E R

5

MOVEMENTS AN D RITUALIZATION

M

ovements ca n indicat e man y things: a scolded dog lowers its ears an d tuck s it s tail betwee n its legs, while a happy dog wag s its tail; a threatening ca t arches its back and stands on tiptoe, hair raised; a duck moves its bill up an d down t o signa l its intention t o fly; male ruff s cour t female s b y dancing in special dancing arenas, or leks. These are only a few examples of the endless array of movements used as signals. The Hand icap Principle give s us new ways of seeing such signals and of understanding how they evolved an d the messages they encode .

DIFFICULT MOVEMENTS Male courtshi p movement s ar e ofte n extremel y elaborat e an d difficul t t o per form. Th e super b bir d o f paradise hang s upside dow n o n a branch, flapping its 61

62 METHOD

S O F COMMUNICATIO N

wings an d spreadin g it s ornamenta l feathers . Hangin g upsid e dow n doe s no t make th e movement s an y clearer, bu t i t doe s mak e the m mor e difficul t t o per form, and thu s stronge r proo f of the male' s physica l abilit y and his motivatio n to cour t females . Ruffs , manakins , and bustard s perfor m elaborat e dance s ver y different fro m thei r day-to-da y movements. Dove s usuall y wal k whe n o n th e ground, bu t whe n courting , mal e dove s ho p o n bot h legs . Th e ho p i s probabl y more difficul t an d demonstrate s the male' s prowess . Male lark s hove r hig h i n th e sk y t o sing . A t th e song's en d th e lar k doe s no t fly down bu t rathe r fold s his wings and plummets like a stone, openin g the wings only at the end of the dive. Some individuals open thei r wings severa l times befor e touchin g the ground; others dare t o mak e a breathtaking , uninterrupte d dive , spreading thei r wings only at the last second. Th e later the wing s ar e spread , th e mor e impressiv e th e dive . This i s very similar t o competition s amon g free-fallin g human sk y divers: a wrong estimat e ca n result i n a disastrous crash . Difference s in th e div e ma y well reflec t the relativ e expertis e o f specifi c mal e lark s an d con vince female s an d othe r male s alik e of the performin g bird's abilitie s a s a mate o r a rival. Not al l movemen t signal s ar e mean t t o conve y informatio n a s importan t a s what a n individual is worth a s a mate or a rival, and th e investmen t i n other such signals can correspondingly b e mor e modest . A dog signalin g friendship wags its tail; this is a reliable indication o f friendliness, since the wagging tail would interfere wit h attac k o r flight . A hungr y nestlin g stretche s it s neck , bea k gaping ; a satiated nestling doe s not. The parent prefers not to exert itself to bring unneede d food; i t demand s tha t a nestling prov e it s need b y making an effor t to o grea t t o be worth makin g unless the nestling is indeed hungry. 1

RITUAL FIGHTING Much o f th e fightin g tha t goe s o n betwee n rival s o f th e sam e specie s i s ritua l fighting; the goal is not to injure or kill the opponent, but to convince it and others that i t i s the weake r an d shoul d withdra w fro m th e contest . Th e initiato r o f th e fight does not even try to surprise its opponent; rather , it signals clearly that a fight is imminen t an d let s it s riva l prepare itself ; only afte r it s opponen t ha s assume d the appropriat e stanc e does th e aggresso r strike—and then onl y a well-protecte d and "legitimate " target, suc h as antlers. The horn s o f a male gazelle ar e strong , sharp , lethal weapons. A male gazelle

Movements and Ritualization 6

3

fighting his rival could surpris e it and run i t through, but h e doe s not. He comes at hi s opponen t slowl y an d let s it ge t int o positio n befor e locking horn s wit h i t and pushing an d pounding it in a contest of strength. The end of such a grueling struggle usually comes when one of the opponent s concedes . It woul d b e a mistake to cal l suc h a struggle a fight. I t i s more lik e a competitiv e spor t i n which contestants tr y t o sho w of f their superiorit y while following fixed rules. In fact , thes e ar e threats that take th e for m o f a physica l contest ; th e winne r does no t pursu e the lose r an d kil l it; it i s satisfied when it s riva l gives up. I n th e rar e cas e when on e of th e contestant s dies , i t i s becaus e i t faile d t o withdraw i n time. 2 Mal e gazelle s hand-raise d b y humans a s pets treat people as members o f their ow n species, and whe n grow n up the y may invite their keepers t o a rivals ' contest. Th e keeper , wh o doe s no t realize that th e gazelle' s initial butt i s an invitation t o ritua l combat , lacks horns to mee t th e attac k and ma y be fatall y wounded . The ritua l battles o f ibexes an d bighorn shee p ar e stunning sights. The rivals rise up o n their hind legs an d hurl themselves with al l their migh t at their oppo nents—who meet the m with a similar charge. The clashe s are repeated again and again, scores of times. Neither contestan t tries to strike at his rival's body. All his force i s aimed at and absorbe d by the rival' s strong horns . These fight s ar e not playacting : they are out-and-out displays of strength an d agility a t th e en d o f whic h th e contestant s ar e exhausted . I f a n ibe x wishe d t o finish of f hi s rival , a s a predato r does , h e woul d hav e approache d i t stealthil y and trie d t o gor e it s bod y wit h hi s horns . Bu t a n ibe x o r a gazell e initiating a ritual fight approaches slowly and show s off his intentions by lowering his horns. He wait s fo r hi s riva l t o prepare , an d i f the riva l doe s not , th e challenge r may butt i t lightl y t o ge t it s attention . B y letting hi s riva l prepare fo r combat , th e ibex i s takin g o n a handicap ; th e handica p pay s off for th e winne r becaus e h e convinces hi s immediat e riva l a s well a s an y onlooker s o f hi s superiority . This saves bot h animal s the ris k inheren t i n a battl e t o th e death , an d on e o r bot h combatants ma y demonstrat e a prowes s tha t save s it fro m havin g t o dea l wit h other rivals . Contests o f force don' t necessarily involve physical contact. Gulls threaten by tearing up grass, chimpanzees break off branches, rhinoceroses and bulls rake up dirt with their hooves. Peacocks fan out their tails and hold them upright, shaking them, which demands a great deal of strength. Since males of these species perform the same actions, they and others can judge reliably which of the rivals is stronger, and thus the contestant s are spared a fight.

64 METHOD

S O F COMMUNICATIO N

RITUALIZATION: DOES I T REDUCE THE INFORMATION CONVEYED ? Huxley,3 who studied th e courtship o f the great crested grebe , suggested the term ritualization for the process by which movements that serv e as signals are derived from movements that originally had some other function. The courting male grebe repeatedly touche s the side s of his body with his beak, a s if to preen hi s feathers, but he is not actually preening himself. When preening, a grebe combs each feather from base to tip. The courting movements are much more stylized; the same feathers ar e touche d repeatedl y bu t ar e no t combed . Thi s formalize d movemen t re minded Huxley o f the formalized gestures humans use when they perform rituals, hence th e term ritualization. Huxley suggeste d tha t ritualizatio n evolve d i n order t o differentiat e betwee n th e original movement— preening—and th e on e use d fo r signaling . H e suggested that th e mos t important functio n of ritualization i s to increas e the clarit y of the signal , in order to differentiat e betwee n i t and the original, functional movement, an d betwee n i t an d othe r signals . Thi s opinion i s still accepte d b y most researchers. 4 Even thoug h movemen t signal s ar e uniform , they are not performe d i n exactly identical way s by differ ent individuals, nor by the same individual in differen t circumstances. This differenc e i s not accidental ; it often show s the intensity of the signal. For example , if stretching the neck indicates fear, then the degree to which it is stretched indicates the degree of fear. Morris5 suggested that movement signals evolve by two conflicting processes: the movement becomes mor e uniform so that the receive r o f the signa l will recogniz e the message , while variatio n emerge s t o show magnitude. H e believe d tha t these tw o competing tendencie s le d to a compromise tha t require d movemen t signal s to b e mor e formalize d tha n th e move ments fro m whic h they had develope d bu t allowe d fo r enough variability to show quantitative differences . Morris's idea is logical, but i t does not addres s reliability in movement signals, and i t doe s no t explai n why certain movements rathe r tha n other s wer e selecte d for a specific message . Both question s ar e answered b y the Handica p Principle. 6 If movement s tha t develo p int o signal s must deman d enoug h effor t tha t cheater s cannot use them o r will not fin d the m wort h thei r while, the n individuals of different abilitie s an d wh o ar e more o r les s motivate d woul d perfor m them differ ently. Th e differenc e i n performanc e reveal s specifi c difference s amon g performers, whic h observer s ar e interested in . The selection o f reliable signals by their receiver s bring s abou t ritualization , and th e cos t o f performing the formal -

Movements and Ritualization 6

5

ized movement s result s i n variability ; ther e i s n o nee d t o posi t tw o conflicting selective pressures . Rather, ritualization an d variabilit y are complementary results of signa l selection . Morris assume d tha t ritualization—th e standardizatio n o f movements fo r th e sake of clarity—reduces the information contained i n the movement messages. But is there reall y a loss of information? We thin k not. In fact , it is precisely the nee d to compar e tw o o r mor e performer s an d se e clearly th e difference s between o r among them tha t lead s to standardization . While working o n thi s chapte r w e happened t o b e watchin g T V coverage of the Olympic games in Barcelona. Th e degree of standardization achieved by sprinters wa s impressive: the y starte d of f a t th e sam e instan t an d wer e n o mor e tha n hundredths o f a second apar t a t the end . Athleti c contests ar e governed b y rules. All contestants perfor m the sam e tasks. Untutored viewer s ma y have trouble de ciding whic h o f the severa l top-notc h gymnasts is the best. After all , experience d judges focu s o n detail s tha t see m almos t inconsequentia l t o th e uninitiated . Fo r example, a great dea l o f a gymnast's success depends o n th e wa y he o r sh e get s off the apparatu s at the end o f a performance: ideally, a gymnast lands with perfect balance, fee t aligne d an d straight . This ma y seem to a lay observer t o b e a n arbitrary, draconia n capric e o f th e judges , bu t close r examinatio n show s tha t th e smooth, rapid transition from movement to a perfect stance at the end of a grueling effort i s indeed on e o f the hardes t part s of the exercise . When on e has merely to tell a good performer from a bad one , judging is easy. Olympic judges set minutely exacting standards because they have to judge among many outstanding, top-notc h performers. Individuals performin g movemen t signal s i n th e animal worl d ar e like competitor s i n a n athleti c con test. The signaler s sho w of f their abilit y t o execut e a certain movement , while th e observer s ar e the judges evaluating the signalers ' performance . The judge may be a femal e lookin g fo r th e bes t fathe r fo r he r offspring; a predator lookin g for prey it can catch; a rival evaluating whether it has an y chance agains t the per former; o r a parent decidin g whethe r it s offsprin g shoul d reall y ge t mor e food . These "judges " forc e the competitor s t o compet e i n a standard manne r in orde r to better evaluat e the difference s amon g them. It is precisely thi s standardizatio n that brings out crucia l differences in performance, which in turn reflect accurately the differen t abilitie s an d motivatio n o f the competitors . The inexperience d ma y not notic e subtl e differenc e i n performance an d may think tha t al l signalers ar e givin g the sam e message in th e sam e manner. Thi s i s not the case. Simpso n remarks tha t experience d observers can predict the result s of a n ongoing fight between tw o Siames e fighting fish: the on e tha t i s holding it s gill covers erect mos t o f the tim e towar d th e en d o f the encounte r i s likely to b e

66 M E T H O D

S O F COMMUNICATIO N

the winner. 7 Thi s seemingl y minute an d trivia l gesture, i n the contex t o f a fight, means tha t th e fis h i s giving up th e ful l us e o f its gills and i s handicapping itself by absorbing less oxygen than it can; this is reliable evidence of its stamina and its chances o f winning the fight . I n nature , just a s in the Olympics , seemingl y negligible differences i n performance can bring about dramaticall y different results . In the Olympics , th e runne r who come s in first gets a gold medal, while th e on e in fourth plac e get s no meda l a t all , even thoug h th e differenc e between the m may be s o small it ca n only be see n with th e ai d of electronic devices. I n th e wild , in species whose males do not participate in the rearing of young, a very small number of top performer s get to father most of the next generation, while many males have no offsprin g a t all.

HOW A RITUALIZED SIGNAL EVOLVES How, then , does ritualization work? How do movement signals evolve from movements that used t o have another purpose ? A signal' s valu e t o th e signale r i s that i t ca n conve y informatio n t o anothe r individual. But the messag e can be conveye d onl y if the othe r individual is interested in the messag e and understands it . The proces s therefor e cannot star t wit h a mutation in the signaler ; because that would requir e tw o simultaneous , coordi nated mutations: one which caused the signaler to perform the signal, and another causing the observer t o take interest in it and understand its meaning. Even in the highly unlikely event o f two such simultaneous mutations , the chanc e of the tw o mutants meeting is practically zero.8 But animal s constantly collect information , including information abou t the behavior of such other individual s as offspring, mates , rivals, predators, an d prey . One rea son is that they need t o be abl e to anticipat e what other individuals wil l do . Mos t action s involv e preparator y movements: a predator ha s to look and ai m before striking; a grazing dee r ha s to shif t positio n befor e fleeing; a bird gather s itsel f t o fly. Observers ca n ofte n anticipat e animals' actions. Cats stalking their prey and dogs preparing to charge watch their prey's activities . A n experience d birder—o r a predator—freezes whe n th e bir d stops feeding and looks at the observer: the fact tha t the bird has stopped feeding and i s stretchin g it s nec k mean s i t ha s notice d th e observe r an d ma y fly away. Likewise, an animal that notices a predator starin g at it intently can anticipate the attack an d flee . Th e observe d movement s ar e no t "signals"—the y ma y convey information, bu t the y are not performe d fo r tha t purpose . Kreb s an d Dawkins 9 called this stage in the evolutio n o f signals "mind reading." Such preparator y movements are made regardless of whether or not anybod y

Movements and Ritualization 6

7

is watching them, because of their intrinsi c usefulness , rathe r tha n t o conve y any message. The abilit y to watch fo r specifi c movement s other tha n signal s can then evolve an d sprea d throug h a populatio n i n th e sam e way that an y other single , useful mutatio n does . Once observer s respon d t o specifi c observe d movements , mutation s tha t exaggerate these movements ma y spread among the ones observed. These mutations are an advantage to their bearer s because they intensify th e movement in the eyes of observers wh o already understand its meaning; at this point th e movemen t ha s become a movement signal. There i s a cost involved: as soon as a movement becomes exaggerated, it ceases to be the optimal movement for its original purpose. Still , the investment is worthwhile t o the signaler, who wants to make sure its actions ar e noticed. When a bird stretches its neck exaggeratedly to ensure that the predator—or birder—knows it has noticed it , it has to stop feeding for the moment and will have more difficult y in launching itself into flight, since in order t o do so, it would have to pull its neck back. Bu t th e bir d als o lessens th e likelihoo d tha t i t will hav e to tak e flight and leave its feeding ground: the potentia l pre y makes it clear to the watcher that th e watcher ha s bee n spotte d an d tha t approachin g woul d b e a waste of time. Both bird an d predator gai n from th e signal : the predator ca n save its energy for other, more vulnerabl e prey , an d th e bir d doe s no t hav e to leav e a feeding o r restin g spot. But why do movements that evolve into signals become exaggerated in one way rather than another? This is where the Handicap Principl e applies. Let's return t o the wolf-gazell e interactio n tha t opene d thi s book. Eve n befor e gazelle s starte d signaling to wolves , individual gazelle s reacte d differentl y t o a wolf's presence — they prepare d t o flee according t o thei r confidenc e i n thei r abilit y t o evad e th e wolf. Wea k individual s fle d righ t away , while confiden t gazelle s coul d affor d t o stand an d watch the wolf, or even to continu e feeding. In fact , Kruu k describes a cheetah burstin g into a crowd o f Thompson's gazelle s and catchin g one o f them. He stresse s that the gazelle the cheetah caugh t was the one who first turned to flee when th e cheeta h began it s charge. 10 A wol f coul d lear n b y experienc e tha t i t ha d a bette r chanc e o f catching a gazelle who fle d righ t away . It migh t eve n pa y off for th e wol f t o advanc e more slowly toward the gazelle s and let them react , so that it could spo t an d chas e the weaker ones. A confident gazelle can make it easier for a wolf—and fo r itself—t o avoid a futile chas e by demonstrativel y postponin g it s escape, or b y exaggerating movements tha t displa y its fitnes s an d self-confidence , like jumpin g straigh t u p (slotting). Individual s wil l perform th e signa l differently; their performance s will differ eve n more than their nonsignal reactions to the wolf's presence, because the wolf pay s most attentio n to thos e movements that mos t reliabl y demonstrate individuals' differing abilitie s to escape it. It is precisely these movements—the ones that prov e t o be reliabl e mean s of communication—that becam e formalized, ritualized movement signals.

68 METHOD

S O F COMMUNICATIO N

Exaggerated movement s use d a s signals are constantly tested by natural selection. The signa l evolves an d endure s onl y as long as watchers gain from i t reliabl e information tha t is beneficial to them, and a s long as the investment in it is worthwhile t o th e signaler . The benefi t in watching another t o anticipat e its actions is what generate s movemen t signals ; th e specifi c characte r o f th e movemen t tha t evolves depend s o n it s specific cost i n th e circumstance s i n which it is used. Fo r example, movemen t toward a rival became a threat, eve n thoug h sideway s movement would have demanded a s much physical effort an d might have given the rival a bette r sens e of the detail s o f the movement . I n thi s case , it is the ris k taken b y approaching th e rival , rather tha n the physical difficult y o f performing the movement o r its specifi c details , that reliabl y displays the degre e of threat . The standardization and increased difficult y of movement s use d a s signals—their ritualization—makes it easier for watchers to determin e the difference s amon g displayin g individuals , and i n th e reaction s o f the sam e individual i n different circumstances . I f eac h signale r dis played i n its own way, observers woul d fin d i t difficult t o compar e them . Movemen t signal s thus become ritualize d no t s o they can be distinguished from th e functional movements ou t of which they evolved, but rathe r to enable observers to make fine comparisons among signalers. The logic of movement signals is the same as that of species-specific pattern s of color and form: uniformity within a species—ritualization—evolve s ou t o f the competitio n member s o f the specie s engage i n t o demonstrat e their differences. I n bot h cases , i t i s the observers , th e ones the signals are directed at , who by their choices drive the evolution o f reliable and unifor m signals. 11

C H A P T E R6

VOCALIZATIONS

M

any animal s mak e decision s based o n informatio n the y ge t from th e vocalizations o f others—thei r mates , thei r collaborators , thei r rivals, members o f othe r species . Wha t i s i t tha t lead s individual s t o trus t information the y get from others' voices? And why isn't vocalization use d to mislead hearers ? It woul d see m eas y to chea t b y vocalization; yet observations an d studies sho w tha t call s usually convey reliably the intention s o f the callers . Do g and cat owners ca n often predict their pets' behavior b y listening t o them, and we don't know of any case in which a dog tried to mislead by using its voice.

THE CORRELATION OF VOICE WIT H POSTURE AND TENSIO N We discovered the key to the reliability o f vocal signal s one evening a t home. We found ou t that we could not make a sigh of relaxation while getting up, nor could we make a convincingly frightened cry while settling down comfortably . To make 69

7O M E T H O D

S O F COMMUNICATIO N

a sound o f fright, on e has to tense one's muscles—which is impossible to do while relaxing in an easy chair. On th e othe r hand, to make a sigh of relaxation one has to rela x one's muscles, and that is impossible to d o while getting up. This simpl e experiment mad e us realize that vocalizations might be faithfu l representation s of the stat e o f the body producin g them. 1 As we sai d i n th e discussio n on rivals , actors ca n soun d convincingl y loving, hateful, threatening , o r depresse d withou t actuall y loving o r hatin g thei r fello w actors o r feelin g despair . I n orde r t o d o so , they have to assum e the stanc e an d movements typica l o f the feelin g the y want t o convey . Opera singers , too, kno w that thei r movements on stag e affec t thei r voice s and lear n t o mov e i n ways that let them soun d the way they want to . If we assume that the body is the resonator, th e sounding board, then it stands to reaso n tha t th e qualit y o f vocalization s i s affecte d b y th e stat e o f tha t body . Vocal pitch , too , reflect s th e tensio n o f the muscle s o f the bod y an d face. 2 Thi s relationship between vocalization s and the body that produce s the m make s them hard t o fake . T o make a deceitful call , one has to adop t th e postur e necessar y to produce th e fake message; since each action has its own optimal startin g position , changing positio n t o produc e a false messag e may make i t extremel y difficul t t o carry ou t th e actio n reall y intended . Thu s what on e gain s b y cheatin g doe s no t make u p fo r wha t on e lose s b y assuming improper posture—an d s o the phon y message is not wort h it s price. In a dolphin researc h center i n Hawaii, dolphin s were taught t o make sound s according t o th e researchers ' request . Whe n the dolphin s wer e aske d t o mak e a threatening call, they made a sound somewhat differen t from true threatening calls. When aske d by their trainers to improve, they made a sound jus t like a real threat call—and als o assumed the facia l expressio n o f a dolphin wh o is angry and ready to bite. They were not angry ; their caretaker, who would no t drea m of putting his hand int o a n angr y dolphin' s mouth , coul d pu t hi s han d i n thei r mouth s wit h impunity. But they had to assume the posture of an angry dolphin i n order to make the call of an angry dolphin. Interestingly, Darwi n observe d tha t whe n w e listen t o singin g an d musi c we interpret th e note s accordin g t o th e action s o f the muscle s tha t produc e them. 3 But Darwi n wa s no t concerne d wit h reliabilit y i n communication . Others , too , have remarked o n the connection betwee n bod y state and voice: Scherer 4 pointed out that vocalizations express the motivation of their producers because their tonal quality is affected b y the stanc e typical of that motivation. 5 But nobody saw in this connection a key to reliabilit y in vocal communication i n general.

Vocalizations 7

1

THE INFORMATION CONVEYED BY VOCAL SIGNALS One ca n often tel l what others are doing merely by listening. This applies both to people an d t o animals . Rowell 6 found she could tel l by listening what was going on i n a confrontation between monkeys . In a radio progra m severa l years ago, it was sai d tha t in the newbor n intensiv e care unit o f the Sorok a Hospita l in Beersheba, Israel , th e nurse s coul d tel l whic h par t o f a newbor n baby' s bod y wa s hurting by the soun d o f the baby's cry. Apparently, pains in different part s of the body mak e differen t muscle s tighten , an d thes e muscle s affec t th e soun d o f th e cry in a way that a trained observe r can interpret . The crucia l importance o f small differences i n vocalizations as a source of information is brought ou t by an experiment that Gaioni and Evans conducted with ducklings.7 When a duckling can't find its mother an d siblings , it emits a series of peeps. When it hears another making peeps, it stops and listens. Individual peeps vary slightly , and th e researchers—wh o wanted t o stud y the mos t effectiv e com ponents in the peeping—compared a natural series of peeps with an idealized one, produced by computer, which repeated a single peep. To their surprise, ducklings quickly lost interes t in the idealize d serie s that was free o f all distortion an d vari ation. The natural series caused the duckling s to sto p agai n and listen. A lost ducklin g probabl y stops peeping in order t o concentrate on listening so that i t ca n get the mos t informatio n from th e peepin g i t hears. It i s precisely th e variations, the ongoin g change s in its sibling's situation , that concer n it : Has th e sibling seen thei r mother an d starte d running to her? I s it hesitating? Is it fright ened? A n artificial , unifor m series of peeps reflect s n o chang e and therefor e conveys littl e information tha t th e ducklin g is interested in . Th e continue d peepin g conveys not merel y the fac t tha t a duckling is lost, o r even which duckling is lost, but rathe r information about the ongoing changes in its circumstances and actions. A simila r logi c govern s th e dail y greeting s w e ex change with on e another . D o ou r coworker s reall y need to be reassured that it is a good, rather than a bad, morning? The greeting enables our colleagues to detect, by the tone o f our voice , fine differences i n our dail y moods. A colleague's moo d i s ver y importan t t o peopl e workin g with him or her—they need to know what to expect fro m their colleague that day. Since mood affects one' s posture and bod y tension , we can tell the moo d o f our acquaintances by listening t o the sound of their voices. It is precisely the standard nature of th e greetin g tha t let s u s detec t fin e difference s i n mood—detectio n tha t i s to the benefit o f both greeter an d listener. This simpl e method seem s to be far more effective an d reliable than would be long, detailed, wordy descriptions of our daily moods. An d indeed , sometime s the respons e t o a good-morning is, "Are yo u all right? What happened? "

72 M E T H O D

S O F COMMUNICATIO N

ANIMAL VOCABULARIES: THE CONNECTION BETWEEN THE MESSAGE AND IT S VOCAL PATTER N If an y sound tha t a n animal makes reflects th e stat e of its body faithfully, wh y d o animals us e particula r vocalization s t o communicate ? Clearly , an y mood—fear , gladness, joy, sadness—will be reflected in any vocalization that we make. But fine differences i n a particular mood—precisel y th e variation s tha t th e listene r i s interested in—see m t o b e displaye d bette r b y som e vocalizatio n pattern s tha n b y others. Different voca l patterns ar e affected differentl y b y changes in posture an d body tensions . I t make s sens e tha t th e voca l patter n mos t suite d t o sho w fin e gradations i n ange r i s differen t fro m th e on e bes t adapte d t o sho w degree s o f relaxation. Som e vocalization s will sho w bette r th e differenc e betwee n sa d an d very sad; others will highlight the difference betwee n simpl e gladness and supreme joy. A sigh of relaxation i s best suite d t o show the differenc e between on e level of relaxation an d another ; th e shar p yel l tha t sometime s goe s wit h a jump i s bes t suited t o distinguis h betwee n differen t jumps . It i s commonly assume d tha t th e voca l patter n typica l of a mood i s meant t o convey the genera l mood o f the vocalize r t o listeners , tha t is , to sa y "I a m sad, " "I a m angry," " I a m pleased." It is indeed possibl e to compile "dictionaries " that describe the vocal pattern typical of each mood: joy, despondency, boredom, wakefulness, appeasement , aggressiveness. 8 But i t doe s no t follo w tha t voca l pattern s evolved t o conve y this dictionary-type information; circumstances are often suffi cient to convey it without an y vocalization at all. The information the listener lacks is the precise degree o r nuance of the mood . A threatener, fo r example, need s t o convey not th e obviou s fact tha t i t is issuing a threat but rathe r th e degre e o f that threat. The degree of threat i s not known in advance and indeed may change fro m moment t o moment. The vocal pattern best suited to show fine differences i n the degree o f threat is the on e that is best adapte d fo r threatening . Obviously, onc e vocalizations optimal for showin g of f fine differences i n specific moods hav e evolved, one can make secondary use of them to identify the type of mood. We know that crying usually indicates sadness or pain; laughter, a happy mood; screaming , anger or alarm. But we believe tha t this function played no par t in the evolution o f vocal patterns. It is parallel, in our opinion, to the use of speciesspecific patterns to identify member s o f a group, after thos e patterns have evolve d to sho w off differences amon g individuals within th e group .

RHYTHM As we saw in chapter 3 , Lambrechts an d Dhondt 9 found that th e more successful great tits—th e one s wh o produce d mor e offspring—ha d song s tha t containe d

Vocalizations 7

3

more syllables and were more rhythmical than the songs of other great tits. Among humans, mor e precis e rhyth m i n threat s tend s t o inspir e greater fear ; i n movies and o n stage, rhythmic music can announce important o r fearful happenings . What i s the connectio n betwee n rhyth m and threat ? There is an inherent conflict betwee n collectin g informatio n and precis e executio n o f vocalization. Both activities demand concentration , but t o collect informatio n on e must concentrate on listening , looking , discerning , an d correlating , whil e precis e vocalization demands that one concentrate on execution. A person who tries hard to catch a weak or unclear sound often close s his or her eyes. And many musicians close their eyes, or assume a dreamy expression, whe n concentratin g o n their playing; they disconnect fro m thei r surroundings. One wh o trie s bot h t o liste n an d t o vocaliz e i s likel y t o falte r a bi t i n th e rhythm—a stumbl e that wil l displa y divided concentration . Therefore , a precise rhythm show s that one is not collectin g information . One wh o is certain of his or her ability to win or is determined not to give in does not have to collect additiona l information. Bu t one less confident, who is still trying to decide which way to act, needs an y information available. Thi s differenc e between th e confiden t an d th e wavering is brought ou t reliabl y by a rhythmic threat.10 Anava studie d th e trillin g call s mad e b y babbler s whe n the y mo b perche d raptors.11 Each trill is composed o f a rhythmic series of short notes an d pauses. A babbler mobbing without interference from other s tends to make loud, prolonged , uniform trills composed o f many syllables. When the mobbing babbler is disturbed by othe r babblers , however , it s call s ar e affected . Whe n a dominan t bir d ap proaches th e callin g bird , th e caller' s rhyth m is appreciabl y less even . I t i s also altered, thoug h les s so , when th e mobbin g babble r approache s grou p member s who ar e subordinat e t o it . The differenc e i s both audibl e t o human s and clearly visible in recorded analyses (se e Graph 6-1). Interestingly, when babblers get nearest the perched raptor during their mobbing, they fall silent. Apparently, assuming the mobbing postur e of raised wings and sprea d tai l interferes with the babbler' s ability to escape swiftly should the perched rapto r try to strike. Thus, the mobbing babbler canno t affor d t o tak e som e o f it s attentio n of f th e rapto r i n orde r t o perform vocalizations. Although those mobbing babblers who come nearest snakes adopt simila r postures, the y make short tzwick calls ; it would see m that snakes are somewhat less dangerous to babblers tha n perched raptors .

VOCAL PATTERNS USED OVER DISTANC E One o f the mos t important factors affectin g vocal pattern s is distance. Babblers standing next t o eac h othe r usuall y communicate with one-syllabl e vocalizations, while thei r communicatio n over greate r distance s take s the for m o f loud, multisyllable calls . Physica l barriers—trees , rocks , water , fog—distor t fin e voca l pat-

74

METHODS O F COMMUNICATIO N

Graph 6-1. Mobbing trills of the male babbler TTZL when alone (a), when disturbed by the dominant male of his group (b), and when disturbing a subordinate (c). Data from Anava, 1992. The left side of the graph shows the length of each syllable within each trill; the right side shows the change in length between each syllable and the one before it. The difference between states a, b and c may seem insignificant on the left graph, but when one focuses on the differences, as in the right graph, they stand out distinctly. Again, these differences are clearly audible to the human ear—not to mention to other babblers.

terns; so does the need to make a call loud enough to be heard over great distances. The differences tha t high volume does not conve y well are often th e very ones that the cal l i s mean t t o communicate . Fo r tha t reason , voca l communicatio n ove r distance usually consists of a series of syllables, and rhythm and changes of volume convey a great dea l of the information. Such a series also gives the listener tim e to locate th e sourc e of the sound , tur n t o it, and thus better perceiv e fin e variations. Indeed, Katsir showed tha t babblers' calls vary according t o the distance they are meant t o cover: the greater the distance, the more constan t the frequency and the greater th e variations in syllable duration. 12

WHY SHOUT ? It isn' t alway s easy to tel l just whom a vocal communicatio n is directed at . As we noted already , the discrepancy between the loud "warning calls " a sentinel babble r makes and the short distance between th e sentinel an d other members of the group led us to understand tha t the calls were actually aimed at the distant predator. But sometimes babblers make loud calls even when clearl y communicating with other s nearby. Humans, too , sometime s shout a t nearby listeners. Everyone know s tha t ther e i s often a correlation betwee n volum e an d th e de -

Vocalizations 7

5

gree of anger or of a threat—the louder the shout , the more intense the ange r or threat. We suggest 13 tha t i n these cases too th e shoutin g is aimed not a t the purported listene r bu t rathe r a t othe r babblers—or humans—wh o ar e farther awa y and ar e no t partie s t o th e conflict . Th e shoutin g make s the m witnesses . Whe n someone threaten s anothe r i n private an d the n doe s no t carr y out th e threat, he or sh e loses standing i n the eye s of that person only . When other s are made witnesses, failure to carry out the threat will cause the threatener to lose standing not only i n th e eye s of th e on e threatene d bu t als o i n th e eye s of th e witnesses . By shouting, the threatener raise s the stakes and makes the threat more reliable; only confident individual s ca n affor d t o shou t thei r threat s befor e th e crowd . Thi s principle is well-known i n politics: a publicly declared intention is likelier t o b e carried out tha n on e agreed upon i n secret .

THE DURATION OF VOCALIZATION : REQUESTS AND COMMAND S The duration of vocalization, like its volume or rhythm, is correlated to its message. The mor e persuasiv e a requester feel s compelle d t o be , th e greate r the duratio n of his or her request , in any language: "Please"; "Could you please"; "If you really don't mind, could you please" and so on. Likewise, the more nestlings repeat their begging, th e more urgent thei r parent s understand the begging to be. But what has duratio n t o d o with requesting? The point o f making a request is to get something that is of value to or demands an effor t fro m th e giver . After all , if the thin g were valueless to its owner, it could be take n without an y fuss. Logically , the give r will consent t o give up th e valued object or spen d the effor t onl y if it either cares about the petitione r or gains from the act of giving. The potential giver already knows how it feels about the requester and abou t the item asked for; what the giver does not know and needs to find out is th e importanc e o f th e requeste d objec t o r actio n t o th e requester . Th e mor e highly the petitioner value s the item or action, the more the giver gains by handing it over . The increasin g lengt h o f th e reques t convince s th e give r tha t th e petitione r really needs an d value s the objec t o r actio n in question , because the ac t o f petitioning i s costl y to th e requester . I t demand s energ y an d tim e an d expose s th e petitioner t o predator s o r rivals . It als o decrease s the standin g or prestige of the requester, an d increase s the standin g o f the giver , in th e eye s of witnesses—who are more likely to notice a longer request , and to note it s significance. Th e longer the petition, th e higher th e price the petitioner ha s paid for the item it wants. A command , o n th e other hand , i s best kep t short . A shor t comman d tha t

76 M E T H O D

S O F COMMUNICATIO N

achieves it s objectiv e means that th e on e ordere d wa s attentiv e to an d read y to obey the commander . Thus a short comman d that is obeyed raises the prestige of the one issuing it.

DIALOGUES AND THEI R SIGNIFICANC E A dialogu e can reflec t reliabl y the relationshi p betwee n thos e who participat e in it. I t has been shown that a male bird interruptin g another's son g with its own is threatening the other. 14 One wh o waits for another t o finish before itself starting to sing demonstrates either submissio n or a willingness to cooperate. The reason, again, seem s to b e th e inheren t conflic t betwee n vocalizin g and listening . By not waiting for anothe r t o finish, a bird show s that it doe s not nee d the information the other provides i n order t o plan its action. Bu t one who is unable to confron t the other o r one who would rathe r not d o so and is willing to collaborate , ha s to pay attentio n t o an y information the othe r offers , becaus e it may be essentia l if a conflict i s to be avoided . Hultsch and Todt 15 remark that nightingale s can mimic a neighbor's son g either i n goodwill o r in rivalry. When th e interaction is friendly, the nightingale starts singing, mimicking its neighbor's song, in the breaks between its neighbor' s choruses ; to signa l rivalry, it bursts into imitativ e singing while its rival is still singing. Very old male s do not respon d t o their neighbors' singing — possibly showing the indifference that attends very superior status . Another wa y t o sho w attentio n i s th e comple x "duets" sung by many songbirds in the tropics. The duet is so well coordinated that listeners find it hard to imagin e that i t i s sun g by tw o bird s rathe r tha n one. Th e due t ha s a fixed pattern. When on e mate starts singing , the othe r ha s to dro p instantl y whatever it is doing an d joi n in. This readiness shows off interest i n th e partne r an d th e willingnes s to inves t in th e partnership . Laughing thrushes, group-living birds relate d t o the babblers, demonstrate the cohesiveness of their group by coordinated group singing on the borders between their territories an d those of other groups. Group singin g by humans, too, often serve s to demonstrat e cohesivenes s an d car e fo r th e group—b e i t b y worshipper s i n church, a student choir , children i n summer camp, soldiers o n the march, or just people getting together in order to sing .

MIMICRY It i s fairly commo n for animals and especiall y for birds to mimic the call s of other individuals or other species and even background sounds. Payne16 and McGregor17

Vocalizations 7

7

found tha t bunting s an d tit s mimi c the song s o f their neighbors ; mimickin g by singing crested larks in Hatzeva have often caused us to stop and look for babblers, and a friend of ours often ran needlessly to the phone, deceive d by the mimicking of a jay. The top marks in mimicking belong t o parrots and mynas, who can mimic human voice s an d us e word s an d sentence s i n appropriat e circumstances. 18 I t would b e fascinating to study the use and value of the imitative faculties of parrots and mynas in the wild . It seem s tha t mimicr y fulfills severa l functions. On e i s to convinc e a listene r that the communicatio n is addressed t o it specifically, a s we saw in chapter 1 . Ofer Hochberg19 tol d u s of a jay near hi s house that would cal l like a cat when i t saw a cat. Alan Kemp of South Afric a severa l times observed a drongo tha t mobbed a Wahlberg's eagl e whil e utterin g mimicrie s of th e eagle' s call s mixe d i n wit h it s own alar m calls . The drong o nea r Kemp' s hom e kne w th e call s of several hawk species an d woul d mimi c them correctl y when the y passe d by. 20 I t ma y be tha t both drong o an d ja y were tryin g to convinc e thes e predator s tha t the y had see n them specifically. Birdsong, amon g its other functions, proclaims th e singer's readines s to defen d its territory. By mimicking a neighbor's song , a singer addresses its threats to that neighbor i n particular. Tits, nightingales , and th e male s of many other territoria l bird species often mimic the rival they are addressing.21 As they grow older, males of many species, such as the lyrebird, the bowerbird,22 an d the nightingale, expand their repertoir e of mimicked sound s an d o f songs. It ma y well b e tha t thei r eve r more variabl e mimicr y let s the m sho w of f their experience , their age , an d thei r ability to learn an d remember . Humans, too, mimic sounds extensively, and Darwin sa w in thi s abilit y th e foundatio n o f th e abilit y t o speak. 23 W e believ e tha t mimicry still plays an important role in human linguisti c changes suc h a s shifts in vowel sounds, the adoptio n o f new words and phrases, changes in the meaning of existing words, and the like.

DO ANIMAL S HAVE VERBAL LANGUAGE? There hav e recently been successfu l experiment s in which researchers taught apes to us e som e sig n languag e an d parrot s t o us e words. Researcher s nex t trie d t o interpret animals ' calls as words. The best-know n o f these attempt s is a study of the vervet monkeys in Amboseli in Kenya.24 It turned ou t that the vervet monkeys used one call when eagle s approached , another for leopards, and yet another fo r snakes. The researcher s also found that these call s caused reactions in others that were appropriat e to the particular predator, even when the calls were sounded by a tape recorder. As we said in chapter 1 , we think that the "warning" call s of babblers are aimed not at other members of the group but a t the predator. Do the monkeys' reactions

78 M E T H O D

S O F COMMUNICATIO N

to th e call s prov e tha t thes e call s ar e meant to warn? And can one compare the calls t o words ? Words—voca l pattern s that serv e a s symbols representin g spe cific meanings—hav e n o inheren t connection t o thos e meanings ; the symbol s are arbitrary. But we believe that all nonhuman animal vocalizations demand a logical connection betwee n the vocal signal and it s message. Babblers, too , mak e differen t call s in respons e t o differen t predators . I n th e case o f babblers , th e call s see m t o bea r a direc t relationshi p t o th e stat e of th e body o f th e babble r makin g them . Rathe r tha n bein g symbol s for a particular predator, th e call s seem to be reflection s o f different physica l reactions, eac h appropriate fo r a specific type of predator o r se t of circumstances. When a raptor appear s in the distance, babblers sound loud, abrup t calls that are like barks. The barkin g babbler i s not afraid ; i t usually goes to the to p o f the canopy and follows the raptor with its eyes. To make the loud barking sound, the babbler ha s to expel a lot of air from it s lungs. Trying to make similar sounds with our ow n bodies, we found we had t o relax our muscles and coul d not repea t th e barks without refilling our lungs with air between calls . If the bark reflects a similar relaxation in a babbler's body—which would make sense—it would seem to empty the lungs of air, which would handica p a babbler attemptin g to dart aside or into shelter. The call also discloses the babbler's location. In other words, a bark reliably shows the babbler' s disdai n for th e rapto r an d it s confidence that it can escape if attacked. When a babbler sees a raptor divin g a t it, its call i s entirely different: i t dart s into th e bushe s with a loud squeak . This "fea r cry " reflect s faithfull y th e tensio n of muscle s used i n th e ac t o f escape. This informatio n may well dete r th e rapto r from furthe r attempt s a t tha t particula r bir d b y showin g of f ho w quickl y an d effectively th e later could dar t into shelter. When a raptor lands on a nearby bush, perhaps i n the hope that the babbler s will become les s aler t in time, the babblers sound a loud trill like a cicada's. This sound ma y last for ten s of seconds , an d repeate d serie s o f such trills may follow one anothe r fo r a considerable time. The trill is a reliable way for the babblers t o signal tha t the y ar e still activel y watchin g th e raptor—sinc e on e who is relaxe d and inattentive cannot make a rhythmic trill. On the other hand, the trill does not overly endanger the triller, since a continuous sound like a trill is difficult t o locate, unlike a bark, which pinpoints the caller's location immediately.23 As if to hammer home the point, whenever the raptor moves, the babblers react with a sharp, tensemuscled tzwick , which tells it yet again that they are following it constantly. Babblers mobbin g a snake make short, shar p sound s tha t reflec t th e considerable tension o f their bodies. Fro m tim e to time one of the babbler s spread s its wings upward, and at that moment the sound it makes is somewhat different. Thes e

Vocalizations 7

9

sounds ar e usuall y calle d "snak e warnin g calls," but if they are warnings, why do the babblers com e neare r th e snak e an d rais e thei r wings? We think that by approaching a snake, babblers sho w off their courage to other members o f their group 26; the chang e in thei r pos ture necessaril y affect s th e characte r o f thei r calls. Babbler s wh o se e a predato r o n th e ground—a wol f o r a cat—mak e a serie s o f short, rhythmic tzwicks, which accompany and reflect thei r jumps from tre e to tree a s they follow the predator. Thes e calls , too, are faithful representations o f the babblers ' movements . In short, babblers, like vervet monkeys, make specific calls in response to specific enemies . It i s easy to sho w a correlation between the type of predator an d its behavior and th e chanc e that a babbler wh o see s it will make one call rather than another: a babbler wil l usually bark a t the sigh t o f a raptor in the distance , utter a shar p squea k a t a raptor divin g at it , tril l a t a raptor sittin g o n a tree an d a t a terrestrial predator in the distance, make "snake calls" at a snake, and emit tzwicks at a predator walkin g o n th e groun d nearby . Like th e monkeys , other babbler s react appropriatel y to these calls: they join the sentinel at the top of the tree when they hear a bark, dart in panic into the brush when they hear a squeak, and gather to mob snake s or perching raptors . But we don't thin k tha t th e various calls are meant t o announc e the nature of the dange r t o othe r babblers . The call s are directed at the predator, and the other babblers merely know from experienc e which circumstances each call accompanies. Interestingly, babblers sometime s see m to "err " i n thei r calls : they may bark at a n approachin g groun d predator , o r issu e snake calls a t a raptor o r a human hand approaching their nest—at which time they also raise their wings, as they do when mobbing snakes . It i s unlikely that the babblers actuall y mistake the rapto r for a snake, however. Rather , i n thes e particula r circumstance s it seem s that re sponses other tha n th e usua l ones ar e appropriat e t o th e specifi c situation , an d that the calls reflect no t the type of enemy but rathe r the particular motions of the babblers in reaction to the enemy. In each case, the call is appropriate to—in fact, is a product of—th e movement . The call s give a faithful voca l representation of the babbler's doings fro m momen t to moment. Researchers repor t tha t vervet monkeys, too, sometime s "mistakenly " us e an "inappropriate" cal l in reactio n to a specific enemy. On e wonders what th e cir cumstances were in those cases, and wha t the monkey s would "say " i f presented with dangers in unfamiliar forms—for example , an eagle in a cage on the ground. The assumptio n that the monkey s label thei r predator s seem s based o n the reac tions o f other monkey s who hear th e call , bu t thos e reaction s d o no t prov e that the callers inten d to let others know what the specific danger is. Of course, others are ver y likely t o b e awar e of th e connectio n betwee n specifi c call s and specifi c

8O M E T H O D

S O F COMMUNICATIO N

dangers: many animals reac t appropriately even to the calls of other species, whic h are clearl y not directe d a t them. Bu t we d o no t thin k tha t suc h call s evolved for the purposes o f letting others know the enemy' s identity. It i s well known tha t animals can correlate thing s with specific sounds, including words. This ability is used by trainers who teach animals to perform in response to commands—to words. Th e trainer teache s the animal to understand fragments of his or her word language. Dogs, dolphins, and parrots can learn persons' names and ca n o n comman d sit , stand , tur n lef t o r right , fetc h specifi c item s an d pu t them in specific places. Some animals can understand simple concepts encoded in words, and can make specific sounds on command. Why, then, did they themselves not evolve a language of words, as humans did, to describe t o their partners where exactly food is located, to alert them to anticipated dangers that are not yet present, or to convey an y other information tha t is best expressed by words? In the final chapter we will discuss the special nature of human language, its strong points, and its drawbacks. Here we shall only say that, with all its potency, human language has n o componen t tha t guarantee s its reliabilit y an d pre vents cheating; nor is it in and of itself well suited to convey exact, fine gradations of feeling and intention. Thus , in no human society does the language of words replace wordless communication—communication by nonverbal sound, intonation, stance , and movements. And this wordless, nonverbal communication is the onl y "speech" of animals.

C H A P T E R7

BODY PART S THA T SERV E AS SIGNAL S

T

he combs of jungle fowl are a good example o f bod y part s tha t serv e purely a s signals . The y hav e n o other practical purpose. The comb is bigger i n rooster s tha n i n hens , i n olde r males than in younger ones, and usuall y bigger in dominant males than in subordinate ones.1 The comb is a delicate, unprotected organ; it is full of blood vessels— which, a s we have seen, advertis e the healt h o f the bird . N o wonde r tha t cocks, when the y fight, try to injur e thei r rivals ' combs: a profusely bleedin g com b may well determine the fight. In th e wild , a large, intact comb o n a cock's hea d i s evidence tha t in spite of such a handicap , no riva l has ye t manage d t o injur e him: th e comb' s owne r is clearly able to overcom e rival s without bein g injured himself. As a result, a cock with a bigger comb ca n ofte n dete r rival s without fighting , sinc e his large, intact comb proves his success in previous battles. Holder and Montgomerie foun d that ptarmigan males attack each other's combs; they also found a correlation between the stat e o f males' comb s an d th e numbe r o f females who chos e t o bree d with them.2 I n th e past , peopl e wh o raise d fighting cocks use d t o cu t of f the cocks '

81

82 METHOD

S O F COMMUNICATIO N

combs an d danglin g facial flesh:3 fighting cocks raised by humans were suppose d to kil l thei r rivals , not dete r the m wit h effectiv e threats . A cock without a com b may be able to fight better, but a s it cannot threaten reliably, it cannot win without a battle .

LONG TAILS : ARE THEY SIGNALS? When a body member has both signal and nonsignal functions, its shape is affected by both. For example , th e peacock' s tai l is clearly a handicap an d thu s a signal. The impressiv e long trai n evolved fro m bac k feathers 4 an d hamper s th e peacoc k both when it walks and in flight. The smalle r supporting fa n of tail feathers holds up th e decorativ e bac k feathers . Bu t peacock s stil l us e thei r tail s a s rudders a s well—though not the most effective rudders . In theory , there i s a clear differenc e betwee n a signal and a feature that is not a signal : on e evolve s becaus e i t enable s th e anima l t o conve y informatio n an d convince other s tha t thi s information is valid, while th e other evolve s because of a straightforward utilitarian benefit t o th e animal . But in practice, it is sometimes difficult t o tel l whether a particular feature is a signal . The lon g tai l o f a n adul t male bird is not necessaril y a signal; such a tail may evolve simply because it offer s an advantag e in fligh t t o stronge r individuals . The longbo w o f the Middl e Ages enabled bowmen to shoot over long distances with great precision—but it required brawny, well-muscled arms. The bow was used because it was an effective weapon, not to prove its user's strength—but clearly, it could als o serve the latter purpose . Females ma y well loo k fo r longe r tail s because the y are evidenc e o f a male's strength, even though a longer tai l that is actually useful t o stronger males evolves under the selective pressure of the direct benefits it confers, rather than as a signal. After all , as Fisher remarked , observation by others of features tha t exist for util itarian reason s precede s th e evolutio n o f the sam e features int o signals . In othe r cases, the long tails are clearly handicaps: Evans and Thomas 5 found recently that shortening th e tai l of the scarlet-tufte d malachit e sunbird helpe d th e sunbird s in flight. Evans6 also found that the length of the tail depended on conditions durin g its development—which is what we have come to expec t with handicaps.

The tail s of many birds, such as swallows and terns, have extralong feathers at the edges (outer tail feathers); in such birds as bee-eaters, yaegers, and tropicbird s

Body Parts that Serve as Signals 8

3

these displa y feathers are in the center o f the tail. They are usually longer i n males than i n females, and longe r i n older male s than in younger ones. Fo r th e matin g season, the male widowbird, a sparrow-size songbird of the savannas of East Africa, grows tail feathers a foot and a half long.7 A widowbird displayin g in flight abov e its singin g post look s lik e a long, black, waving ribbo n wit h a small bulge a t th e front—the bod y of the bird . Mal e widowbirds, lik e peacocks , have to bear suc h extreme handicaps because they do not participate in rearing the young: their only contribution to the nex t generatio n is thei r sperm . Thus , female s do not min d sharing them with others and can afford t o search for the few most superior males, who end u p fathering mos t of the next generation.

BRISTLING HAIR OR FEATHERS : AN ILLUSIO N OF SIZE , OR A HANDICAP ? In egrets, birds of paradise, and other birds , some of the feathers of the back (the mantle) hav e evolved int o long , threadlik e decorativ e feathers. Most o f the tim e these feather s are hidden amon g other, more utilitarian feathers; but o n occasion they ar e held u p an d displaye d to mates or rivals , framing th e bird. I t seem s that the actua l cos t of developin g and carryin g such feather s is small; but ther e is a handicap involve d in displaying thes e special feathers. They probably function like the long hairs on hyenas' backs or the long, soft quills on porcupines' backs, which stand u p i n a situation when th e anima l is making a threat o r i s excited. In orde r to make the feathers, hair, or quills stand up, the animal must tense certain muscles that may well interfere wit h attack o r flight, which makes its threats more reliable. The usua l explanatio n fo r such bristlin g hai r o r feather s is that the y increas e th e apparen t size o f th e animal. 8 Bu t a n ani mal's body , o r it s head , i n fac t looks smalle r withi n a fram e o f bristling hai r than it does without such a frame. We are familiar with this effec t i n everyda y life : th e head o f a perso n wh o i s bald o r who shave s his hea d look s large r tha n tha t of a person wit h a ful l hea d of hair . Similarly, an egret bristling its threadlike feathers, a porcupine its long quills, or a hyena it s hair makes itsel f appea r smaller tha n it actually is . Animals displa y bristling hai r o r feather s no t t o fake siz e but t o demonstrate it reliably ; only a large individual ca n affor d t o mak e itself look smalle r i n the eye s of rival s or collabo rators.

84 M E T H O D

S O F COMMUNICATIO N

MANES AND CREST S The conventiona l explanation for manes, a s for bristling hair o r feathers , i s tha t they deceiv e rival s abou t th e siz e o f a n animal' s head . I f mane s were mean t t o convince rival s that an animal's head is larger than it really is, however, on e would expect the m t o b e th e sam e color a s the head. 9 Yet tha t i s rarely the case . Th e langur monke y ha s a golden man e an d a black face . On e ca n easil y distinguis h between th e skull , with its jaws and jaw muscles, which are the langur's weapons, and th e mane , whic h reflect s th e socia l standin g o f th e male . The n too , i f th e purpose o f the mane is to enable th e animal to cheat, why don't youngsters cheat? After all, they would gain more from pretending t o be larger than full-grown adults would. If, on the other hand, the mane is actually a handicap, onl y monkeys wit h larger skulls could affor d t o make their head s look smalle r in this way. 10 Crests of feathers similarly reduce the appearanc e o f the beak, especially if the crest stand s up, lik e a hoopoe's. Th e bill is an important weapon for birds, an d a larger bil l i s a greater threa t tha n a smaller one . Again , only a bird wit h a large beak ca n affor d a large cres t tha t make s its beak loo k significantl y smaller . Bot h manes an d crests , however , presen t othe r handicap s a s well: crest s make the di rection o f a bird's gaze evident, an d manes interfere with vision .

HANDICAPS THA T INTERFERE WITH VISIO N Sight i s the mos t importan t sens e i n man y species, an d anythin g that interfere s with a n animal' s abilit y to se e is a very serious handicap. Compare , for example , the field of vision of a young orangutan with that of an adult male: the head of the former resemble s tha t o f a human child , an d th e youngste r ha s a wide field of vision. Adult male s hav e sunken eye s an d a fleshy fram e aroun d thei r face s tha t impairs their peripheral vision; the older the male, the narrower hi s field of vision. Only a large male can afford t o advertis e his social standing with a handicap that prevents hi m fro m collectin g informatio n fro m al l around him. Such a handicap i s too much for a young male. Th e mane s o f adul t lion s an d o f male langurs, baboons, an d other monkey s restrict visio n a s well. In ou r discussio n o f sexua l display, we describe d the bulge that the courting whit e pelican develops on the bas e of its beak just in front o f its eyes. The bulg e blocks the pelican's view offish nea r the tip of its beak, proving t o potentia l mate s tha t i t i s a n excellen t enough fisher to succeed even when thus handicapped; the bulge disappears by the time the pelican has to fish for it s young. One ca n als o fin d bulge s tha t interfer e

Body Parts that Serve as Signals 8

5

with vision in some plovers and in Cracidae, large birds that gather fruits from th e tops o f tropica l trees . Asiati c hornbills liv e a t th e to p o f tropica l forests . Thei r gigantic bill s ar e adapte d fo r pickin g fruit fro m th e en d o f branches. Bu t som e have large horny cask s on their beaks , almost like anothe r beak , which overhang their eyes. The cas k impairs their ability to see birds of prey flying above and their ability to peck efficientl y a t fruit righ t in front o f them, and als o adds unnecessary weight to the beak. Bustards are large fowl o f the ope n deser t an d savanna . The neck o f the male is covere d wit h lon g feathers ; when h e displays , he stretche s his nec k fa r backward, an d hi s nec k feather s stand up , formin g a large ball. H e canno t se e what is aroun d hi m an d dance s blind . Onl y a mal e secur e i n hi s knowledg e o f hi s surroundings an d confiden t in his abilit y to overcom e predators an d rival s alike can affor d t o danc e fo r minute s a t a time without bein g abl e t o see , in ful l vie w of al l comers. In fact , the bustard's blind dancing might b e too great a handica p for i t to bea r i n environments other tha n th e ope n flatland s i n which it lives; the specimen a t th e Te l Avi v University zoo repeatedl y bumped int o it s fence whil e displaying.

BODY PARTS THAT EMPHASIZE THE DIRECTION OF GAZE Another aspec t of vision that signal s often sho w is the directio n a bird o r animal is looking in . Th e California quail has a long blac k cres t tha t curl s forward. We were puzzle d b y it: the quai l we were familiar with, th e Europea n quail , spend s practically all its life o n th e groun d i n dens e vegetation. What coul d suc h a bird show off with a crest of feathers? The myster y was solved when we saw the California quai l i n its natural habitat . Unlike the European quail, th e California quai l gives it s lou d cal l while standin g o n suc h high observatio n post s a s the top s of rocks or bushes. Its blac k cres t i s visible fro m afa r an d show s th e directio n of its gaze. When the cres t appear s a s a straight lin e extending upward from its head , the quai l is gazing directly at an observer; when the crest appears arched, the quail is looking to the side. But by the same means, an observer can tell where th e quai l i s not looking. The shap e o f its crest allows the quail to display its confidence; plumes on guardsmen's helms an d Roman centurions ' helmet s may have evolved t o fulfil l th e sam e function . In Israel , ther e i s only on e lar k tha t ha s a crest ; th e crested lark is also the only lark in Israel that sings standing on a hillock, a post, a fence, o r a bush, a s well a s in flight. The cres t shows clearly what direction i t is facing ; in flight, a crest is not seen and is not an effective showing-

86 M E T H O D

S O F COMMUNICATIO N

off signal . In human s and i n som e monkeys, noses sho w the directio n o f an individual's gaze. The nos e is more develope d i n males than i n females, an d mor e in adults than i n th e young—a n indicatio n o f its use fo r showin g off. Small tufts of feathers o n bot h side s o f the head , a s in owls an d othe r birds, an d shor t horns , such as the bony protrusions on the heads of giraffes, als o let an observer see from afar whic h way the animal's head is turned. The arche d horns of gazelles, antelopes, and other suc h animals also show off the directio n o f their gaze ; we found tha t ou t a t the Shushlu i Nature Reserv e in South Africa. Till then we had always toured African reserves by car. At the Shushlui, a student doin g researc h offered Amot z a chance to sightse e on foot. Unpro tected and unarmed among lions, rhinoceroses, and buffaloes, Amotz couldn't hel p but liste n t o ever y sound an d kee p a constan t watc h o n hi s surroundings . Th e animals were no t use d t o th e sigh t o f humans walking; antelope s barke d a t th e sight. Becaus e of the distance , neithe r thei r eye s nor eve n th e marking s on their faces coul d b e seen, but thei r horns always looked eve n and distinct to Amotz and his guid e an d followe d the m around , proo f tha t th e antelope s were lookin g a t them. The rea l adventur e o f th e tou r cam e i n a n encounte r wit h a square-lippe d rhinoceros. Square-lippe d rhino s have two horns: the larg e frontal hor n serve s as a weapon; Amot z had a hard time trying to figure out th e function of the secon d smaller horn behind it—unti l the rhin o cow charged. The guide had assure d Amotz that square-lipped rhinos are usually peaceful, and that even if they charge, one can usuall y deter the m a t th e las t minute , o r escap e be hind a tree. They came within a hundred yards of a rhino cow and her young calf. The guide, who probably wanted to sho w off , suggeste d gettin g nearer . Amot z ha d n o choice bu t t o com e along ; being lef t o n hi s ow n was an even more frightening prospec t than the rhinos. The rhinocero s co w snorted, stamped , an d charged . Amotz an d th e guid e clappe d thei r hand s to dete r her , and indeed at about twenty yard s she swerved aside. The instant she turned, the small horn coul d be seen : unti l then it had been hidde n behind th e large, threatening fronta l horn . I t was only then tha t it became clea r that th e functio n o f th e rhino's smaller hor n i s to show what direction its owner is facing. The smal l horn lets one tell whether a rhinoceros i s confident or hesitating. A rhino with a well-developed back hor n canno t hide it s intentions; an y hesitation, any change of direction or glance sideways i s unmistakable. Such horns function this way only in an open landscape , where they can be see n fro m afar . Th e horns of Asiatic rhinos, which live in dense vegetation, do not sho w the directio n of the rhino's gaz e in a similar manner, and indee d thei r bac k horn s hav e degenerate d into small horny bumps.

Body Parts that Serve as Signals

87

BODY PARTS THAT HANDICAP FIGHTING The special snout or proboscis o f the elephant seal hangs down and covers its face. The snout prevents mature males from seeing objects right in front of their mouths, and the y canno t fee d o r bite rivals without swingin g it aside . Th e elephan t seal' s snout, like a cock's comb, can be grabbed and torn by a rival. The larger the snout, the bette r proo f i t i s of th e male' s status : only a large, strong , an d experience d male can afford t o fight burdened wit h suc h a snout. The puffi n and a few other member s of the Alcida e family nes t i n burrows ; thes e bird s develo p a specia l horny layer around their beaks during breeding season. Richard Wagner, who studied th e razorbill—a member of thi s family—foun d tha t th e residen t o f th e burro w protects it from withi n by presenting its beak to a rival in suc h a way that th e intrude r ca n grasp it. 11 The de pressions i n the rough , horny sheath o f the bea k le t rivals get a good gri p o n it . If the beak s were smooth, as they are outside breedin g season , rivals could no t ge t a good hol d o n th e bea k o f a puffi n insid e it s burrow . Only a bird sure of its strength ca n afford t o let its rival grab it effectively . Th e siz e of the beak' s horn y sheath, th e numbe r o f grooves in it, an d th e decoration s tha t sho w i t of f t o bes t advantag e al l increase with age . Members o f the Alcida e family wh o do not nes t in burrows d o not develo p suc h sheaths. The bear d of the mal e ibe x puzzle d us for many years. Clearly it is a signal: adult males have them, bu t not female s or young . But what doe s th e bear d sho w off? I t i s not heavy ; i t ma y chang e somewhat th e ap pearance o f it s owner' s hea d bu t i s dwarfe d i n thi s respect by the adult male's gigantic horns. The solution came from a totally unexpected direction . Giora Hani, who studied leopards at the Em Gedi Nature Reserve, found tha t ibexe s ar e a n importan t par t o f the leop ard's diet . A leopard ca n overcome even a large ibex: it grabs the ibex by its face, cover s the ibex's nose and mouth wit h it s ow n mouth , an d suffocate s it. 12 Anything , suc h a s a beard, tha t makes it easier for a leopard to grab hold o f the ibex's face makes it more difficul t for th e ibe x to escape . An ibex that grow s a beard is showing off its contempt fo r leopards an d its confidence in its ability to escape them.13 Another such body part is the loose skin on the lower jaws of adult male moose; wolves go for this dangling skin when the y attack moose. The heav y branched antler s o f deer an d th e heav y arched or curlin g horns of

88 M E T H O D

S OF C O M M U N I C A T I O N

goats an d ibexe s ar e not efficien t fightin g weapons . I n fact , Darling 14 note s that the lunge of an antlerless deer is far more dangerous than that of an antlered deer. Yet it is clear that the larger and more branched a deer's antlers, the more attractive they are to females and the more of a deterrent t o other males. 15 From tim e to tim e on e finds deer withou t antler s (hummel ) or with straight antlers (pronghorns) . Darling remark s tha t i n battle , straight , sharp antler s ca n wreak a great deal of harm. But such deer disappear from natural populations: the straight antlers may be better weapons but ar e less effective when used to threaten and to display one's strength. It may well be, then, that heavy, cumbersome horns and antler s evolved no t simpl y as weapons, but rathe r a s handicaps that show off the strength of the animals that carry them. Females see m to be looking for males that ar e stronger overall, not fo r males who ca n kill better.

CAN BOD Y PART S EVOLVE TO REDUC E THE COST OF SIGNALS? The Handica p Principl e states that ever y signal imposes a cost on the signaler — and this is true of the body parts we have just discussed. But can body parts evolve to reduce th e cos t o f signals ? During breedin g seaso n mal e skylark s perform a s many as ten singing flights per hour. Moller 16 found that both the number and the duration of these flights vary greatly; a flight can take from hal f a minute to twenty minutes. H e als o foun d a revers e correlatio n betwee n wingload—th e rati o of weight t o win g area—and th e duratio n o f each bird's singing flights. Molle r cu t off som e o f th e win g feathers of certai n males , reducin g thei r win g are a b y 2 0 percent an d increasing their wingload markedly . He found that such males made fewer and shorter display flights. He therefore suggeste d that male skylarks evolved larger wing s tha t lessene d thei r wingload s a s "cost-reducing " features , to mak e display flights easier. The question is, are the large wings and smaller wingloads actually an advantage to males otherwise, suc h as when they are escaping raptors, or during migration? If so , then they are the equivalen t of the longbow—goo d for their own sake, and also good indicator s o f overall strength. In tha t case , the singin g flights show off the larger wings to advantage : after all , according to the Handicap Principle , signals—such as display flights—are more affordable t o stronger individuals than to weaker ones . If, o n the other hand, the larger wings are an advantage in display flights only and ar e a handica p otherwise , the n bot h th e wing s an d th e singin g flights are signals. I n tha t case, the cos t o f the displa y flights would consis t no t onl y of th e bird's investment in the flights themselves, but its investment in growing the larger wings and developing th e strength to use them, and the liability larger wings pose

Body Parts that Serve as Signals 8

9

in regular , nondisplay flight . I n eithe r case , cuttin g of f some of the win g feathers surely affects the displays flights but proves nothing about the purpose of the larger wings of male skylarks.

THE EVOLUTION O F HORNS AN D ANTLER S The Handica p Principl e ca n explai n on e o f the dilemma s o f evolution : th e first stages in the evolutio n o f new body members , such as horns, antler s o r feathers. The laws of natural selection asser t that only a mutation that benefits the individual who carrie s i t wil l sprea d throug h th e population . Bu t ho w ca n we explai n th e chain o f events tha t le d t o th e appearanc e o f horns o n a previously hornles s un gulate? I t doesn' t see m likely that a single mutation coul d creat e a pair o f sharp, hard, symmetrical horns tha t serv e their beare r efficiently . The firs t stag e in th e evolutio n o f horn s an d antler s ma y have been th e ap pearance of small, soft bulges, like the present-day giraffe's, which served as signals showing the directio n o f an animal's gaze . This explain s the locatio n o f horns on the forehead, as well as their general shape and symmetry. At this early stage horns would hav e been useles s i n a fight—i n fact , the y would hav e been a hindrance . Female Grimm' s duiker s hav e n o horns ; the y figh t eac h othe r b y butting thei r heads. The male s have small, slender horn s an d fight their rival s by kicking with their front legs . If they tried t o butt thei r heads , thei r daint y horns would break . Such delicat e horn s ar e clearl y not weapon s bu t rathe r a signal showin g of f th e male's quality . Once horns attained their place and general shape as signals showing th e directio n o f gaze , they coul d the n evolv e int o efficien t fightin g tool s b y means of mutations tha t improve d thei r strengt h an d sharpnes s (utilitaria n selection). For example, the straight , hard, sharp horns of the females of several gazelle species are a useful, effectiv e weapon against small predators. Bigger, stronge r male s ca n us e stronger , bigge r horns , an d tha t i s jus t wha t utilitarian natura l selectio n produce d i n the expecte d way . Then signa l selectio n took ove r again , and mad e horn s heavier and more elaboratel y spiraling , curled , or branchin g in orde r t o sho w of f such males ' strengt h an d stamina . The resul t was the heavy, back-curved horns of the male ibex, the spiraling horns of the kudu, the branching antlers of male deer. Such heavier, convoluted horns and antlers are

9O M E T H O D

S O F COMMUNICATIO N

better indicator s o f each male' s overall quality, and bette r instrument s for ritua l fighting, eve n though the y are less effective a s conventional weapons . Changes occurring in this later stage of signal selection were confined to males: Females, a s we have seen, have to inves t more i n eac h of their offsprin g tha n d o males, since eggs are larger and fewer than sperm; female mammals have to invest even mor e i n pregnanc y an d suckling . Thu s the y have less to inves t i n showin g off. The y als o have less to gain : the numbe r o f offspring eac h female can bear in her lifetim e is far more limited, whereas the numbe r o f offspring a male has de pends far more on the number of partners who mate with him. Thus females have to g o fo r qualit y an d concentrat e o n carefull y choosin g th e father s o f thei r offspring. Of course, thi s means that males are much likelier tha n females to end u p with no offsprin g at all. Thus male s show off considerably more than females do , particularly whe n th e female s d o no t nee d thei r hel p i n raisin g offspring . Thi s explains the great size and weight of male horns, and the wealth of complex shapes they take ; by contrast , females, when horne d a t all , ten d t o hav e small , sharp , efficient horn s for protection agains t predators. But a further interestin g stage in the evolution of antlers occurred in reindeer : their heavy , branching antlers evolve d a n extra branch, which turned ou t t o be a useful too l fo r clearin g snow of f the lichen s the y eat i n winter . This i s probabl y the reaso n female reindeer—unlike other femal e deer—gro w antlers. The evolutio n o f horns an d antler s thus consiste d o f several alternating stages of natural selection. Originally, structures that signaled the direction o f an animal's gaze evolved by signal selection, an d their shap e and position wa s determined b y this function. These structure s became progressively stronger , evolving by utilitarian selection int o weapons. They continued t o change, becoming larger and more elaborate a s the y evolve d int o handicap s an d instrument s o f ritua l rathe r tha n actual combat . An d wit h reindeer , antler s furthe r develope d int o food-findin g tools. Still, throughout the saga, in all their various permutations, horns and antlers continued t o serve as signals that showe d the directio n o f an animal's gaze.

SIGNAL SELECTION AND TH E EVOLUTION OF FEATHER S Feathers, like horns and anders, present a puzzle to researchers of evolution. Birds clearly evolved fro m reptiles , an d feather s from scales . But i t is unlikely that on e mutation create d something as complex an d a s beautifully functional as a feather. Nevertheless, i n orde r fo r an y chang e t o sprea d i n a population, i t mus t b e a n improvement; one can readily see how a body member tha t has a particular function o r purpos e ca n evolv e t o bette r serv e tha t purpose , bu t ho w ca n a bod y member lik e a reptilian scal e evolv e graduall y into a totally differen t on e lik e a

Body Parts that Serve as Signals

91

feather? Th e transformatio n o f a scal e int o a feathe r ca n onl y hav e occurre d through a series of countless tiny changes, one after another ; each of these changes could spread among the population onl y if that change, on its own merit, enhanced the fitnes s o f an anima l tha t carrie d it—tha t is , the numbe r o f reproducing offspring tha t i t had . Ye t obviously , the change d scal e became a less efficien t scal e long before it turned int o a n efficient feather . One day we encountered a n article about the amazing structure of the feather, with stunning photographs taken throug h a n electroni c scannin g microscope . Could i t be, aske d the authors , that suc h a marvelous structure evolved through gradual changes? Or rather did a higher Providence creat e the feather for the function i t ha s today ? I t turne d ou t tha t th e articl e was published by a fundamentalist foundatio n t o disprove evolution an d to present creationis m as a scientifically valid paradigm . Of course , present-day natural scientists dismis s creationism , but w e stil l confron t th e question : Ho w coul d a scale evolve into a feather? Signal selection works counter to utilitarian selection; it therefore favors a series of changes that decreas e rathe r than increase the straightforwar d usefulness o f a body part . Th e selectiv e advantag e of each o f these change s is not efficienc y bu t rather effectiveness a s a signal tha t convey s reliabl e information. I t would be entirely logical to posit tha t some scales changed gradually into complex featherlike decorations. A heron nowadays survives well even though some of its feathers have become threadlik e decorations—signals—an d n o longer hel p th e heron t o fly or to insulate its body. In just the same way, some of the scales on an ancient reptile's front leg s might hav e evolved into loose, long, decorativ e plates that impaired its ability to walk and climb , an d wer e favored by signal selection precisel y because such a handicap was reliable evidence of the reptile's walking and climbing ability. Such decorative scales might eventually turn out to be useful for gliding from tre e to tree—opening the doo r t o the evolution of feathers by utilitarian selection. This i s not a new idea. Man y before u s have recognize d the crucia l rol e o f "sexua l selection " i n producin g nove l trends i n evolution—an d a s w e hav e seen , wha t Darwi n termed sexual selection is a subset of the process we call signal selection. Our discover y is not the process itself but rathe r the logic behind it . We sa y that b y its very nature, the reliabilit y required i n signalin g militates against efficiency . Handicap s increase the reliabilit y of signals not despite th e fac t tha t they make an animal less efficient, but because they do. Any improvement i n a signal must be accompanie d b y a cost t o th e signaler—that is , it mus t mak e th e signal' s bearer less welladapted t o its environment .

92

METHODS O F COMMUNICATIO N

Handicaps ar e no t random . Specifi c handicap s evolv e a s signals in orde r t o demonstrate specifi c qualitie s an d abilitie s reliably , an d thi s channel o f natural selection—signal selection—runs contrary to simple selection fo r efficiency—util itarian selection . Throug h thi s channe l ca n come the first stages in the evolutio n of new body members, which start out by serving as signals, and which then acquire new functions b y gradual selection for efficiency. W e believe that it is through this alternating proces s that th e deer got its antlers and the bird its feathers.

C H A P T E R

8

THE US E OF COLO R FOR SHOWING OF F

C

olor ma y camouflage a n animal, in which case its benefits are obvious an d need no explanation. Colo r ma y also affec t hea t absorptio n o r reflection. There ar e case s i n which colorin g itsel f signal s health , a s d o th e re d o f cocks' combs , the colo r o f human lips, an d similar structures that show off blood circulation, an d th e colo r o f flamingoes and othe r bird s tha t ge t their hu e fro m foods rich in carotenoids.1 Carotenoid s ar e active molecules an d can cause damage within th e body ; we speculate that hig h levels o f carotenoids i n the blood a t th e time the bird i s growing its feathers may well be a handicap tha t can be sustained only by a high-quality bird. But color , unlik e markings , rarely convey s a specifi c message . Rather , colo r determines the conspicuousness and precision o f messages expressed by markings, decoration, shape , an d movement . Th e messag e may be bol d an d striking , clear even fro m a distance, o r cryptic and subdued , readabl e onl y at close quarters . Traveling i n differen t habitat s ove r severa l continents , w e watched bird s an d animals an d constand y aske d ourselves : Wha t i s it tha t w e see in each ? Wha t i s the anima l showin g off ? Wha t woul d hav e change d i n ou r perceptio n o f tha t particular individual under these particular circumstances if the colors of its markings were different ? W e tried t o understand ho w certain colors appearin g over an 93

94 ME

T H O DS O F C O M M U N I C A T I O N

animal's whole body , o r o n larg e parts o f it , affec t th e distanc e a t which a shape or pattern ca n be seen , and the clarit y of the message. 2

BLACK I N TH E DESER T The first inkling we had that there might be a logical reason tha t a particular bird was a certai n colo r cam e fro m a pape r publishe d i n 195 0 b y tw o outstandin g ornithologists, May r and Stresemann. 3 They pointe d ou t tha t wheatears living in open habitats , such a s the desert wheatear and the isabelline wheatear, are cryptic in coloring, while wheatears that live in hilly country and inhabit deep gorges, such as the mournin g wheatea r an d th e white-crowne d blac k wheatear , have conspicuous black and white coloring. They also pointed ou t that the tails of species living in open habitat s tend to be black, while the tails of those living in mountains have more white in them . We soon notice d tha t othe r deser t bird s ten d t o have black body parts tha t in birds of more temperate habitats are usually white: skylarks have white on the sides of thei r tails, while th e side s o f desert larks ' tails ar e very dark (thoug h not quit e black). The Cretzschmar' s bunting in the Mediterranean area s of Israel has white tail feathers, while the deser t buntin g ha s no white in its tail.

This rule applies to desert mammals as well. Male ibexes have huge black horns and blac k beards , bot h ver y conspicuou s agains t th e beig e backgroun d o f th e desert; sometime s th e black horns an d beard ar e the only clu e to the presenc e of this otherwise ver y cryptic animal. During courtship , th e male ibex grows magnificent black fur on its chest. Black shows up clearly in the desert in all lights, except for th e ver y rare cases in which the deser t backgroun d itsel f is very dark. Black is more clearl y visible than any other colo r eve n in the middle o f the day , when th e hot air blurs one's vision. At dusk, when the gray body of a blackstart merges with the twilight , the spreadin g black fa n of its tail can still be seen . In the desert s o f the Nege v an d the Sinai , the black garment s and black goats of the Bedoui n ar e conspicuous from fa r away; it occurred to us that this conspicuousness might be precisely the reason these people wore black clothes an d chose black goats . Whe n we mentioned thi s ide a to researcher s who stud y the physiological connection betwee n color s an d temperatures, the y smiled an d pointed ou t that i n th e Sahara , the Bedoui n dres s i n light colors ; wouldn' t thi s indicat e that the Bedouin selec t th e colo r o f their clothin g arbitrarily? When peac e with Egyp t allowe d u s to visit th e Western Desert , th e Egyptia n

The Use of Color for Showing Off 9 5 part of the Sahara, we realized that there was a crucial difference between i t an d th e deser t w e knew in Israel and th e Sinai . The Sahara is far dryer , an d som e parts o f it see no rai n whatsoever for year s on end . I n th e Nege v an d i n th e Sinai , Bedouin liv e with thei r herd s al l over th e desert ; i n th e Sahara , they live i n crowded oases . They ma y go into the deser t t o trave l from on e oasis to the next , or to mak e use of temporary grazing ground s that appea r afte r unpredictabl e rains ; but the y don't live in th e open desert . Since the ecological and social conditions in the Sahara Desert are s o differen t fro m thos e i n th e desert s o f the Nege v an d th e Sinai , it follows that th e background colors , the distanc e a message must carry, and perhaps even the actua l messages conveyed ar e also different. I n short , the fac t tha t dweller s in these two differen t desert s favo r differen t color s actuall y supports our clai m that there is a relationship between th e environmen t and th e color s o f advertisement. One needs to know the advertiser, the message conveyed, and the conditions unde r which i t i s conveye d befor e drawin g hast y conclusion s abou t th e us e o f certain colors.

BLACK AND WHIT E IN OPEN SPACES Black and white are the dominant colors in open spaces: in the deserts, on steppes, at the seashor e and o n th e ope n sea , and i n the sk y above the canop y of tropical forests. It is unusual in these spaces to see red, yellow, blue or orange birds (though there ar e some exceptions , lik e bee-eaters and rollers) . Black and white ar e both used for long-distance advertisement; each of them has its benefits and drawbacks. What i s the differenc e between blac k an d white , an d wha t i s each good for ? We hi t upo n th e answe r whil e drivin g o n a highway. Along th e roa d stoo d tele phone pole s cappe d wit h smal l porcelain insulators . Some of the insulator s were black, others white, in no particular order . We thus had a chance to look at black and whit e object s of the sam e shape an d i n the sam e light. Sometimes the background was the sky , sometimes dark rocks, sometimes greenery. We soon realized that we could se e the white insulator s from a greater distance , but tha t we could see the shap e o f the blac k one s more clearly . A white surfac e i s seen over longe r distances, sinc e i t reflect s most o f th e ligh t hittin g it ; ye t that ver y reflectiveness seems t o caus e th e edg e o f a white body t o look somewha t blurred. Th e edg e of a black object, on the other hand, appears bolder, probably because it absorbs all the light that hits it, thus creating the greatest possible contrast between the object and its background, which usually reflects som e light . Birds that advertis e their presence by perching for long periods in high places tend t o be black: ravens , drongoes, blac k eagles , and so on. Since black provide s

96 METHOD

S O F COMMUNICATIO N

the greatest contras t with most backgrounds, it shows these birds' shape s to perfection. Th e blurrin g o f the edge s o f white bodie s ma y explain wh y hardly any small bird s ar e completel y white. Presumably , such birds a s swans or egret s ar e large enough that a minor blurring of their outlines is insignificant fo r them. White is especially effective in advertising movement: a soaring flock of pelicans or gulls is revealed b y the light reflected off the white patches on their wings and bodies a s the y turn . Th e reflecte d ligh t ca n revea l speed , win g movement , an d other factors . I n fact , man y birds have white in their wings or tails that i s visible only when the y are i n flight. The blurrin g cause d by white i s not a n issu e here, since these birds' motio n itself blur s their outlines . In Australia , we visited friend s whos e daughter came home cryin g fro m he r firs t balle t lesso n becaus e th e teacher scolded he r fo r not wearin g the traditional black leotard. The teacher explained tha t without the black leotard, the exact position and line of the girl's body could not b e see n a s clearly and s o were harder to correct . I n performance, o n the other hand, ballet dancers are more likely to b e attire d i n whit e o r i n ver y ligh t colors : thi s increases the audience's ability to appreciate the dancers' movements rather than isolated positions .

COLORS IN FORESTS AND O N CORA L REEFS In the tropical forest, one finds red, yellow, and blue birds; few are black or white, and i t usuall y turns ou t tha t thos e fe w spend significan t amount s of time abov e the canop y or o n th e ground , outsid e th e patchwor k o f ligh t an d shado w tha t characterizes the canop y itself. In th e dens e vegetation, birds see each other only over shor t distances ; white and black thus lose most o f their advantages . Indeed, the patchwork of gleaming rays of sun and dee p shad e within the canop y distorts white, black, and black-and-white shapes and patterns. But ther e i s still a need t o sho w of f both shap e an d movement . Yello w an d red ar e easily distinguished bot h fro m patche s of sunlight and fro m th e shadow s of leave s and branches . Yellow has most o f the reflectivenes s of white, reflectin g all light color s othe r tha n blue , ye t it i s differen t enoug h fro m whit e no t t o ge t mixed u p wit h ray s o f sunlight ; various shades o f yellow show of f movement in the forest . Re d contrast s bes t wit h th e gree n canopy . It reflect s littl e ligh t and , while dark , i s differen t enoug h fro m patche s o f shad e no t t o b e mixe d u p wit h them; forest bird s tha t advertis e by perching are red. O n th e forest floor, where the light i s dim an d ther e ar e very few patches of sun, white an d blac k are again the dominant colors for advertisement. We also find a great deal of green in forest birds—but thoug h thi s colo r look s brigh t an d prominen t i n zoos , i n th e birds '

The Use of Color for S

97

natural habitat it is actually cryptic and conceal s the bird agains t the gree n background o f the forest . The influenc e of environmen t o n th e selectio n o f colo r ma y explai n a phe nomenon note d b y Diamond 4; mixed-specie s flock s o f bird s i n tropica l forest s are usuall y made u p o f specie s tha t hav e simila r colors . H e describe s flock s o f birds who are mostly black an d brown o n the forest floor; green and yellow birds, mostly insect-eaters , wander te n t o twent y feet up . Eac h suc h flock roam s a certain leve l o f the jungle , and w e think the birds ' color s sui t th e par t o f the fores t in which they are active: the to p o f the canopy , halfway down , or the forest floor. Since bird s o f whatever specie s o n a give n "story " hav e t o dea l wit h th e sam e conditions o f light an d distance , they use simila r colors . We find the same pattern amon g fish. In open water most fish are gray , black , o r white ; o n th e cora l reef s an d i n kel p forests, wher e long-rang e visibility is obscured an d shadow s and patches of light abound, bright colors enable fish to show off a t shor t range. Once we figured out thes e "rules, " we made up a game, using the description s o f birds' color s in guidebooks to guess the habita t an d som e o f th e behaviora l traits o f specie s un known t o us . When w e compare d ou r prediction s wit h th e knowledge o f local birders, we usuall y found a good match. We wer e therefor e surprise d when tol d tha t th e golde n bowerbir d display s near the groun d i n th e dee p rai n forest . W e woul d hav e expected thi s bright yello w bird t o displa y in the well-lit upper part s of the green canopy. Fortunately, we were able to visit Gerald Borgia at his study area near Atherton, Australia. We waite d fo r a long tim e in blind s nea r th e bowe r o f a golden bow erbird, hearing its calls without seein g it . Suddenly the bowe r was lit by a ray of sunlight. Th e mal e cam e flying down th e sunbea m a s though h e wer e par t an d parcel of it, and landed on his perch, a horizontal branch joining the two pyramids of stick s tha t mad e up hi s bower. Th e ligh t was reflected ont o th e bir d b y light gray lichens, whic h he ha d place d o n th e sam e branch. Th e bowerbird' s golde n color now made sense; but ho w could sunligh t penetrate tha t deep into the fores t floor? W e late r learne d tha t mal e golde n bowerbird s meticulousl y pick of f th e leaves above their bowers t o let the light penetrate at the angl e they desire.

THE USE OF TWO COLOR S Blackbirds, ravens , an d cormorant s are all black; egrets and swan s ar e all white. But more often birds are two colors or more. Two contrasting colors side by side— often o n tw o distinc t component s o f the sam e body part, suc h a s a tail or wing, as we saw in chapter 5—can show off shape and relativ e dimensions. Parts of the

98

METHODS OF COMMUNICATION

body that are seen while the bird is perching are most often black ; white and light colors ar e ofte n hidde n whe n th e bir d i s at res t an d reveale d onl y when i t i s in flight. Som e species, though, have markings o f both colors that can be seen when the bird i s at rest. This allow s the bird t o advertis e in all lights: the white belly of the great gray shrike and of the wheatear glisten in the light when the sun is at the observer's back , but whe n the bird is perched between th e sun and the observer , the blac k a t the sid e of its wings helps defin e it s outline. Sometime s the messages are aime d onl y at nearby observers: the zebra' s stripes, which accentuat e bodily features, merge into a cryptic gray at a distance. Two contrastin g color s o n a bod y reduc e th e overall impression it makes and the distance at which its shap e ca n be see n distinctly ; it ma y also distor t the overal l shape . Fo r tha t reason , w e ofte n fin d a "compromise color" : on e tha t i s dar k enoug h t o show the animal' s shape, yet reflects mor e light than black o r dar k red . I n ope n habitat s an d abov e th e forest canopy the most common compromise color is gray; insid e th e fores t i t i s mostly beige, orange , or brown. O n th e forest floo r one finds a great dea l of reddish-brown, whic h i s th e colo r o f man y fores t mammals includin g squirrels , the commo n re d do g of India , an d th e sambar , a n Indian fores t deer , all of whom liv e in th e depth s o f the jungle . The sambar's undertail "flag" is light brown, unlike the white "flag" of deer in open spaces. Another recurring pattern is the presence of two similar but not identical colors to advertise one message over a short distance and another over long distances. A nearby observer sees the details; over distance, both colors merge into one message. The distance that the detaile d message travel s varie s wit h th e siz e of the patche s of colo r an d wit h th e intensit y of the contrasts . Th e dar k re d patc h o n the red winged blackbird's wing is very prominent a t close rang e but doe s not lessen th e effect o f the bird's silhouett e when it is seen from a distance.

GLOSSY COLORS AND MOVEMEN T Movement i s very effectively shown off by the glossiness that occurs when incident light is reflected in a single direction instead of being widely diffused. Gloss y colors are produced not by pigments but by physica l properties of the feathers , scales , or cuticle. Because of physical constraints, most reflective colors in nature are blues and greens—the shortest waves in the visible spectrum. Since the reflection of light is directional, a surface tha t look s gloss y from on e directio n reflect s n o light an d therefore look s dar k from al l other directions . Because of this attribute o f glossy

The Use of Color for S

99

surfaces, th e sam e coloring ca n advertis e shap e an d movement : small , fast-flyin g birds suc h as hummingbirds an d sunbird s ar e glossy; the dar k colo r define s thei r shape clearly , while the glossiness provides information about their movement.

EXCEPTIONS TO TH E RULES There are exception s to the rule s we hav e outlined : bee-eater s and roller s are colorful bird s tha t live in ope n habitats ; many finches an d bunting s (calle d sparrows in the Ne w World) tha t live in open habitat s ar e red an d yellow , and ther e are black and white flycatchers that live in woodland an d forests. Once one knows the bird, an d what its habitat is like in all seasons, however, one can often explai n such exceptions . Fo r example , th e yellow-bille d finc h (Melanoderma xanthogrammd] tha t lives in open habitats in the Patagonian Mountains of South America has a black chi n wit h a yellow outline, rathe r tha n a white outline, a s one might expect a bird living in the open to have. Watching the finch in early spring, when the birds arrive to stake their territories, however, we realized that at that crucial time the bird s often stoo d and san g on rocks where there were many patches of snow. Whit e woul d hav e bee n les s effectiv e agains t suc h a background; yello w worked muc h better . A relate d species , M . melanoderma, whic h live s a t lowe r altitudes, below the sno w line, has a black chin outlined wit h white. Another example: the black and white flycatchers (Picedula spp.) arrive at their breeding grounds when many trees of the forests and woods are still bare of leaves. A close r look a t the place s i n which the y display may well sho w that a t the star t of breedin g season , when effectiv e displa y is crucial , th e ligh t i n thei r habitat i s similar t o tha t o f open country—eve n thoug h the y are forest birds ; thus th e us e of black an d white . Another possibl e complicatio n i s th e effect s o f ultraviole t light , t o whic h some bird s ar e sensitive , unlik e humans. 3 I t i s ver y possibl e tha t som e color s look differen t t o bird s sensitiv e to U V tha n the y look t o us . Thi s ma y explai n the color s o f kingfisher s an d bee-eaters . O n th e othe r hand , th e amoun t o f UV light i s usually minimal, especially in forests . This subject requires more detaile d research.6 The hypothese s presente d i n thi s chapte r ar e base d o n ou r subjective impressions of the birds we watch. Still, the search for logica l explanation s regardin g the color s of particula r birds an d animals has taught us to see their lives and habitats with new eyes. Our observation s tend to raise questions other than thos e deal t wit h i n most studie s o f color , an d w e hope our suggestion s will lead to new and fruitfu l research .

This page intentionally left blank

C H A P T E R 9

CHEMICAL COMMUNICATIONS

M

any organisms, bot h mono- and multicellular, communicat e by secreting chemica l molecules and mak e decisions based o n chemica l molecules secreted b y others. Suc h chemicals , which are produced b y one individual an d influenc e th e action s of othe r individuals, are calle d pheromones. In order to carry reliable messages, pheromones hav e to meet the three conditions we have established fo r signals : they have to b e costl y to th e individua l sending them; th e cos t ha s to be more of a burden t o a dishonest communicato r tha n t o an honest one; an d ther e must be a logical relationship between the specifi c cos t of the signal and the message conveyed by the signal .

PHEROMONES IN BUTTERFLIES AN D MOTHS : CHEMICAL HANDICAPS We touched briefl y on chemical communications in chapter 3, when we discussed arcteid and danaid butterflies. Man y other butterflies of the Danaidae and Arctei101

1O2 M E T H O D

S O F COMMUNICATIO N

idae families secrete pheromones that contain chemicals—often poisonous ones — that accumulat e in thei r bodie s fro m th e plant s the y eat , o r tha t the y collect i n other way s a s a defens e agains t predators. 1 Suc h poisonou s chemical s o r thei r derivatives in a male's pheromone s testif y tha t he is able t o deal with th e poison s and utilize them. Male pheromone s var y greatly from specie s t o species , reflectin g to a degre e the varieties of plants they eat. Female pheromones, on the other hand, are much more similar. 2 Researchers assert that female pheromones evolve d in order t o enable males to locate and identify females of their species who are ready to copulate. But i f that i s so, it i s strange that the femal e pheromones o f differen t specie s ar e so similar. At the least , one would think , female pheromones woul d var y as male pheromones do . In th e discussio n o f species-specific markings in chapte r 4, we explained ou r conviction tha t marking s that ar e common to a group, an d hav e therefore been thought to identify th e group, actually evolved out o f competition members of the group engag e i n t o establis h th e superiorit y o f on e individua l ove r others . W e believe the sam e principle applies to pheromones that evolve in females of a given species. I t make s sense that female s compete wit h on e anothe r ove r males , each trying to attrac t male s and mat e with th e bette r ones . Sinc e males indeed follow pheromones t o fin d females , the pheromon e molecule s presumabl y give reliable information o n th e qualit y of th e femal e disseminating them. I f tha t i s the case , then ther e must be a true correlatio n betwee n th e qualit y of the femal e and th e quality or quantit y of the pheromon e sh e secretes; the cos t of the signa l has to be high enoug h t o preven t would-b e cheater s fro m abusin g it . I n short , th e phero mone has to be a chemical that would har m a female who overproduced it—an d the level a female can safely produce should correlate to qualities that are of interest to males. What kin d o f harm can overproduction o f pheromones caus e the female? We are now conducting theoretical researc h to answer this question. A t this stage, we believe that th e pheromone s concentrat e around th e female that produces them , penetrate her sensor y cells, and impair them. Female pheromones ar e mixtures of molecules, mostly long, unsaturated chains consisting of ten t o twenty carbon ato m groups , that usually contain a t their en d an activ e chemical group tha t i s water-soluble: an acetate , acid, alcohol, or alde hyde. It makes sense that the long chains of the pheromone, which are fat-soluble, can penetrate the fatt y membranes around the female's sensory cells. The sensory receptors on these membranes are proteins with a particular spatial configuration based o n a precise folding of protein chains ; we speculate that th e active segment of th e pheromones—th e acid , alcohol , o r aldehyde—ma y damag e the receptor s that let females sense various chemicals around them and thus impair the female's ability to react to her surroundings. The secretion of too much pheromone woul d thus dul l th e female' s senses. This dullin g o f the sense s would impai r her abilit y to evaluat e the male s courting her—and wha t i s th e poin t o f attractin g males if

Chemical Communications 1O

3

one canno t distinguis h amon g them? Indeed, Elli s an d hi s colleagues foun d tha t female cotto n leafwor m moth s who were placed in an environment high in female pheromones stoppe d reactin g t o mal e courtship ; th e numbe r o f copulation s fel l almost t o zero. 3 Mal e activity , o n th e othe r hand , wa s increase d unde r simila r conditions.4 What i s the connectio n betwee n th e secretio n o f female pheromone s an d th e quality of the female? It may well be that a female who is better able to manufacture new membranes to replace thos e impaired by the pheromones ca n therefore affor d to secret e more pheromones . I t ma y also be that th e abilit y to replac e membran e is correlated t o th e fa t reserves in a given female's body, which i n turn correlate s with her abilit y to lay eggs. In tha t case , the productio n of more pheromones, or of pheromone s wit h a higher concentratio n o f active molecules, would testif y re liably to the female' s ability to lay eggs, making her desirabl e t o males. The evolution of female pheromones probabl y parallels that of any other signal: to begi n with , female s secreted variou s fatty acid s o r relate d fatt y substance s o n the cuticle, like many other organisms. These chemicals were costly to a greater or a lesser degree , partly because of the damag e they inflicted o n living cells; but th e protection the y offered wa s worth th e cost . Still , th e amount s each female coul d afford t o secret e woul d vary , and i t is reasonable t o assum e that there was a correlation betwee n th e quantit y of these chemica l secretions an d th e qualit y of th e female secreting them. Quite likely, males evolved receptor s sensitive to the chemical mixture s tha t bes t reflecte d a female's quality. (Thi s would b e analogou s t o the "mind-reading " i n the developmen t o f ritualization, at which stag e observers note functional motion s or postures that precede certain actions but which are not intended t o signa l thos e actions. ) Onc e tha t happened , the n mutation s causin g overproduction o f these chemicals would benefit their bearers. At that point, signal selection took over, and picked a s pheromones th e chemicals that impose the most telling handicaps and thus provide th e most reliable information about th e quality of th e female. Although there is a great deal of similarity between th e pheromones o f females of various species, each species has its own special mix of chemicals, which can be used to identify it. The male's brain has receptors adapted to the specific chemicals typical of a female o f his species; a few molecules of chemical are enough to cause a reaction i n the male's brain . This would see m to support th e idea that the pheromones evolve d i n orde r t o identit y the specie s o f the female , bu t w e think oth erwise. Afte r all , th e female s themselve s hav e receptor s tha t reac t t o chemical s secreted b y the plant s the y lay eggs on, plant s that thei r offsprin g consume ; yet it is highly unlikely that the plant s secret e the chemical s in order to announce thei r presence an d their identit y to their pests ! We thin k that th e typica l mix of pheromones fo r eac h species is a by-product of th e species ' ow n distinctiv e metabolism . We ca n draw an analogy to alcoholi c beverages that humans consume. Each nationality has its characteristic drink: Russians typicall y imbibe vodka ; Serb s prefe r slivovitz , and Japanes e lik e sake . But

1O4 M E T H O D

S O F COMMUNICATIO N

nobody claim s that th e various alcoholic beverages evolve d s o that the nationality of their consumers could b e determined. Rather, each develope d ou t o f a people' s simple desire fo r a n alcoholic beverage . The specia l characte r of each beverage is a by-produc t o f th e differenc e i n th e ra w material s an d technology availabl e in each culture. 5 The specia l physiology and die t o f each species , then , affect s th e compositio n of the femal e pheromone s o f each species . Bu t tha t substanc e evolves as a pheromone i n th e sam e way that othe r signal s evolve : i t deliver s a message about th e genuine qualit y of the individual that secretes i t in a way that cannot be profitably faked. We assume that male moths choose their mates according to females' ability to secrete harmful—that is , handicapping—components of their pheromone. Ob viously, as in other case s of selection, moth s do not rel y solely on the compositio n of pheromone s t o choos e amon g candidates ; othe r factors , suc h a s the rat e an d timing o f pheromone emission , an d th e female' s movements, probably affec t th e final selection o f one femal e ove r another .

YEAST SEX PHEROMONES AND PROPHEROMONES: THE ROLE OF GLYCOPROTEINS Thus we see that primarily , pheromones advertis e an individual's quality , rather than taggin g him o r he r a s a member o f a specifi c group . Thi s le d us , togethe r with Edn a Nahon , Daniella Atzmony, and David Grano t to develop a new interpretation o f sexual pheromones i n yeasts. 6 That interpretation i s still largely speculative. But it deserves to be put forth here because it illustrates how the Handica p Principle ma y appl y t o communicatio n amon g single-celle d organism s jus t a s it does to signal s amon g more complex animals . Yeasts are unicellular organisms. Yeasts of the specie s Saccharomyces cerevisiae (baker's yeast ) hav e two genders , alpha an d a , that mat e with on e another . Eac h of the m secrete s a specia l peptide , an d ha s a specia l recepto r fo r th e peptid e secreted b y th e other . Detectin g th e peptid e o f th e opposit e gende r lead s th e individual receiving it to stop growing and begin th e processes tha t prepare it for mating. Matin g involve s mergin g th e tw o cell s togethe r t o for m on e organism . Mating in yeasts, as in other unicellular organisms, is an all-or-nothing proposition : each has only one chanc e to mat e in its life . Up to now it has been assumed that the peptide simply announces the presence of a n individual o f on e o r th e other gende r tha t i s read y t o mate . Bu t recentl y Jackson an d Hartwell 7 found that yeasts select the cells they mate with, and prefer to mate with on e that secretes more pheromone. The y suggested tha t the amoun t of pheromone secrete d b y each yeast cell advertises its quality.

Chemical Communications 1

05

The fac t tha t yeast s select thei r mates rather than matin g at random supports our assumptio n tha t eac h organis m advertises its quality, rather tha n its identity. But we doubt tha t the amount of peptide constitutes a reliable test of quality. The short peptid e tha t comprise s th e pheromon e i s not a harmfu l chemical ; it s raw materials ar e readil y available; and ever y yeast cell i s genetically programmed t o produce i t an d supplie d wit h th e necessar y information. The proces s o f protei n production in the cel l is an ongoing one, an d the pheromone peptid e constitutes only a minuscule part o f the yeas t cell' s output . Thus, i t is hard to imagine how a short, simple peptide ca n reflect th e pheno typic quality of the advertisin g cell. In addition , the concentratio n o f a chemical is not a goo d indicatio n o f quality , since a nearby cel l producin g smal l quantities would b e perceived a s equal to a more distant cel l producing much higher quantities. S o we starte d lookin g fo r indication s tha t other molecule s ar e involved in the communicatio n betwee n yeasts , molecule s tha t ca n testif y t o th e phenotypi c quality of the cel l by means other than shee r quantity. It turned out, when we surveyed existing literature, that the yeast pheromone— the peptide—is cleave d ou t o f a larg e protei n molecule , calle d a propeptide o r propheromone. Th e propeptid e o f the alph a gende r i s a glycoprotein—a protei n that carries sugars. We believe tha t glycoproteins are better suite d to testify to the quality of a cell than simpl e proteins , or peptides. Unlike the manufactur e of the amino acid backbone of proteins, all of which are produced by the same enzymatic machinery, the synthesi s of the specia l sugars of glycoproteins occurs by means of special enzymatic pathways that are activated only under specific conditions, which depend o n th e physiologica l state o f the cell . Fo r example , i n human s th e composition o f th e variou s sugar s on th e gonadotrophi c hormon e FSH , whic h i s a glycoprotein, varies with change s in the female menstrual cycle.8 The compositio n of sugar s on a give n glycoprotei n an d th e numbe r o f unit s o f suga r vary . Thi s microvariation, a s it i s termed, ca n reflec t specifi c condition s i n a way that pur e proteins cannot . I n othe r words, the complemen t o f sugar units on a specific glycoprotein ca n reflect reliabl y the conditio n o f a cell. In yeasts , it turns out tha t th e alph a propheromone i s essential to the proces s of mating , not jus t necessary to the manufactur e of the pheromone : i n lab experiments, yeas t cells that ar e manipulated s o that the y can produce th e alph a pro pheromone bu t los e the abilit y to cleav e the pheromon e fro m i t ca n still mate, if pheromone is added to the mixture ; but whe n scientist s prevent yeas t cell s fro m producing the propheromone, th e cells cannot mate even if supplied with the alpha pheromone.9 Two genes produc e alph a propheromone, an d eac h produces a different version: fou r molecule s o f pheromon e ar e cleave d fro m on e versio n o f th e alph a propheromone; a n identical pheromone molecul e and one that is slightly different are cleave d fro m th e othe r version . Al l natural population s o f yeasts have both genes, eve n thoug h th e pheromon e cleave d ou t o f th e tw o version s o f th e pro -

1O6 M E T H O D

S O F COMMUNICATIO N

pheromone i s nearly the same . It would see m that the orde r an d number o f molecules o f pheromone i n eac h of the tw o variants of the alph a propheromone are essential an d important, not accidental . What i s the rol e of the alph a propheromones i n the communicatio n that precedes mating between yeast cells? One en d of the propheromone ha s a hydrophobic segment— a segmen t tha t dissolve s i n fats , bu t no t i n water . Th e other en d contains the pheromon e segments . The sugar s of the glycoprotei n ar e connecte d to a middle segment. The pheromone segment s are cleaved off the propheromon e by enzymes and are secreted by the cell. The pheromone molecules reach the other gender, ar e sense d b y its receptors, an d trigge r the receivin g cell int o sendin g a copulating offshoo t towar d th e cel l secreting the pheromone . There is indirect evidence that complete molecules of the propheromone—that is, propheromon e wit h it s pheromone segment s still attached—ca n b e foun d on the outer membran e of the yeast cell that secretes the pheromone. Ou r conjecture is that th e alph a propheromon e i s anchored t o th e cel l membran e by its hydrophobic segment in such a way that the pheromone section s attached to it protrude outside th e cel l membrane, and tha t the y are held i n a specific spatia l configura tion—in which the sugar s in the middle segmen t of the molecule probably play a role. We thin k tha t these pheromone segment s may bind t o receptor s i n the othe r cell's offshoot , jus t a s free pheromon e does . I t may even be that both type s of the alpha propheromon e are needed for a good link : a bond form s betwee n the two alpha propheromones ( a heterodimer), an d then thei r peptide s link with a group of receptor s o n th e membran e of the receivin g cell. This woul d explai n why the pheromone i s manufactured by tw o gene s rathe r tha n one . I t make s sense tha t impairments i n th e geneti c o r phenotypi c qualit y of th e advertisin g cel l woul d affect th e number o f sugar molecules attached to the propheromone an d thus the spatial configuration of the propheromone i n that particular yeast cell's membrane. A change i n th e spatia l configuratio n coul d wel l affec t th e strengt h o f the bon d between th e propheromone s o n the signalin g yeast cell and th e receptor s o n th e other yeast cell, and thus affect mat e preference: an individual cel l weakly bonded to its would-be partner might detach itself to bond strongl y to a preferred mate. If this conjecture proves true, then the role of the pheromone peptide itself— a relatively small, mobile molecule—would be merely to draw the attentio n of cells of th e othe r gende r to the cel l that secrete s it, so that the y would com e an d "in vestigate" by sending a mating offshoot towar d it s membrane. The critica l "deci sion" whether to go ahead and mate with that particular partner would b e based not o n th e pheromon e bu t o n contac t betwee n th e propheromon e o n th e membrane of the one yeast cell and the receptors o n the other cell' s membrane. We hav e describe d ou r interpretatio n o f th e communicatio n betwee n yeas t genders in detai l because we see it as a general model: i t illustrates how cell s can advertise their phenotypic quality by means of peptides an d their propeptide pro genitors, which hold them in specific spatial configurations. Indeed, recent findings

Chemical Communications 1

07

show that the propeptides of other peptide s used in communication als o bond to the sam e receptors t o which the peptides themselve s attach.10 We believ e tha t mos t i f not al l chemical communicatio n betwee n individual s contains reliabl e testimon y about th e phenotypi c qualit y of the on e sendin g th e signal. Molecular structures ca n serve this purpose no less than bodily adaptations do i n highl y developed organism s suc h a s mammals and birds . Needles s t o say, we do no t clai m that yeasts make conscious decisions ; bu t th e proces s o f natural selection sa w to it that individuals who followed better lif e strategies, by means of whatever mechanisms, survived and multiplied, while those who did not perished.

CHEMICAL COMMUNICATION WITHIN THE MULTICELLULAR BOD Y Unlike unicellula r organisms , al l the cell s i n a multicellular bod y hav e the sam e interest. I t woul d see m a t firs t glance , therefore , tha t withi n suc h multicellula r organisms there is no need t o test the reliabilit y o f communication between cells , that such communication need only be clear and efficient. Ye t it turns out that the molecules use d i n communicatio n betwee n cell s withi n th e body—hormones , growth factors , and neurotransmitters—are similar to, and sometimes identical to, the molecule s o f pheromones use d in chemical communication between separat e organisms. Many of these hormones ar e glycoproteins or harmful chemicals—th e same kinds of chemicals that, in our opinion, serv e as the handicaps that guarantee that chemical communication between individuals is reliable. The question is, what is the rol e of such handicaps in communication within the multicellular body? 11 The multicellula r bod y has a division o f labor: the specia l role of some cells is to receive information from their surroundings, process it, and pass on instructions that change the behavior of other cells in the body. Thes e instructions ar e sent by chemical signals—hormones . I f for an y reason thes e instructions are wrong, th e resulting harm will affec t al l body cells, including both the sender and the receiver of th e signal . The genes in all the cell s of a given individual's body are basically identical, of course, an d s o the gene s for manufacturin g all chemical signal s used in the bod y exist i n al l body cells . Th e differenc e betwee n th e cel l whose rol e i s to sen d th e signal an d other cells is not on e of genotype—all cells in the body are of the sam e genotype—but rather of phenotype: where the particular cell happens to be in the body, an d wha t its state is. A mistake in sensing, in decision making, or transmission coul d caus e a cell to manufacture the wrong chemica l signal or to sen d tha t signal off at th e wron g time . Thus, i n orde r t o preven t mistakes , the signa l itself should testif y i n a reliable wa y to th e phenotypi c stat e o f the cel l sendin g it . A signal within the body can be reliable only if it cannot be produced b y cells of the

1O8 M E T H O D

S O F COMMUNICATIO N

wrong phenotype. In other words, we find a use for the Handicap Principl e withi n the multicellular body. It was recently foun d that free radicals lik e nitric oxide and carbon monoxide, which ar e very poisonous, ar e used i n communication between cell s in the bod y and ar e secrete d b y nerv e cell s i n importan t bod y part s lik e th e brai n an d th e heart.12 Thi s us e o f highl y toxic chemical s i n th e mos t crucia l bodily messages supports ou r view . O n th e othe r hand , the fac t tha t man y hormones an d many neuropeptides tha t conve y signals within th e brai n ar e shor t peptide s seem s to contradict the Handicap Principle: a s we said in the discussio n of yeasts, we don't think that short peptides have the capacity to provide reliability in communication. It ma y be, though , tha t i n a t leas t som e o f th e peptid e signal s in th e body , reliability i s provided not by the short peptide itself bu t rather by the prohormone (propeptide) t o which it is attached. 13 The propeptide , i n other words, may well provide reliabilit y i n th e sam e way that w e sugges t th e propheromon e doe s i n communication between yeas t cells. Another metho d fo r ensurin g reliabilit y in signalin g is used, i t seems , in th e EGF system . EGF i s a peptide tha t enhance s the divisio n o f cells in the wall s of blood vessels . Thes e cell s divid e whe n a damage d blood vesse l need s repair . I t turns out that EGF causes cell s to divide only when it is attached t o special sugars . EGF an d th e sugar s attac h to on e anothe r i n the intercellula r substanc e i n th e walls o f blood vessels; whe n th e tissue s are injured, EG F an d the sugar s ar e released together . The presence in the blood of EGF attache d to sugars is a reliable indicator tha t bloo d vessel s hav e indeed bee n injure d an d tha t ther e reall y is a need fo r repair. Like an y signals , chemical signal s between competin g organism s have to b e reliable—otherwise they will not be accepted by potential receivers . Chemical signals withi n th e multicellula r bod y als o deman d reliability : no t becaus e o f conflicting interests , bu t rathe r t o preven t mistakes. 14 I f w e ar e right , the n al l chemical signals—includin g signal s withi n th e body—wor k accordin g t o th e Handicap Principle . I t shoul d therefor e be possible t o find a logical relationshi p between th e structur e of molecules used in signaling and the messages they carry. This premise guides our current research into the messages encoded i n molecules that ar e used in chemical communication.

P A R T I I

I

THE HANDICAP PRINCIPLE I N SOCIA L SYSTEMS

This page intentionally left blank

C H A P T E R 1

0

TESTING TH E BOND

A

t the end of each workday, when we arrived at home, our little daughters used t o com e runnin g t o welcom e us . They would jum p up o n u s and demand tha t we take part in their game s or a t least tell them a story, no matter how exhausted we were. Our dog , too, would leap up and insistently push himself agains t us . An d i n th e morning s th e do g woul d mak e eve n mor e o f a nuisance of himself: he would li e down i n the doorway , forcing us either t o walk around him or to push him aside every time we went from the kitchen to the dining room. All of us, including the dog , knew that we belonged together an d were willing to d o a lot fo r eac h other. Bu t eve n in the mos t loving family, th e degree o f willingness an d abilit y to cooperat e ma y well chang e fro m da y to day . Thus, on e of the mos t important question s fo r eac h membe r o f a partnership i s the degre e t o which other partner s ar e willing to d o things fo r it at any given time. The eterna l questions "D o you still love me?" and "How muc h do you love me?" are especially urgent when on e o f the partner s ha s been awa y for a time, or i s getting read y to leave. Our childre n and our dog revealed t o us the existence of behavioral systems that enable individuals to collect information about the social bonds between the m and their partners.1 111

112 THE HANDICAP PRINCIPLE IN SOCIAL SYSTEMS

TESTING B Y IMPOSITION How ca n on e ge t reliabl e informatio n abou t th e qualit y an d strengt h o f a social bond? Action s tha t benefi t anothe r ar e likel y t o b e accepte d b y hi m o r he r fo r their own sake. The only way to obtain reliabl e information about another' s commitment is to impose o n that other—to behave in ways that are detrimental to him or her . W e ar e al l willing to accep t another' s behavio r i f we benefit fro m it , bu t only one trul y interested i n the partnershi p i s willing to accep t an imposition. As we shal l see , al l mechanism s use d t o tes t th e socia l bon d involv e imposin g o n partners. People who don' t lik e dog s will no t le t a dog jump up o n the m o r lic k thei r hands. Almos t anyon e would fin d i t too muc h o f an imposition t o be jumped o n by a large dog like a pointer o r a German shepherd . Suc h dog s us e their weigh t to test th e bond : the y approach an d stan d nex t t o a visitor in a friendly manner, leaning agains t th e visitor' s legs , graduall y transferring more an d mor e o f thei r weight, unti l the y ar e pushed off . How muc h o f this treatmen t a visitor will tak e before pushing the dog off enables the animal to assess the visitor's attitude toward it. When a dog gets in its owners' wa y in the morning, the way they move it aside or g o around i t when the y are in a hurry gives it the informatio n it seeks. And a t the en d o f the workday , a dog ca n fin d ou t reliabl y withi n a minute ho w muc h attention it s owner s ar e willing t o giv e it jus t then , b y how willin g o r unwillin g they are to be jumped up on. Information abou t th e importanc e of a partnership t o eac h o f its members is useful o n severa l levels . I n th e mos t extrem e cases , it ma y lead t o a decision t o break u p th e partnership . Mos t o f th e time , though, i t help s individual s decid e how much t o inves t i n a partnership an d ho w much ca n reasonably be expecte d from th e othe r partner . Clearly , there i s no poin t i n asking for somethin g fro m a partner wh o has no interest in the partnership . But eve n amon g willing partners ther e ar e ofte n conflict s o f interest . A cake eaten by one is not available to another. A moment of attention given to one partner is a moment not devoted to a third. And even when all members of the partnershi p benefit equall y from a n action, if one partner carrie s out th e actio n when anothe r might have, it ca n be see n as a loss to th e on e and a gain to th e other . The importanc e o f a partnership t o it s members an d th e relationship s amon g partners depen d o n man y factors that ca n easil y change , which make s frequen t testing o f the socia l bon d essential . It i s important t o eac h t o b e awar e o f other partners' attitude s at any moment: it can be costl y to make a request a t the wrong time, because onc e it has been denied , i t may be less likely it will ever be granted . It is better t o asses s things first and wait until the odds ar e on one's side .

Testing th e Bond 11

3

AGGRESSION IN COURTSHIP An individua l can test a social bond only by imposing upon another . The degre e of impositio n depend s o n th e importanc e o f the informatio n sought an d o n th e state of the bond. During earl y courtship, th e ris k involved in a decision t o mate is very high, and the preexistin g investment i n the partnershi p small . Under such conditions, th e impositions ten d to be high. Some courting males behave aggressively toward females; most females respond by leaving , an d th e mal e thu s find s ou t whic h one s ar e trul y interested i n him. Such aggression is especially common among birds when both the male and female invest a lot in their offspring. In many such species, males first establish territorie s and the n wait for females to join them. The male attacks every female who shows up i n his territory, and th e females fly away. Some of the females come back again and again . Gradually, the male's aggression toward one of the females lessens and then disappear s completel y o r almos t completel y fo r th e duratio n o f their lives together. Israelis are familiar with pairs of wagtails, small European songbirds that winter in Israel; the wagtails go through this process every fall, though they do not breed together; they only cooperate in defending their feeding territory,2 and each of them travels north alon e in spring to breed in Europe with a new partner. As w e se e it , thi s hig h leve l o f aggressio n early on is the simplest way for the male to test the intentions of females who are proposing t o share his property (his territory) and his future. A female might find it convenient t o settl e in a territory on a temporary basis , so as to hav e a secure plac e t o ea t an d resid e while searchin g for a permanent mat e amon g the neighboring males . Such a temporary arrangement would benefi t the femal e but could harm the male considerably, because during that time other desirable females who migh t otherwis e choos e hi m woul d no t approac h him . B y chasing a newly arriving female , th e mal e test s he r intentions . A femal e who regard s hi m a s a temporary convenience will not put up with being chased as much as a female who has alread y decided that this territory and thi s male are her permanent choice . In textbook s thi s aggression of males toward female s is explained a s resulting from th e fac t tha t male s have to be highly aggressive in order to chase rival males away from thei r territory; 3 they cannot ris k accepting females, whose appearanc e is suspiciousl y simila r to tha t o f th e riva l males . I n tim e the y learn t o accep t a particular female and sto p attacking her. This is no more than a description of the facts presented a s an "explanation," and is an insult to the birds' perceptive ability. Male babblers, too , chas e females who tr y to join their group . The y continu e t o attack female s t o a lesse r exten t throughou t thei r lon g live s together ; bu t the y hardly ever attack adult male members of their ow n group. 4 There i s no reason t o suppose that a male is unable to differentiate between male s and females and limit

114 TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

its aggression to male s only. In fact , i n nonmonogamous species, where male-onmale aggression is especially high, there is very little aggression towar d females. 5

HIDE-AND-SEEK: GENTLER TESTING IN COURTSHIP Among some bird s th e only collaboration betwee n th e genders is copulation; th e males of such species hardl y ever exhibit aggressio n toward visiting females. Th e males mate with many females and thus don' t hav e to be very choosy about thei r partners—yet eve n here , som e selectivit y i s calle d for . No t ever y femal e wh o comes to suc h a male's dancin g aren a decides t o mate with him; some come just to watc h him closel y and compar e hi m with others . Males , especiall y thos e wh o are much sough t afte r b y females, have to be carefu l no t t o spend to o much time with an y female wh o i s no t intereste d i n mating ; using aggressio n ma y frighte n away female s who migh t otherwis e b e willin g t o copulate , though , s o the mal e has t o d o hi s testing peaceably. Peacocks' courtshi p include s suc h gentle testing: whe n a peahe n approache s a peacock , h e turns his back to her and shows her the undecorated bac k of his tail. In order t o mate, the femal e has t o g o aroun d i n fron t o f the male . We thin k that thi s i s a tes t b y whic h th e mal e determine s the extent of the female' s interest . A female who does not bother to go around in front o f the male is no t intereste d i n examinin g hi s fa n o f feather s closely—an d i s probabl y no t interested i n mating either. Bowers built b y bowerbirds ar e known t o be a means of male sexual display.6 The size and decoratio n o f the bower clearl y testify to the male's ability. But what was i t tha t le d t o th e developmen t o f a certai n structur e fo r th e bower ? Afte r talking with Geral d Borgi a and viewing videorecordings he' d made, 7 we came to realize that the bower's specific structure may well have developed largel y because it enable d th e mal e t o determin e whethe r a female actuall y intends t o copulat e with him or not . When a female arrives at the bower, th e male hides behin d th e bower o r in it and watches the femal e t o see whether sh e attempts t o follow him. Throughout th e courtship , th e mal e repeatedl y peek s ou t an d then hide s again . An intereste d femal e follow s hi m a s in a dance , revealin g he r intention s t o th e male. This ca n explain how the bowe r evolve d ou t o f a bare dancin g arena: even the first stick placed o n the dancing floor could make it possible for a male to hide and thu s more reliabl y test th e interes t o f females. Of course , onc e the structur e evolved, then th e siz e of the columns , their number , order , an d decoratio n coul d serve an additional purpose—to enable the male to sho w off his quality. The male s of many bird specie s clea n an d maintai n a dancing arena . A clean

Testing th e Bond 11

5

arena i s evidence tha t it s owner ha s th e leisur e t o spen d tim e o n th e aren a an d tend it . There i s one species of bowerbird—the Australian catbird—that does not build a bower bu t instea d clean s a dancing arena . Unlik e othe r bowerbirds , th e male o f thi s specie s display s considerabl e aggressio n towar d visitin g females . I t may be tha t sinc e it canno t use the bowe r to tes t the female' s intention s and to compete wit h othe r males , this bowerbird use s aggressive behavior t o show off its strength an d to test females' intentions . Face-hiding game s are a commo n way of testing mutual interest between a human guest and a small child. Visitors who don't know how interested youn g children are in their compan y ofte n hid e thei r face s be hind somethin g t o see a child's reaction. A child wh o i s intereste d i n a visito r wil l move i n orde r t o se e th e visitor' s fac e again, s o a s t o determin e ho w intereste d the visitor is. Both childre n an d adult s love this gam e of peek-a-boo, without a s a rule being awar e that the y are collecting information .

CLUMPING, AND PREENIN G OTHERS (ALLOPREENING) Another kin d o f testing o f the socia l bond is seen amon g babblers. I t i s crucially important t o babbler s t o b e awar e of their partners ' commitmen t t o the m a t all times becaus e o f th e danger s involve d i n fight s betwee n groups . Mos t borde r clashes between group s o f babblers don' t g o beyond th e exchang e of threat call s or perhap s a few chases , bu t no t infrequentl y the y becom e rea l fights in whic h participants ar e wounded an d sometime s even killed . Such grou p fight s ar e fa r mor e dangerou s tha n th e one-on-on e conflict s o f solitary birds. I n one-on-one conflict i t is very difficult fo r one bird to kill another. The victor is usually satisfied when the loser flees. Group fights are different. Once a babbler grapple s with another, clutching it and being clutched by it, other grou p members may come to the ai d of their partner—and a group of babblers gangin g up o n a single bir d ca n easil y kil l i t with minima l ris k t o themselves . A babble r charging an enemy group risk s this fate, an d its life depend s o n the willingness of others of its group t o com e to its aid. Confrontations betwee n neighborin g group s ar e frequent , sometime s occur ring severa l time s a day . On e neve r know s whe n a confrontatio n wil l en d wit h an exchange of threat call s and when a rival group will decid e t o fight in earnest. Thus a babbler ha s to b e read y to figh t a t al l times. It s groupmates ' willingnes s to figh t fo r i t varies ; i t ca n b e affecte d b y th e arriva l o f new babblers i n th e re gion, interna l conflicts , an d s o on. Therefore , i t is important fo r eac h babbler t o

116 THE HANDICAP PRINCIPLE IN SOCIAL SYSTEMS test it s groupmates ' commitmen t t o i t often . A great dea l o f th e babblers ' dail y behavior serve s thi s purpose : clumping , preenin g other s (allopreening) , an d group dancin g al l evolved t o enabl e babbler s t o tes t th e strengt h o f the socia l bond amon g group members . Babblers ofte n clum p together , bu t thi s i s not th e norm among birds: individuals of most bird species tend to keep a certain distanc e from others. Swallows or starlings perched o n a power lin e are always a set distanc e apart, like musica l notes on a staff. I f a lower-status individual gets too near, they threaten it; if a higher-status one approaches , the y mov e away . Thi s insistenc e o n individual spac e make s sense : bird s nee d a certai n amount o f roo m t o sprea d thei r wing s and fl y off; one who does not keep its distance from others cannot take immediate flight t o avoid a predator or catch prey that may appear. This freedom is given up by birds who clump, a s do babblers and other species that depend on strong social bonds. Many group-living bird s clump. Bulbuls, parrots, and bullfinches, al l of whom have a strong pair bond, clum p in pairs. We think that clumping reflects reliabl y birds' readiness to invest in the partnership, becaus e it hampers freedom of movement. It is no coincidenc e that girlfriends press against their boyfriends just when the boyfriend is talking to another girl: that's just the time to pressure him to prove his interest in the relationship . Another method babblers use to test their social bond is allopreening, or preening one another's feathers. Babbler s ofte n clum p togethe r an d preen eac h other . The babbler bein g preene d put s itself a t the preener' s beak , feathers fluffed and eyes ofte n close d t o let th e othe r preen it s facial feathers fully . Th e recipien t of the preenin g ca n hardly see in that stat e and obviousl y is not prepare d t o fl y or flee. The preene r test s th e other' s will t o cooperat e with it , an d th e on e being preened show s of f that i t trust s the othe r t o warn it i n cas e of danger. Both, by clumping and allopreening , tak e on a greater risk than birds restin g alone. It use d to b e assume d that the functio n o f allopreening i s to preen feather s a bird canno t preen on its own, such as head feathers. But this does not explain why it is that only social birds pree n eac h other, and why they preen als o parts of th e body tha t eac h bir d ca n easil y reach on it s own. Preenin g eac h other' s head s is simply a useful bonu s o f a behavior that serves primarily to test the socia l bond.

GROUP DANCES AND SIMILA R RITUALS The most impressiv e socia l activity of babbler groups is the morning dance , whic h was researched in Hatzeva by Roni Osztreiher. 8 W e believ e thi s ritual is another

Testing th e Bond 11

7

means by which babblers test the social bond amon g group members. The morning dance takes place almost exclusively at the first light of day . It happens only once every several days, and it is not yet clear what makes the babblers decide to dance on a particular day. One of the babblers suddenl y stops at a suitable "dance floor " and start s preenin g itsel f nervousl y o r sprawlin g wit h it s throa t touchin g th e ground. Sometimes there is no response, an d the babbler give s up an d goes off to look for food. At other times, though, another babbler comes and clumps with the first, preening its own body. Two babblers clumpin g and preening ar e an almost irresistibl e invitation, an d i n most case s th e res t of the grou p soo n join s in . The danc e ca n last up to half an hour and sometimes even more. During the dance, babblers press against eac h other , squeezin g under an d ove r an d betwee n thei r partners . The y dance in the open, near a bush, even though the y could dance more safely under trees. The danc e almos t alway s take s plac e befor e sunrise , when th e dange r fro m raptors i s greatest, since they can exploit th e low light an d surpris e th e babbler s with relative ease. This is also the best time for babblers to feed, since many night creatures are still active—including termites, a favorite babbler food. In short, the dance doe s no t happe n a t th e easies t tim e and plac e but, o n th e contrary , at an inconvenient tim e an d place , using inconvenien t movements . A n individual who is willing t o undertake all this with it s partners shows reliabl y it s commitment t o the group. The dancing of babblers reminde d us of stories our parents' generation told about the dancing of members of pioneering kibbutzim in Israel in the earliest, most difficul t an d exhaustin g years : "Fo r month s w e had nothing to eat, but we danced al l night... ."9 Group ritual s that test the social bond ar e not unique to babblers. Other social animals hold the m too , ofte n befor e other grou p activities . African huntin g dogs jump on each other before embarking on a hunt.10 The hunt is risky to the hunter: when it grabs the leg or neck of large prey and hang s on, its safety depends o n its partners' cooperation . Africa n dwar f mongooses hold simila r rituals. 11 The nee d to test the bond may even explain why dogs jump all over and around their owners when settin g ou t o n their dail y walk. An d there are the dail y greeting s betwee n humans. Kisses , handshakes, and th e exchang e of nod s an d verba l greeting s all help us assess the attitud e of our partners, friends, associates , and coworkers during the openin g moments o f each encounter . I n chapte r 1 8 we shall se e how th e sexual ac t became a means of testing the bon d betwee n partner s for various animals.

This page intentionally left blank

C H A P T E R 1

1

PARENTS AN D OFFSPRIN G

A

parent and a child are the simplest reproductive coalition there is. At first glance, it seems obvious that both parent and child have the same interest: the goo d o f the child . Bu t i f that i s so, why ar e ther e s o many conflicts between parent s and children? According to Trivers,1 the picture is less simple than it first appears. The child's interest is to take care of itself, while the parent ha s to take care of other offsprin g as well. The chil d aims to get from it s parent whatever it can get, even if the parent shortchanges an d perhap s eve n lose s it s othe r offspring . Th e paren t want s th e child t o succeed , but onl y if its investment i n that one chil d doe s no t reduc e th e total number o f successful, reproducin g offsprin g i t can have. A good exampl e of a mother-child conflict i s the nursin g and weaning of baby mammals. Trivers explains the crie s of a baby who wants to nurse as a psychological weapon aime d a t making the mother nurs e it against her will. He assert s that the baby does al l it ca n to forc e th e mother t o nurse it, while the mothe r trie s to avoid nursing when nursing is against her ow n best interests. Everybody know s tha t a screamin g bab y ca n forc e a mothe r t o action s sh e would not otherwise take. But why exactly are mothers moved by a baby's screams? Labeling th e scream s a "psychologica l weapon " doe s no t explai n what it i s that

119

12O

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

makes them effective. Trivers did not take into account that behavioral mechanisms themselves evolve by natural selection.2 If mothers who did not respond to screaming, o r who did not respond a s much, had had more successful offspring than ones who responded, then th e tendency not to respond woul d hav e spread in the population, an d the "psychologica l weapon " woul d hav e lost its potency. Behavioral mechanisms are only proximate factors in the evolution of traits; the ultimate factor in evolution i s successful reproduction .

THREATS OF SELF-INJURY : THE WEAPON O F THE WEAKER PARTNE R Babbler fledglings, like human babies, cry loudly when hungry. An observer senses intuitively that they cry in order to be fed, and it was a long time before we realized what was wrong with this explanation. After all , the cries pose a real danger. They can be hear d ove r a great distance—hundreds of yards. They revea l to predator s the fledglings' location—and a t that ag e the latter ca n barely fly, or eve n hop. I n fact, thes e loud crie s often hel p u s find fledglings' hiding places. The commo n explanatio n i s tha t th e crie s tel l th e parent s wher e th e fledglings ar e an d tha t the y are hungry. But the crie s are often loudes t whe n the parents ar e nearb y and kno w jus t wher e t o fin d th e fledglings. These strikin g fact s tell u s who m th e call s ar e actuall y directe d at , an d wh y they forc e th e parent s to atten d t o th e young : we believe tha t th e crie s ar e actuall y meant to b e hear d by predators . Th e fledgling s say , as i t were , "Cat , cat , com e an d ge t me ! I a m here an d I don' t car e who know s i t unti l m y parents fee d me. " Onc e the y are fed, th e fledgling s sto p calling . Th e fledgling s ar e forcin g thei r parent s t o fee d them b y endangerin g themselves . The weaker member o f a partnership ca n blackmail the stronger by threatening to bring injury on itself. 3 Of course, this threa t i s effectiv e onl y i f th e stronge r partne r ha s a n interest in the well-being of the weaker party. This is exactly the situatio n with mothe r an d young . By investing more in an offspring who cries, the mother risks reducing slightly the total numbe r o f he r futur e offspring , bu t b y silencin g th e complainer sh e improve s th e chance s tha t a n offsprin g in which sh e has already invested a great deal will survive, and she reduces th e ris k of losing it and its nestmates to a predator a t one fel l swoop . Th e mothe r is moved b y her baby's screams because th e scream s constitute a real threa t t o he r own interests . Most o f us have heard a n "abandoned " kitten wh o has

Parents and

121

climbed t o the top o f a tree an d crie s all night long. Kindly neighbors tr y to hel p it, assuming that i t has lost its mother. We think the kitten i s trying to blackmail its mother into continuing to nurse it, to prevent the kitten fro m wanderin g off to dangerous places . Th e dange r tha t th e kitte n wil l expos e itsel f t o predator s i s greater tha n th e potentia l benefi t t o th e mothe r o f weaning it early. And a s long as the benefi t of additional nourishment is greater than the ris k of predation, it is in the offspring' s interes t t o cry. Yet th e ris k involve d i n th e crying is , an d ha s t o be , real—otherwis e th e mother would no t submi t t o th e blackmail . S o from tim e to time , a fledgling o r a kitten may be eaten. The young of hares and gazelles, who have many predators, cannot affor d t o us e cryin g as a weapon; th e ris k is too great . They lie quietly in a hidin g place , an d th e mothe r come s fro m tim e t o tim e t o nurs e them . Obvi ously, th e mothe r doe s no t nee d t o hea r he r babie s cryin g to fin d them . I f th e young of hares and gazelle s have any way of blackmailing their mothers, we have not foun d i t yet. The ris k of predation i s not th e onl y factor tha t influence s parents . Even hu mans, who live far from predators , give in to their children' s cries . Of course , the lack of predators and enemie s "allows" babies to cry as much as they care to with no risk, which according to the Handicap Principle should greatly reduce the value of cryin g as a signal . S o why d o parent s respond ? Protracte d cryin g by a baby harms th e adult' s prestig e b y makin g a ba d impressio n o n hi s o r he r partner , neighbors, an d friends . Prestig e ha s rea l value, a s we shall see in a later chapter . It may well be in the parent's interest t o give in rather than lose face .

OTHER METHODS OF BLACKMAIL Crying is only one of the ways offspring extort help from parents . Sometimes young are more brightly colored tha n their parents: baboon babies ar e dark, while their parent's lighter color s merge much better with their habitat. Wild piglet s are decorated wit h prominen t stripes , whil e thei r parent s ar e grayish-brown , like thei r environment. Fledglin g babblers are more striking than thei r parents: their beaks are black, a prominent colo r i n the desert , with a yellow base, and their breasts are decorated wit h brown . These colors, together with such attention-drawing behaviors as waving their wings, make it easy to find the young babblers. Lik e their cries , this behavior ca n be explained b y th e fac t tha t th e dange r t o whic h th e fledgling exposes itself forces its parents and caretakers to devote more attention to it. Babbler egg s ar e a brigh t emeral d green , whic h stands ou t clearl y agains t th e gray-brow n twigs tha t

122 THE HANDICAP PRINCIPLE IN SOCIAL SYSTEMS make up the nest. This may also be extortion: the mother may be "using" the color of the egg to extort her group partners' hel p in incubating them. Brightly-colored eggs ar e mor e likel y t o b e take n b y nest-raidin g predators i f no t continuousl y incubated—and precisel y fo r tha t reaso n ar e mor e likel y t o b e continuousl y incubated b y other group members ; this benefits both mother an d young. Great creste d grebe s spen d al l their tim e o n th e water ; duck s spen d par t of their tim e on land . Yet i n a seeming paradox , th e fledglings of ducks have thick plumage tha t let s the m spen d lon g period s o f tim e i n col d water , whil e th e fledglings of creste d grebes have thi n plumage. 4 Grebe s carr y their offspring o n their backs . It ma y be precisel y because baby grebes' plumag e is thin tha t greb e parents don' t leav e their offsprin g alon e i n th e wate r eve n for a short perio d o f time. Becaus e they have thin plumage , the fledglings cannot survive in the water for long , bu t th e benefi t ma y be th e car e the y get. Similarly , the bare belly of the bab y chimpanzee encourage s the mothe r an d its other caregiver s t o kee p i t nex t t o thei r bodies . Th e sam e principle ma y account for th e tota l helplessness o f human babies. Feldman and Eshel assert that it is difficult t o see how the mechanism o f blackmailin g parent s coul d evolve , sinc e offsprin g wh o carry thi s tendenc y will suffe r fro m i t later—th e tendenc y will b e passed on to their own offspring, wh o will blackmail them in turn.' But thi s is a difficult y onl y if one assume s that both th e degre e o f extortion an d the respons e of parents ar e fixed. We think tha t th e abilit y t o extor t help fro m parents i s a benefit when use d "wisely, " i n a way that i s appropriate t o existing circumstances. A behavioral mechanism suc h as we describ e make s sense onl y if it is sensitive to outside conditions—i f th e young endanger themselves only when the extra car e they receive is worth the ris k they take. And indeed, when babble r fledglings hear their parents' alar m call announcing a predator nearby, they freeze and d o not utte r a sound. And when an offspring reliabl y shows its need for care, there is a direct benefi t to th e parent ; otherwise , th e paren t migh t expen d unnecessar y effort providin g more foo d tha n i t ca n eat , fo r example . O f course , th e paren t woul d prefe r a demand mechanis m tha t woul d no t endange r th e offspring—tha t woul d leav e more leeway for the parent to decid e ho w to spen d it s resources. But there i s no clear lin e of demarcatio n betwee n a n offspring' s need t o reliabl y inform parent s how much care it requires, and its ability to use the same mechanism to extort too much care. 6 It make s sense for parents to giv e in when the y know tha t offsprin g really d o nee d help—whic h i s proven b y the ris k th e youn g take i n demandin g that help—an d i t eve n make s sens e for the m t o giv e the youn g somewha t more care than is necessary, in order t o minimize the risk. Natural selection keep s both the extortion an d the parental response in balance. The benefit t o the offspring of the extra care, the increased likelihood tha t it will survive and reproduce, is a part of this balance; s o is the effec t o f the parent' s extra investment in the demandin g offspring o n its potential to produce other successful offspring .

Parents and Offspring 123

EXPLOITATION O F OFFSPRIN G B Y PARENTS It i s no les s interestin g t o se e cases i n which parent s us e their powe r t o exploi t their offsprin g fo r thei r ow n gain. Gain i n evolutionar y terms means an increase in th e numbe r o f reproducing offspring : al l that is really meant by evolution and natural selection is that some individuals have more offspring who themselves manage to reproduce successfully , an d that therefore the traits of these parents are the ones that will prevail in future generations. Thus, in evolutionary terms, exploiting one's childre n mean s usin g som e o f one's offsprin g t o increas e the chance s tha t others of them will mature and reproduc e successfully. Among animal s who tak e car e of their young , the offspring' s qualit y depend s on their parents ' effort s t o guard, warm, and fee d them . Parents often prefe r one sibling ove r others . Man y birds—such a s owls , herons, and parrots—start incubating when they lay th e firs t egg . Thi s eg g hatches first , an d th e nestling tha t emerge s fro m i t wil l naturall y b e bigger an d stronge r than its younger siblings. If food i s scarce , th e firs t nestlin g hatche d ca n snatch foo d awa y fro m it s siblings . Thi s com petition ove r food may lead to the death of one or more of the younger nestmates, but the parents do not interfere. 7 Under some conditions, it is better for the parents to en d u p wit h a smaller numbe r o f stronger offsprin g tha n a larger numbe r of weaker offspring wh o will be unable to compete agains t the offsprin g o f others of their species . Similarly , Diamond 8 suggest s in hi s book Th e Rise and Fall o f th e Third Chimpanzee that menopause cam e abou t t o enabl e wome n t o inves t more care in their existing offspring, who still need their car e to survive, rather than put their effor t int o producin g an d rearin g additiona l offspring , whos e chance s fo r survival after thei r mother' s deat h would b e very small. The rule s o f inheritanc e i n som e human societies , pas t an d present , wor k i n the same way: the bul k o f th e famil y fortun e i s passed o n t o on e offspring , t o ensure that a t least that on e will succeed . Splittin g the legacy equally between all offspring woul d mea n that each would receiv e only a small portion, and al l might fail a s a consequence . Usuall y the preferre d hei r i s the firstbor n son , who i s in a stronger positio n t o begi n with , bu t sometime s i t i s th e younges t wh o get s th e largest share . In th e latter cas e it may be that th e parents have a better chanc e of raising children an d grandchildre n if the firstbor n so n leaves the hom e and seeks his fortune elsewhere, withou t competin g fo r resources with parents who can still bring up more children . Among macaques, a kind o f monkey, it is the youngest daughter who inherits her mother's socia l position; olde r daughter s are subservient to the youngest one. 9 The younges t daughter' s preferre d statu s come s no t fro m he r ow n abilit y bu t from th e fact tha t the mother, an d sometimes other higher-ranking females, intervene i n conflict s among daughter s i n favo r o f th e youngest . This preferenc e fo r

124

E HANDICAP PRINCIPLE IN SOCIAL SYSTEMS

the younges t ma y lessen somewha t th e chance s o f olde r daughters , bu t i t als o prevents older daughters from competin g with their mother or younger sister, thus increasing the mother's overal l reproductive success. Such exploitation i s practiced t o the greates t degre e amon g the socia l insects; we wil l sa y more abou t tha t i n chapte r 13 . But eve n i n les s extrem e cases , any child-parent situation combines conflic t wit h common interest. The child is interested in getting the bes t star t i n life; th e paren t i s interested in using hi s or he r resources in the most effective way, not necessarily to produce the greatest number of offspring, bu t rathe r to maximize the numbe r tha t will themselves successfull y reproduce. Each offsprin g trie s to persuad e it s parents t o favo r i t ove r it s siblings; eac h parent tries to ensur e tha t it s resources are utilized in the best way, whether by devoting th e mos t effor t t o th e offsprin g likelies t t o succeed , b y helping th e on e to who m tha t hel p wil l mak e th e greates t difference , or eve n b y exploitin g on e offspring t o benefi t it s sibling. I n th e las t case , of course, eac h siblin g wil l try t o ensure that i t is the beneficiary of such treatment, rather than the victim. The specifi c detail s o f eac h cas e depen d o n a species ' lif e strateg y an d o n circumstances o f each particular situation. I n mos t cases , there i s room fo r com munication betwee n th e parties : parent s gaug e th e need s an d th e vigo r o f eac h offspring, an d offsprin g d o their best to convinc e their parents of the same , so as to ge t al l the hel p the y can . Indeed , child-paren t relationship s provid e fertil e ground fo r communication , an d reliabilit y i s n o les s essentia l her e tha n i n an y other communication among individuals. As we have said, the only means we know of t o guarantee that reliability ar e handicaps.

C H A P T E R 1

2

BABBLERS, COMPETITIO N FOR PRESTIGE , AN D TH E EVOLUTION O F ALTRUIS M

W

e have been studying babblers sinc e 197 0 and continu e to do so as of 1997. Babblers ar e the onl y birds i n Israel that we know of that live permanently in groups. A s a rule, mos t groups hav e three t o twelv e members, thoug h som e groups have reached twenty . Babblers sta y in their territories year-roun d an d cooperat e t o protec t the m agains t neighboring group s of babblers an d nonterritoria l individuals . I n breedin g seaso n al l members o f th e group—breeding adults , nonbreedin g adults , an d young—tak e car e o f a single nest in the territory. The birds' dealing s with each other includ e a great deal of assistance to others. Such altruis m is common amon g group-living organisms and poses a major prob lem in the study of evolution: it seems to run counter to Darwinian logic, according to which the individual pursues its own interest and that of its own offspring. Ou r study of the babblers , combine d with ou r understandin g of signal selectio n an d of the Handica p Principle , le d us to develo p a new and differen t solutio n t o this evolutionary enigma. Babblers ar e songbirds o f the many-membered tropical family Timaliidae. Th e babblers i n Israel belon g t o on e branch o f this family , whic h sprea d fro m Indi a northwest across the desert bel t an d into Morocco. Th e species we study is found 1 25

126 THE HANDICAP PRINCIPLE IN SOCIAL SYSTEMS

in th e Arabia n Peninsul a an d th e Sina i Desert. I n Israe l i t ca n be foun d i n th e Dead Se a region, th e Arav a Valley , and th e war m wadis (seasona l watercourses that ar e dry except fo r a few rainy days a year) in the borderin g hills. Babblers ar e th e siz e o f blackbirds o r grackles , abou t 2 t o 3 ounce s (70-8 5 grams). Their tails, at about 6 inches (150-160 millimeters), are slightly longer than their bodies . Th e bird s ar e grayish-brown , like thei r deser t environment . Thei r wings are short; they are slow flyers, but their strong legs and long tail enable them to maneuver in thickets wit h great dexterity . In the open , suc h raptor s a s hawks, peregrines, an d harrier s ca n catc h the m easily ; the babbler s prefe r t o sta y near trees an d shrub s int o which they can escape. They usually seek their foo d i n an d near bushes, on and under the ground, and under tree bark. They eat any creature they can swallow or tear to pieces: insects, snails, scorpions, and small snakes and lizards. We have observed two cases in which babblers kille d small birds, but they could not ea t them: their beaks are dull, adapted t o digging in the soil rather than tearing the flesh of birds. Babblers als o eat the juicy berries of some desert shrubs and occasionall y eat flowers and drin k flower nectar. We stud y th e babbler s i n th e Sheza f Natur e Reserv e near th e Hatzev a Fiel d Study Center , which is run by the Societ y for the Protectio n of Nature in Israe l (SPNI). W e follo w som e 30 group s ove r a n are a of 1 5 square miles or so . Each individual babbler i n the stud y area—ther e are about 25 0 of them at the tim e of this writing—is tagged with a combination of three colore d bands and one numbered one , an d ca n thus be easil y identified. Babblers readil y become habituate d to th e presenc e o f humans who behav e calml y and giv e them smal l bits o f food; such gifts ar e especially welcome during winter, the lean season. They also remember wel l an y har m don e t o them , however , an d the y sta y awa y from whoeve r perpetrated it . The babblers in the stud y area are used to our presence . The tiny amounts of breadcrumbs w e give them occasionall y do no t chang e their foo d suppl y in any meaningful way , but a s a result o f these "donations " they do no t fea r u s an d le t us wal k amon g the m freel y an d watc h al l thei r doing s a t clos e range . W e ca n hear the soft , widel y varied calls they use to communicate , calls that don' t carr y more tha n a fe w yards . Som e individual s ar e talkative, other s taciturn , an d stil l other s grumbl e al l day long. When a higher-rankin g individual approache s a lower-ranking one, th e latte r ofte n make s a sof t soun d t o acknowledg e th e other, a s if saying "Yes, sir. " Adul t babbler s who stan d guar d do s o quietly , bu t youngster s tryin g thei r han d a t i t vocaliz e softly, a s i f callin g their nearb y comrades ' attentio n t o th e ef fort the y are making. In th e spars e desert countr y around Hatzeva, we move easily among the babblers; we see what they are looking at an d become awar e of their intentions. Fo r example, a babbler wh o finds food may not swallo w it right away but instea d may hold i t in it s beak an d loo k aroun d t o se e whom i t ca n feed . If it see s only indi-

Babblers, Competition for Prestige, and the Evolution o f Altruism 1 2

7

viduals whose rank is higher than it s own, it may waver briefly, the n swallo w th e food. Whe n th e bet a male , the number-tw o mal e of a group, goes up t o th e to p of a tre e t o stan d a s sentinel , w e ofte n se e th e alph a male , hi s superior , busil y looking fo r food; he then gives it to the sentine l in full vie w of the other member s of the group and replaces hi m on guard duty. In many cases, we can tell when the sentinel notice s th e alph a male' s preparation s t o fee d hi m fro m th e directio n of the sentinel' s gaze , and ofte n becaus e the sentine l abandon s his lofty pos t befor e the alph a male arrives to displac e him . Our tea m check s on each group o f babblers i n the researc h area about onc e a week, bu t ofte n mor e frequently. Thu s w e closely follow changes in the compo sition an d routin e activities o f each group . Over the year s we have learned their life histories an d social structure. Detailed studie s have focused on specific aspects of babble r behavior : sentine l activit y (Tirza Zahavi , Ton y Larkman , Nir Faran), territoriality (Arno n Lotem), the feedin g of young (Thams i Carlisle) , allofeeding among adult s (Ami r Kalishov) , playin g (th e late Ori t Pozis) , mobbin g (Zahav a Carmeli, Avner Anava), allopreening (the late Andres Gutman), the morning dance and feedin g a t th e nes t (Ron i Osztreiher), shout s (Zoha r Katsir) , wate r balanc e (Avner Anava), and competition ove r mating (Yoel Perl).1 Kim Lundy and Patricia Parker carried ou t DN A studies , Dietmar Tod t an d his students ar e studying vocalizations, and Jonathan Wrigh t i s studying feeding at the nest. 2

TERRITORIES, GROUPS, AND NONTERRITORIA L INDIVIDUALS Each babble r grou p ha s it s ow n territory . A grou p protect s it s territor y against neighboring group s in daily or almost daily border encounters, whic h occasionally develop into clashe s and a t times escalate into all-out raid s on neighboring terri tories. Group s als o protec t thei r territor y agains t nonterritorials—babblers who have been throw n ou t o f a group o r whose group has lost its territory. As a rule, any area with enoug h foo d an d shelte r t o suppor t babbler s i s occupied by a group of the birds. Babblers who are not members of a territorial group have a hard time finding foo d an d shelter, let alone breeding, because the group s that protect eac h territory attac k them o n sight. The greatest danger to such nonterritorial babblers is not predators—whom they can avoid by diving into a bush— but th e territory-holding babblers , who are equally adept at maneuvering in dense vegetation, an d who m the y ca n escap e onl y by fleein g int o th e shelterles s ope n desert. Nonterritorials ar e never a large part o f the population ; w e estimate them at about fiv e percent o r les s a t an y given time . They fac e a high ris k o f predation , and most do not survive long. Still, the life of a nonterritorial, with all its hardships

128 TH

E HANDICA P P R I N C I P L E I N SOCIA L SYSTEM S

and dangers, is not totall y hopeless. Occasionally , on e of them will live for months or even years near its old territory, hiding from th e territory holders, until it finds a grou p that accept s it a s a member o r manage s t o fin d a vacant are a an d join s with other nonterritorial s to for m a new group. Babblers ar e long-lived and ma y reach the ag e of twelve or fourteen. In years of drought , when foo d is scarce, babblers d o not breed . I n rain y years, when th e Arava Desert get s a few centimeters o f rain at the righ t intervals, a group of babblers ma y produce tw o or three successive broods. Becaus e of the combination of breeding succes s and lon g life , man y individuals wh o canno t fin d a territory of their ow n t o bree d i n g o on livin g wit h thei r parents ; thei r parents ' territorie s provide enoug h food t o support the m all , and those parent s ar e still able t o raise new offspring successfully. Tha t arrangement improve s th e lif e chances o f all parties. Parent-offspring group s are stronger than pairs without offspring; a s a result, even i f an area becomes vacant , new settler s i n it have to for m group s i f they are to protect themselve s against the families aroun d them. As a rule, new groups that form includ e a t least three adult babblers .

RANK, AVOIDANCE O F INCEST , AND TH E LIFE STRATEGY OF MALES AND O F FEMALES Within eac h grou p ther e i s a clea r orde r o f rank . Older male s outrank younger males, an d olde r female s outran k younge r females . Al l adul t male s outran k al l females, especiall y females tha t joi n the group i n order t o breed. When fledglings come ou t o f the nes t togethe r the y figh t amon g themselves an d establis h a rank order without regard to gender. A s a rule this order is set by the end of their first week outsid e th e nest , thoug h aggressiv e behavior amon g the m ma y still erup t during the first few months and may in rare cases lead to a switch in rank. Beyond these firs t fe w months, ran k withi n a brood doesn' t chang e fo r bird s o f eithe r gender. Babblers d o not bree d with thei r parent s o r with an y other babble r who was a member o f their group when they hatched. Thi s rule is all but absolute : in more than twenty-fiv e years o f study , w e hav e see n onl y fou r exception s t o it . Som e babblers may live in their parents ' group a s long as six, or in one case eight, years without trying to breed, althoug h the y are physically able to reproduce when the y are two years old. Male s normally stay on in their parents' territor y and take their chances there, ofte n fo r life; there are territories tha t have seen five generations of the sam e male line since we began observing them an d that ar e still goin g strong. If a particular group breeds enoug h adul t males, and i f the males in a neighboring group becom e weak, th e bigger grou p may conquer the territory of their weake r neighbors an d the n divid e int o tw o groups. O f th e male s who eventuall y breed,

Babblers, Competition for Prestige, and the Evolution of Altruism 129 some 50 percent d o so in their birth territory ; some 30 percent breed in a territory that border s it—usuall y afte r takin g i t over—an d abou t 2 0 percen t bree d on e territory farther out. Females canno t mat e with their fathers and brothers an d so have to leave their parents' grou p i n order t o breed. Som e are accepted by a neighboring group even before the y ar e capabl e o f reproducing, whe n the y are a s young as one yea r old, while other s liv e with their parents til l they are four and in rare cases even five or six. O f the female s who eventuall y breed, a few manage to d o so in their parents ' territory, but onl y when that territory is taken over by males from elsewhere. About half th e female s who breed d o s o across the border fro m thei r parenta l territory , and the rest no more than three territories away. Young males sometimes overtake their sister s in rank when the y reach one or two years of age. It i s hard t o tel l whether th e males ' dominance—the y may now hit o r shove their sisters—i s a result of their ow n increased powe r or because th e females choos e no t t o asser t themselves agains t them. Afte r all , females have no future i n thei r fathers ' groups ; the y have t o joi n othe r group s an d bree d there . Thus there i s no great conflic t o f interest between the daughter s of the group an d their male siblings, and no real reason to compete with them past childhood, whe n they compete t o receiv e food.

THE COMPOSITION OF GROUPS; COALITIONS OF MALES AND O F FEMALES The simples t babble r grou p consist s o f a pair o f parents an d thei r offspring . In such a group onl y tw o individual s breed . Othe r group s ar e mor e complex . I f a mother dies , the father , wh o canno t mat e with hi s daughters, is joined by females from elsewhere . When a female joins a group, all adult males in the group ar e able to copulat e wit h her ; the fathe r now competes wit h hi s sons, an d with an y other adult males in the group, for a chance to reproduce. The dominant male copulates most; his chance s o f copulating wit h a female when sh e is ovulating, and thu s of becoming a father, ar e therefore bette r tha n those of other males . But the other s can tr y too an d sometime s succeed . All the female s in a group wh o la y eggs do s o in a single, communal nest. Females wh o joi n group s together—two , three , o r eve n fou r o f the m a t a time — compete wit h on e anothe r bot h t o la y egg s an d t o la y th e firs t egg s i n thei r group's nest ; the first nestlings t o hatch will have a head star t on their nestmates, which mean s a bette r chanc e o f surviva l an d o f achievin g hig h ran k withi n th e brood.3 Suc h contest s ar e comple x an d fascinating . Each i s a unique story , reminiscent o f historical epics , Shakespearea n dramas, an d biblica l tableaux . I n on e case three males were copulatin g with fou r females, all of whom lai d eggs . It was

13O

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

impossible fo r anyone—huma n or babbler—to tell which o f the fledglings in the common nes t wer e whose offspring ; i t would hav e taken DN A analysi s for us — the babblers themselve s had n o idea . As a rule, such a partnership between females who join a group together last s onl y unti l the y brin g forth on e successfu l brood . After wards th e dominan t femal e be comes th e onl y breedin g female . She either kill s her rival s or expel s them, o r the y submi t an d becom e nonbreeding helpers . Sometime s one of the latter eventually casts out the dominant female—or kills her— and take s he r place . Male s ma y continue copulatin g wit h th e sam e femal e ove r severa l breeding cycle s and eve n years, bu t th e competitio n fo r breedin g privilege s continues , an d th e loser s eventually leave , ar e expelled , o r ar e kille d b y thei r peers , unti l onl y on e i s left. Th e on e remainin g male then father s th e broo d o f the femal e who survive d the struggl e with he r femal e peers ; th e breedin g coupl e i s now supporte d b y a growing number o f adult offsprin g wh o canno t mate with their parents or among themselves.

STRUGGLES WITHOUT AGGRESSIO N (WELL, ALMOST ) The mos t fascinating aspect o f these intens e internal struggle s is the almos t complete lac k o f aggressio n fo r mos t o f thei r duration . Actua l fightin g i s extremely rare, an d eve n ritualize d aggressio n is very unusual between tw o adul t male s o r two adul t female s o f the sam e group. W e hav e seen onl y a few score full-fledge d fights in over twenty-five years of study, in contrast to the daily border clashes with other groups , th e violen t chasin g of nonterritorials, the fairl y commo n aggression of males toward females, and the constant squabbles amon g youngsters in the first few month s of their lives. An adult fights seriously with another adul t of its group only once, if ever. That one fight is all-or-nothing: the loser either is killed or flees and becomes a nonterritorial. Not surprisingly , such a fight almost always occurs between the most dominant mal e o r femal e an d th e on e nex t i n rank : ther e i s n o poin t i n riskin g everything for anythin g less than th e top position . Suc h a fight is savage. It starts all of a sudden, and mos t observers fail t o se e any sign that it is imminent. It is as though al l grievances an d suppresse d antagonism s between tw o brothers o r be tween a father an d so n who hav e lived togethe r peacefull y fo r years burst ou t i n

Babblers, Competition for Prestige, and the Evolution o f Altruism 13

1

this on e life-and-deat h fight . Th e struggl e is strictl y between th e two : the othe r members o f th e grou p observ e i t closely , bu t w e hav e neve r see n an y of the m intervene. This is in stark contrast to battles between groups, in which all members of the grou p may join in an attack against a stranger or come to the ai d of a group mate.

THEORIES THAT EXPLAIN ALTRUISM: CROUP-SELECTION THEORY AND IT S FAILINGS Members o f a given group hel p t o defen d thei r territor y and t o tak e car e of th e young. It might at first seem logical to assume that babblers help defend the common territor y because it is this territory that ensures their survival . But each babbler coul d tr y to le t others d o th e fighting. And why do thes e birds inves t effor t in raisin g offspring tha t ar e no t theirs ? This question , an d simila r issues among other species , hav e le d i n th e las t twent y year s t o a wav e o f studie s o f group breeding bird s an d animals. There ar e group-breedin g organism s in almos t al l orders o f animals . The lis t includes no t onl y group-living birds like the babblers, 4 bu t als o animals ranging from suc h mammal s a s lions , chimpanzees, 5 spotte d hyenas, 6 Africa n huntin g dogs,7 dwarf mongooses,8 and naked mole-rats9 all the way to slime-mold amebas10 and communa l bacteria. 11 Natura l selection mean s the sprea d o f traits that hel p individuals to reproduce mor e than others. I n this contest, the only measure of an individual's succes s i s successfu l offspring . I t woul d see m tha t investin g i n an other's offspring undermines one's own reproductive success and contradicts nat ural selection. Ho w ca n one explain this contradiction ? The evolutionar y mode l know n a s group selection onc e seeme d t o offe r a n explanation fo r th e behavio r o f socia l animals like the babblers . Thos e who be lieved in group selection began with a fact that we accept: large groups can protect their own territory better than smal l ones, and they can take over the territories of smaller groups . Bu t grou p selectionist s g o o n t o sa y that i t i s therefor e i n th e interest of each group member to help its parents have more offspring. By helping its group grow larger, they say, the individua l increases the odd s that it will itself survive and reproduce . Indeed th e odd s ar e bette r tha t member s o f large r group s wil l surviv e an d reproduce than that members of smaller groups will. But this explanation, like all explanations base d o n grou p selection , i s not stable . Th e fla w i s inherent i n th e logic o f group selectio n itself : What prevent s som e members o f the grou p fro m becoming socia l parasites? Group selection , i n other words , woul d rewar d individuals who exploit th e willingness of others t o strengthen an d enlarge the group, and wh o sav e their ow n strengt h fo r th e struggl e t o reproduce ; thus , precisel y those who work hardes t for the goo d o f the grou p woul d b e penalized . And th e

132 THE HANDICAP PRINCIPLE IN SOCIAL SYSTEMS parasites woul d rea p anothe r benefit : they would spare themselve s th e effor t o f caring for th e offsprin g o f others. 12 Once it was recognized that group selection doe s not provide an answer to the riddle o f altruis m in group-livin g animals, two othe r theorie s wer e brough t forward: the theor y of kin selection, 13 which we will discuss in the next chapter , an d the theor y o f reciprocal altruism. l4 Th e theor y of reciprocal altruis m i s the mor e general one, sinc e it doe s no t assum e the nee d fo r genetic closeness between th e helper an d th e one being helped: on e ofte n encounter s groups of unrelated individuals both amon g babblers an d among other group-livin g birds an d mammals.

THE THEORY OF RECIPROCAL ALTRUISM AND TH E PROBLEM OF ENFORCEMENT With hi s theor y of reciproca l altruism , Trivers suggeste d that i f one individual's investment i n anothe r i s eventually reciprocated b y a n investmen t o f the secon d individual in th e first , o r i n the offsprin g o f the first , the n i t makes sense for th e two t o inves t i n eac h other . Triver s kne w tha t reciproca l altruis m ca n onl y b e maintained i f there is a mechanism that ensures reciprocity . Without such a n enforcement mechanism , reciprocal altruism is no differen t fro m grou p selection — that is , th e theor y woul d demonstrat e nothin g mor e tha n tha t a grou p whos e members help each other is more successful than one whose members do not. That reciprocity must be guaranteed is a crucial element o f Trivers's idea , but h e doe s not specif y ho w it can be guaranteed in the anima l world. Let's take a very simple example of the problem of enforcement, on e that most of us have encountered in everyday life: littering. Those who carry a candy wrapper or a n empt y soda ca n t o a tras h receptacl e ar e altruists : the y take o n a n extr a burden s o as to preserv e th e qualit y of the environmen t fo r all . Those wh o flin g away a soda can the minute it is empty are selfish individual s who save themselves some trouble a t the expens e of all. It follows, then, that one who reprimands a litterer i s rendering a service to the public—but ma y pay a price fo r doin g so : he o r sh e ma y be verball y abuse d o r even hi t by the litterer. Indeed , mos t peopl e avoi d reprimandin g others, because it is "awkward," o r because they fear bein g abuse d o r attacked . Thus, thos e who volunteer t o deman d altruisti c behavior ar e themselves altruists; it takes altruism to enforce altruistic behavior in others. Trivers himsel f though t tha t amon g humans , and i n ape s an d othe r animal s with well-developed brains, individual s wh o do not reciprocate are discriminated against by other grou p members, and that this discrimination ensures reciprocity. The problem i s that the very action group members take against those who do not reciprocate altruis m i s itsel f a n effor t fo r th e benefi t o f th e group—tha t is , a n

Babblers, Competition for Prestige, and the Evolution o f Altruism 13

3

altruistic action. Trivers' s solution to the question o f altruism, then, begs the question, sinc e his enforcement mechanis m itsel f demand s altruism . Axelrod 15 ra n a computer simulation that found that punishing individuals who do not reciprocate will work, but onl y if any individuals who do not punis h are themselves punished as well. Th e proble m i s that punishin g demand s effor t i n it s own right . Axelro d concluded tha t th e mos t efficien t wa y to preserv e social norms is to have professional law enforcers suc h a s police, inspectors, an d judges—who are paid to d o the punishing. Of course , such professional enforcers do not exis t in nature. Trivers's theory of reciprocal altruism was accepted by th e scientifi c communit y as an important explanation of altruism between nonrelate d individuals. There were two reasons for this: first, there was no other theory tha t explaine d th e man y case s i n whic h grou p members clearly invest in the goo d o f the group, even though the y have n o geneti c relationshi p t o on e another; an d second, Trivers's theory seemed to be supported b y the many instances in nature where altruism was indeed reciprocated . The theor y of reciprocal altruism led to the devel opment o f comple x mathematica l models , lik e th e "tit-for-tat " model. 16 Thes e models investigate d through compute r simulation s under what condition s o f reciprocity altruis m would b e desirable . Wit h al l their sophistication , none o f these models manage d to resolve the greatest problem pose d by the theory of reciprocal altruism: How ca n one who does a favor fo r another ensure that the favo r wil l be returned?17

COMPETITION OVER ALTRUISTIC ACTS I N BABBLER S Other difficulties with th e theory o f reciprocal altruism emerg e when we observe babblers' behavior . Babbler s engag e in many altruistic activities. When the y eat, one o f the grou p act s as sentry. The sentr y is clearly hungry—when offere d foo d by human observers, it often eat s eagerly—but still , it stands guard when it could be feeding . Babblers giv e food the y find to othe r adul t members of their group . Again, they clearl y feed thei r comrade s befor e they are full themselves : when offered a crumb of bread right after they have fed their fellows with a similar crumb, they eat with relish—somethin g that satiated babblers d o not do . They endange r themselves by mobbing raptor s and snakes. They imperil themselves by coming to the rescu e o f grou p member s wh o ge t caugh t i n a ne t o r b y a predator , o r b y enemy babblers durin g a fight. They fee d th e offsprin g o f other grou p member s and take care of them. Not onl y ar e babblers , b y al l accounts , a t leas t a s altruisti c a s othe r group -

134 THE HANDICAP PRINCIPLE IN SOCIAL SYSTEMS breeding birds; 18 close, detailed observatio n show s that babblers actuall y compete with one another for the "right" to be altruistic. Instead of expecting their partner s to retur n ti t fo r tat , the y attemp t t o prevent the m fro m doin g thei r share . Th e theory o f reciproca l altruis m canno t explai n wh y individual s compete fo r th e chance t o hel p other member s o f the group , le t alone why they prevent others from helpin g in return. If we assume that the function of altruism is to benefit others , the n i t make s n o differenc e wh o the altruis t is , and ther e is no sens e i n competing for the "right" to help. On the contrary, the helper should prefer to have the benefit come from someone other than itself; the benefit to the ones being helped woul d b e the same , and th e cos t t o itself would b e less . But i f the helpe r benefits directly from th e ac t of helping, an d th e benefits t o other s ar e incidental—a sid e effect—the n i t makes sense t o compet e for th e opportunit y t o help . A s we shall show , the helper—th e "altruist"—doe s reap a direct gain—in prestige.

Sentinel Activities Let's take th e babblers ' sentine l activitie s as an example. When babbler s fee d i n the morning, on e of the group often perches on a tall branch o r a treetop. Whe n this sentine l see s danger , i t issue s "alar m calls" 19—which, a s we hav e said , ar e directed a t th e approachin g predator , bu t whic h als o aler t th e group . Standin g guard means exposing oneself to danger and forgoing food until the watch is over. Individual babblers may spend up to two or three hours a day standing guard. All group member s tak e part , a fac t whic h a t firs t seem s t o suppor t th e theor y of reciprocal altruism. But more detailed observation shows that the babblers ar e not at al l interested in having others do a fair share—tha t is, in reciprocity . Tirza Zahavi 20 foun d tha t higher-rankin g babbler s stan d guar d mor e tha n lower-ranking ones do, and that they interfere in the guardin g done by their sub ordinates; the y often expe l their subordinate s fro m th e guar d post an d tak e thei r place. When a lower-ranking guar d refuses to leave its post, a higher-ranking bab bler wh o come s t o replac e i t may shove it an d eve n hit it . On th e othe r hand , a babbler tryin g to replace a guard higher in rank than itself does s o by sitting on a lower branch—sometimes on a lower branch in another tree—an d waiting for the guard to leave; only then does the lower-ranked replacement go up to the dominant babbler's pos t (se e Graph 12-1) . As a rule, the guard clearly sees the one attempting to replace it: one can observe it preening its feathers, shifting around, at times climbing to an even higher branch. The sentinel is certainly in no hurry to take advantage of the other's offe r an d leave

Babblers, Competition for Prestige, and the Evolution of Altruism 13

5

Graph 12-1. A typical sequence of sentinel activity in the morning in one group; a downward arrow shows replacement. The dominant male (Ml) and the dominant female (Fl) both stood guard first thing in the morning. The third male (M3) came up to replace the dominant female only after Ml went down; he was several times replaced by the beta male, M2, and once by the alpha male, Ml. M2 in his turn was repeatedly replaced by Ml. When the second female (F2) tried standing as a sentinel on another tree, she was immediately replaced by the top female (Fl). Throughout this time, a young female (F3) stood guard intermittently; none of the others bothered to replace her.

its post. A romantic view would sugges t that dominant babblers ar e volunteering to serve in the place of tired comrades who are younger and less experienced, bu t detailed observatio n show s tha t thi s is not th e cas e a t all . Eac h make s th e mos t effort t o replac e the on e just belo w i t in rank, who i s likely to b e closes t to i t in age; individuals ar e considerably les s likely to replace much younger group members. If guardin g were based o n reciprocity, there would b e n o point i n striving to do more guard dut y tha n others . Eve n i f on e assert s that suc h competitio n i s necessary t o ensur e tha t th e grou p i s neve r withou t a sentinel , on e woul d stil l have t o explai n wh y each bir d interrupt s th e watc h o f the on e neares t t o i t i n rank, rathe r tha n attemptin g t o replac e younger , more inexperience d babblers . The dominan t male' s interferenc e wit h th e guardin g don e b y th e on e nex t t o him i n ran k ca n become s o intense that that second-ranke d bird, th e bet a male, ends u p guardin g much less than th e thir d i n rank , with who m th e top-ranking bird interfere s less (se e Graph 12-2) . This does not happen because of any lack of desire to guard on the part of the beta male : in fact, th e beta male goes up to guard more frequentl y tha n th e third in rank ; th e reaso n h e end s u p guardin g les s i s tha t th e alph a mal e keep s in terrupting hi s guar d duty . I t s o happened tha t i n tw o o f th e group s i n whic h detailed studie s of guarding were done , th e dominan t males disappeared durin g the study . In bot h cases , once the dominan t disappeared , the one next to him in

136

THE HANDICA P P R I N C I P L E I N SOCIA L SYSTEM S

Graph 12-2. Number of shifts and duration of watch of males Ml, M2, and M3 in one group. M2 came up to stand guard more often than M3 (M2 took more shifts), but his total sentinel activity was less, because the alpha male replaced him frequently.

Graph 12-3. Duration of watch for males M2 andM3 in several groups, relative to the duration of watch for the alpha male (Ml); note the reversals in groups G and H after their alpha males disappeared (G' and H'). Ml = 100%. Maximum duration: 2.5 hours/day in group H. Minimum duration: 0.5 hour/day in group C.

rank—now the top male—immediatel y started guarding much more tha n before the chang e (se e Graph 12-3 , groups G' an d H') . Befor e the dominan t mal e disappeared, th e bet a male s in these groups ha d bee n guardin g less tha n th e third ranked males ; after th e change , when th e dominan t wh o represse d th e bet a was gone, th e sam e birds wer e guardin g far more than th e formerl y third-ranke d individuals (se e Graph 12-3) .

Babblers, Competition for Prestige, and the Evolution o f Altruism 1 3

7

Feeding of Nestlings Babblers als o compet e ove r the feedin g of nestlings. Thamsi Carlisle 21 foun d tha t higher-rankin g one year-old babbler s actuall y interfere with othe r bab blers o f th e sam e ag e when thos e bird s tr y t o tak e care of the group' s young . This ca n reach surprising extremes: Carlisl e observe d case s in which a group of a dozen babbler s ha d onl y one surviving nestling, and four or five babblers would come to the nest with food i n thei r beak s a t th e sam e time. Eac h ha d t o wait—with food in its beak—until all those that outranked it had finished feeding the sole nestling. Sometimes the lower-ranked birds would not even dare approach the nes t (se e Graph 12-4) . If th e onl y function o f feeding nestlings is to nurtur e the nestlings , why on eart h d o babbler s preven t other babbler s fro m doin g jus t that?

Feeding of Other Adults (Allofeeding) Adult babblers , too , fee d eac h other . Usually , withi n eac h gende r suc h feedings go onl y on e way : a dominan t feed s its subordinate . Exception s t o thi s orde r o f rank and gender ar e rare, and even rarer among males. The dominan t femal e may feed the dominant male without challenging him, and other females may feed other males without dir e consequences . But we have seen case s in which lower-ranked individuals who attempte d t o fee d higher-ranking ones o f the sam e gender wer e beaten u p fo r thei r trouble. I n a t leas t on e case , suc h reverse d feedin g wa s a declaration o f open, aggressive revolt—an extremely rare event, as we have seen— which ended with the lower-ranked rebe l being thrown ou t o f the group . Babblers feedin g others draw attention t o themselves: they emit a special trill and lif t thei r bea k abov e th e bea k o f the on e the y are feeding. When ther e i s a large differenc e in age between the feede r and the one bein g fed , the latte r is sometimes eage r t o accep t th e food , an d ma y even approac h th e feeder , making begging sounds . But in som e 1 5 percent o f cases, babblers tr y to avoi d being fe d by anothe r bird. 22 Whe n a guard see s a higher-ranking comrade approachin g it with food, to feed it and replac e it on sentinel duty, it may sacrifice its guard post to avoid being fed. In other cases, the sentinel may close its beak tightly and refuse to accep t th e food being offered , eve n though it is hungry; it eagerly accepts a dry crumb o f brea d fro m u s immediatel y afte r refusin g a juic y insec t fro m anothe r babbler. I f th e ai m of allofeedin g is to suppor t weake r grou p members , why d o babblers try to feed individuals who do not want to be fed, and why do the latter resist? In thei r first months o f life, fledglin g babblers are still aggressive and hi t each

138

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

Graph 12-4. Visits to nests by yearling helpers, as percentage of mean visits. Data and graphs by Thamsi Carlisle (Carlisle and Zahavi, 1986). Left: Visits to unattended nests. Summary of 598 visits to 4 nests by 3-9 yearling helpers. When there was no other bird at the nest, there was no apparent relationship between the rate of visits to the nest and the helper's rank. Right: Visits to nests where the incoming helper had to replace another at the nest. Summary of 187 visits to 3 nests by 3-9 yearling helpers. The solid line indicates that the lower the rank of the incoming bird, the less frequently it came to the nest when another helper was there. other frequently. The y may even switch ranks during their first year of life. Thamsi Carlisle23 recorde d 9 4 cases of fledglings and yearling s feeding each other . I n 8 6 cases a yearlin g offere d foo d t o it s subordinate . I n 2 5 o f those—wel l ove r a quarter—the subordinate refused th e offere d foo d or only tasted it. In 7 of these 25 case s of refusal , th e would-b e feede r chase d th e subordinat e an d sometime s pecked at it. Only in 8 cases did a subordinate yearling offe r foo d to one rankin g above it. In full y 6 out o f these 8 instances, the dominan t yearlin g refused to eat and attacke d the would-be feeder. In Carlisle' s research , most feeding was don e with breadcrumbs given to the birds by the observer. More recently, Amir Kalishov studied allofeedin g amon g adul t babbler s withou t offerin g the m any food ; his findings were similar to Carlisle's. 24 In one case in Carlisle's study, an eight-month-old female tried to feed her tenmonth-old sister; the latter stood up , snatched the food out of her younger sister's beak, forced he r t o crouc h lik e on e begging fo r food, an d then stuffe d the foo d down he r throat . Onc e th e younge r bird ha d swallowe d th e food , the dominan t sister pecke d he r unti l sh e fled. The frustrate d younger sister then too k anothe r crumb and went to feed her subordinate younge r brother, who was hunting quietl y for foo d some thirty feet away. In anothe r case—one of the tw o instances in Carlisle's stud y in which the gif t of food was accepted b y one higher in rank than the feeder—the dominant fledg-

Babblers, Competition for Prestige, and the Evolution o f Altruism

139

ling accepted th e foo d but the n attacke d th e feeder and kicked i t in the face, a s if to say "I'll tak e this food, bu t I still outrank you." Fa r from rewardin g altruistic behavior an d returnin g ti t fo r tat , babblers who ar e offered foo d b y their subor dinates actually punis h them fo r making the offer. I f the function o f feeding other s were to benefit the ones being fed, why would th e recipients o f the benefit punis h their benefactors ? Babbler s behave as though i t is the ac t o f giving, rathe r tha n the benefit given , that matters .

Mobbing Babblers' mobbing is generally simila r to that of other vertebrates. A babbler who notices a snake or a raptor will sound a tzwick and then make barking an d trilling sounds. This alerts th e other members of the group, wh o arriv e an d approac h th e raptor . When a babbler comes within four to six feet of th e raptor , i t spread s it s tai l an d lift s it s wings, fannin g the m out . I t holds that stanc e for severa l seconds; it may turn sideway s and circle the raptor , tai l an d wing s stil l spread . These display s last a minute o r two , an d th e babblers' movements chang e accordin g to the degree of danger. Zahava Carmeli, who studied babblers' mobbing of snakes, found tha t the babblers were well aware of the risk: she determine d tha t o n averag e mobbing babbler s kep t a distance o f 1 0 to 15 inches from a snake's tail , but the y kept some 20 inches from its head, whic h would indicat e tha t th e bird s kne w ful l wel l whic h en d o f th e snak e pose d a danger.25 If the babblers attacked the snake at all, they aimed at its head. Mobbers' behavior varie s a great deal : som e div e headlon g int o danger , other s hol d back . Different babbler s stop at different distance s from th e predator. The tendency o f a babbler to mob correlate s to the social makeup of its group. Avner Anav a studied babblers ' mobbin g o f raptors , ofte n usin g stuffe d models . Two of the groups he studied comprise d one adult male living with several females and year-ol d fledglings; in thes e groups , th e dominan t mal e di d no t participat e much i n mobbin g (se e Grap h 12-5 , grou p HAK) . I n group s wit h severa l adul t male members, all those adul t males participated muc h more. In these groups, the dominant mal e alway s mobbe d longe r tha n th e other males , cam e close r t o th e raptor, stoo d guar d during mobbin g mor e than others , and actively disrupted th e mobbing activitie s o f th e other adul t male s (se e Graph s 12- 5 an d 12-6 , group s ZEH an d MZR). The activitie s o f the second - an d third-ranking males depended mostl y on the relationships betwee n the m and the first male. When such males accepted unquestioningly the position of the dominant male—fo r example , if they hardly made any

Graphs 12-5 and 12-6. Competition over mobbing and interference with mobbing. Data from Anava, 1992. At the time, group MZR comprised a father (Ml), a mother (Fl), an adult son of the father but not of the mother (M2), and a son and a daughter of both father and mother (M3 and F2). Group ZEH comprised three sibling males and a female who was not related to any of them. Ml and M2 were from the same brood, and during the course of the study there was intense conflict between them; eventually, Ml disappeared —probably ousted by M2, who then took his place. Group HAK was composed at the time of a father (Ml), a mother (Fl), their daughter (F2), and a stranger female who had recently joined the group (F3). Graph 12-5 (top): Position relative to the raptor: nearest = 1, farthest away = 4. In the MZR and ZEH groups, the dominant males were almost always the ones nearest the raptor, and the dominant females, who were the only adult females in each group, stayed farthest from the raptor. In group HAK, which comprised one adult male and three adult females, the dominant male stayed farthest from the raptor, and the dominant female came nearest. Graph 12-6 (bottom): Interference during mobbing—who interferes with whom. In the MZR group, no one interfered with the dominant male (Ml). The dominant male himself interfered a great deal in the mobbing activity ofM2. N o one interfered with M3, but M2 interfered with the dominant female (Fl), and M3 interfered with his sister, F2. In the ZEH group, Ml interfered a great deal with M2, but M2 also interfered with Ml. Both Ml andM2 interfered with M3—Ml more than Ml. M3 interfered with the female only. In the HAK group, Ml did not interfere with Fl but did interfere somewhat with the stranger female, F3. None of the other females interfered with Fl. Both Fl andF3 interfered with F2, and both Fl and F2 interfered a great deal with mobbing by F3.

Babblers, Competition for Prestige, and the Evolution o f Altruism 14

1

attempt t o copulat e wit h th e breedin g femal e a t laying time—they mobbed les s than he did: the y stayed farther away from th e predator, mobbe d fo r shorter periods, stoo d guar d less during mobbing , and sounde d fewe r an d less orderly trills (see Graph 12-5 , grou p MZR) . At th e othe r extrem e wer e tw o group s i n whic h ther e wa s a mor e intens e conflict betwee n th e dominant and the beta male s over copulation . I n both cases , the firs t an d secon d mal e participate d almos t equall y in mobbing , an d th e first interfered a great deal with the mobbing o f the second . The interference was such that in on e group th e bet a mal e ende d u p mobbin g eve n les s tha n th e thir d in rank. One o f these groups was the only one in which the beta male dared interfere with the mobbin g o f the first (see Graph 12-6 , group ZEH) . I n both groups th e first male disappeare d eventually , probably kicke d ou t b y the secon d mal e who then took his place: in one case this happened durin g the study, in the other, right after i t ended. The difference s i n mobbing between females als o correlate d wit h competition amon g them. A dominant babbler's interferenc e in the mobbing activitie s of its subordinate was usually subtle; the dominant might come and sit right next to the lower-ranked mobber, pree n th e subordinate' s feathers , o r pus h i t lightly . In suc h cases , th e subordinate woul d ceas e fo r a tim e t o utte r th e trillin g sound s babbler s mak e during mobbin g an d woul d crouc h i n plac e o r back off. Sometimes the dominant was rougher: it might flank its rival abruptly, cross the rival's path, or lan d nea r o r i n fron t o f it . Th e othe r woul d react wit h a frightene d jum p accompanie d b y a tzwick an d sto p mobbin g temporarily , o r eve n stop altogether . Again , the dominan t mal e intervened i n mobbin g b y th e second-ranke d male , who is usually an experienced adult, far more than in the mobbin g o f youngsters. In othe r words , it is no t t o protec t inexperience d grou p member s that the dominant bird interferes . Incidentally, mobbin g crow s hav e als o bee n foun d t o compet e fo r prestige . Slagsvold26 showed tha t higher-ranking individuals mobbed mor e than others and even attacke d an d chase d awa y lower-rankin g individual s during mobbing . H e suggested, a s we have, tha t th e higher-rankin g crows advertis e i n thi s wa y their control ove r the flock.

ALTRUISM A S A SUBSTITUT E FO R THREAT S In short , babblers compet e to perform altruistic acts, and often the y even attempt to preven t thei r subordinat e comrade s fro m actin g fo r th e goo d o f th e group .

142

HE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

Evidently, each babbler care s mainly about providing the benefit, rather than about the actua l benefit s th e grou p receives . I f w e wan t t o understan d th e babblers ' altruistic behavior , the n w e hav e to determin e wha t th e giver—th e "altruist" — gains from th e ac t of giving, rather than what th e grou p a s a whole gains . It i s important to remember that grou p members of the same gender compet e with eac h othe r ove r th e utmos t biologica l need—th e chanc e t o reproduce . I n spite of this rivalry, they hardly ever fight one another. In the vast majority of cases, conflicts within th e grou p are settled by means of the competitio n ove r altruism. How ca n altruism replace physica l aggression? This brings us back to the substitution o f threat s fo r actua l fightin g (se e chapter 3 , o n rivals) . Because o f th e danger an d cos t inheren t i n fighting , t o bot h lose r an d winner , other way s o f resolving conflicts evolved. Actual, all-out physical struggle was replaced by threats that showe d wit h a reasonable degre e o f reliability how likely the threatene r wa s to win an actual fight. But threats have their own price. When threat s are not effective, the threatene r does have to fight—or forfeit. Th e stakes are highest when an animal issues a threat in th e presenc e o f witnesses; when a n individual's threat s go unheeded, th e individual ma y fai l t o dete r no t onl y th e riva l bu t th e witnesse s a s well. Thus , th e presence o f witnesses tends to make threateners less likely to compromise. More over, witnesses who see a contestant injure d or spot som e weakness in one or th e other may seize the opportunity and take over. This is especially true among highly social animals like babblers, wh o ar e always together. An y threat not followe d u p by action , an y unsettled conflict , an y failure of a dominant membe r i s noticed b y the entire group—comrades who are also potential rivals . Yet conflict s d o occur an d deman d resolution—s o a substitut e for threaten ing—which itself evolved a s a substitute fo r fighting—i s highly desirable. Actions that ar e not direc t threat s but ar e closely relate d t o a n individual's abilit y to win a struggle can take the plac e o f threats. By showing off how much they can invest in standing guard, in feeding their comrades, in taking risks, and in other altruistic acts, babblers show off their ability to win in a fight and their desirability as groupmates. The altruist's investment in the altruistic ac t offer s a reliable , concret e inde x of that individual's ability. Altruistic acts have another benefit : they prove th e giver' s interes t i n th e receiver , even whil e the y proclaim th e giver' s domi nance. By investing in the good of the group, the dominan t mal e show s of f both hi s su periority an d hi s willingnes s t o giv e t o hi s subordinates. Thi s make s thes e subordi nates les s likely to leave and thu s helps th e dominant bir d remain at the head of a large, strong group.

Babblers, Competition for Prestige, and the Evolution of Altruism 1 4

3

Other grou p member s hav e good reaso n t o pay attention whe n an individual acts for the group's benefit: what they learn about that individual helps them avoid a high-ris k fight against a rival stronger than the y are. All members of the grou p must constantl y be assessin g thei r prospects . Afte r all , eac h bir d ha s t o decid e whether to sta y on in the grou p as a subordinate, leave and tr y its luck elsewhere, or ris k everythin g in a fight for th e to p position . Altruisti c acts that benefi t th e group help each babbler answer these very questions.

RANK AND PRESTIG E Rank within the group is very important in the life of babblers. Every social activity, from fightin g other group s t o dancin g an d mobbing , i s affecte d b y an d reflect s social rank . Babblers slee p clumpe d togethe r alon g a tree branch: the dominan t male alway s sleeps a t on e end , an d th e second-ranke d mal e usually at the othe r end. More dominant birds usually stand guard more than others. They often tak e over fro m thei r subordinates , while thei r subordinate s avoi d tryin g to tak e over from them . Each babble r feed s its subordinates, whil e th e latte r ver y rarely feed their superiors . Subordinates sometimes make a special sound when a more dominant bir d passe s b y them . Dominant s ca n pree n thei r subordinates ' back s an d tails as well as preening them elsewhere; subordinates preen babblers that outrank them almost exclusively in areas other than the back or tail. Why is rank expressed so vigorously when for all practical purposes it does not change as long as a babbler stays in its original group? In time we discovered tha t the relationship between the first and second males is no t th e sam e in al l groups. I t depend s a great dea l o n th e particula r group' s makeup. Eve n whe n th e tw o ar e brothers , a s is often th e case , this relationshi p varies a great deal from grou p to group. At one end of the spectrum, the dominan t male is an old, experienced babble r and the second is five or six years younger. In such a case , the secon d neve r dare s fee d th e firs t durin g th e dominant' s guar d duty, alway s makes th e submissiv e soun d whe n th e dominan t approaches , an d opens hi s beak t o as k the superio r fo r food . A t th e othe r extreme , th e secondranked bir d i s o f th e sam e brood a s the firs t an d becam e hi s subordinat e a s a consequence of childhood squabbles . I n that case, the second stands guard a great deal, the firs t feed s hi m onl y infrequently and i s often refused ; o n rar e occasions the secon d will even feed th e first. The social rank is the sam e in both cases—one is clearly the first and th e other clearl y the secon d i n rank—but the differenc e i n the degre e of control is very significant. To reflec t thes e difference s in th e qualit y of relationship s within groups, we use the ter m prestige, which fo r us is different fro m an d complementar y to rank. (In previous writings we used social status. 21) Prestig e is the respec t accorde d an individual by others. I n socia l systems where the stron g have litde interest i n the

144

THE HANDICA P P R I N C I P L E I N SOCIA L SYSTEM S

well-being of others, prestige can be won by unprovoked aggression , whether physical attack or threat. Such aggression is so common that social ranking is also called "pecking order." 28 Prestige ca n als o b e acquire d b y "wasteful " showin g off ; the peacock's tail, the deer's antlers, an d the bowers built by bowerbirds are examples of showin g of f tha t i s neither harmfu l no r usefu l t o others . Babbler s an d othe r altruistic animal s gain prestig e b y investing in thei r fello w grou p members ' well being. Their partner s benefit, bu t th e reaso n for their altruis m lies elsewhere—in the real testimony to their abilit y that it provides . Social rank is easy to discern; prestige, o n the other hand, is complex and harder to measur e precisely. Prestige reflect s th e degree o f a superior individual' s dominance, as recognized by subordinate members of the group. In othe r words, prestige is gauged by others. The dominan t ma y claim prestige, but fo r the prestige to be real it has to be accepted by subordinates, and it is this acceptance that actually determines a n individual's prestige . Prestige ha s real value: a dominant male with higher prestige ca n get with ease what a dominant mal e with les s prestig e gets only with grea t effort , o r no t a t all. Babblers mak e constant , unceasin g effort s t o stres s thei r ability , no t t o ris e i n rank—an adult babbler ca n only "promote" itself by killing its superior or chasing it from th e group, o r by leaving the group itself—but t o persuade their comrade s to recognize tha t abilit y and grant them prestige . A babbler who ca n stand guard longer tha n it s comrades , giv e them par t o f its food, approac h a raptor, tak e th e risk of sleeping at the exposed end of the row—and can also prevent others from doing such deeds—proves daily to its comrades its superiority over them. By doing so, that individual increases its prestige an d ha s an easier time exerting control . Greater prestige means that subordinates will avoid getting into a fight for the higher position . Greate r prestig e convince s the female to stay faithful t o the dominant mal e an d not t o mate wit h hi s subordinates. In short , a babbler's altruistic acts ar e a n investmen t i n it s prestige , an d tha t prestig e ha s real , concret e value . The difficulty—th e cost , or handicap—of this altruistic advertising, whether foo d sacrificed o r danger incurred, i s what makes it a reliable indicator o f the abilit y of the advertiser . Subordinates hav e their ow n prestige, whic h is shown by their abilit y to resis t their dominants and by their superiority over those below them. Babblers replace d during guar d dut y o r fe d b y on e mor e dominan t tha n the y ofte n tr y t o regai n prestige a s bes t the y ca n b y replacin g a lower ranking guard, or by attacking a bird of a differen t species, lik e a bulbu l o r a blackstart . Th e fre quency of these behaviors shows clearly both that individual babbler s ar e ver y awar e o f thei r ow n prestige an d tha t man y of their actions , especiall y altruistic actions , ar e done to gain prestige . We als o find abundant evidenc e that othe r grou p member s ar e well awar e of their comrades ' claim s to prestige. Subordinat e male s often approac h a female or

Babblers, Competition for Prestige, and the Evolution o f Altruism 14

5

a lower-ranked male who is standing guard, feed it , and take its place as a sentinel; in som e groups, this almost alway s provokes a reaction fro m th e dominan t male, who hurries to feed the subordinat e guar d and replace him. Sometimes the dominant male stands guard only for a moment before leaving the guardpost—as if to say that hi s only intention wa s to make his rival stand dow n an d asser t his superiority, which may have suffered slightl y from the subordinate's displa y of prestige.

"SHYNESS" OVER COPULATION A S A TEST OF MAL E PRESTIGE Babblers' copulatio n ritual s ar e tests an d demonstration s o f the dominan t male' s prestige. Mos t animal s copulate freel y i n th e presenc e o f other member s o f thei r species. No t s o babblers. Eve n thoug h babbler s spen d mos t of their live s within sight o f their fello w grou p members , they do not copulat e i n their presence : they hide under or behind bushe s and copulat e in privacy. The other group members know what the pair are doing in their hideout, since the invitatio n t o th e ac t i s often mad e i n public : a male invitin g copulatio n ap proaches the female with a bit o f food or a twig in his beak, and instead of feeding her, lift s u p hi s gif t an d move s awa y fro m th e group . I f th e femal e wishe s t o copulate, she follows. If she does not follow, the male may beat her and chase her, but thi s is usually ritualized, show-off aggressio n only . We hav e never yet seen a case of a female being thrown out o f a group because she refused to copulate with the dominan t male , though we often se e females refusing an d being harassed. Sometimes on e of the other males follows th e pair. In such cases the dominant male stops , turn s an d approache s th e subordinate , "reprimanding " hi m vocally and sometime s circling him. In mos t case s the subordinat e mal e halts, acknowledges his inferior status, and return s to the group . But in a significant minorit y of cases he is not deterre d b y the dominan t male's warnings. In som e of the group s we have observed, th e secon d mal e continue d followin g the pair , preventing th e dominant mal e from copulatin g fo r a number o f days . Such a conflict sometime s ends in a real fight in which the defeated party is killed or thrown out of the group; when i t does not , i t proves t o the female that this particular dominan t mal e does not wis h t o o r doe s no t dar e figh t hi s secon d i n rank , presumabl y becaus e h e cannot affor d t o los e him a s a group member . Th e abilit y of the secon d mal e to interfere is a clear measure of his prestige. Indeed , in such cases we often find that the female copulates later in her fertile period wit h the second mal e as well as with the first. 29 Each male's reproductive chance s depend on cooperation fro m the female. The female needs to gauge as precisely as she can the current prestige of her suitors—no simple feat , sinc e th e competitio n ove r breedin g wit h he r i s likel y t o b e mos t

146 T H

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

intense within day s of her arriva l in the group. At that point, the new female doe s not yet know the group well; moreover, the senior subordinate males in the group are not he r offspring , an d thu s al l of them ar e eligible t o copulat e with her . Th e new female can learn the males' ranks quickly and easily, but ran k alone does not tell her th e degre e o f control th e dominan t mal e has over his subordinates. Ho w much doe s h e need them ? How much do they recognize his dominance? The new female has to split her male suitors' chance to breed according to the balance o f powe r amon g them , bu t ho w ca n sh e fin d ou t wha t tha t balanc e of power is if they live together in peace, with no overt aggression whatsoever? If the female make s a n erro r i n grantin g o r withholdin g he r consent , sh e ca n en d u p with a powerful, dissatisfie d male groupmate who wil l break he r eggs . We hav e observed several cases of males breaking eggs in the nest under such circumstances. The mai n victim i s the female , wh o invest s more i n producing egg s than d o he r consorts. The "shyness " tha t prevent s th e femal e fro m copulatin g i n th e presenc e o f other babbler s help s he r determin e th e prestig e o f eac h male—or , t o b e mor e precise, it allows her to see how much prestige each male is accorded by the others. This enable s he r t o shar e he r favor s accordin g to th e relativ e importanc e o f her partners. Th e dominan t male's prestige is demonstrated whe n other males refrai n from followin g him an d the femal e eve n at a distance. Babblers who let the dom inant mal e retir e wit h th e femal e withou t eve n tryin g t o observ e the m fro m a distance show their acceptance o f the dominan t male' s superio r prestige. By contrast, a male who follows with impunity displays to the female his own high prestige as a subordinate . A few days before egg-laying, the dominant male starts "guarding" the female. 30 He follow s the breeding femal e closely and keeps her in sight; at whatever time of day, h e i s neare r t o he r tha n i s an y other grou p member . Whe n anothe r mal e approaches her , th e dominan t mal e doe s s o too . I n mos t case s hi s approac h i s enough t o caus e the subordinat e t o mak e the submissiv e sound o r show his discomfort b y fluffin g hi s feathers and preenin g himself. The dominan t mal e alway s copulates wit h th e femal e more tha n an y o f th e other males do. But other males, too, may copulate with the femal e if they are not her offsprin g an d if she cooperates. I f the female wishes to copulat e with anothe r male, she can always find the one or two minutes needed when the dominant male is absorbed i n something else , such as a border conflic t or nest-building. I n som e groups, subordinate male s manage to copulate with the female even on her fertil e days an d eve n i n the mornings , which see m t o b e th e bes t tim e fo r fertilization. Yoel Perl found that i n some cases the dominan t stops guardin g the femal e eve n while she is still fertile, befor e the ovulation of the fourt h egg ; the secon d i n rank then copulates with her and presumably fertilizes that egg.31 In other groups, males of similar rank do not copulat e a t all. An individual's reproductiv e chances , then, cannot be predicted by rank alone.

Babblers, Competition for Prestige, and the Evolution o f Altruism 1 4

7

When the dominan t male allows his subordinates t o mate with the female , he does so because he needs them in the group an d must therefore make concession s to them. We think that th e position o f each grou p membe r depend s both on the balance o f power withi n th e grou p an d o n th e group' s neighbors . Som e groups have stron g neighbor s an d nee d mor e an d bette r fighter s fo r protection , whil e others hav e weak neighbors . Som e dominan t male s ar e strong an d d o no t nee d much assistanc e in fighting; others depen d more on their subordinates . Another factor tha t ca n affec t th e balanc e of power withi n a group is that th e more partners the dominan t mal e has, the les s dependent h e is on any single one of them. And a dominant male with adult sons by the curren t dominant female— who do not mate with their mother—depends less on the assistance of other males who might compete with him over copulation. Al l these factors affect cooperatio n among males. Every group of babblers has one male who outranks all the others, but som e dominant males let their comrades copulate, while others threaten them or eve n pec k a t the m whe n the y try . Clearly , rank i s onl y par t o f th e picture ; prestige, too, is a crucial element.

REASONS FOR AND CONSEQUENCE S OF LIVING IN GROUPS The imperative to live in groups is powerful amon g babblers and is reinforced i n various ways. Babbler groups make it difficult fo r solitary pairs to succeed o n their own. When a pair tries to breed nea r a large group, thei r neighbor s giv e them a hard time : we have observed severa l raids tha t cause d a breeding pai r t o deser t their nest or led to the killing of their fledglings. Even after years of drought, which make it impossible fo r babbler s t o rais e young and leav e some territories empty, subordinate adult babblers in neighboring territories do not usually hasten to settle on thei r ow n i n th e unoccupie d lands . Instead, the y ofte n sta y i n thei r origina l territory and compete with their brothers for the chance to reproduce. Presumably, being th e secon d male , or eve n th e thir d o r fourth , in a large, secur e group with a good territory is better than being the dominant in a marginal territory, or trying to live without supporting subordinates . It seems that male babblers prefer to split off an d tr y t o for m thei r own grou p onl y when the y have a t least one helpe r t o accompany them, an d when two or more females fro m other group s are available to join them. It may well be that, over time, some organisms develop characteristics that help them liv e efficientl y i n group s an d los e other s tha t migh t bette r enabl e the m t o live singl y or i n pairs but ar e detrimenta l i n a group setting . Fo r example , in an unfamiliar situation, a group member ca n afford t o refrain fro m action and observe

148 T H

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

its experienced comrades , rather than experimenting and possibly failing. Babblers indeed benefi t fro m a n extende d perio d o f learning, durin g whic h the y acquire both group-living skills and abilities that enable them to find food, avoid predators, and dea l with strange babblers. Likewise , in a group setting, it pays to be inclined to compromis e rather than quic k to fight over one's rights; we have seen that it is exactly this temperament that characterize s babblers . Adaptations such as these may compensate fo r the disadvantage s of group life ; this may explain why the score s of species o f babblers in the Indian subcontinen t all live i n groups , althoug h the y live i n man y differen t habitats , which ar e als o occupied successfull y by solitary songbirds.32 After all , some adaptations are likely to b e irreversible , sinc e the y involv e a comple x se t o f interconnecte d changes . Reptiles an d mammal s wh o lef t th e lan d an d returne d t o th e wate r continu e breathing through thei r lungs , even thoug h i n their ne w habitat i t would benefi t them to breathe through gills. If babblers could all at once regain the characteristics they los t i n adaptin g t o grou p life , the y migh t d o bette r a s pairs i n som e ne w environments; as it is, they do better living in groups . Some species of birds live in groups or in pairs depending o n conditions . Seychelle s warblers , which usuall y live in groups , were transferre d i n one cas e t o a n islan d withou t warblers . The y formed pairs, nested, and increased in number until all suitable territories were occupied, an d then they went bac k t o grou p living. 33 Fo r thes e warblers, grou p livin g seem s t o b e a relativel y new development: unlike the babblers, they don't seem to have group-living relatives. Group livin g has develope d i n differen t spe cies for different reasons , and there are many different kinds of collaboration. Some provide a n edge i n the competitio n fo r resources , whethe r wit h othe r specie s o r with othe r group s o f th e sam e species . Other s affor d bette r protectio n agains t predators o r adverse environmental factors. Among some birds, like the scrub jays of Arizona , coalitions ar e limited t o parent s and adul t offspring ; i n suc h specie s the adult offsprin g d o not shar e in reproduction. 34 Some other jays , som e Africa n starlings , an d long-tai l tit s for m coalition s t o protect thei r territory . Each grou p i s made up o f several pairs, bu t eac h pai r has its own nest. Individuals whose nest fails or who do not find a mate assist the other pairs at their nest s and help protect th e commo n territory, but the y do not themselves reproduce. Th e groove-bille d ani s of Central Americ a for m ne w coalition s every year, with several pairs sharing a single nest.35 Simpler coalitions may consist of severa l female s wh o la y egg s i n th e nes t o f a singl e male , a s i n th e cas e o f ostriches, o r o f severa l male s wh o hel p a singl e femal e rais e he r young , a s d o Galapagos hawk s and the harrier s of Scotland. 36

Babblers, Competition for Prestige, and the Evolution o f Altruism 14

9

PRESTIGE AND TH E EVOLUTION OF ALTRUISM: ALTRUISM AS A HANDICAP In al l cooperativ e animals , whatever th e primar y reason fo r formin g the group , conflicts amon g partners ar e settled to one degre e o r another b y the competitio n for prestige . A n individual's prestig e is reflected in—indeed, define d by—th e respect accorde d hi m o r he r b y others. Thi s respec t allow s the dominan t male t o keep other s a s his partner s an d t o contro l hi s grou p withou t punishin g them . Greater prestige ensures an individual a bigger share of the partnership's "gains"— that is, a better chance to reproduce successfully. Increased prestige for one partner means a loss in prestige fo r another . In other words, it is a zero-sum game within the group . Scientists increasingly recognize that many altruistic activities observed among animals canno t b e explaine d b y th e theorie s o f ki n selectio n an d o f reciproca l altruism. Some have suggested recentl y that even though altruisti c behavior is not reciprocated b y other s directly , i t i s rewarde d b y "indirec t reciprocity" : other s "reciprocate" by according status to the altruists. 37 But this concept stretche s the definition o f reciprocity to the point o f meaninglessness. That the altruis t gains in status, or prestige, is precisely what we have been suggestin g for years. And onc e one accepts that altruisti c activities bring those who perform them a gain in status, then no furthe r explanatio n for altruism is needed . Both th e competitio n ove r altruis m an d th e effor t b y others to avoi d receiving benefit s hande d ou t b y altruist s ru n counte r t o th e theor y of "reciproca l altruism." Indeed , on e ca n eve n vie w altruisti c act s a s implie d o r surrogat e threats, sinc e th e prestig e th e altruis t gain s allow s hi m o r he r t o achiev e wha t other animal s gai n b y threats . Observatio n suggest s that altruis m i s a goo d in dicator o f a n individual' s quality , both a s a partne r an d a s a rival—an d thu s it can substitut e fo r over t threats . Needless t o say , a grou p whos e member s compet e fo r prestig e b y demon strating thei r altruis m will b e bette r equippe d t o compet e agains t other group s than a coalitio n whos e member s vi e fo r prestig e b y harassin g thos e o f lowe r rank o r engagin g i n wastefu l display . Bu t i t i s importan t t o remembe r tha t what motivate s th e individua l i s not th e benefi t t o th e grou p bu t th e enhance ment o f hi s o r he r ow n prestige ; th e benefi t to th e grou p i s a side effect . I f in dividuals di d no t gai n directl y i n thi s way from thei r investmen t i n th e goo d o f the group , ther e woul d hav e been no stabl e foundatio n for the developmen t of altruism. Logically , w e assum e tha t i f circumstance s change i n th e futur e an d babblers o r othe r socia l animal s fin d i t mor e beneficia l t o gai n prestig e b y wasteful display , intragroup fighting , o r threats , altruis m wil l disappea r i n spit e of it s benefit s t o th e group . Once one views altruistic acts as signals that show the ability and the intentions of th e altruist , the n altruis m n o longe r pose s an y evolutionary enigma. Th e in -

ISO TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

vestment i n th e altruisti c behavio r i s the facto r tha t ensure s th e reliabilit y o f th e signal.38 To th e altruist, the cost—the waste—of altruis m is no differen t tha n th e waste of, say, growing and carryin g a large, heavy, decorated tai l like the peacock , or building an elaborate bower like the bowerbird. Th e cost of these and any other signals i s th e handica p tha t ensure s tha t the y ar e reliable—tha t th e signale r is indeed wha t h e o r sh e claims to be . Thu s ther e i s no nee d t o loo k fo r a specia l evolutionary mechanism to explain the evolution of altruism. On the contrary, we believe that ou r explanation o f altruism as a signal holds true not only for babbler s and other birds , but als o for mammals,39 including humans ; for social insects; and even for one-celle d organisms .

C H A P T E R 1

3

THE SOCIA L INSECTS: WHY HEL P TH E QUEEN ?

I

n natura l selection , "success " mean s succes s in reproduction . Ou t o f tens of thousands of seedlings on a forest floor, only a handful survive to become fruit bearing trees; out of thousands offish eggs, few develop into reproducing adults. Most individuals die before they ever have a chance to reproduce. This tiny survival rate does not conflic t with the theory of natural selection but rathe r displays it in all its potency : faile d seedlings , o r fis h tha t hav e not reache d maturity , die an d disappear. Among bees , too , onl y on e i n ten s o f thousand s i n a hiv e ma y brin g fort h descendants. That in itself would not caus e a raised eyebrow. But seedlings or fish eggs simpl y disappear; the femal e worker s i n a hive—that is , the eventua l losers in reproductive competition—rather than focus all their efforts on trying to survive and reproduc e themselves , actuall y expen d effor t t o hel p th e quee n reproduce . How coul d suc h behavior evolve by natural selection? This question has been seen in different light s at different periods . For Darwin, the main question was a technical one: How can a queen, so different fro m workers, pass o n t o he r offsprin g th e characteristic s of workers? Afte r all , both th e bod y structure and behavior of a worker bee i s very distinct from thos e of a queen bee , and amon g ants and termites , th e differen t type s of workers vary greatly not only 151

152 TH

E HANDICA P P R I N C I P L E I N SOCIA L SYSTEM S

from th e quee n bu t als o from on e another . Darwi n di d no t se e any difficulty i n the existence o f sterile individuals who assist the group : he simply assumed that a queen whose daughters were better workers had an advantage over a queen whose daughters were less well-adapted t o their task. 1 But worker bees, wasps, and ants of man y species ar e no t inherentl y sterile , an d som e actuall y d o reproduce . So researchers now pos e a different questio n a s the centra l issu e in the evolutio n of the socia l insects: Why do the workers, who retain the ability to reproduce, work for th e hive , rather tha n devotin g their effort s t o whatever chance they may have to produc e thei r own offspring ?

THE EVOLUTION O F SOCIAL STRUCTURE S IN THE SOCIAL INSECTS Conditions that Favor Collaboration: Food Storage and Helpless Offspring First, w e need t o find out why the socia l hymenoptera—bees, wasps, and ants — band togethe r a t all . Th e answe r ma y be foun d i n thei r particula r strateg y fo r raising offspring. Unlik e the larvae of butterflies or grasshoppers, who ca n move and see k foo d by themselves and whos e succes s depend s o n thei r own ability t o find foo d an d avoi d enemies , th e larva e of socia l hymenopter a ar e fe d b y thei r mother or by other caregivers as they grow. This allows them to use food gathered over much greater distances, far beyond th e reach of such larvae who get to their food b y crawling—thus , mor e foo d i s availabl e t o them . I t als o save s them th e need to move: in fact, the y are eyeless an d legless. An offspring's dependenc e o n food supplie d b y the adult does not necessarily cause th e developmen t of a social structure, but i t was a major facto r i n the de velopment o f social organizations in hymenoptera. Both the stationar y larvae and the food gathered for them attrac t parasites, predators, an d same-species robbers. These dangers make it desirable to build a nest for the eggs and the larvae, and to protect that nest. Partnerships ca n provide better protection : one partner can seek food an d leav e the othe r guarding the nest. In evolutionar y history, thi s fac t was the basi s of collaborations—but i t als o mad e it possibl e fo r stronge r partner s t o exploit weaker ones. The larvae's dependence o n food supplie d by adults allows the latter to contro l each larva' s size an d destiny . Th e egg s fro m whic h queen s an d worker s develo p are the same. In order t o develop int o workers rather than queens, larvae have to be represse d b y their caregivers—throug h undernourishment o r pheromones or with physical force. Futur e workers get small amounts of inferior food; larvae that are designate d b y th e adult s t o develo p int o queen s receiv e larg e quantitie s of

The Social Insects: Wh y Help th e Queen? 15

3

superior food—lik e th e honeybee' s roya l jelly. The caregiver s thu s divid e thei r resources unequall y and depriv e som e of the offsprin g in a way that lessen s th e latter's chance s to reproduce . Parent s exploi t som e o f their offsprin g b y making them int o helpers who will assist them o r their other offsprin g t o reproduce . The theor y of "parental manipulation, " a s it was called by Alexander,2 assumes that the parent is concerned wit h th e su m tota l o f it s reproductiv e suc cess—but no t necessaril y wit h th e fat e o f on e o r another offspring. 3 Thi s explain s th e benefi t t o th e parent in exploiting som e of its offspring, but i t does not explai n the other sid e of the question: Why does the shortchange d offsprin g hel p it s parent? Alexander assume d that the theories of kin selection an d of reciprocal altruis m explai n adequatel y why the represse d worker s assis t thei r re producing kin . We disagree.

The Haplodiploid Mechanism of Gender Determination Another facto r played a part i n th e evolutio n o f insect socia l systems: the haplo diploid mechanism , a system that determine s gende r i n hymenopter a (a s well as in some othe r insect groups, but no t i n termites). Males come out of unfertilized eggs and ar e haploid—that is , they have only one se t of chromosomes, inherite d from their mother. Females hatch out of fertilized eggs and have a diploid (double ) set o f chromosomes , hal f fro m th e mothe r an d hal f fro m th e father . A fertilized female—a queen—keep s th e sper m sh e receives i n a special receptacl e near he r oviduct an d ca n lay either fertilize d eggs that hatc h female s o r unfertilize d eggs that hatc h males . Unfertilize d females—workers-—ca n onl y lay unfertilized eggs that hatch males.

Selection Through Queens Only, or Selection Through Workers Too? If workers were completely sterile, then workers' traits could change only through a mutation in the queen that she bequeathed t o her daughters, both workers, who would exhibi t th e change d trait , and new queens, wh o would pas s the trait on to their daughters. In that case, a hive could evolve in the same way as a multicellular body, wit h th e quee n a s the orga n of reproduction an d th e workers a s the othe r body parts. But in fact ther e are a great many reported case s of female workers of all orders o f the socia l hymenoptera—bees, wasps, and ants—layin g eggs. Sometimes, indeed, a large percentage o f the males come out o f eggs laid by workers.4 The fac t tha t many workers ca n and d o lay male eggs brings forth anothe r possi -

154

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

bility. I s i t possibl e tha t b y laboring fo r th e hive , workers ar e actuall y doing th e best the y can to raise their ow n sons? This woul d ope n u p a completel y differen t avenu e fo r th e evolutio n o f th e hive's social structure . Let's assume that by working for the hive , a worker could increase her own chances to have successful sons. Then any mutation that improves the worker's abilit y to labo r fo r the grou p wil l enabl e he r bot h t o produc e an d raise more son s an d to pass the mutation o n through the m t o the following gen erations of queens, workers , and drone s (males) . The fac t tha t caretaker s ca n manipulat e femal e larva e an d rais e worker s o r queens a t thei r discretio n o n th e on e hand, an d th e workers ' abilit y to la y male eggs o n th e other , mad e possibl e th e creatio n o f asymmetrica l partnerships be tween large , fertilize d females—queens—an d unfertilized , mostl y smaller , fe males—workers. A worker canno t establis h he r ow n colony, since she cannot lay fertilized egg s an d brin g fort h daughter s wh o woul d assis t he r i n guardin g he r colony an d takin g car e o f he r offspring . Bu t sh e ca n sta y i n a colony, assis t th e queen to raise more workers, and try to lay eggs that will hatch males. That colony may b e eithe r he r birt h colon y o r another ; amon g stingles s bee s ther e ar e eve n cases in which queen s join colonies establishe d by workers. In ou r opinion , th e chanc e the worker ha s to reproduc e withi n th e colon y is the cement that permitted th e creation of large, stable partnerships, encompassing thousands of individuals, in which the reproductive success of the individual workers depends on the succes s o f the queen . The fac t tha t the queen is able to raise daughters smalle r an d weake r tha n hersel f make s i t possibl e fo r he r t o exploi t them; i t i s th e inequalit y betwee n quee n an d worker s tha t limit s th e workers ' options an d makes the asymmetrica l partnership s o stable.5

HOW INSEC T COLONIES FORM Colonies o f hymenoptera ar e formed mainly in thre e ways. First, by coalitions of queens o f th e sam e age ; thes e ar e usuall y sister s o r ar e otherwis e related , bu t sometimes they are strangers. Second, by coalitions of a mother and her daughters. Third, b y th e breaku p o f a n existin g colon y int o severa l daughte r colonies . T o illustrate the workings of such partnerships, let' s examine several cases. In man y species o f wasps an d ants , several queens join together to form a new colony. 6 In the beginning , th e partner s nee d eac h other , and ofte n mor e tha n on e quee n lays eggs. But once smalle r worke r female s hatch , th e to p queen ca n affor d t o ge t ri d o f her partner-ri vals. An d indeed , a t tha t poin t th e foundin g queens usually start fighting among themselves,

The Social Insects: Wh y Help th e Queen? 15

5

and th e winne r kill s o r expel s he r partner s o r enslave s them. I n mos t case s the colony ends up with one queen only, who goes on laying eggs, while her daughters and the daughters of her former partners take care of the offspring. These workers, who ar e not fertilized , cannot compet e wit h th e quee n effectivel y sinc e they are smaller, an d sinc e she is the onl y one wh o ca n lay fertilized eggs to creat e more workers. In th e temperat e region s ther e ar e man y specie s o f hymenopter a tha t for m annual colonies i n the spring, which di e out with the onse t of winter. This means that i n th e sprin g there ar e no larg e colonies, an d a single queen ca n establish a new nest on her own. For example, young queens of the wasp Vespa orientalis and its relativ e Vespula germanicd 1 hatc h i n the fal l an d liv e for a while i n thei r birth colony. A t tha t time , the y d o nothin g bu t ea t an d rest . Later , when th e colon y nears the end of its life, each queen goes on a mating flight, then passes the winter in hibernatio n i n a crack in th e soil , ofte n wit h siblings . I n sprin g sh e wakes u p and start s building a nest, laying eggs, and gatherin g food for her offspring . The firs t offsprin g o f such queens ar e females. They cannot mate , since there are n o male s a t tha t tim e o f th e year ; s o the y function a s workers, helpin g th e queen t o rais e additional offsprin g throughou t th e summer . Late in the summer things change . Th e quee n start s layin g unfertilized eggs—mal e eggs—an d th e workers star t giving some o f the queen' s fertilized eggs large quantities o f superior food ; thes e egg s develo p int o th e nex t year' s queens . A t th e sam e time , some of the mos t dominan t worker s manage to la y eggs: each take s over several cells, whic h sh e protects , an d i n whic h sh e lay s unfertilize d (male ) eggs . Thus, some of the male s tha t hatc h i n the colon y ar e offspring o f the workers, not th e queen. The laying workers ar e very aggressive toward each other and toward th e queen. The y ea t egg s laid b y their rival s and b y the quee n an d destro y the cell s of thei r rivals. Why d o thes e worker s sta y in th e colony ? Fo r tw o reasons . First, a worker, being small , has no prospects o n her own against the other colonie s around, each with a large queen an d a number of workers. The workers are small because their mother ha s raise d the m t o b e small , and th e first workers ar e particularly small because their mother has raised the m on her own and had a hard tim e bringing in enoug h foo d fo r he r larvae . Second, a worker canno t la y female egg s to raise daughter workers who would help her protect and take care of her nest. Thus, the only way a worker that hatches in the spring can have viable offspring i s to survive in a n establishe d colon y until th e fall—th e tim e fo r raisin g queens an d males — and then attemp t to lay male eggs. Since she has to sta y in a colony, she may well choose to stay in the colony she is familiar with and hope conditions will allow her to bring up successful sons when the time comes. But she does have the option of moving t o anothe r colony , an d i n fact , man y workers d o mov e fro m colon y t o colony,8 presumabl y seeking a place where the y will hav e a better chanc e to re produce. Discussions about colonies usually center on conflicts of interest between work-

156

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

ers an d queen , but i n fac t eac h worke r ha s her ow n individua l interests ; a large part o f th e lif e o f a colon y make s n o sens e unles s conflict s o f interes t amon g individual workers ar e taken int o account. 9 There i s an order of rank amon g the workers. If the top worker were to kill the queen an d take her place, that worker might b e abl e to lay her ow n unfertilized male eggs freely, bu t th e othe r worker s would b e represse d b y that upstar t worker . An d thei r colon y would b e weaker; only th e fertilize d quee n ca n strengthe n th e colon y b y layin g female egg s an d creating more workers. Each worker' s chanc e t o hav e sons, an d thos e sons ' chances , depend o n th e well-being o f the colony , an d th e colon y ca n be strengthene d onl y by the queen . So it is better fo r most workers to hav e a queen a t the hea d of the colon y than to have another worker in that position. Each worker's chanc e to reproduce successfully conflict s mor e with that of the other workers than with that of the queen. T o get th e bes t sho t a t havin g successful son s later, then , younge r worker s hav e t o cooperate with the queen and suppress egg-laying by the highest-ranking workers. No wonde r tha t al l summer long the colon y grows by raising more workers: the workers themselve s preven t th e raisin g of princesses (youn g queens ) an d males , and they suppress an d kill workers who tr y to lay eggs. In th e fall , th e cas e is different: th e colon y is nearing the en d o f its life . Th e time is ripe to rais e new queen s an d males . The quee n begin s layin g unfertilized male egg s a s well a s fertilized female eggs, and th e worker s star t givin g some of the queen' s fertilize d egg s the treatment tha t will make them into futur e queens . At thi s seaso n a struggle often break s ou t ove r the layin g of male eggs, both between workers and queen and among the workers. The winners among the workers manage to lay some of the colony's male eggs—sometimes even most of them. The workers who d o not manag e to lay male eggs help the quee n an d their egg-laying comrades raise their young. A simila r logi c applie s whe n existin g colonie s brea k u p int o ne w ones . T o withstand neighborin g colonies, a colony o f wasps has t o avoi d breakups, whic h would sprea d it s strength to o thin . In tha t situation , it is in the interes t o f workers t o suppres s th e raisin g o f princesse s an d th e layin g of mal e egg s b y othe r workers—since, when grown , these would swar m and spli t th e colony . But onc e the colon y is large enough t o becom e tw o viable ones, a split is in the interes t of both quee n an d workers : th e quee n ensure s tha t a t leas t on e o f her daughter s will have a colony o f her own , an d a t the sam e time, each worker's reproductiv e chances increase , because sh e ha s hal f a s man y othe r indi viduals to compet e with . Why can' t a worker strik e ou t o n he r own , buil d a few cells, store food, and wait for the appropriate tim e to lay male eggs? Becaus e hyme n op tera gathe r foo d fo r thei r offspring , such smal l nest s would b e destroye d b y members o f neighboring colonies , wh o would stea l anythin g stored in them to feed thei r ow n larvae. This dange r create s the nee d t o coop -

The Social Insects: Wh y Help th e Queen? 15

7

erate in guarding and the opportunity for the strong to exploit th e weak. Robbery and predation preven t th e workers from reproducin g outside large colonies ruled by queens .

WHY D O TH E WORKERS WORK FOR THE COLONY? We hav e seen why workers sta y in a colony waiting for a n opportunity to repro duce, but tha t doe s no t explai n why they actually work for the colony . There are many workers, and most of them ar e willing to work for the group; why not leave the labor to others?10 The colony would not be harmed appreciably, and the "lazy" worker would save her strength for the final struggle and the egg laying. One would think that such social parasites would b e in a better positio n i n that final struggle and would have more sons than their competitors, an d thus such parasitism would spread i n th e population ; ye t we d o no t se e that happening . W e believ e tha t a worker or a lower-ranking queen enhances her own chances to reproduce by working for the colony. 11

Altruism and Prestige We have found a similar relationship in babblers, of course. In babbler groups , a babbler invests in altruism, we believe, because it reliably demonstrates its ability as a riva l an d it s valu e a s a partner , i n way s that othe r member s o f th e grou p recognize. The gain to the "altruist" is not th e benefit that the group derives, but the recognitio n of other individuals—the prestige he or she wins. As w e sa w in th e previou s chapter , a n individual' s reproductiv e chance s are determined i n large part b y prestige. Th e mor e a babbler doe s for its group, the greater it s prestige , becaus e it s abilit y t o d o fo r th e grou p prove s it s abilit y in general. And high prestige—the recognition o f its ability by others—spares it the need to fight or even threaten its groupmates when the time comes to compete for the right to reproduce . The "altruistic " individual thus serve s its own interests—it expend s effor t t o demonstrate its quality reliably. Because it does so by engaging in altruism, rather than b y pure showin g off, as a peacock doe s with it s tail and a bower bir d wit h its bower, the babbler is also showing its interest in continued collaboratio n with the members of its group. Groupmate s pa y attention t o the dual message because this information enables the m to make their own decisions. At the same time, the whole group benefits fro m th e action s that conve y the altruist's message. In both a group of babblers and a colony of social insects, one has to cooperate to survive, and at the sam e time one competes with others in the group or colony for th e righ t to reproduce . I n suc h situations, anything that allows individuals to

158

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

demonstrate and asses s their own and eac h other's abilit y without a physical clash or direc t threats , and anything that will help them evaluat e a partner's interest in continued collaboration , wil l increase their chances of success. Thus ther e is a ready audience fo r messages that reliably combine evidenc e of an individual's ability—as show n by its doing deed s that demand ability—an d of its interest i n continued collaboration—a s demonstrated b y its willingness to devote thes e show-of f effort s t o it s collaborators . B y helping, th e signale r demonstrates its ability, which is recognized by others; this recognition translates into the rank an d prestig e tha t enabl e i t t o reproduc e mor e successfull y tha n individuals who receiv e less recognition. This ca n become a powerful selective force fo r be havior that helps the group; the evolution of what looks at first glance like straightforward altruisti c behavior is in fact drive n purely by the mechanism of individual signal selection .

Queen Pheromones and Prestige A grou p o f babbler s i s smal l enoug h tha t th e partner s ca n observ e eac h othe r regularly, an d th e deed s o f eac h fo r th e benefi t o f th e grou p ar e know n t o all . Things ca n b e th e same in smal l colonies o f social hymenoptera: 12 al l individual s know eac h other an d ar e aware of the wor k don e by each , and thu s the relativ e prestige o f each . Bu t i n a colon y o f severa l thousand s o r ten s o f thousand s o f individuals, it is highly unlikely that an y individual coul d observ e and remembe r the deed s o f each of the others. If indeed workers do serve a large colony in order to wi n prestige an d acquir e rank, there mus t be a mechanism that demonstrate s each individual' s abilit y reliably. The mechanis m that seem s likeliest t o us is pheromones.13 W e sugges t tha t in a large colony, a worker's ability—her worth—is measured by her capacity to carry queen pheromone . W e thin k tha t tha t capacit y reflects tw o factors: the amoun t the worker's bodily condition allows her to carry, since the pheromone is a harmful chemical; and the amoun t she manages to acquire, either by force or in return for her service s t o others . Th e situatio n wil l b e mor e comple x an d interestin g i f it turns out, a s we think possible, that by working, not onl y does a worker acquir e queen pheromone fro m the queen or from other workers in return for her services, but sh e actuall y increases her capacit y to carr y queen pheromon e a s well, sinc e working increases her metabolism. It i s known tha t th e pheromon e secrete d b y the quee n motivate s workers t o serve her. Bee s ar e very eage r fo r th e pheromon e an d lic k i t of f the bod y o f th e queen. Worker s ca n ge t th e pheromon e directl y fro m th e bod y o f th e quee n when the y serv e her , o r indirectly , fro m worker s wh o hav e serve d th e queen . They may also get the pheromon e from th e larvae they serve, who, i t seems, produce simila r pheromones. 14 Still , ther e seem s t o b e som e facto r tha t limit s th e amount o f pheromone eac h worke r collects . Whe n th e pheromon e i s dispense d

The Social Insects: Wh y Help th e Queen? 15

9

by researchers , al l individuals ar e attracte d t o it , bu t onl y som e o f the m ea t i t freely, whil e other s ea t non e o r ver y little.15 In som e respects , one ca n liken th e quee n phero mone to money in human societies. In a large society, individuals d o no t kno w eac h othe r personally , bu t they ca n tel l eac h other' s economi c statu s b y th e amount of money each has. Only persons of means can hold o n t o larg e amount s o f money . On e carryin g money als o ha s t o protec t i t fro m thieve s an d rob bers—in othe r words , mone y testifies no t onl y to it s holder's abilit y to acquir e it, either through work o r by force, bu t als o to hi s or her abilit y to protect it.

The Handicap in Queen Pheromone It is known that there is a correlation between a queen's ability to lay eggs and her ability t o produc e pheromone. 16 I t i s also know n tha t worker s kil l queen s wh o secrete onl y a small amount of pheromone. 17 I f queens coul d separat e egg-laying from th e secretio n o f pheromone an d concentrate when old or in ill-health on the latter t o avoi d being killed , surel y they would d o so . Apparently, they cannot— and thus the amoun t of pheromone the y secrete is a reliable indicator of the number o f eggs they lay. According to the Handicap Principle, queen pheromone can be a reliable indicator o f eg g laying only if overproduction o f it—deception—woul d caus e th e queen mor e har m tha n good . W e believe , i n othe r words , tha t th e pheromon e comprises intrinsically harmful chemicals . Just a s this pheromone advertise s reliably the qualit y of a queen, smaller quantities of the same harmful pheromone can serve a s reliabl e indicator s o f th e abilitie s o f worker s a s well. Thu s a worker' s ability to handle pheromone sh e receives from th e quee n would be recognized by others and would translate into rank and prestige, which would enable workers to settle their conflict s ove r reproduction wit h fewer physical struggles. It i s well known tha t queen pheromon e inhibits the developmen t o f workers' ovaries. I f th e pheromon e wer e jus t a conventiona l marker , with n o significan t inherent properties , on e would expec t t o find an inverse correlation between th e quantity of queen pheromone a bee carrie s and the stat e of her ovaries . In fact , it turns ou t tha t th e mor e highly developed th e ovarie s of a worker are , the more queen pheromone she carries. Rather than serving as a conventional marker, then, the quantit y of queen pheromon e carrie d b y an individual seem s to indicat e her ability t o withstan d it s harmfu l effects . W e believ e tha t th e capacit y to tolerat e queen pheromone, then , varies in individuals according to their physical condition, which i s just th e wa y the Handica p Principl e woul d hav e predicted a chemical signal would work.

1 6O TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

Roseler an d Hon k sho w that in bumblebee nests , the ovarie s of the stronges t workers, who are nearest to the queen, are well developed eve n though their proximity to the queen exposes them to high levels of queen pheromone. Th e researchers remar k tha t thes e worker s d o no t hol d thei r hig h ran k because o f thei r well-developed ovaries ; the worker s maintai n their ran k eve n if their ovarie s are removed surgically . We thin k tha t thes e worker s ar e physically superior t o thei r colleagues, and that this superiority gives them both their high rank and the ability to develo p thei r ovarie s in spit e o f the hig h concentratio n o f quee n pheromon e they are exposed to. 18 In honeybees, the queen pheromone contain s unsaturated fatty acids with keto groups. I t i s known tha t fatt y acids , especiall y unsaturate d fatt y acid s with ket o and alcoho l groups, are active, harmful chemicals . They are similar in structure to the femal e mot h pheromone s w e hav e alread y discussed. I t ma y be tha t quee n pheromone i s absorbed b y th e bee' s expose d sensor y cells an d cause s a state of intoxication, a s we suggested for moths i n chapte r 9 . If we assume that a worker who carrie s quee n pheromon e i s harme d b y i t i n thi s o r other ways , the n th e amount a worker ca n carry is limited not onl y by her abilit y to acquir e it by forc e or in exchange for services, and her abilit y to protect it , but als o by her abilit y to withstand tha t harm. Showing off by consuming harmful chemical s like alcohol, tobacco, betel nut, opium, and the like is common among humans. In some societies, men even demonstrate their vigor by drinking naphtha. In his book The Rise and Fall of the Third Chimpanzee, Diamon d uses the Handicap Principle to explain this phenomenon.19 Veblen,20 the sociologist who coined th e phrase "conspicuous consumption," lik ened the men who hang out in bars and pubs, drinking and buying drinks for each other, to American millionaires wh o sho w off by funding colleges , hospitals, and museums. On e wh o ca n drin k quantitie s o f alcoho l withou t apparen t ill-effec t shows reliabl y his or he r goo d physica l condition ; on e i n poor conditio n woul d get visibly drunk. Laborers who do hard physical work are notoriously heavy drinkers in many societies. In sum , we think tha t a worker servin g the colon y collect s quee n pheromon e in exchange for her services. Since the capacity of a bee to carry queen pheromon e depends o n her being in good physical shape, the pheromone is a reliable indicator both of her ability to acquire pheromone an d of her current bodily condition. He r tolerance fo r pheromone i s evident t o he r comrades , and a s threats do, i t allow s her t o exhibi t he r abilit y and achiev e her end s wit h les s actua l violence. In fact , possessing the pheromone generall y makes even overt threats unnecessary. Moreover, we wonder whether the very act of working for the hive increases the worker's ability t o carr y pheromone, jus t as , afte r a yea r o f constructio n work , a colleg e student may "hold his liquor" better than he did in his fraternity-party days. The suggestio n that quee n pheromon e i s the handicapping signa l that proves rank an d prestig e i n a large hive or colony , and tha t a worker's capacit y to carry queen pheromon e reflect s he r qualit y both a s a partne r an d a s a rival , is stil l a

The Social Insects: Wh y Help th e Queen? 16

1

hypothesis. I t fit s th e fact s tha t w e know, bu t th e pictur e i s far fro m complete . Today instruments exist that can measure the quantity of queen pheromone carried by a n individual insect . It shoul d therefor e be possible t o carry out researc h tha t tests our hypothesis .

KIN SELECTIO N THEORY AN D IT S DRAWBACK S Most evolutionary biologists wh o discuss the social insects take it for granted that workers in a hive or a nest would have a better chance to reproduce if they didn't squander thei r effort s helpin g others . I n Englis h an d other languages, altruism is commonly defined simply as devotion to the welfare of others, or regard for others as a principle of action, but evolutionar y biologists defin e it as "a feature that helps another a t a cost t o the helper." Sinc e the onl y evolutionary definition of success is successful reproduction, "cost " is defined as a decrease in reproduction. I n other words, the evolutionar y theories tha t have been develope d s o far to explain altruism assume that altruistic behavior decrease s reproduction b y its very definition. 21 Now, ho w ca n a feature that reduce s th e reproductio n o f its bearers evolve , when evolutio n b y it s basic natur e is the selectio n an d promotio n o f traits tha t enable individual s t o brin g fort h more descendants ? N o explanatio n o f the evolution o f altruism makes sense unless i t show s that altruisti c activit y will actually improve the reproductive chance s of its carriers. 22 In other words, any theory that attempts t o explai n th e evolution o f altruism mus t identif y some mechanism tha t compensates the altruis t for its supposed los s o f reproductive "fitness. " At first glance, group selectio n seem s to explai n th e altruis m o f social insect s easily: by that theory' s logic, if colonies whos e worker s assis t the quee n compet e successfully with colonies whose workers do not, then the latter will become extinct and only colonies in which workers do in fact work will survive. There is no doubt that th e former are more successful; but the question i s whether tha t is enough t o support th e evolutio n of such "altruism. " Logically, in a society in which some individuals ar e inclined t o assist, a selfish individual woul d benefi t by bein g selfish—i t woul d enjo y th e prosperit y o f th e colony that arise s from th e work of others while conserving its own energy for its own reproduction. Eventually , more and more of the group's members will be the descendants of such selfish individuals, and presumably they will be selfish as well, even thoug h thi s woul d i n the en d caus e a breakup o f the colony . Sinc e lik e al l mutations, those for selfishness—tha t is , the tendency not t o help—appear regularly, on e woul d expec t t o fin d group s i n differen t stage s o f thi s process , wit h varying percentages o f selfis h individuals . But observatio n show s that thi s is no t the case. When explanation s based o n group selection wen t out of fashion, the questio n came up again : Ho w coul d creatures evolve who impair thei r own reproductio n

162

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

by assistin g other s t o reproduce ? Thi s questio n recur s i n ever y stud y o f socia l animals, an d mos t dramaticall y in studie s of the socia l insects . The theor y of kin selection wa s offered b y Hamilton a s a solution t o this problem. 23 Central to the theory of kin selection i s the fact that evolution ca n be measured by variations i n the prevalence o f particular genes , which carr y traits, in a population. Dawkins in fact went further, seeing evolution as competition between genes rather than between individuals; this is the key assumption in his well-known book The Selfish Gene. 24 Kin-selection theory states that one can increase the prevalenc e of one's own genes in future generation s not onl y by one's own reproduction bu t also b y assistin g one' s relative s t o reproduce , sinc e eac h specifi c gen e tha t on e carries is likelier t o b e carried als o by one's relative s tha n b y others. Eve n if , by assisting, on e harms one's ow n reproductive chances , according t o th e theorists, as long a s one improves the chance s that th e genes one carrie s will appea r i n th e next generation, on e is successful . According t o kin selection theory , an individual that invests in helping it s sibling t o rais e two additiona l childre n woul d bequeat h t o th e nex t generation , on average, the sam e number o f its own genes a s it would contribut e b y raising one child o f it s own. 25 B y this logic , if a n individual' s hel p enable s a sibling t o rais e more tha n doubl e th e offsprin g th e individua l ca n hav e o n it s own , then tha t individual i s better of f aiding that siblin g tha n devotin g it s effort s t o it s ow n re production. This kin d o f calculation , which measure s a trait's effec t no t onl y on its carrier' s reproductio n bu t als o on th e reproductio n o f its carrier' s relatives , is called "inclusiv e fitness. " Scientist s distinguis h thi s from simpl e "fitness, " whic h takes into account one's descendants only.26 Hamilton suggeste d kin selection a s an explanation for the evolution of altruism in animals in general, and he suggested tha t hymenoptera have some special trai t that supports th e evolution of social systems by kin selection. Afte r all , true social (eusocial) systems , in which most or al l workers ar e almost sterile , evolve d in this group som e doze n time s independentl y (severa l separat e time s i n bees , severa l separate time s in wasps, and i n ants); in al l other insect group s the y evolved only twice, in termites 27 an d in some scale insects. Hamilton suggeste d tha t th e haplodiploi d mechanism of gender determinatio n i n hymenoptera plays an important role in the evolutio n of their social systems. Male hymenoptera have only on e se t o f chromosomes , inherite d fro m the mother , whil e female s hav e tw o set s o f chromosomes, on e o f whic h comprise s hal f their mother' s chromosome s an d th e other all of thei r father's . Th e geneti c relatednes s o f a female bee, wasp, or ant to her siste r is 75 percent, since they share half their mother's genes and al l of thei r father's . Accordin g t o Hamil -

The Social Insects: Wh y Help th e Queen? 16

3

ton, i t wa s thi s geneti c relatednes s tha t facilitate d th e evolutio n o f altruis m i n hymenoptera: the workers who take care of the young are raising sisters who share 75 percent o f their genes , while thei r ow n daughter s would hav e shared onl y 50 percent o f their genes . The theory of kin selection too k the scientific world by storm.28 Researchers of social behavio r i n animal s starte d measurin g th e degre e o f relatednes s betwee n individuals. Bu t soo n i t was discovered tha t altruisti c behavior exist s even wher e there can be no genetic relatedness betwee n th e altruist and the one being helped , such as in those species of ants, wasps, and birds in which altruistic social systems link unrelate d individuals . What i s more, i t is known toda y that ofte n mor e tha n one mal e fertilizes a queen,29 making for a far lower degre e o f genetic relatednes s among workers an d queen s i n a hive: siste r bees wh o shar e a mother bu t no t a father hav e on averag e a genetic relatedness of only 25 percent . It was also found that no t infrequently, worker bee s and ants leave one colony and mov e to another , where the y work a s usual, even thoug h th e quee n an d th e other worker s ar e not thei r relatives at all. 30 All this made it necessary to develo p another explanation for altruism among unrelated individuals—Trivers' s theory of reciprocal altruism , which we discusse d i n chapte r 12 . Still, nobod y dare d chal lenge th e theor y of kin selectio n a s the basi s o f the evolutio n o f altruism . It was hard t o giv e u p th e magi c o f a theor y tha t seeme d t o enabl e on e t o calculat e mathematically the desirabilit y of altruism.

Parasitism Among Kin, or Haldane's Other Brothers We believ e tha t th e basi c premises o f th e theor y o f ki n selectio n ar e a s weak as those of group selection . The idea of kin selectio n was first proposed b y Haldan e in 193 2 i n hi s boo k The Causes o f Evolution? 1 I f a n altruisti c trai t benefit s a n individual's descendants an d relatives, Haldane wrote , then that trait increases the individual's fitnes s (define d a s th e individual' s productio n o f successfu l descendants); the trai t thus ca n spread i n the populatio n b y natural selection. Haldan e offered a s an example two brothers walkin g b y a river: one o f them fall s int o th e water an d start s drowning ; th e othe r ca n jum p i n an d sav e him . Accordin g t o Haldane, i t woul d mak e sense , evolutionaril y speaking , for th e brothe r standin g on the shore to jump in and save his brother, even though he thus lessens somewhat his own chances to survive and reproduce . The trait to sacrific e onesel f for others thus cam e abou t throug h ki n selection , accordin g t o thi s theory . (Interestingl y enough, Haldan e himsel f sai d that he had twic e in his life jumpe d into a river t o save another person , an d that i n neither cas e did he stop t o reflec t tha t it did not make sense fo r him to d o so , since the on e being save d was highly unlikely to b e his relative.) Haldane's example , however, ha s the sam e weakness that the theory of group selection does : i t is vulnerable t o social parasitism. The basi c fallac y i n Haldane' s

164

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

argument can be illustrated by a variation on his story about the drowning brother. Let's assum e that instead o f two brothers walkin g by the riverside , we have three or four . On e o f them fall s int o th e river , and anothe r jump s in to sav e him. Th e others stan d by , doing nothing . Al l brothers wil l receive the sam e genetic gai n if the rescu e is successful; bu t th e rescue r risk s injury o r even death , while th e one s standing by risk nothing—tha t is, the slackers ' tota l gai n wil l be greater . The altruist, then, will benefit less than a selfish brother , an d o n average , the numbe r of altruisti c genes in the nex t generation will decrease rather than increase . Motro an d Eshel formulated a mathematical model t o solve this dilemma, but they di d s o by adoptin g a n unrealistic premise : tha t n o brothe r ca n assume that another brothe r wil l jum p i n th e wate r i f h e himsel f refrain s fro m doin g so. 32 According t o th e theor y o f kin selection , w e would predic t tha t eac h woul d en courage his brothers to jump in rather than jumping in himself. In reality, we know it is much likelier that all the brothers wil l jump in to sav e their drownin g sibling , and tha t eac h wil l d o hi s utmost t o tak e on th e ris k an d perfor m th e rescue ; n o brother wil l consider that becaus e the other s ar e struggling to d o the same , he is risking his genes needlessly . Thus, grou p selectio n theor y and kin selection theor y have the same flaw: they invite social parasitism. In fact , ki n selection i s simply group selectio n amon g relatives.33 W e fin d i t strang e that despit e this , man y researchers who rejec t grou p selection see kin selection a s a viable, stable mechanism, like individual selection. 34 Basing an explanation o f altruism on kin selection theory presents tw o further problems. Research on group-living birds shows that one or two helpers can indeed increase the success of a reproducing couple , but tha t more helpers than this have no effec t o n th e succes s of reproduction. 35 By the logi c o f kin selection , t o invest in providing unneede d hel p i s a pure loss both to the helpers an d to all their kin . By that sam e logic, i n fact , al l the helpers ' kin woul d hav e benefited i f these un necessary helper s ha d save d thei r effort s fo r thei r ow n reproductio n o r trie d t o form ne w groups. Ye t the reproducin g coupl e in group-breeding birds often get s assistance from mor e helpers tha n they need. And i f the drivin g force behind th e evolutio n o f altruism is the benefi t gained by kin, shouldn' t th e sam e logic militat e strongly against aggression toward kin ? Yet we often se e violent struggle s between ki n collaborators . Indeed , these struggles frequently end with on e or more of the collaborator s bein g wounded o r even killed. Th e theor y of kin selectio n doe s no t explai n why it is that relative s d o no t avoid harming each other in their struggles . Gadagkar an d Josh i describ e a colon y o f wasps tha t spli t int o tw o colonies ; before the breakup, the level of aggression in the colony was so high that it caused a sharp decrease in the number o f offspring. Onl y after th e breakup di d the agression subside, an d the number o f offspring the n increased a great deal in both new colonies.36 As Darwin remarked, it is precisely the social insects that are noted for excessive enmity to their closest relatives: mother kills daughter, sister kills sister.37

The Social Insects: Wh y Help th e Queen? 16

5

In thei r monumenta l boo k o n ants , Holldobler an d Wilson 38 remar k that lif e i n an an t colon y is a constant struggle for supremacy.

Are Offspring "Kin"? The debat e between supporter s o f kin selectio n theor y and it s critics, like many scientific debates , i s not fre e o f emotio n an d i s very hard t o conduc t wit h pur e logic. In the heat of argument, Dawkins39 alleged that critics of kin selection refus e to "count " the kinshi p of offspring an d parent s in kin-selection calculations ; th e inclusion of direct offspring , Dawkin s insisted, lends plausibility to the theory. In fact, everyone grants the obvious kinship of parents and offspring; but this is beside the point. The debat e ove r ki n selectio n i s a debat e ove r th e differenc e betwee n tw o measures o f an individual' s success : straightforward Darwinian natural selection, which measure s th e success , o r "fitness, " o f a n individua l b y ho w man y of it s offspring successfull y reproduce ; and kin selection, which measures "inclusive fitness" b y adding t o thi s th e individual' s effec t o n th e reproductiv e succes s of its relatives. Ther e is no poin t i n arguing over what th e tw o theories agre e about— that is , the individual' s own offspring . Th e critica l issue lies i n th e difference be tween the theories. The question is whether the reproduction of those among one's relatives who ar e no t one' s direc t descendant s ha s an y effect o n th e evolutio n of traits.

PARTNERSHIPS AMONG KIN: WHY I T MAKES SENSE T O JOIN THE FAMILY BUSINESS As we have said, we believe that helpers generally stay in their home group, nest, or colony because that is where their own reproductive chances , slim as they may be, ar e greatest—an d becaus e th e mor e the y prov e themselve s b y helpin g th e group, the better those chances become. Why do animals collaborate in particular with thei r ow n relatives ? Becaus e they know them : th e ran k orde r amon g them has alread y been established . If for various reasons one's best chanc e is in a partnership, i t ofte n make s sense to ban d togethe r wit h individual s on e knows well, whose ran k relative to one' s own is already established. West describe s th e formatio n o f a colony o f the was p Polistes canadensis by a group o f seven queen-sisters. 40 In ou r opinion , he r descriptio n show s what each partner gets out of the initial partnership, and what each gains by banding together with her sister s rather than with strangers. Since they were sisters, the social rank-

166 TH

E HANDICA P P R I N C I P L E I N SOCIA L SYSTEM S

ing amon g the m ha d alread y been establishe d th e previou s fall , whe n the y were living togethe r i n thei r mother' s nest. 41 In th e ne w nest , a division o f labor ap peared: the top quee n staye d in the nest, guarding and taking care of it. The three sisters next to her in rank went out gathering food. The three lowest-ranking sisters stayed b y th e nes t withou t takin g par t i n an y activities . Whe n th e to p femal e disappeared, th e secon d too k it s place an d on e o f the inactiv e sister s joined th e food gatherers . This arrangemen t enable s eac h siste r t o fin d ou t wher e he r bes t chanc e t o succeed lies. The secon d i n rank alread y knows from experienc e tha t sh e canno t overcome her senio r sister . By going out to collec t food , sh e can meet wasps fro m other nest s who ar e als o seekin g food. I f sh e finds herself stronge r tha n mos t of them, there i s a good chanc e sh e coul d protec t a nest effectively . I n tha t case , it will make sense for her to strik e out on her ow n and establish her own nest . On th e othe r hand , i f the secon d siste r find s hersel f weake r than othe r foo d seekers—each o f whom ha s a n eve n stronge r siste r guardin g th e nes t i t come s from—it will be better for her to go on being a helper an d bring food to her home nest. In tha t case , her best chanc e will be the possibility tha t her senior siste r will disappear an d that sh e will take her place in an established nest , with helpers and possibly workers . A s fo r th e weaker , inactiv e sisters—a s lon g a s the y se e thei r stronger sister s coming back to the nest rather than striking out on their own, there is little chance that they, being even weaker, would manage alone; there is no poin t in trying their luc k elsewhere . West's seve n sisters had known eac h other from th e time they hatched an d had established a n orde r o f ran k amon g themselve s befor e the y formed th e nes t to gether; the y didn' t nee d t o spen d tim e an d effor t an d ris k injur y t o establis h a hierarchy a t the beginnin g o f the season , a critical time. We believe , i n fact, tha t the familiarit y an d establishe d ran k amon g relative s i s th e mai n reaso n fo r th e prevalence o f coalitions among relatives in the anima l kingdom.

THE KI N EFFECT There is no doubt that among social animals, other thing s being equal, an individual wh o belongs t o a group whos e member s collaborat e reproduce s bette r tha n one who belongs t o a group whose members fight each other. This is true even of collaborations between differen t species , as in symbiosis. I t i s also clear that if for some reason it makes sense to collaborate with relatives, then in future generations a scientifi c tall y may sometime s sho w a n increas e i n th e representatio n o f thos e relatives' genes within the general population. We suggest calling this phenomeno n "kin effect." 42 Kin effec t i s different fro m ki n selection . I n ki n selectio n theory , the selective factor tha t drives the evolution o f altruistic traits is the benefit the altruist's relatives

The Social Insects: Why Help the Queen?

167

gain fro m th e investmen t tha t individual makes in them. By contrast, we say that altruistic traits evolve by natural selection; that is, they evolve because i n fact the y improve a n individual' s reproductiv e chances . Investment i n th e grou p benefits the investo r directly ; ofte n th e gai n t o th e altruis t i s increase d prestige , whic h facilitates his or her own reproduction. Th e kin effect —the impac t these altruisti c traits may have on the reproductio n of the individual's kin—canno t be the facto r that selects the tendenc y to invest in altruistic acts .

This page intentionally left blank

C H A P T E R 1

4

THE PARENTIN G COUPL E

F

or animal s who reproduc e sexuall y to hav e offspring, tw o individuals must collaborate. I n chapte r 3 , we discussed th e conflic t o f interests betwee n th e genders durin g courtship , when eac h ha s to choos e a partner t o mat e with and must convince that partner to mate with it. But it is not just the selection of a mate that create s conflict between th e genders. 1 The pair must the n cooperate — mate—in order to bring forth offspring. Som e of their roles are innately different : only female s ca n supply eggs , an d amon g mammal s the y ar e the onl y one s who can bear and suckle the young; only males can fertilize. But either or both can take care o f the offspring, and that ver y possibility create s mor e potential conflicts . Both mates gain equally from reproducing , but the y need no t invest equally in the offspring . Ofte n one adult can take care of the young on its own; indeed, each mate stand s t o benefi t i f it ca n exploi t th e other' s interes t i n th e offsprin g an d leave that mat e to d o the job alone. An individual ca n increase the number o f its successful offsprin g b y choosin g th e bes t availabl e mate , by having many sexual partners, or by combining both strategies, which complicates the picture. 2 In many species, one of the pair leaves th e offspring in the car e of the other. 3 Usually, the male leaves the young to the car e of the female 4 and goes off in search of anothe r female—i n other words , h e practice s polygamy . Sometimes, thoug h 169

1 7 O TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

more rarely , the femal e i s the on e wh o goe s of f to fin d anothe r mal e an d leaves the car e of the offsprin g to the father—o r she may mate with severa l males who all assist he r i n th e car e of their commo n offspring . I n suc h cases , she practice s polyandry. Even whe n raisin g offspring take s both partners, their investment need not be the same . True, i f one stint s and thei r offsprin g suffer , bot h partner s lose; but if , by deserting, on e o f the tw o mates can form a new or a n additional matin g partnership, i t may well gain more than it loses . But a deserter doe s not necessarily gain. In general it has been foun d that one is more likel y to successfull y raise a brood wit h a familiar mate , wit h whom on e has succeeded i n raising offspring before , than by starting over with a new mate.5 Becaused o f the pair' s familiarit y wit h eac h other , the y are probabl y spare d th e effort o f checking each other ou t and coordinating their actions . Also, research on songbirds ha s show n tha t youn g hatched earlie r i n th e seaso n ar e usuall y mor e successful.6 Thu s an individual who deserts its mate and shortchanges their young may wel l stan d t o los e considerabl y more , i n term s o f it s overal l reproductiv e success, than i t ca n gain in another partnership ; moreover , ther e is no guarantee that the deserte r wil l find another partner. 7

PATERNITY AND MATE-GUARDIN G The femal e usuall y selects he r offspring' s fathers . Now tha t paternity ca n be de termined by genetics, scientists are discovering that many nestlings are not fathered by the males who take care of them. These findings explain the effor t male s make to guar d thei r femal e partners durin g thei r fertil e period. 8 In chapte r 12 , we described ho w th e dominan t mal e babble r begin s followin g th e breedin g femal e closely a few days before sh e lays eggs ; he continue s unti l th e da y the las t eg g is fertilized. I n th e morning, whe n the female flies to the nest an d lays her egg , the dominant mal e immediately follows her; he sits on the edg e of the nes t o r near it and tries to copulate with her afterward—presumably the best time to fertilize the egg that will be laid the next day. Needless t o say, during that time the dominan t male tries t o preven t an y other male, an d i n particular hi s secon d i n rank, fro m approaching th e female. 9 In a thorough stud y of house sparrows, Moller 10 found that during egg-laying, a female who is not escorte d b y a male is likely to be attacke d by groups of males who tr y to rap e her— a phenomeno n tha t i s known i n duck s a s well. Th e "hus band" tries to guard his mate and chase away any male who approaches the female or th e nest . But , as genetic tests show , eve n clos e supervisio n canno t ensur e paternity. Molle r als o report s tha t man y female sparrow s copulat e willingl y wit h males who are not their mates. The chosen paramour is usually a high-status male with a large black bib , a "status badge " that, as we saw earlier, is a reliable indi-

The Parenting Couple 17

1

cation o f its bearer' s quality . Morton 11 found tha t olde r pair s o f purple martin s often le t younger pairs settle near them; not infrequently the younger females were seen copulatin g wit h thei r olde r mal e neighbors. Zilberman 12 observe d sunbird s in Israel copulating with males, usually their neighbors, who were not their mates. Extrapair copulation , a s it is termed, turn s out t o be quit e commo n in nature. Why doe s a female copulate wit h male s other tha n he r mate ? Man y reasons come to mind. There is the ris k that males from neighborin g territories o r solitary males will destroy her nest and kill her offspring , forcin g her to lay another round of eggs in the expectation that they will be the ones to fertilize them.13 To prevent that fro m happening , th e femal e can tr y t o gai n th e protectio n o f stron g males living near her, who migh t be willing to shiel d he r offsprin g i n exchange for th e chance t o fathe r som e o f them . Thi s phenomeno n o f killin g offsprin g was first observed with langur monkeys in India.14 In Israel, Giora Han i and Yotam Timna found tha t femal e leopards mat e with severa l males, and that males kill the young of female s who hav e not copulate d with them. 15 It i s also possible that females mate with additional males in case their chosen mates are not fertile : by copulating wit h anothe r male, a female raise s her chances of layin g fertile eggs. She might als o improv e her offsprin g b y copulatin g with a high-quality male who might not agree to desert his mate for her but is quite willing to father a few additional offspring o n the side. In short , a male who pair s up with a female does not necessaril y father all her offspring. A male who wishes to be sur e that th e offsprin g hi s mate bears ar e his own ha s t o persuad e th e femal e of his qualit y and o f his willingness to inves t in her an d i n their commo n offspring , s o that sh e will not tr y or agre e t o copulat e with other males.

TAKING CARE OF THE YOUNG T O GAI N PRESTIGE The nee d t o convince one's partne r o f one's commitment an d qualit y is so strong that often , rathe r tha n exploit eac h other, the members of the breeding pair compete to provide for their offspring , o r even take care of offspring tha t are not their own. Such competition is evident amon g herons, egrets, gulls, plovers, and others. For example , when one of the pair is incubating the eggs, and the other comes to replace it , the forme r may not giv e up it s task willingly; the othe r ha s t o pus h i t of f the nes t in order t o d o its share. According t o Trivers, 16 performin g a tas k that would otherwis e b e don e by another is altruism. Theorie s o f reciproca l altruism , however, a s w e sa w i n chapte r 12 , fai l t o explai n why one of the pair would g o so far as to push

172

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

the other off the nest. It would seem that this competition ove r caregiving is meant to demonstrat e th e qualit y and motivation of the caregive r and to strengthe n the relationship between the parental pair. In fact, when we discuss parasites in chapter 16, we will sugges t tha t by taking car e o f cuckoo chicks , crow s sho w themselve s off a s good mate s in order t o preven t th e breaku p o f the pair . Competition ove r th e car e of offspring , lik e th e competitiv e altruis m o f babblers, show s both a n individual's qualit y and it s commitment t o the partnership . Single males often assis t widows i n raising their offsprin g fo r similar reasons; they thus prov e thei r qualit y an d th e seriousnes s o f thei r inten t an d increas e thei r chances t o pai r wit h th e widow s late r on . The widow s ar e desirable mates: the y have alread y proved themselve s b y havin g offspring. A singl e mal e ma y assis t a pair t o rais e thei r offsprin g i n orde r t o increas e hi s chance s o f mating with th e female shoul d he r partne r die . Sometime s th e female' s mate is kicked ou t b y th e "helper" himself— a phenomeno n tha t ha s bee n observe d amon g monogamou s species as well a s among species that live in groups. 17 An extrem e example o f an anima l proving its quality by caring for the offsprin g o f others is provided b y fish. In th e Japanese fish Alcichthys alticornis, fertilizatio n occur s withi n th e female' s body. Afte r th e female's egg s have been fertilized by one male, she chooses another and lays her eggs in hi s nest. Onl y afte r sh e has filled his nest wit h the first male's offspring doe s the female agree to copulate wit h th e nest' s owner ; sh e then goe s off with the eggs he has fertilized and lays them in the nest o f ye t anothe r male . Most male s o f thi s fish thus tak e car e o f th e offsprin g o f others , a s ha s been prove d b y DNA analysi s of the offspring. 18 Females prefer males in whose nests there are already large egg deposits; thus, the investment of each male in others' offsprin g increase s his attractiveness to females. Indeed, th e females of many fish are attracted to males who are taking care of nests that contai n plent y of eggs. 19

OTHER MEANS OF SHOWING OFF TO ONE' S MATE Showing off to one's mate does not necessarily involve caring for offspring. Blackstarts and wheatears nest in small caves and holes in rocks. They pave the entrance to the nest with smal l stones—often score s and eve n hundreds o f them. Fishman, who studie d the white-crowne d blac k wheatear in the Sina i Desert, observe d females tryin g again and agai n to lif t stone s to o large for the m t o carry , and i n th e end making do with the larges t stones the y could handle. 20

The Parenting Couple 1 7

3

Moreno an d hi s colleague s studie d stone-collectin g b y blac k wheatear s i n Spain.21 In that species, both males and females collec t stones. In the days befor e egg-laying, the y may bring u p t o three or four pound s (1.5- 2 kilograms) of small stones to the nest. The researchers found a clear statistical correlation between the number of stones the pair gathers and the male's participation later in feeding the young, and also between the number of stones and the pair's reproductive success. They checked to see whether the stones themselves were beneficial—whether, for example, they deterre d predator s o r modified the microclimat e aroun d th e nest. They could not find any such effect, however. They concluded that stone-collecting lets a pair of black wheatears show off their quality to each other and thus strengthens the bond betwee n them. Leader, wh o studied blackstarts , foun d that female s collect stone s to pave the entrances o f their nests . He di d no t fin d an y correlation between th e numbe r of stones collecte d an d th e qualit y of th e femal e an d suggeste d tha t i n th e cas e of blackstarts, the stone s may serve a utilitarian purpose—probably as an alarm system tha t warn s o f approachin g predators. 22 I f so , thi s ma y wel l hav e bee n th e original, utilitaria n behavior tha t i n black wheatear s evolve d int o a signal an d a test o f quality.

DOMINANCE BETWEE N MATE S Cooperation betwee n gender s allow s each to evolv e specific traits . A partnership between parties who complement on e another is more effective tha n one between two identica l collaborators . Thi s means , o f course , tha t eac h o f th e tw o ha s t o search fo r a partne r tha t wil l complemen t it . Ofte n th e evolutio n o f on e trai t conflicts with th e demand s of another: for example , fighting demands a muscular body, while egg-laying demands reserves of fat an d protein . A femal e duck, who lay s som e te n egg s i n a clutch , spends mos t o f he r da y during that period straining food out of the water to accumulate th e nutrients she needs fo r he r eggs . The wa y ducks feed , an d th e larg e number o f eggs that sh e lays, preven t her fro m becomin g a good fighter : sh e is unable t o protec t th e ter ritory sh e an d he r offsprin g need , an d whe n sh e puts he r hea d i n th e wate r t o collect food , sh e is exposed t o predators . Th e male , who i s not unde r th e same pressure to gathe r food, ca n afford t o evolv e th e muscula r mas s he needs to b e able to fight neighboring pair s over territory, to chas e off other male s who tr y to copulate with hi s mate, and t o guard her agains t predators. The stronge r o f the tw o partner s in a n anima l pair—usually th e male—ha s a degree o f freedom an d o f control ove r share d resource s that th e weake r partner doesn't. His strengt h mean s that h e has the abilit y to establis h th e best territor y he can , the n choos e th e bes t femal e tha t h e ca n find; the weake r gende r ha s t o either choos e a territory with th e mat e that come s with it, o r choos e a mate and

174 T H

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

settle for whatever territor y tha t mat e ca n hold. The stronger gender can choose a partner, o r kick a former partner ou t o f the territory. If he ca n hold a large and high-quality territory, he may even be able to get an additional female. 23 His greater strength ofte n allow s him to dominat e the female .

CONDITIONS FOR FEMALE DOMINANC E In som e species , however , th e femal e i s stronge r an d large r tha n th e male , an d dominant ove r him. Female dominanc e evolves i n specifi c ecologica l conditions : for example , whe n o n th e on e han d tw o adult s ar e neede d t o tak e car e o f th e offspring, an d o n th e othe r han d th e "bottleneck " in reproduction i s not th e female's abilit y to bea r youn g or la y eggs but occur s a t a later stag e in th e breed ing cycle . The chick s of jacanas and some sandpipers24 eat different foo d tha n adult s o f these specie s do . Th e food adult s ea t i s plentiful, yet even thoug h the females hav e abundan t resources , the y canno t us e them t o produc e mor e eggs . Fo r th e foo d tha t is available t o thei r chicks i s not plentiful , thi s is th e bottleneck restrictin g reproduction . I n a case lik e this, where th e ke y constraint o n reproductio n oc curs afte r th e youn g are hatched, female s are freed fro m devotin g mos t o f their resource s t o eg g production. Thus , female s ca n affor d t o becom e physicall y strong an d ca n becom e bigger an d stronge r tha n males. In bird s o f prey, too, female s ar e ofte n th e dominan t partners . Raptor s wh o catch thei r prey in flight have to be strong an d must b e quick flyers—which also makes them abl e fighters. Neithe r doe s such a raptor female have to devote all her time t o feeding ; on e pigeon o r grous e is enough t o satisf y he r fo r a n entire day. Thus, sh e is not adapte d t o collec t foo d i n ways that interfer e with he r abilit y to defend a territory. Food for th e chicks is a different story. Many raptor chick s die in th e nest , an d th e parent s canno t rais e more tha n on e or tw o successfully—a n indication tha t i n this case as well, the territor y provides mor e than enoug h foo d for the adults, an d that it is the scarcity of food for the chicks that limits the number that ca n be raised . This explain s why a female rapto r ca n affor d t o becom e big and strong ; bu t why should the male not be as large, or even larger and stronger? In many species, after all , ther e i s a grea t dea l o f overla p betwee n th e siz e o f male s an d tha t o f females. I n mos t specie s o f birds , larg e females see k ou t eve n large r males , who are likely to hav e better territories , eve n thoug h the y could choos e male s smaller

The Parenting Couple 1 7

5

than themselves an d be dominan t over them. Among raptors, by contrast, bigger males seek even bigger females, though they could choose smaller females and lord it over them . A male raptor wh o select s a female bigger than himself—who i s presumably a strong and efficient hunte r and has been feeding well—is assuring himself of better offspring. Th e female contributes more to the offsprin g tha n the male does, since she provides no t onl y genes, a s he does , bu t als o the eg g itself, that is , the envi ronment i n which the young develops.23 Under such conditions, i t seems that th e higher qualit y of the femal e compensate s the mal e for sacrificin g dominanc e and control over their common property, the territory. The female is the dominant not only in raptors, but als o in several species of plovers and frigate birds, and in quite a number o f fish. Sometimes, a show of dominance may not b e al l it seems. Female finches lord it ove r thei r mate s before egg-laying , while th e male s ar e dominan t a t al l other times. It may be that the female's aggressiveness toward the male, which looks like dominance, is actually a test of his commitment to her. The mal e has to convince the female of his sincere interest i n forming a partnership with her; if he does not submit to her wishes, he might find himself raising the offsprin g o f others .

THE PARENTAL COUPLE AS A PARTNERSHI P The parentin g couple is a partnership. As in any other partnership , the choic e of a partner and the relationship between th e partners depend bot h on external conditions and on the individuals involved. For example, one may give up dominance in order to gain a better partner and better offspring. And as in any partnership, the two parties have some interests that coincide an d others tha t conflict . Each member of the partnership—of the breeding couple—ca n gain from exploiting the other; an d indeed, sometimes a breeding partne r exploits the other' s interest in their commo n offspring, forcin g the other t o provide th e car e that the offspring need . Ye t it is precisely because of the parenta l pair's conflictin g interests—because each can take actions that will not be in its partner's best interest— that eac h of them als o seek s information on th e other' s abilit y and commitment, and provides suc h information to the other in a reliable manner.

This page intentionally left blank

C H A P T E R 1

5

SOCIAL AMEBA S (CELLULAR SLIM E MOLDS )

M

ost scientist s who stud y social system s and communication s i n multicellular organisms ceased long ago to use group selection a s a model— though the y still revert t o that theor y occasionally without realizin g it. Microbiologists, b y contrast, still routinely use models of group selection to explain the behavio r of one-celled organisms such as bacteria an d viruses. This approac h is justified, the y say, because mos t unicellula r organism s reproduce asexually , by division, an d live in groups that ar e genetically uniform; therefore the populatio n counts a s a singl e uni t i n term s o f evolution , th e interest s o f it s member s ar e identical, an d trait s that benefi t th e grou p ar e likely to flourish eve n i f they harm specific individuals . But thi s reasonin g ignores th e fac t tha t th e larg e number s o f one-celled indi viduals an d thei r rapi d reproductio n naturall y brin g fort h mutants . Any mutant that can take advantage of others in the population is likely to increase more rapidly than thos e othe r cells . W e say , therefore, tha t microorganism s to o hav e t o b e studied fro m th e poin t o f view of individual selection .

177

178

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

THE LIF E CYCLE OF CELLULA R SLIM E MOLDS Social amebas, or cellular slime molds, of the genus Dictyostelium live in the ground as unicellular organisms , each o n it s own. Bu t when foo d is scarce or condition s are otherwis e stressful , the y gather togethe r int o aggregate s o f hundreds o r even tens of thousands o f individuals. A few hours later th e aggregat e starts migrating; at tha t stage , thi s "slim e mold," which ca n reac h th e lengt h o f one millimeter, looks t o th e nake d ey e like a small, slim y slug. In a petri dis h i n the laboratory , where mos t experiment s o n slim e mold s tak e place , thi s migratio n ca n las t fo r several hours; when i t stops, th e aggregat e gathers together agai n and th e "slug " changes into a fruiting body— a stalk with a ball of spores on its tip. 1 During this process, mos t o f the amebas a t the front o f the "slug"—no t quite 20 percent o f the total—differentiate int o a prestalk type; then they form the stalk and die . Th e ameba s farthe r back—no t quit e 8 0 percen t o f th e tota l "slug" — differentiate int o a prespor e type , clim b th e stal k an d becom e spores ; an d th e amebas i n a last smal l group, whic h u p t o no w were th e ver y end o f the "tail, " also change into th e prestal k type ; these becom e th e basal disk a t the foo t o f the stalk and die. The process by which the population i s differentiated into two types of cell s i s considere d a model fo r th e stud y of differentiatio n in primitiv e multicellular organisms and i s studied a s such in many laboratories aroun d th e world . Detailed informatio n on the process, and on the chemica l signals facilitating communication among the amebas and coordinating their behavior, has been collecte d in these studies .

FORMING THE STALK: ALTRUISTIC SUICIDE? The ameba s that di e in forming the stalk ar e considered altruistic , since by dying they help th e res t o f the populatio n t o survive . The stal k seem s t o improv e th e chance tha t the ball of spores will manage to hitch a ride on a passing insec t and thus reach a new supply of food; it may also be that the stal k protects th e ball of spores fro m predator s i n the soil, and that an y single spore is likelier t o survive in the ball than it would on its own in the soil . But why do some of the ameba s "sacrifice" themselve s t o form th e stalk? This does not presen t a problem for someone wh o believes in group selection; but a s we hav e alread y stated , grou p selectio n model s ar e inherentl y unstable . I f th e mechanism tha t cause s the ameba s t o for m th e stal k evolve d becaus e th e stal k benefits th e spores , the n wha t happen s when a chance mutation lead s eve n on e ameba to disregard the chemical command to change into the prestalk type? That ameba will change int o a spore in defiance of the command; i t may survive, unlik e the others, and such defiance will spread in the population. But we don't find such a trait spreading in any population tha t has been studied .

Social Amebas (Cellular Slime Molds)

179

But if we try to explain the phenomenon o f stalk creation i n terms of individua l selection, we are faced wit h anothe r problem : Why do the amebas that end up in the stalk develop feature s that eventuall y kill them? Why don't the y instead try to get a new lease on life by becoming spores ? This question is all the more interesting if we note—as experiment s show—tha t th e proces s b y which particula r ameba s become stal k ca n b e arreste d an d th e fat e o f those ameba s altered . Whe n a researcher cuts one slime-mold "slug " in two across its width, each piece on its own usually produce s a smal l fruitin g body , whic h ha s th e sam e for m a s a regula r fruiting body . When th e slim e mold i s cut in two, most of the ameba s that were in the fron t of the origina l slug , which woul d have formed it s stalk an d died, become spores instead. O n th e othe r hand , 20 percent o f the other cu t piece—amebas that were in the middle and back o f the original slug, which would all have developed into spores in that slug—end up becoming stalk , even if they show signs that they were in the process of turning int o spores when th e original slug was cut. The destin y o f individua l amebas , then , i s no t seale d a s soo n a s th e slim e mold forms—s o again , why d o almos t on e fift h o f thos e ameba s giv e their live s to for m a stal k tha t wil l hel p thei r fellow s survive ? Ca n individua l selectio n ac count fo r th e evolutio n o f th e fruitin g bodie s o f socia l amebas ? Thi s questio n led u s t o conduc t researc h o n th e matte r wit h Daniell a Atzmon y o f Te l Avi v University an d Vidyanan d Nanjundia h o f th e Nationa l Researc h Institut e o f Bangalore, India. 2

THE INDIVIDUAL SELECTION HYPOTHESIS We have reache d two related hypotheses. First , that these stalk-formin g ameba s are adoptin g a strateg y tha t unde r som e conditions , howeve r limited, actuall y increases thei r chance s o f survival . Second , that i f these ameba s ha d attempte d to becom e spores , thei r chance s o f surviva l woul d i n fac t hav e bee n eve n smaller. Thi s i s essentiall y th e sam e explanatio n w e offe r fo r "altruism " amon g babblers an d i n th e socia l insects : th e assistanc e offere d b y nonreproducing in dividuals t o reproducin g individual s actuall y increase s th e former' s chance s of reproducing. The individual ameba s gather together whe n food or water is scarce, and their chances of surviving on their ow n in that plac e and a t that time are nil or close t o it. Aggregates—slugs—hav e a bette r chanc e t o survive : the y ca n mov e quickly , bridge gap s between soi l particles , an d reac h a new source o f food, o r clim b u p to the surfac e o f the soil , where thei r spore s wil l hav e a greater chanc e to reach a better environment. Experiments show that the probability that a particular ameba will become a spore depend s o n th e amoun t of nutrients i t has stored compare d with other s i n th e sam e aggregate : th e one s riche r i n store d nutrient s becom e

ISO T H

E HANDICA P P R I N C I P L E I N SOCIA L SYSTEM S

spores. Amebas poor in nutrients—for example, ones that have just divided, whose resources are therefore depleted—adopt the prestalk type. This effec t i s very pronounced when on e mixes populations o f amebas with artificially altered amounts of stored nutrients : amebas raised on sugars, and thus rich in nutrients, are much more likely to become spores than amebas raised on bacteria, which ar e poorer in nutrients—even though almost 80 percent of each population would have become spores if each had bee n allowe d t o form a fruiting bod y on its own. We believe , then , that ameba s forming a slime mold differentiat e b y th e following process. When ameba s gather into a n aggregate, there is no way to tell in advance ho w eac h wil l compar e wit h others . A s a result, they all behave i n th e same way: they broadcast their interest in getting togethe r by waves of a chemical called cAMP, 3 and they all home in on the principal sourc e of those waves. Once they get together, i t is apparently too late for an y of them to change course. They cannot leav e an d see k anothe r aggregate , and i t seem s that a t tha t point , thei r chances on their own are nonexistent. Once ameba s enter an aggregate, their best chances ar e i n tha t aggregate—eve n fo r thos e wh o tur n ou t t o b e les s well nourished tha n others ; the odd s i n the aggregat e might be agains t them, but th e odds outside it would be worse still. If th e onl y way to surviv e when foo d i s scarce were to become a spore in a fruiting body, it would be pointless for any individual ameba to adopt th e prestalk strategy . However, nutrient-poo r amebas have other chances , however unlikely. By joining a migrating slug , they increase their odds of survival, since th e "slug " provide s som e protectio n agains t drying out , an d sinc e i t move s more quickl y than an ameba can move on its own. One wa y to survive is to become a spore, an d indeed, as we predicted , it ha s recentl y been found—afte r w e forme d ou r hypothesis—that a few o f the ameba s that star t ou t a s prestalk actuall y end u p becoming spores. 4 One ca n imagin e other luck y events tha t woul d enabl e a prestalk ameb a t o survive. The migrating slug may meet with another slug that contains even weaker amebas. Whe n tha t happens, th e weakes t ameba s i n th e firs t slug , whic h wer e headed towar d stal k formation, may well end u p amon g the almos t 80 percent of the combined slug that ar e stronger than the res t an d becom e spores . Moreover , the amebas tha t do not become spore s do not necessarily di e right away . As long as the y are aliv e they have some chance to surviv e if the slu g quickly gets to a n environment rich in food. For stalk-formin g amebas, these various chances of survival, however slim, are better than the odds would have been had they tried fro m the outse t t o becom e spore s i n the presenc e o f their stronge r peers . Obviously , this subject requires further researc h not onl y in the artificia l environment of petri dishes in the lab but i n the field.

Social Amebas (Cellular Slime Molds)

181

DIP AS A POISON The differentiatio n of amebas into the prestal k type follows the emissio n of DIP, a chemica l produced b y the prespor e amebas . And i t is here that ou r hypothesi s that the slim e mold is a product o f individual rather than group selection becomes as well a question about th e functio n o f DIP as a signal. Most scientists consider DIP to be a "signal," which "tells" some of the amebas that they have to develop int o the prestalk type. Group-selection model s have no difficulty wit h this: DIP causes some amebas to differentiate int o prestalk amebas, they say, and thes e prestalk ameba s end u p assistin g the prespore ameba s for the good of the group. But if one assumes the process evolved by individual selection, then two questions come up. First, why do the prestalk amebas "obey" the signal and differentiat e into the prestal k form , a process likely to en d with their death? Second, why do all prespore ameba s bother to emit DIP? After all , if the amebas around them are already emitting DIP, a single ameba would still reap the benefi t of that emission even if it did no t emi t any itself—and i t would save itself the cost of producin g an d emittin g the DIP . Ho w doe s eac h individual prespore ameba benefit b y emitting DIP? DIP eventually causes the prestalk amebas to die; for that reason, we suspected that it was a harmful chemical— a poison. This prediction turne d out to be true. 5 We believe that DIF's poisonous properties are its very reason for being. We think that the DIP is a poison use d by spores to avoid predation, and that the prespor e amebas manufacture it for their own individual benefit—probably to protect themselves. Chemicals that ar e secreted to keep other s away—allelopathi c chemicals—including antibiotics, 6 are usually understood t o be a tool that enables a population resistant to those chemical s to get rid o f populations o f bacteria that are sensitive to them. This model, again, is based on group selection: any given individual could save itsel f th e troubl e o f manufacturing the chemica l and rel y instead on th e secretions of others in the population . We, on the other hand, ask rather how the individual cell benefits by secreting an antibiotic or a toxin. Perhaps th e antibiotic helps the individual cell protect its living spac e and it s immediate surroundings against other individual s tha t might prey o n i t o r caus e it othe r harm . If so , the n althoug h antibiotic s ma y help a n entire population against another, thi s is not wha t drives the evolutio n o f antibiotics bu t i s rathe r a sid e effect . I n fact , i n natura l populations , th e amoun t of antibiotics secreted by cells is minute and canno t kill off other populations . Cells that manufactur e large amount s of antibiotics 6 ar e produced onl y by means of a strenuous selection proces s conducted b y researchers in the lab. It i s no t uncommo n fo r microorganism s to us e poison s t o protec t dorman t spores, whic h woul d otherwis e b e eas y prey. The effor t neede d t o produc e th e poisons, an d th e dange r an d inconvenienc e the poison s themselve s pose t o th e pre-spore cell s tha t produc e them , ar e compensate d fo r b y th e protectio n th e

182

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

poisons provide . W e suspec t tha t thi s is the primar y reason fo r the manufacture and emissio n o f DIF; th e effec t o n lower-qualit y amebas—the one s wit h fewe r stored nutrients—i s secondary . If this hypothesis turn s out t o be true , the n DI F is not a signal but rathe r a straightforward tool. Fo r example , rain causes people to take cover, but nobod y would sugges t that rain is a signal that appears in order to cause people to take cover. Since we suggested thi s hypothesis , Atzmon y and Nanjundia h have found indications throug h bioassay that DIF i s indeed presen t i n spores. 7 If hig h concentration s o f DIF kil l amebas , lower concentration s ar e probably harmful a s well. The use of DIF force s amebas to devote resources to surviving its influence and breaking it down. In weaker amebas, we believe, these efforts create the prestal k phenotype , whic h increases thes e amebas ' abilit y to surviv e the poison's influence in the short term; we have suggested that by trying to form a spore, a weaker ameba actuall y reduces its chance s of survival, and i t is the presenc e of DIF more than anything else that creates this stark reality. Collaboration i s crucial to all the amebas—if they don't for m th e slug none will survive—and that is what makes it possible for the stronge r ones , thos e that can withstand the poison and form spores , t o exploit the weaker ones. Cellular slim e molds ar e not unique . Myxobacteria , too, for m communa l fruiting bodie s fo r their spore s i n reaction to hunger. 8 In the process of forming the fruit ing body, some 80 percent o f the cell s are killed by fatty acids emitted by the bacteria themselves; apparently the stronger bacteria emit the fatty acids and kill the weaker in th e proces s o f becoming spores . I n thi s case too, i t seems t o us , i t i s because collaboratio n i s necessary t o all that stronger individual s are able to exploit weaker ones; again, to understan d the process, we have to understand how each spore-forming bacterium cell benefits by emitting the specia l fatty acid s that kill its comrades. 9

THE DIFFERENCE BETWEE N PRESTALK AN D PRESPORE AMEBA S What cause s the differen t reactio n t o DI F i n prespore and prestal k amebas? As we said, the difference between the two is not in genotype, but rather in phenotype: the prestalk amebas are relatively short of stored nutrients. We think that in order to breathe, the y have to break dow n protein s fro m thei r own bodies, unlik e their prespore siblings , whic h us e thei r store s o f glycogen . Coul d i t b e tha t th e DI F interferes someho w wit h th e breakdow n o f proteins ? W e starte d explorin g thi s

Social Amebas (Cellular Slime Molds)

183

hypothesis wit h th e ai d o f Dori t Arad , a chemis t wh o studie s th e relationshi p between a given molecule's structure and its activity. But whil e w e were doin g so , Shaulski an d Loomis 10 found tha t DI P impairs the activity of mitochondria—the organelles, or intracellular organs, that produc e energy by breathing. They also found that in the process o f differentiating int o the prespore type , prespor e ameba s produc e additiona l mitochondria . Perhap s th e new mitochondria develo p t o supplemen t thos e weakene d by DIF. If so , it may be that the DIF prevent s th e weaker amebas from becomin g spore s becaus e they do not have the resources to build new mitochondria. It may still be that DIF als o interferes with the breakdown of proteins; we are still awaiting conclusive evidence on the matter. In either case , DIF has indeed bee n show n to be a poison. If DIF were simply an arbitrary, conventional signal, then a mutation that made an individual ameb a disregard it would b e beneficia l to tha t ameba . But if , as we think, DI F i s a poison tha t actuall y prevents weake r ameba s fro m turnin g int o spores, then a weak ameba cannot under an y circumstances afford t o try and for m a spor e i n it s presence . I t woul d mak e sens e t o thos e weake r ameba s t o g o th e prestalk rout e an d surviv e as best the y can, for a s long a s they can. In fact , som e of th e specifi c propertie s o f prestal k ameba s see m t o hav e evolve d precisel y t o enable these cell s to withstand th e poisonou s effec t o f DIF: the y emit an enzyme that break s dow n th e DIF , an d the y move awa y fro m th e prespor e ameba s that emit DIF an d into the front en d of the slug, where the DIF concentration is lowest. The emission of DIF, then , provide s informatio n to the weaker amebas; but DI F emission di d no t evolve in order t o d o this , but rathe r t o enhanc e eac h stronger ameba's chance s o f survival.

SOME REMAININ G QUESTIONS If DI F i s a poison, tha t woul d explai n wh y prestalk ameba s attempt t o brea k i t down an d mov e awa y fro m it . Bu t i t doe s no t explai n th e considerabl e effor t prestalk ameba s devote t o moving the slug and building th e stalk—the behaviors that hav e traditionally bee n calle d "altruistic. " The fas t movemen t to the surfac e of th e soi l benefits th e prestal k ameba s overall, but tha t in itself does not explai n the effor t tha t eac h individual ameb a devote s t o th e movement . Our mode l wil l not b e complete until we can explain these aspects of the social amebas' life cycle. The movement of the slug involves the emission of cAMP, which attracts amebas. When th e amebas first gather together, eac h ameb a reacts to cAMP by emitting cAM P itself , an d thu s th e larg e aggregate s o f amebas ar e formed. But wit h the formation of the slug, it seems that the prespore ameba s stop emittin g cAMP, while the prestalk ameba s at the front ti p o f the slug continue emitting it and thus get th e res t o f th e ameba s t o follo w them . T o understan d th e movemen t o f th e

184

THE HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

aggregate o r slim e mold, on e ha s t o understan d no t jus t th e effec t o f cAM P o n the ameba s that receive it and respon d t o it, but mor e important , how each individual ameba that emits cAMP benefits by doing so. Unfortunately, we don't have answers to these questions yet, but we hope that by posing them we will have made a useful contribution .

WHEN IS A CHEMICA L A SIGNAL ? Until now, th e approac h o f most scientist s ha s been t o explain th e valu e of each chemical "signal " accordin g to it s effec t o n th e individual s reactin g t o it , rathe r than the benefit it brings to the individual emittin g it , or the reliable informatio n it provide s abou t tha t individual . Thi s ha s led , i n ou r opinion , t o a lumping to gether o f chemica l signals —chemicals tha t evolve d t o facilitat e communicatio n between individual s o r betwee n bod y cells—an d chemical s tha t evolve d fo r straightforward utilitaria n purposes , whos e effec t o n othe r individual s i s eithe r incidental o r coercive . In orde r t o understan d chemica l signaling, one mus t differentiate between thes e categories . We believe that chemical signals conform to the rules laid out in earlier chapters for al l other signals: the signal must benefi t bot h the sende r an d the receiver ; th e signal must b e reliable—whic h means it must impos e a real cost o n the signaler , a cost to o hig h for a cheater to pay; and the specifi c cos t o f the signal must relate logically to the content of the message it conveys. Investigations of chemical signal s among one-celled organisms , and chemical signals in general, carried out with these principles i n min d ma y lea d u s t o strikingl y ne w understanding s o f ho w suc h signals work.

C H A P T E R 1

6

PARASITE AN D HOS T

AN ARM S RAC E OR A STAT E OF EQUILIBRIUM?

W

hen a parasite attache s itself to a new host species, both species rapidly develop new traits. The parasite tries to exploit th e host as much as it can, and the host tries to protec t itsel f agains t the parasite . Dawkin s and Kreb s compared thi s relationship t o a n arms race between tw o superpowers, with eac h manufacturing ne w and mor e sophisticate d weapon s t o counte r thos e the other develops,1 and it is common fo r scientists nowadays to portray hosts and parasites as being in the midst o f such an arms race. We se e things differently . Ther e ar e undoubtedly "arm s races, " bu t w e think they quickly reach a stalemate. We begi n b y assuming that unles s there is a specific reaso n t o think otherwise , mos t host-parasite systems that exist today are in a state o f equilibrium: howeve r hard i t may have tried, the host has not manage d to ge t ri d o f the parasite , an d obviousl y the parasit e has not exploite d th e hos t to extinction . T o understan d thi s balanc e i n eac h particula r case , on e need s a thorough, detaile d knowledg e o f the live s o f both hos t an d parasite . What i s it that stop s the host fro m evolvin g the ability to prevent th e parasite from exploit ing it? What i s it tha t prevent th e parasite from exploitin g eac h host t o the point of zer o reproduction , o r fro m exploitin g mor e tha n a certai n percentag e o f it s potential hosts? 185

186 TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

EUROPEAN CUCKOOS AN D REE D WARBLERS Let's take the European cucko o and the reed warbler a s an example. The cuckoo lays egg s i n th e nest s o f ree d warblers . Immediatel y afte r hatching , th e cucko o nestling pushes it s host's egg s or nestlings ou t o f the nest; thus, a nest parasitized by a cuckoo does not produce even one warbler. In response, reed warblers evolved the abilit y t o identify th e cuckoo' s egg and push it out. Thi s ability to recognize and rejec t cuckoo egg s evolved, i t seems, in direct reactio n t o the cuckoos ' para sitism: in Australia, where ther e ar e no European cuckoos , none of the ree d warblers rejecte d cuckoo eggs placed experimentall y in their nests. Once the warbler s evolve d thei r new metho d of self-defense , the Europea n cuckoos in Europe and Asia, in their turn, evolved eggs simila r i n size and colo r to the egg s of reed warblers. European cuckoo s parasitize other songbirds as well and ar e divided int o different subgroup s (genets ) according to the host they parasitize. In most cases, the color , size , and shape of the cuckoo' s eggs are similar to those o f the host species . In area s inhabited b y cuckoos , som e specie s o f songbird s reject all cuckoo eggs, and cuckoos thus cannot parasitize them. Not so with reed warblers: in almost all areas inhabited by both reed warbler s an d cuckoos , som e ree d warbler s rejec t cucko o eggs an d other s d o not . Davie s an d Brook 2 assume d that thi s current situatio n i s a temporar y stag e i n a lon g evolutionar y process—an "arms race"—at the end of which all reed warblers will rejec t cucko o egg s an d th e cuckoo s wil l no t b e abl e t o parasitize reed warbler s a t all. Is thi s reall y the case ? As we sai d in th e discussio n o f prey an d predator s i n the firs t chapte r o f this book, anythin g ca n b e explaine d awa y a s not ye t having evolved t o its logical conclusion. Yet all reed warble r populations, bot h in Britain and in Japan—and, presumably, anywhere in between—include some individuals who rejec t cucko o egg s an d other s wh o accep t them . Thus , th e "mutation " t o recognize cuckoo egg s and reject them has apparently already appeared in all reed warbler populations . Sinc e th e benefi t o f rejectin g cuckoo egg s i s s o great, wh y has thi s "mutant " no t ye t sprea d throughou t th e population ? Wh y i s it tha t al l over Europ e and Asia, from Britai n to Japan, w e do no t kno w o f even one pop ulation o f reed warblers al l of whose members rejec t cuckoo eggs? What i s it that prevents some reed warblers from doin g so? Lotem, wh o studie d cuckoo s an d ree d warbler s i n Japan wit h Hirosh i Nakamura and with us, suggested that the difference betwee n ree d warblers who reject cuckoo egg s and thos e who accep t the m i s not a genetic trai t i n the forme r that has "not yet " sprea d throughou t th e population; rather , it is a phenotypic differ ence—a difference in age. Most reed warblers who reach their breeding area early reject cucko o eggs. All these birds have new tail feathers, which indicates that they molted the previous fal l an d thus that they are two years old or older. By contrast,

Parasite and Host

187

reed warbler s who arriv e at their breeding are a later, most of whom d o not rejec t cuckoo eggs , have ver y worn tai l feathers . Bot h th e stat e o f their tail s an d thei r late arrival indicate tha t they are young, and tha t thi s is their first year of nesting. What differenc e doe s age make? The eggs of reed warblers, like the eggs of many bird species, vary; each female lays eggs that hav e their ow n distinct color , pattern, an d shape. Som e ar e darker, other s lighter ; some have larger spots, other s fin e dots . Cucko o egg s may look lik e ree d warbler egg s generally, but the y cannot loo k exactl y like the egg s of a specific femal e ree d warbler. 3 Lotem suggest s tha t a ree d warble r wh o ha s neste d before ha s learned t o recogniz e her ow n eggs and s o can safely reject cuckoo eggs; inexperienced reed warblers, on the other hand , cannot . If they were to reject any egg that looked suspicious , the y might inadvertently destroy their own eggs—which is apparently a greater ris k than th e 5to-20-percent chanc e tha t the y wil l fal l victi m t o a cucko o i n tha t firs t nesting . Lotem conducte d experiment s tha t showe d tha t indee d ree d warbler s lear n to recognize thei r ow n eggs. The relationshi p betwee n ree d warblers an d cuckoo s is thus no t a n ongoin g arm s race , i n whic h som e warbler s ar e a t a disadvantag e because the y "haven' t yet " acquire d a geneti c trai t enablin g the m t o ge t ri d of cuckoo eggs ; rather , the y have reache d a n equilibriu m i n whic h eac h individua l does th e best it can to succee d i n reproduction . Of course, on e still has to explain why reed warblers who don't destroy cuckoo eggs take care of the cuckoo's nestling after i t is hatched.4 The parent warbler doe s this both i n the nest an d ou t o f it, for some four t o five weeks, by which time th e young cuckoo' s size , shape , color , an d call s ar e al l very differen t fro m thos e o f reed warble r nesdings .

GREAT SPOTTED CUCKOOS AND CROWS Crows, unlik e ree d warblers , neve r try to rejec t th e egg s or nesdings o f the grea t spotted cuckoos that parasitize them.5 Crows lose only a part of their reproductiv e potential i n a parasitized nest: the laying cuckoo usuall y breaks o r ejects onl y one of thei r egg s when sh e lays her ow n an d leave s th e res t alone . Fo r it s part , th e nestling o f the grea t spotte d cucko o doe s no t pus h th e crow' s youn g ou t o f the nest, but i t does compet e wit h the m fo r food an d thu s cause s them considerabl e harm. Yora m Shpirer , wh o studie d cuckoo s an d crow s i n Israel , foun d tha t a n average unparasitized crows' nest fledges two successful offspring, while on average a nest that contain s a cuckoo nestlin g fledges only one successfu l crow . Crows hav e a reputation fo r bein g amon g th e mos t intelligen t o f birds. The y

188 TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

distinguish between differen t food s a s well as between differen t people , and they recognize enemies . On e woul d thin k tha t crow s coul d increas e thei r breedin g success a great deal by learning to discriminate between their own eggs and cuckoo eggs and gettin g ri d o f the latter . But they don't d o this. A common explanatio n is that cuckoo s hav e just starte d parasitizin g crows, and th e crow s "haven' t yet " evolved the ability to fight against them. But we think this is unlikely. Great spotted cuckoos parasitiz e various population s an d specie s o f th e cro w family (Corvids ) all over Asi a and Nort h Africa , an d throughou t thi s larg e regio n n o populatio n shows any resistance to them, with one exception: magpies in Spain. Many magpies in Spai n reject eggs of the grea t spotted cuckoo , bu t no t al l of them do. 6 We ca n think of two alternative models that can explain the constraints that prevent crows from resistin g cuckoos: the prestige mode l an d the Mafia 7 model .

THE PRESTIGE MODEL The prestig e model suggest s that th e host devote s effor t t o rearin g the parasite' s offspring t o gai n prestig e i n the eye s of its ow n partner an d t o practic e workin g in tande m wit h tha t partner . Takin g car e o f th e offsprin g o f others t o improv e one's ow n chance s at reproduction i s not s o unusual; 8 we discussed more tha n a few such cases in chapter 14 . If an individual can improve its reproductive chances by raising the offsprin g o f others of its own species, it makes sense that individuals might als o gain by raising parasites. It i s well know n tha t pair s who hav e already successfully raised offsprin g to gether ar e likelier to succeed tha n those who have not; this is true partly because pairs wh o d o no t succee d i n raisin g offspring ten d t o brea k up. 9 I t ma y well b e that whil e a new partnershi p offer s bette r prospect s tha n on e tha t ha s failed , it represents mor e o f a gamble tha n doe s stickin g with a n old partne r wit h who m one has alread y established a successful workin g relationship. The research of Yom-Tov and others on crows 10 suggests that the reproductiv e success o f crow s depend s largel y on thei r abilit y to ge t hold o f a good territor y and kee p it . We assum e that a well-coordinated pai r of partners i s much likelie r to achieve this. What, then, are the options for a pair of crows in whose nest a cuckoo egg has been substitute d fo r on e o f thei r own ? Onc e that happens, they cannot raise as many young crows i n tha t nes t a s the y otherwis e woul d have, for one of their own eggs has already been destroyed. And becaus e the y hav e bee n les s than fully successful in raising their young, they also run the risk of splitting up. Yet by devoting care t o th e cucko o nestling , the y may be abl e to convinc e eac h other o f their qualit y as par-

Parasite and Host 18

9

ents an d thus stay together. I f they do not brea k up , the y stand a good chanc e of having a fully successfu l nes t next time. If they "divorce," on the other hand, each partner woul d hav e t o fin d a ne w mat e an d migh t fai l altogether ; a t best , eac h would b e likely to have difficulty raisin g offspring nex t yea r as well. Fledgling cuckoo s tha t com e ou t o f crows' nest s behave as though the y know that the onl y reason thei r hosts take care of them is to gain prestige. They beg for food ver y loudly an d pursu e thei r host-parents fa r mor e aggressivel y than thei r crow nestmates. Rather than try to imitate the crow nestlings' behavior, the young cuckoos stand out by noisily importuning. We believe that the crows' real offspring don't nee d to beg as much, since their parents derive a direct reproductive benefi t from feedin g them . Bu t th e cro w doe s no t gai n th e sam e way from feedin g th e cuckoo, of course; the only gain it gets from feeding that "pest" is proving to other crows—including its mate—how good a provider it is. Thus, th e cuckoo nestling has to draw others' attentio n to the act of parenting in order to get fed. Of course, this noisy behavior is dangerous: the young cuckoo is more likely to be noticed by predators. Thi s ris k is the price i t has to pay for its food. Needless t o say , we woul d no t drea m o f assertin g tha t th e cucko o an d th e crows ar e awar e of an d consciousl y calculat e al l thes e factor s of ris k an d gain . Rather, fledglings who strik e the best balanc e between risk y begging and making do without begging are more likely to grow up, become successful adults, and have offspring wh o will spread their traits in the population. Similarly, adult crows who invest th e righ t amoun t in convincing other s of their qualit y as providers ar e th e ones mor e likel y to fin d an d kee p bette r mate s an d t o hav e more successfu l offspring. Ultimately, as with any trait, cos t an d gai n must be balanced—and young cuckoos hav e t o inves t mor e i n beggin g tha n youn g crow s do , i n orde r t o ge t sufficient car e out o f their host-parents.

THE MAFIA MODE L The Mafia model is very different fro m bot h th e "arms race " model and the prestige model. I n thi s case , the parasit e is not tricking its host bu t rathe r forcing th e host t o tak e care of its offspring. I t i s well known tha t cuckoo s prey on the egg s and nestling s of other birds; 11 thus, it would see m they could enforc e their will as effectively a s any mafioso. Suppos e a cuckoo were to revisi t nest s it has laid eggs in, leavin g alon e th e one s wher e i t find s it s nestlin g livin g bu t eatin g th e entir e contents o f nests where its offspring has been rejected. I n tha t case, the discriminating host will com e ou t th e loser , an d the abilit y or propensity t o discriminate will not spread in the population. 12 What woul d a cuckoo gai n fro m "punishing " hosts tha t recogniz e and rejec t her eggs ? She cannot be doin g s o in order t o help her rejecte d offspring, whic h is already dead. If we cannot find some other direc t benefit to the individual cucko o who doe s th e punishing , then th e Mafi a mode l wil l b e based o n group-selectio n

1 9O TH

E HANDICA P P R I N C I P L E I N SOCIA L SYSTEM S

arguments—that is, it would b e goo d fo r cuckoos a s a group but woul d deman d that the individua l make an effort fro m whic h she herself doe s not gain . We have already explaine d wh y we canno t accep t model s base d o n group-selectio n logic . In this case, though, several direct benefit s to the cuckoo who does the punishing come to mind. Perhaps the cuckoo is forcing the hosts to lay again, which will give her anothe r opportunit y t o la y an eg g in thei r nest . O r perhap s sh e just gain s a good meal. Lotem di d no t fin d evidenc e o f suc h behavio r amon g Europea n cuckoo s i n Japan; neithe r the y nor th e grea t creste d cucko o tha t parasitize s crows i n Israe l seem t o punis h unsatisfactor y host s accordin g t o th e Mafi a model . Bu t Soler' s research in Spain seems to point in precisely that direction: in Spain, great crested cuckoos return t o th e nest s o f magpies, and i f their egg s have been rejecte d they harm th e magpi e nestlings, 13 presumabl y to teac h magpie s that nes t i n their ter ritory not to reject cuckoo's eggs. This fascinating phenomenon is now under study by Soler an d his colleagues. The relationshi p betwee n th e parasit e be e Nomada marshamella an d it s host, th e be e Andrena sabulosa, 14 illustrate s th e logica l nex t ste p i n th e Mafi a model. Th e parasit e bee enter s th e nes t o f it s hos t withou t resistanc e from th e owner an d lay s he r eg g at th e sid e o f the host' s pupa . I t i s highly probable tha t the hos t recognize s th e parasite' s eg g a s such, bu t sh e continue s bringin g foo d to th e nest , take s car e of he r ow n offspring , an d doe s no t tr y t o ge t ri d o f th e stranger. Why doesn' t th e hos t ge t ri d o f th e parasite' s egg ? Apparentl y the parasite , who doesn't nee d t o collec t food, remains near the nest almost all the time. Once she has lai d he r ow n egg s in it , it i s worth he r whil e t o protec t th e nes t against robbers an d othe r parasites . Fro m tim e t o time , sh e threatens the hos t wit h th e stinger a t the en d o f her abdomen , whic h ca n fold t o th e fron t lik e a scorpion's . The parasite , wh o doesn' t nee d t o carr y food, ca n affor d t o evolv e an abdome n flexible enoug h t o be a n effective fightin g too l insid e th e tunnel s o f the nest ; th e host, wh o ha s to carr y pollen an d necta r fro m plant s in the field to the nest, does not have such an abdomen. Since th e parasit e ca n figh t bette r tha n th e nest' s owner, she is able to prevent others of her species, and individuals o f othe r species , fro m robbin g th e nest . The hos t accept s the damag e done b y this on e parasite an d acquire s i n exchang e a partner who protect s her agains t others. 15 Thi s i s supposed t o b e th e rela tionship o f the Mafi a t o it s clients, and thi s was ofte n the relationship betwee n knights and their serf s in the Middle Ages: the knights levied taxes, often very heavy ones, an d demande d service , bu t the y protecte d th e serfs agains t the eve n greate r damag e that other knight s an d robber s migh t have inflicted.

Parasite and Host 19

1

ACCEPTING A PARASIT E TO MINIMIZ E DAMAG E Many parasiti c bird s pec k o r pus h ou t o f the nes t onl y som e o f the host' s egg s when they lay their own . Once this significant bu t limite d damage is done, getting rid o f the parasite' s egg will no t repai r th e damage . Offspring hatched earlie r in the seaso n are usually more successful, an d it is preferable to continue taking care of th e nes t an d t o mak e d o with fewer, mor e promisin g offspring tha n t o deser t the nes t an d attemp t t o buil d a ne w one—whic h itsel f migh t fal l pre y t o th e parasite. Sorensen describe s suc h a relationshi p betwee n tw o specie s o f ducks , th e canvasback (Ay thy a valisinaria) an d th e redhea d (Aythya americana), a parasit e that lay s i n th e former' s nests. 16 Th e hos t recognize s th e parasit e a s such , an d when th e parasit e approache s he r nes t sh e treat s he r a s a n enem y an d attack s her. Bu t onc e th e parasit e settle s o n th e nest , th e hos t stop s resistin g an d th e two female s en d u p layin g egg s together . Th e resul t i s a mixe d nest , wit h n o more duckling s o n averag e tha n a normal brood . Th e hos t incubate s th e com bined egg s alon e an d seem s t o hav e n o troubl e lookin g afte r al l the duckling s and leadin g the m t o foo d sources . Accordin g to Sorensen , fightin g o n top o f the nest woul d hav e caused mor e damag e to th e host' s egg s than th e presenc e o f a few parasit e eggs does. Song sparrows are parasitized by brownheaded cowbirds ; older son g sparrows seem actively to help th e cowbird find their nests by mobbin g th e parasit e whe n i t ap proaches. As a result, female son g sparrows two years old or more ar e parasitized more heavily by cowbird s tha n are younge r song sparrows, wh o d o not mo b th e cowbirds . Ye t olde r son g sparrow s ar e more experienced tha n younger ones an d o n averag e reproduce mor e successfully. Obvi ously, it is to the parasites' advantage to know who these better surrogat e mothers are—but why do the older sparrow s help the cowbird find their nests? Smith and his colleagues , wh o describe d th e phenomenon , suggeste d that th e syste m is not yet in balance, and that eventually evolution wil l lead song sparrows to stop mobbing cowbirds. 17 We suggest a different explanation : when a cowbird lay s her eg g in the nest of a song sparrow who i s herself stil l laying eggs, the cowbir d doe s not destro y all of her host' s eggs . Bu t whe n sh e find s egg s i n th e late r stage s o f incubatio n i n a sparrow's nest , then th e cowbir d tend s t o peck more or even all of them—apparently in order t o make the hos t lay again, so that the cowbird' s egg s will be incubated fro m th e da y the sparro w start s incubatin g he r own . Older , experience d song sparrows star t nesting earlier i n the seaso n than younger ones. I f we assume that eventually the cowbird will find most of the song sparrow nests in her territory, including replacemen t nest s a sparrow ma y build afte r he r first nest fails , the n i t

192 T H

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

makes sense for song sparrows to expose their nests to her early, so that the parasite will lay in their nest s rather than destro y all their eggs . Thus i t seem s tha t olde r son g sparrow s wh o mo b th e cowbir d an d thereb y reveal their nests ' locatio n woul d rathe r accep t a certain amoun t of damage than pay a heavier pric e fo r attemptin g t o evad e o r rejec t the parasite . If th e damag e caused b y th e on e foreig n nestlin g i s not to o great , an d i f i t i s bette r fo r son g sparrows to fledge offspring earl y in the season rather than later, as is the case with many birds , then i t seem s that son g sparrow s wh o mo b cowbird s ar e using th e best strateg y available to the m when cowbird s ar e present, a s in western Canad a today. Why, then , don' t younger female sparrow s mo b cowbirds ? I t ma y be that they are not abl e to rais e both thei r ow n young and the parasite' s nestling. Thei r only recourse i s to lie low and hope th e cowbir d wil l miss them. In ou r opinion , th e coexistenc e o f parasites an d host s ofte n depend s o n th e fact tha t a host stand s t o lose more in an all-out struggl e with the parasit e than it loses b y toleratin g th e curren t leve l o f parasitism . I n eac h cas e th e resul t i s a dynamic balance, an d an y change in conditions—an y weighting o f the scale s on one side or the other—migh t jar it out o f equilibrium.

NEUTERING THE HOST Many parasitic worms an d fung i attac k the host' s reproductiv e organ s and make it sterile or nearly so.18 Then even if the host later manages to get rid of the parasite, the host canno t reproduc e an d pass on to its descendants th e trait s that enable d it to free itself . We believe that this strategy of sterilizing the host ca n evolve only if th e individual parasite gain s directl y fro m it ; again , if th e specie s a s a grou p benefits rathe r than the individual, we are left with a model based on group selec tion. But the individual parasite stand s to gain plenty in this case: because the host is prevente d fro m devotin g itsel f t o it s ow n reproduction , mor e o f its resource s are available to the parasite . I n this case , the onl y defense a host specie s ca n pass on to futur e generations—th e only defense that ca n spread amon g the host pop ulation b y evolution—is strategies that prevent th e parasite from gainin g even the first foothold, o r that preven t i t from sterilizin g the host .

FROM PARASITE TO COLLABORATOR Schwammberger19 studied the relationship betwee n th e parasite wasp Sulcopolistes artimandibularis and its host, the wasp Polistes biglumis. This parasite, it turns out, takes over a main nest, where she lays eggs; she then takes over neighboring nests, where sh e does no t la y eggs but rathe r steal s larg e larvae and pupae , whic h sh e

Parasite and Host 19

3

brings to the mai n nest as food for her own and her host' s offspring . The parasite protects he r mai n nes t agains t other robbers , a very real benefit t o th e host . Schwammberger found that the presence of this parasite was beneficial to the host. True, whe n a t the en d o f the seaso n h e compare d th e succes s o f nonparasitize d colonies wit h thos e tha t ha d thi s parasite , the forme r produce d mor e offspring . But whe n h e took int o accoun t al l the colonie s tha t faile d durin g th e season , h e found tha t th e chance s that a nest would surviv e were much higher i f it had suc h a parasiti c collaborator. I t turn s ou t tha t th e parasit e provide s crucia l hel p earl y in th e season , when th e hos t doe s no t ye t have daughter s t o shar e the load : b y protecting the nest, the parasite enables the host to undertake longer hunting trips. In fact , i n this case the parasite fulfill s th e function same-specie s cofounders do in other species . Sometimes a host ca n choose amon g parasites and pic k th e on e least likel y to damage it; that one might even protect it against the others. As we saw above, this is a logical outcome of the Mafia model and can lead to symbiosis between species , like the partnership betwee n human s and dogs . Whe n i n distant prehistori c day s humans starte d bringin g foo d t o thei r camp s and storin g it , rodent s an d insect s started gathering there, and in their wake predators that lived both on the humans' leftovers an d on the rodents an d insects they attracted. Among these larger predators was the wolf, who, because of its preexisting social traits, was easily integrated into human society. The human-allie d wolf refraine d fro m preyin g on weaker humans suc h a s young children an d chase d other predators , including other wolves, away from th e territory it shared with the human beings. In exchange, the humans shared their food with th e wolf-dogs . A do g fe d b y humans has a n advantage over it s wild kin : i t is protected an d fe d an d thu s ca n affor d to becom e large r an d to tak e risk s in fights that i t coul d no t afford t o take if it had t o find food or catc h prey on its own . By cooperating with humans and feeding at its owner's table, the do g gave up som e of its ability to live independently .

THE LES S VIRULENT PARASITE AS A COLLABORATO R AGAINST IT S VIRULENT VARIANT The partnershi p betwee n a host an d a nonvirulent parasit e i s like tha t betwee n human an d dog . Th e suppor t provide d b y the host let s th e nonvirulen t parasit e evolve traits that enable it to compete successfull y with virulent variants of its own species. It makes sense, then, for the host to invest in the nonvirulent parasite and support i t against the virulent parasite . Sometimes a parasite can deman d suppor t b y holding th e threa t o f virulence

194 TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

over its host's head . Th e bacterium tha t causes diphtheria exist s in the throats of most human s in a nonvirulent form . A toxin gen e resides in plasmids insid e th e bacterium,20 whic h ca n manufactur e a killing toxin ; bu t i n mos t case s a special represser protein, manufacture d by a gene in the bacterium's chromosome, keeps the bacterium's toxi n gene inactive. To manufactur e the represse r protein , th e bacteriu m need s iro n tha t i t gets from it s huma n host . Whe n th e bacteriu m doe s no t ge t enoug h iron , i t stop s manufacturing th e represser protein, an d the plasmids star t producing th e diph theria toxin , causin g th e deadl y disease . A s lon g a s the bacteriu m receive s th e nutrients it needs—including iron—it does not produce th e toxin, so it is worthwhile for a human body that can spare them to discharge some nutrients to supply the bacteria. The bacterium thus "blackmails" it s host int o supporting it . What doe s the individual bacterium gain from emittin g the toxin? The starved bacterium ca n probably feed on nearby body cells the toxin disintegrates; perhaps the toxi n als o protect s th e starve d bacteriu m agains t othe r diphtheri a bacill i nearby, who are just as starved as it is (see the discusssion on antibiotics i n chapter 14). I f the host does not cooperate, it is to the parasite's advantage to act virulently; it is only the host's suppor t tha t enables th e nonvirulent phenotyp e t o overcom e the virulent one . In diphtheria , thi s tria d o f host , nonvirulen t parasite , an d virulen t parasit e involves differen t phenotype s rathe r tha n differen t genotypes o f the parasite . I t ma y well b e tha t similar relationships occur between host s and various strains, or genotypes, of fungi, bacteria, or viruses, some more virulent an d som e les s so . Her e again , th e hos t ca n assist th e nonvirulen t parasit e an d us e it a s a partner against its virulent relative. Reliable communication is as necessary between such partners as it is between any others.

THE IMPLICATIONS OF ASSUMING A STATE OF EQUILIBRIU M To su m up, ou r approac h t o parasitism is different fro m tha t of researchers who assume that they are witnessing "arm s races " between parasite s and hosts. As we have stated , an y adaptation—includin g an y socia l system , whethe r i t operate s within the species or between species , as parasitism does—can be explained awa y as not "yet " havin g reached it s logical evolutionar y endpoint. Suc h assumptions, by thei r ver y nature , ca n accommodat e an y finding s an d d o no t generat e ne w predictions o r open ne w avenues of research . Our ow n researc h hypothesi s regardin g parasitism is always based o n th e as-

Parasite and Host 19

5

sumption that what we see today is a dynamic state of equilibrium, a balance that results because each party is doing its best t o reproduc e a s successfully a s it can. In case s where w e do no t understan d thi s equilibrium , we assum e that w e have not yet found all the element s of it, and that we have to look for the missing ones. Just a s the assumptio n that animal communication is necessarily reliable open s one's eyes to the richness of communication actuall y taking place in nature, so also the assumptio n that social system s are in a state of equilibrium enable s one to see and appreciat e th e incredibl e complexit y an d beaut y o f thes e systems . This ap proach lead s u s to as k new question s an d devis e new hypotheses t o test , an d so ultimately it turns ou t t o be far more fruitful tha n its alternative.

This page intentionally left blank

C H A P T E R 1

7

INFORMATION CENTER S

FOOD SOURCES AN D SOCIA L ORGANIZATION : THE WHITE WAGTAI L

T

he white wagtail is a common winter migran t in Israel. It s socia l behavior varies.1 Some individual s wande r aroun d i n flock s an d look for suc h sporadic source s o f foo d a s a tractor plowin g o r reapin g o r a herd o f shee p grazing and stirring up insects. Others find a more permanent source of food, such as a barnyard, a collection o f trash containers, the corner of a landfill, or a vegetable garden an d establis h a territory tha t they defend agains t other wagtails. Each territory is defended by one male—but not necessarily the first to stake a claim, since stronger males often expe l weaker ones and take over their territories . In som e of the territories , a female joins th e male . Th e tw o behave a s if they are going to establish a nest together, but the y do not copulate . Th e female helps defend th e territor y an d chase s awa y other females . Sometime s th e relationshi p between the two lasts all winter, but ofte n it does not. Sometimes the female moves on to another territory or rejoins a flock; sometimes sh e is expelled by a stronge r female; an d sometime s one female establishes a relationship wit h mor e tha n on e male and feeds i n both thei r territories . Males , too , leave their territories i f thei r food source s dr y up; they then rejoi n flocks. In fact , w e have found tha t individual s will claim territories fo r themselve s o r form flock s dependin g o n th e wa y we distribut e foo d i n th e field . W e tagge d 1 97

198 TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

members o f one flock with colore d le g rings. When w e then presente d tha t floc k with small heaps of food, eac h heap was taken over by one individual, who chased others awa y from it . Bu t when w e scattered on e suc h hea p ove r a larger area , an individual coul d n o longer defen d it effectively; h e stopped trying to and fed side by sid e wit h others . Whe n w e stoppe d supplyin g food , al l the bird s wh o hel d territories wen t bac k t o the flock. As far a s we know, it was the first time that an experiment i n the field showed tha t environmenta l factors can chang e socia l behavior. As much a s the socia l behavior o f individual wagtail s varies, these birds hav e one thin g i n common : ever y evening, al l wagtails gathe r i n communa l roosts — sometimes thousands of them together. Bot h flocks and territor y holder s congre gate in the roosts, and some flocks fly fifteen to twenty miles to join them.

COMMUNAL ROOSTS A S INFORMATION CENTER S Communal roost s ar e not uniqu e t o wagtails . In winter , starling s gathe r b y th e millions t o roos t i n wood s an d ree d thickets ; some o f them trave l sevent y miles every evenin g t o joi n th e roost . Sparrows , egrets , gulls , swallows , crows , doves , and many other specie s gather to roost together. What d o birds gai n from makin g the long daily commute to the centra l roost? One migh t thin k a t first that th e birds selec t a resting place tha t i s protecte d from predators , say , in th e hear t o f a cit y o r a swamp ; o r perhap s on e tha t i s warmer than its surroundings, such as an urban are a or a warm valley; or perhap s a spo t sheltere d fro m th e wind . Bu t warm , protected place s ar e not scarce , and one need no t trave l grea t distance s t o fin d them . Som e hav e suggeste d tha t th e communal roos t itsel f i s a defens e agains t predators; bu t fo r tha t purpos e a few hundred bird s woul d b e enough . Why , then , d o birds gathe r in such hug e numbers? Communal roosts ca n be seen during the da y as well. Ward, 2 who studied th e huge midday roosts of black-faced diochs , suggeste d tha t they enabled individual s who ha d no t foun d enoug h t o ea t that mornin g t o joi n a flock that ha d locate d food. Suc h a theory was first proposed b y Darling3 to explain the value of breeding colonies: h e suggeste d tha t gull s tak e thei r cue s fro m thos e neighbor s the y se e returning to the colon y with food; when th e successfu l providers go out agai n to forage, the observer s can then follo w them to their sources . When w e firs t tackle d th e issu e wit h Ward, 4 w e bega n b y considerin g th e species who congregate in large roosts. We found that those who do tend to search for foo d tha t appear s i n short-ter m concentrations , suc h a s school s o f fis h o r swarms of insects, or in unexpected place s like freshly plowed fields or fish ponds that ar e being drained . Fo r individual s o f such species, i t is sensible to mak e use of th e knowledg e o f others . Th e foo d o f specie s wh o ten d t o roos t alon e o r i n

Information Centers 1 9

9

small groups , o n th e othe r hand , i s generally spread sparsel y over larg e area s in quantities tha t d o no t suffic e fo r mor e tha n a few. For suc h species, ther e i s no point i n following others to a food source, and thus no point in communal roosts.5 Since w e firs t publishe d th e ide a tha t communa l roosts serv e a s "informatio n centers " i n 1973 , man y studies an d experiment s tha t suppor t i t hav e bee n done.6 Th e theor y is now widely accepted. Brow n ob served swallows coming to their nests in nesting colonies with food in their beaks , and othe r swallow s followin g them when the y went back t o get more food. 7 Heinric h studied crow s i n th e northeaster n Unite d States. 8 I n winter, thes e crow s fee d o n carcasse s they find on to p of the snow . When Heinric h pu t suc h carcasses in particular places , h e foun d tha t th e mornin g afte r a singl e cro w discovere d them , many crow s cam e to fee d o n them . H e note d tha t thes e bird s flew straight t o a carcass they had no t know n o f previously; most of them cam e directly from thei r communal roost, following the roostmat e who had foun d it the da y before. In our original article on roosts and gatherings, we suggested that all gatherings of animal s evolved t o provid e individual s with informatio n abou t foo d sources . Today it is clear that som e gatherings provide other benefits.9 Animals also gather to find mates: the communa l leks of ruffs an d o f black grous e that were describe d in chapte r 3 hol d hundreds , sometime s eve n thousands , o f birds , an d th e larg e concentration o f males enables female s to choose the best of them. Wagner found that razorbills— a monogamou s bir d specie s i n whic h bot h mate s tak e car e of offspring—gather i n special mating arenas to copulat e both with their mate s and with other s (extra-pai r copulation). It woul d mak e sense that gatherin g togethe r in breedin g colonie s an d i n communa l roosts, whether durin g th e da y or fo r th e night, also gives birds the opportunity to get to know potential mates and thus to choose thei r mate s better , thoug h ther e i s no proo f o f tha t yet . Som e breedin g colonies primaril y offer protectio n fro m predators , an d ye t other gatherings, usually small ones, enable individuals to pool their bod y warmth on cold nights. 10 There ar e man y additional example s o f information centers . Sandgrouse , for instance, ar e bird s o f th e stepp e an d desert ; they ar e abou t th e siz e o f pigeons , and the y ea t grain s an d greenery , which ar e distribute d ver y unevenl y i n thei r desert habitat . I n th e summe r they fly large distances t o source s o f water. Som e species drin k in the morning, about a n hour after sunrise , and others drink in the evening. The sandgrouse come to the water sources in couples or small flocks; once there, however , the y gather int o large flocks and spend tim e together chattering , chasing each other, courting, an d preening thei r feathers. They are very cryptic in color, an d th e flock looks lik e a field of stones—which suddenl y turn int o bird s when they rise and fly together t o the water. When th e sandgrouse rise up in flight they are visible an d audible a t a great distance. No doub t these gatherings, which can be seen in several places in Israel's Negev desert, serve as information centers.11

2OO TH

E H A N D I C A P P R I N C I P L E I N SOCIA L SYSTEM S

INSURANCE AGAINST EVI L DAYS: WINTER GATHERINGS OF ROOKS Studies of rooks at Aberdeen, Scotland , seemed a t first to contradict the theory that communal night roosts are information centers. Researchers at Aberdeen University have been observing local rooks for decades. Rooks are crowlike birds that nest i n communa l rookeries , fro m whic h the y dispers e t o see k food . I n winter , rooks fro m severa l rookeries congregat e togethe r an d for m communa l roost s in marshes. In the morning, each flock flies to the trees in which it nests in the spring, stays there for one or two hours, then roams around seeking food. In the evening the rook s retur n ove r lon g distance s t o th e communa l night roost . Even thoug h they spen d th e nigh t together , member s o f differen t rookerie s wer e neve r see n feeding i n eac h other' s areas . Moreover , th e foo d rook s ea t i s more plentifu l in winter tha n i n summer , but i n the summe r the differen t rookerie s d o no t gathe r in the communal night roost . I suggeste d t o Feare , wh o was then a student a t the researc h station a t Aberdeen, that when unusually severe weather strikes, the rooks might lose their regular food sources ; if that shoul d happen , informatio n provided at th e centra l roos t b y members o f other rookerie s migh t prove crucial. As it turned out , the next winter was exceptionally harsh. In the past, researchers had not gone out to watch rook s i n foggy weather . Thi s time , Fear e did—an d he was rewarded with a sight never before seen by members of the Aberdeen researc h station: a flock of rooks from on e rookery joinin g a flock fro m anothe r rooker y i n th e day time. They al l flew together t o a farm where a pile of corn jutted out of the snow, and there they fed. 12 The snow and fog continue d fo r severa l days. The usua l food sources exploited b y the rook s were covere d wit h snow, and the fog made i t har d fo r the m t o fin d ne w ones. The y migh t hav e starve d t o deat h ha d they not followe d th e members of other rookeries , who knew where to find food. Because they made the daily investment in the long-distance flight to the communal night roost, the rook s had insuranc e against emergencies.

FLOCKS AND LONERS : THE COMMUNAL ROOS T O F KITES IN GOTO DONANA An individual ca n get other important informatio n about foo d by joining a night roost. Heredia, Alonso , and Hiraldo13 studied red kites in Spain. In their researc h area, Goto Donana, the kite s feed o n chicken carcasses. A carcass, once kite s dis-

Information Centers 2O

1

cover it , doesn't las t long, and they are widely scattered; by the tim e kites gathe r in the night roost, carcasses found that day have already been eaten. Even so, most kites leave their communa l night roost together an d fly over the marshes and sand dunes in a scattered flock. Once a kite finds a carcass, the other s joi n it and fight over it . Interestingly enough , the flock the researchers observe d wa s composed mostl y of migratory , wintering kites . Local kites , wh o kne w th e are a well, di d no t a s a rule come to the roost an d did not join the flock but rathe r went out on their own, usually in their regula r territories. Th e researcher s also found that kite s tha t had fed well the previous da y also preferred to strike out on their own, apart from th e flock. A group o f kites can find an isolated chicke n carcas s more easily than a single individual can . But once a carcass is discovered, its finder has to fight for it. Very hungry kites, for whom eve n a mouthful would mak e a difference, d o no t hav e a choice: the y g o with th e flock , whic h i s likelier t o brin g the m t o a food sourc e sooner, eve n a t the pric e o f having to struggl e at that point with their flockmates to ea t some of it. On th e other hand, a kite who knows the area well or who is well fed can affor d to search longer an d then ea t at its leisure, far from othe r kites. Only if it does not find food ove r a stretch o f days would i t benefit from joinin g the flock. The com munal roost, then, enable s kites to search fo r food in a group when the y need to ; in addition, by checking the number of other kites who join the flocks, an individual ca n estimate ho w abundan t foo d i s i n th e genera l area . I t ca n the n decid e whether to stay or move on to another region.

BRIGHT ADULTS AND DUL L YOUNGSTERS: HANDICAPS IN FOOD SQUABBLES Gulls als o searc h fo r foo d i n a scattered flock , an d lik e th e re d kites , the y ofte n fight ove r what the y find . A bright whit e gul l diving a t a piece o f food i s visible from a distance and capture s th e attentio n of other gulls , who know that a diving gull means a meal. But the summon s may well end i n a squabble, an d ther e is no certainty that the one who discovered th e food will get to eat it. A gull must both find food an d defend it . The gull's bright colo r help s other gull s but i s clearly a burden to its bearer: if it were not s o visible from s o far away when i t dived, the gul l might enjoy its prize in peace. An d indeed, onl y adult gulls are so conspicuous. Young ones' backs are a dull gray-brown, probably for camouflage, and their underparts are grayish rather than brilliant white . I f a particular gull grew to adulthood withou t losin g the dull , cryptic colors o f its youth, then while it could trac k its bright whit e fellows to see

2O2 TH

E HANDICA P PRINCIPL E I N SOCIA L SYSTEM S

where the y found food , it woul d no t disclos e it s ow n resources . On e coul d say, then, that adult gulls are altruists, and the young are selfish. But we say, of course, that the individual who bears the brilliant white feathers must benefit b y doing so or the trait could no t have evolved. What i s th e poin t o f th e adul t gull' s brigh t plumage ? Again , the Handica p Principle provides a n answer. Gulls nest in colonies ; this means that they have to b e abl e t o find food in th e presenc e o f other gull s an d t o defen d i t successfully . A gull looking for a mate has to make sure that the mate it chooses is able to stand up for itself one who i s bright i n colo r show s of f the fac t tha t i t ha s managed t o compet e successfull y wit h other s wh o can se e its every move clearly; one who has a youngster's crypti c colorin g doe s no t provid e tha t proof , and indeed unti l it loses its dull plumage, a gull is far less likely to fin d a mate and rais e a family. The same principl e applie s t o certai n other bir d specie s tha t fee d i n loose flocks . Egyptian vultures, for example, gather t o feed o n carcasse s and o n piles of garbage; the adult s have striking col ors—black an d whit e i n thi s case—whil e youn g individual s ar e a dull , crypti c brown.

HOW INFORMATIO N REPOSITORIES WORK Wagtails in a night roos t tha t draw s bird s fro m a distance o f some fifteen mile s pool together informatio n about th e availabilit y of food in a n area exceeding 400 square miles. I n a night roos t o f starlings, which draw s birds fro m u p t o seventy miles away , informatio n abou t foo d i n a n expans e o f mor e tha n 12,00 0 squar e miles—a huge store o f information, by any measure—is available. It is quite likely that the birds who come to such a roost learn more than simply where to find food th e next morning. Fo r example , migrator y birds mus t decid e when to move on. The number of birds who need help locating food in the morning is an indication o f how availabl e food i s in the area . A starling that arrives a t th e night roos t hungr y probably wait s a while th e nex t mornin g an d watches other s going off to their source s of food. When resource s ge t thin, one would expect th e number o f morning loiterers t o rise, and that very increase can tell other starlings that it is time to move on to a new region. Once a starling notices that th e total number of birds coming to a night roos t is down , i t woul d mak e sense fo r i t t o joi n one o f the flock s migratin g t o othe r regions. Even if the individua l bird stil l knows of good foo d sources, it will suffe r

Information Centers 2O

3

if they dry up later; finding new ones will be difficul t o n its own or in a small flock. In unfamiliar places, large flocks that cover wide areas can find food more easily— especially i f some members of the flock know the are a from years past. Bucher recentl y suggested that th e famou s passenger pigeon (Ectopistes migmtorius) becam e extinc t partl y because it s traditiona l roost s wer e disrupte d b y hunting and development, and the reduced populations could not gather the information the y needed t o fin d th e mas t crops o f breech trees , th e scattere d an d unpredictable foo d supply on which they depended.14 Wynne-Edwards describe d th e gatherin g o f bird s i n communa l night roost s and suggested that such roosts evolved in order to enable a population of birds to spread itsel f over a n are a according to th e distributio n o f food i n it. We d o no t accept Wynne-Edwards's explanatio n o f the evolutio n o f communal night roosts, but i t should be noted tha t the collective information in these centers does indee d indirectly help adjus t th e distributio n of birds accordin g to food distribution. 15

HUMAN GATHERINGS Human gatherings , too, serv e a s information centers . Immanue l Marx followe d festive gatherings (zaharas) at saints' tombs among Bedouins in the Sinai Desert.16 He foun d tha t whe n th e politica l situatio n was stable, fe w came to suc h gatherings; when condition s were volatile an d th e futur e uncertain , the numbe r o f participants ros e markedly . Whether o r no t i t i s the mai n reaso n th e festivitie s are held, i t i s clear that on e functio n o f the gatherin g is to provid e participant s with information. Amotz himsel f foun d ou t ho w beneficia l communal praye r ca n be . H e wa s called u p fo r arm y reserve duty. None o f the soldier s in his unit kne w what they had bee n summone d for , where the y were going , what the y would d o there , o r when the y would b e released . Th e uni t was taken in the dar k of night to a n unknown location , an d the soldier s went to sleep . I n the morning, the pious among them ros e earl y an d wen t t o prayer , whil e th e other s staye d i n bed fo r anothe r precious half-hour. Amotz wok e u p amon g the latter , who wer e al l as clueless a s they had bee n the previou s night. But the guy s coming bac k from morning praye r knew everything. The congregatio n had include d me n from othe r unit s a s well—military intelligence, transportation , operations , an d s o on—an d chat s befor e an d afte r prayers lef t th e devou t full y informe d abou t th e unit' s prospects . Th e nee d fo r knowledge i s on e o f th e majo r reason s peopl e gathe r i n clubs , pubs , an d othe r places17—even though what gets a person out o f his or her house may be the wish to go to a sporting event, watch birds, have a drink, or fulfill religiou s obligations . Veblen, who lived at the end of the nineteenth century , explained why printing

2O4 TH

E HANDICA P P R I N C I P L E I N SOCIA L SYSTEM S

press workers were particularly kee n on drinking contests at that time: they had a specialized skil l that with the rapi d expansion of publishing was in great demand, and thus they could move from on e place to another an d work for whoever would pay th e most . Thes e wanderin g workers made transient connection s wit h peopl e they ha d no t know n before , wh o wer e thei r professiona l peer s an d occasiona l competitors. B y drinking together, the y coul d ge t to kno w eac h othe r an d brag about thei r abilitie s to their ne w friends. Thus , bars enabled eac h to establish his social an d professiona l statu s and probably als o served as information centers for those seekin g work.18

COMMUNAL DISPLAY S AT GATHERINGS: PROMOTING TH E ROOST O R MUTUAL TESTING? One o f the typica l features o f a communal nigh t roos t i s the attention-grabbin g behavior o f its participants: sparrows an d starling s sen d u p a loud choru s from the tree they have chosen to roost in; egrets and wagtails stand on high, prominen t perches. Breathtakin g aeria l maneuvers are als o characteristi c of suc h congrega tions: flock s o f starlings , wagtails, or swallow s wh o hav e gathere d togethe r ris e suddenly an d tightl y circl e th e roost , makin g sudden, shar p turn s an d climbin g and plunging abruptl y before descending agai n into th e tree. Wynne-Edwards suggested that the purpose o f such showing off is to advertise the site of the roost, to the benefit of all: the more bird s at a site, the more information availabl e there.19 But this, again , is a group-selection argument . It may be best fo r th e grou p i f the sit e i s advertised , bu t tha t doe s no t explai n wh y each individual should inves t in the advertising, since those who do not take the trouble still benefi t fro m th e bod y o f knowledge assemble d throug h th e effort s o f thos e who do.20 Another theor y is that flocks perform flight displays t o confuse raptors, and indee d flocks approaching communa l roosts tend t o reac t to raptors by executing sharp turns. But the aerial maneuvers above communal roosts are too many, too lengthy, and too elaborat e for this to be a convincing explanation. Perhaps eac h bir d participate s i n suc h maneuvers to compar e its own ability with that of the flock.21 It is reasonable to assume that some individuals ar e more skilled in flight than others. Ideally, the individuals in a flock will be of comparable ability, so that, when necessary, they can fly in harmony. If one is significantly less skilled tha n the res t of the flock, then when ther e is danger, one may find oneself alone, exposed t o raptors, 22 or may lag behind in the search for food. On the other hand, i f an individual's abilit y i s much greater than tha t o f others i n the flock, it may miss out o n ric h sources o f food tha t weaker birds might not reach . Perhaps by testing itself against others i n aerial maneuvers, each individual decide s which flock it should joi n the nex t day.

Information Centers

205

It i s difficul t fo r researcher s t o follo w individua l behavior i n a huge congre gation o f birds. Onl y whe n technolog y enable s u s to mar k many individuals an d track them closel y within suc h a large gathering will we begin t o understand th e mechanics o f flocks better. Still, we believe tha t individuals in flocks and in roosts are aware of each other's doings , and that many of them know and remember one another. Althoug h w e don' t kno w exactl y how individual s benefi t fro m calling , performing aeria l maneuvers, and s o on, we do think it possible tha t showin g off "in public " increase s an individual's prestig e in flocks and communa l roosts a s it does in other interactions among animals . A s always, we are guided by the logic of individual selection—even when seeking to understand remarkable phenomena of collectiv e social activit y like roosts and flocks .

This page intentionally left blank

P A R T I

V

HUMANS

This page intentionally left blank

C H A P T E R1

8

HUMANS

H

uman social life, like that of all other organisms, reflects the interaction of cooperation an d competition among collaborators. We don't mean to suggest that human social systems are not vastl y more complex tha n those of animals. Still, we believe that the same principles guide both; the behavioral mechanisms tha t surviv e ove r generation s ar e thos e tha t increas e th e numbe r o f th e individual's childbearin g offspring. Thus, we will be guided by the Handicap Principle as we examine the logic behind mechanisms of social behavior an d method s of communication amon g humans. Some people objec t when human behavior i s compared to that of animals. Yet we do this routinel y in human physiology. Research on the heart an d circulator y systems, the kidneys, and the immunological systems of animals has taught science a great dea l about th e functioning o f the human body. Why, then, should we not seek the sam e kinds of insights when we consider bod y parts and traits that serve social purposes?

2O9

21O HUMAN

S

INNATE BEHAVIOR IN HUMAN S One exampl e is the tabo o agains t incest. I n mos t huma n societies , marriag e between clos e kin—between parent an d child , o r brother and sister—is avoide d o r forbidden. Thes e rule s are usually considered a part of the local moral or religious code. But whatever the ostensible reason for them, these taboos prevent inbreedin g that might cause genetic damage to offspring. Animal s too commonly avoid incest. It is interesting to see the mechanisms that serve to prevent such behavior, both among human s an d amon g animals . Babblers , fo r example , avoi d matin g wit h individuals that they helped rea r or ones that were present i n the group when they were fledglings . This i s not a perfect mechanism, sinc e it ca n preven t a babble r from matin g with on e who i s not reall y its relative , bu t wh o happene d t o b e its groupmate whe n i t gre w up . I t turn s ou t tha t unrelate d person s wh o gre w u p together fro m infanc y in the same communal children's house on an Israeli kibbutz seem to avoid marrying one another, althoug h they are perfectly free to do so. 1 In more traditional settings , growing up so close together almos t always meant growing u p i n th e sam e extended geneti c family . I t seem s tha t amon g humans , too , there is an unconscious mechanism that prevents close relatives from mating; legal and religiou s prohibition s only give expression t o an d reinforc e inbor n behavio r that evolved throug h natural selection . Another unconsciou s mechanis m operate s whe n w e siz e up strangers . Hes s conducte d a n experiment wit h two identical photographi c portraits: 2 h e retouche d th e photograph s to make the subject's pupils a tiny bit larger in one and a tiny bit smalle r i n the other . H e then aske d peopl e wh o did no t know what he had don e t o describe th e person i n each portrait. When show n the portrait wit h the dilated pupils , they consistently describe d the subjec t a s a nice, pleasan t person ; whe n show n th e othe r portrait , the y said the subject looked vicious, o r dangerous. This clearly demonstrates ou r faculty to collect an d evaluat e information without bein g awar e either o f the proces s o r of the reasons for our responses . Why should pupil size make a difference? A contracted pupil , like a contracted aperture in a camera, increases the sharpnes s of the image. A person wh o intend s to attack needs to see the opponent ver y clearly and to keep the opponent's image in shar p focu s a s they move toward o r awa y from eac h other—this i s what pho tographers cal l a "high dept h o f field. " A person wit h n o intentio n o f attacking can affor d t o se e less clearly and t o let another' s imag e become "softly " focused ; to suc h a person, th e handica p involve d i s trivial, while a n attacker would find it unaffordable. On e ca n assum e that a person wit h dilate d pupil s i s not goin g t o attack. The same study showed that when people ar e at ease with individuals they love, thei r pupil s dilate . Thus , unknowingly , we provide onlooker s wit h reliabl e information abou t ou r emotions an d intentions .

Humans 21

1

We ar e all familiar wit h th e difficult y o f reading o r driving with dilate d pupil s following a n ey e checkup. Th e ophthalmologis t use s belladonn a drop s t o dilat e the pupils . I n Italia n bella donna mean s "beautifu l lady" ; wome n use d t o appl y belladonna drop s to their eyes before going to a ball, in order to make a doe-eyed impression an d arous e th e affection s o f suitors . The y didn' t consciousl y realiz e what it was that made that impression "count " in the eye s of their beholders. Th e handicap they took on—thei r dilated pupils—mad e it difficult t o see their suitors very clearly, or evaluate their intentions well.

THE HUMAN BODY AND IT S DECORATION S Hair Like most animals , humans have body parts that serve as signals. The most prominent o f them i s the hai r o n th e scalp . Human s ar e the onl y animals whose hair grows beyond a certain finite length; thoug h ther e ar e som e human population s with shor t hair , the hai r o f some peopl e ma y even reac h th e lengt h o f the bod y itself. What messag e can be expressed by long hair? What handica p does it entail, and why is it unique to humans? Two of the important traits that distinguish huma n from ap e are a higher level of intelligence and hands capable of fine manipulation. Long hair emphasizes both these traits. Without dexterou s hand s to tie, braid, o r cut one's hair, and without the understanding that long, unkempt hair impairs one's vision and movement and should b e take n car e of, one coul d no t surviv e with constantl y growing hair. N o other animal has the mental and manual dexterit y to make long hair into an asset rather tha n a liability. Ancient human s probably live d i n famil y groups , eac h with a territory that it defended agains t neighbors. As a result, humans often ha d to assess potential rivals from afar . Eve n mor e important , sinc e marriag e be tween clos e kin could b e detrimenta l to the offspring , people had to find mates in other, rival family groups— and to evaluate potential mate s from a distance, without revealin g themselve s prematurely . Hai r ca n pro vide considerabl e information . A well-kep t an d well-dressed hea d o f hai r testifie s t o th e amoun t o f time its owner ca n affor d t o spen d takin g care of it; it can proclaim, both at close range and at a distance, the patience, skill, and imagination of its bearer. The hair's condition an d luste r als o bear witnes s t o it s owner' s physiological state and general health.

212 HUMAN

S

Eyes, Eyebrows, and Eyelashes Eyes, eyebrows, and eyelashes convey a wealth of reliable information in the course of social relations . W e are so used to watching th e eyes of people we interact wit h that i t makes us feel uneasy , even threatened, t o talk with someon e wearin g sunglasses; an d indeed , wearin g sunglasse s indoors , wher e th e eye s d o no t requir e protection fro m th e sun , is considere d unpleasan t an d impolite . Th e siz e of th e pupil, a s we have seen, convey s information abou t friendl y or hostile intent. Th e colored iri s and whit e eyebal l mak e the movemen t o f the ey e easy to follo w an d reveal the direction o f a person's gaze , and thus the object of interest; the eyebrows accentuate the directio n o f gaze, making it easy to discer n fro m farthe r away. Eibel-Eibelsfeldt3 observed that , in most human societies, raised eyebrows convey friendly greeting, an d puckere d eyebrow s a threat. Puckere d eyebrows , lik e contracted irises , indicate th e sharpl y focused stare of somebody gathering infor mation—quite possibly with ill intent. Raising the eyebrows involve s some loss of focus—try t o rea d thes e lines with raise d eyebrow s an d see . This cos t i s small t o an individual wh o is honestly friendl y but heav y to a cheater who wishes to signa l friendliness while intending to do physical harm; the signal is therefore difficul t t o fake. Again, the signal , thoug h reliable , is often unconscious. Men's eyebrow s ar e usually thicker tha n women's , an d i n both gender s the y get heavier with age . What is the purpose of thicker, heavie r eyebrows ? I n huma n society, a s i n tha t o f man y animals , the olde r adult s ar e i n charge . A dominan t person—just lik e a dominant animal—benefits from displayin g intention s clearly ; this way, others ca n acced e t o th e dominant' s wishe s an d avoi d conflict . A subordinate, on the other hand, has something to lose by disclosing intent too clearly , because a dominant ma y intervene an d prevent him or her from carryin g out tha t intent. Subordinates , then , benefi t fro m probin g their way carefully an d findin g the path of least resistance . This is probably th e reason why younger persons ten d to have fine, inconspicuous eyebrows : such eyebrows stil l help display thei r bear ers' intent t o nearby observers, but the y draw less attention t o it. Lowered eyelid s an d blinking als o serve as signals. Lowerin g one's eyelid s mean s givin g up th e collectio n o f visual informa tion—which a n individual wh o is poised to attac k or to moun t a vigorous defense cannot affor d t o do . A brief glanc e followed by lowere d eyelid s therefor e signal s nonthreatenin g interest , a well-known come-hithe r look. Paintin g th e eyelids increases the distance from whic h others ca n see their movement. Blue is best for thi s purpose : i t i s differen t enoug h no t t o blen d wit h th e eyeballs o r the res t o f the face , an d reflectiv e enough t o convey the fine r detail s o f movement.4

Humans 2 1

3

Nose and Facial Wrinkles Noses an d facia l wrinkle s also convey emotions an d intentions , whether friendl y or not . Th e longe r th e nos e an d th e deepe r th e wrinkles , th e easie r i t i s for a n observer to deduce intentions. A wrinkled face with prominent features is considered "expressive" an d "communicative, " mor e so than the smooth face and short and rounded nos e of a baby or a child. An adult with a smooth face that expresses few emotion s i s said to b e "baby-faced"—no t a wholly complimentary term. The famous actres s Sophi a Lore n rejecte d i n a n intervie w th e ide a o f cosmeticall y stretching her facial skin to eliminate wrinkles; she would sacrifice her expressiveness, she said. A nose makes it easier to tell what directio n the perso n i s looking in. Indeed, busybodies are said to "stic k thei r noses " into others' affairs . Lik e eyebrows, the human nose gets larger with age , probably for th e sam e reason—to enable more dominant individual s t o displa y thei r inten t mor e clearly. Maker s of mask s are well awar e of the effec t of noses: witches and evil persons are given long noses, while clowns , who ar e harmless , ar e give n rounde d ones. The rounded nose does not point directly at any particular perso n an d enable s th e clow n t o jok e an d ridicule the audience without insulting anyone in particular. We do not believe that the makers of masks in various culture s ar e awar e of the reason s behind thi s effect; the y simpl y know, fro m traditio n an d experience , ho w bes t t o conve y a particular character.

Chin and Beard Another huma n body part tha t is clearly a signal is the beard . Someon e showing defiance ofte n raise s his o r he r chin , exposing i t to blows , henc e the expressio n "leading with one's chin"—that is, making oneself defiantly vulnerable. A bearded man who doe s thi s is thrusting his beard eve n close r to his opponent. A s we saw in chapte r 2, a beard can make a man vulnerable in a fight; the Bibl e itself tell us this, i n th e stor y of Joab, wh o grabbe d Amas a b y his beard an d kille d hi m with his sword. 5 Growing a beard, then, is a way of showing off male confidence. This also explain s why women d o no t gro w beards: neither now nor earlie r in evolutionary history have women as a rule bragged about their ability as fighters. Handicapping themselves with beards would not serve to advertise any important female trait. The bear d convey s other message s as well. It combine s with the sideburn s t o frame th e fac e like a mane, reducing the face' s apparen t size. 6 A bushy beard also reduces lateral vision. In older men, graying beards often develop patterns of color

214 HUMAN

S

that, lik e wrinkles , increas e th e expressivenes s o f th e face . Th e wa y a beard i s groomed provide s informatio n about it s owner's skill s and preferences . And lik e scalp hair, the beard's colo r an d natural luster provide informatio n on age and on physiological condition .

Red Cheeks and Lips The re d colo r o f lip s an d th e pin k o f ros y cheeks com e fro m periphera l bloo d circulation clos e t o th e skin . Especiall y i n col d environments , bloo d vessel s i n exposed skin caus e heat loss , which waste s energy . Indeed , when peopl e are sick or sufferin g fro m exposur e t o cold , thes e bloo d vessel s contract , an d lip s an d cheeks lose their red color; in extreme cases they turn blue. I n cold climates , such as in Scandinavia and Russia, therefore, red lips an d cheek s are a reliable signal of health an d ar e considered beautiful : person s wh o ca n afford suc h "wasteful" display prove tha t they are healthy and vigorous. Morris suggeste d tha t re d lips remin d men of women's nethe r labia , which in contemporary human societies ar e concealed b y clothing. But this does not explain why the lips o f young children ar e red, or why cheeks ar e red. 7 What abou t lipstick ? Ar e women who wear lipstick lying about thei r health ? We don' t thin k so . People do no t tr y to hide th e us e o f lipstick, an d i t i s eas y t o distinguis h i t fro m the natural color o f lips. Th e lipstick enhance s anothe r func tion of the color of lips: it brings out the shape of the lips and the fin e detail s o f facia l expressio n an d make s both clearl y visible from farthe r away . This is why stage actors, who have to project their emotion s to the audience , use heavy makeup and stron g lines of color t o heighten thei r facial expressions . This i s not a message in itself , bu t a mechanism that enable s one to transmit facial message s more clearly and ove r greater distances.

Menstruation Body processes , too , ca n be signals ; a case in point i s menstruation. Women ar e unique among mammal s i n the amoun t o f blood an d body tissue they discharg e every mont h i n thei r menstrua l flow . Thi s i s no t require d fo r fertility , fo r mos t mammals conceive without it. What, then, is its purpose ? Menstruation i s a reliable indicato r o f a woman's physical condition. Whe n a woman is sick, either bodily or mentally, or when she expends great physical effort, as in sports competitions, he r menstruation may cease or become disordered. Preg nancy stops menstruation. Menstruation thus informs a woman's mate or potentia l

Humans 2 1

5

mate tha t sh e i s in goo d enoug h shap e t o bea r children—a s witnesse d b y th e amount of blood and tissues she can afford t o "waste" monthly on menstruation— and als o that sh e is not pregnant a t that time, which is very important to one who wants to ensure that his chosen mate's child will be his own.

Breasts and Body Fat Women's breasts ar e primaril y signals. Most o f the breas t is fatty tissu e that ha s nothing to do with feedin g babies; the babies of most other mammals suckle suc cessfully fro m nipple s o n almos t fla t mammar y glands. Larg e breast s hampe r a woman's freedom of movement an d wast e energy by increasing heat loss . What , then, are they good for ? Large, heav y breast s sho w clearl y tha t thei r beare r di d no t lac k fo r foo d when the y developed : fa t concentrate d i n breast s i s easie r t o gaug e tha n fa t spread al l ove r th e body . O f course , suc h showin g of f i s meaningfu l onl y i n societies tha t experienc e foo d shortages: 8 thu s i n western society , where foo d is not scarce , a big , plum p woma n wit h heav y buttocks an d larg e breast s i s n o longer th e idea l o f female beauty. Breasts also underg o periodi c change s in size and textur e an d provid e informatio n abou t hormona l state s an d abou t pas t childbearing. Body fat, both in males and in females, signifies affluence even today i n man y societie s aroun d th e world . Fa t show s tha t it s bearer is a successful person, who has had food in plenty without having t o wor k har d fo r i t physically . In developin g countries , this ca n b e importan t information . I n societie s wher e foo d i s comparatively cheap and plentiful, though, like much of modernday Western Europ e o r th e Unite d States , body fat has lost it s value as a signal; it is now considered a drawback. Its opposite , a trim body, shows off in these societies that its owner has enough self-control to eat a sensible diet and to exercise .

Clothing We sa w in chapte r 4 how animals ' marking s accentuate bod y part s or trait s that are important t o othe r members o f their species . Th e sam e can be sai d of much human decoration . Clothing , especiall y when no t neede d fo r protectio n agains t cold, wind, o r sun , is used a s decoration an d fo r showin g off. Like the markings of animals , style s o f clothing evolve whe n member s of a set compete to show off their quality in areas important to that group. The styles that evolve emphasize the points of beauty, proper conduct, and economics that are important to the set and enable observers to compare one member o f the set with another. People who do

216 H U M A N

S

not bothe r t o dres s accordin g t o th e cod e o f their se t ar e als o showin g off : the message may be that they do not belong to the group and do not care to be judged by it s standards , o r tha t thei r strength s can b e effectivel y displaye d eve n i f they don't conform to a dress code. Ever sinc e propert y becam e a n importan t criterio n o f socia l success , clothe s have serve d t o sho w of f wealth—whether the y ar e mad e o f expensiv e materials or requir e a lot of skilled work. Lately, modern productio n technique s hav e lowered th e cos t o f clothing t o suc h a n exten t tha t i t is difficul t fo r most t o tel l th e difference betwee n expensiv e clothing and imitations. Designer label s are no help, since they are absurdly easy to imitate. It seem s that now, more than a t any other time i n hundred s o f years , clothes ar e becomin g les s importan t i n showin g off. Instead, peopl e show of f their bodies , wit h shor t clothing , cutoffs , an d revealin g necklines. Sometimes standard s in clothin g ar e imposed fro m above , a s is the cas e with army or school uniforms. Of course, such prescribed clothin g does not reflec t th e personal tast e o r wealth o f its wearers, but i t enable s superior s an d commander s to asses s differences i n personality traits that are important to them, such as neatness an d precision . Thes e trait s are reflected by the wa y a uniform is maintained and worn, an d the y are easier to evaluat e precisely because th e individuals bein g compared al l wear the same uniform. For th e same reason, the first few months at the Israeli ai r force's school of aviation are devoted t o drills, polishing equipment , and precis e maintenanc e o f one' s uniform . These activities , which ma y seem t o have nothin g to do wit h flight , in fac t test eac h would-b e pilo t wit h regar d to personality trait s that ar e critical t o hi s or her commanders—orderliness , obedience, and precis e executio n o f commands . Clothing an d othe r decoratio n modifie s or enhances signal s that are conveyed by specifi c part s o f the huma n body . Th e neck , fo r example , support s th e hea d and has to bear its weight. The longer the neck—provided it is also strong enough to support th e head properly—the more one is able to move the head aroun d and the bette r us e one ca n make of sensory apparatus in the head . Me n have neede d a stron g nec k fo r fighting , however , an d therefor e coul d no t affor d a long, thi n neck. But women ar e not usuall y fighters and ca n afford t o possess and sho w off a longer neck. Women, especiall y young ones, often decorat e their necks with close-fittin g necklace s o r ribbons . A lin e acros s a lon g structure makes it appear shorte r than it really is, so it is precisely a long nec k tha t on e ca n advertis e reliabl y b y tyin g a ribbon aroun d it ; th e perso n wh o wear s the ribbo n i s proclaiming that , despit e th e handica p sh e imposes o n herself , her nec k ca n be see n to be longer tha n those of her compet itors. A girl with a short neck canno t afford t o wear a ribbon; such decoratio n woul d mak e he r appea r ridiculousl y short-necked.

Humans 2 1

7

A slende r wais t ca n sho w tha t a youn g woma n ha s probabl y no t ye t gon e through pregnancy ; slim waists and tri m ankle s are considere d beautifu l and attractive in women . Indeed , i t i s difficul t fo r suc h slende r structure s t o bea r th e weight of the whole body—and that may be precisely why their slimnes s shows to advantage the high quality of die woman's body tissues and structures. Both waist and ankles thicken durin g pregnancy when a woman carries a child, to enable her to support an d balance the added weight . For men , broa d shoulder s ar e a n advantage : the broade r th e shoulders , th e greater th e forc e availabl e t o th e arms , other thing s bein g equal . Lon g necktie s handicap the display of the dimensions of men's shoulders: the line formed by the tie, whic h i s perpendicular t o th e lin e o f the shoulders , decrease s th e apparen t breadth o f the shoulders—jus t lik e th e blac k ban d o n a great tit' s breast, a s we saw i n chapte r 4 . A ma n wit h broa d shoulder s wil l appea r impressiv e despit e wearing a tie, and eve n one with unimpressive shoulder s ca n use a tie to show off how much better h e looks tha n one who i s even more frail . Whe n attemptin g t o downplay th e exhibitio n o f manl y strength , me n wea r a bow ti e instead . Goo d examples ar e waiters , classical musicians, an d Europea n male s a t a very forma l social occasion , wher e th e competitio n t o displa y masculinity is intentionally re duced. Some people suppos e tha t one of the functions of clothing is to cover defects, to "cheat " the observer , as it were. However , mos t article s of clothing that might seem a t first to d o thi s tur n ou t t o d o nothin g o f the sort . A short woma n may make herself look taller by wearing high heels, but her taller rival will appear taller still when sh e wears them. Likewise, when al l wear corsets an d lace them a s tight as possible, the difference s betwee n th e thinner- and the thicker-waisted are even more pronounced tha n before. A style of clothing that is actually designed to cover defects wil l no t b e interestin g an d informative , an d i t will no t prevail . A fashio n in dres s is accepted, i t seems, only if it accentuate s a n important characteristic in a reliabl e way.

TESTING THE HUMAN SOCIAL BON D A person's socia l life involve s a complex se t of partnerships an d cooperations . At the foundatio n i s th e famil y partnership , forme d fo r th e raisin g o f children . I n addition t o this , on e collaborate s wit h other s t o procur e food—eve r sinc e th e prehistoric communa l hunt—for protection , fo r economi c gain , for recreationa l and religiou s purposes, and so on. At each level the individual must choose his or her collaborator s an d must test their willingness to invest in the collaboration an d to shar e their assets . As we saw in chapte r 10 , the onl y way to tes t a social bond is to impose on the individual tested. No t surprisingly, this turns out to be the key to som e of the more peculiar behavior s we humans engage in.

218 HUMAN

S

When we happen t o meet a good friend whom we hav e no t see n i n a while , w e ma y approac h from behind , sla p th e perso n o n th e back , an d utter "offensive " expression s like , "Wher e hav e you been hiding so long, you old rascal?" Why are we slapping an d "insulting " a good friend , espe cially afte r a long separation ? Unconsciously , we are testing t o se e whether this person i s still our friend . Only a friend wil l accep t such behavior , tur n t o us , an d gree t u s happily—as on e learn s t o one' s chagri n after behavin g thi s way toward someon e who turns ou t t o be a stranger. Only by imposing on and challenging our collaborators can we acquire reliable information about thei r readiness to collaborate wit h us further . We hu g ou r love d one s tightly, sometime s s o much s o that they have to hol d their breath. Onl y a person wh o love s u s back will agree—will in fact b e glad — to bear such imposition. Indeed , al l our love signals are impositions of one sort or another: kisses , hugs, and pettin g intrud e o n ou r persona l spac e an d impai r ou r freedom o f movement. In spit e of this, we like them: these behaviors demonstrat e that the teste r love s us enough t o test u s and enoug h eve n to invit e us to impos e a similar test on him or her. Even lovers who just hold each other's hand for hours at a time are each giving up th e us e of a hand fo r that time, a pretty heavy imposition. If signal s of love were meant onl y to identif y love , cleve r humankind— Homo sapiens—could easil y establis h conventiona l signal s t o replac e th e repertoir e o f impositions we use now. Bu t one of the most important issues for lovers is reliability—how sincere is the other? It seems that natural selection coul d find no other way to test for this information. Ethologists say that the function o f indications of love is to strengthen the bond between partners . If that wer e true, then on e coul d strengthe n a failing bon d by holding hand s or by reciprocal petting. But when lov e is gone, such forced closeness will probably only hasten the breakup. We think rather that love signals serve to tes t th e bond . Th e investmen t in love signal s is very high, especially when th e relationship i s new and not yet well established, because that is the time when th e information i s most needed, when one can still dissolve an ill-considered partner ship wit h minima l harm, an d whe n th e wron g decisio n ca n be disastrous . Th e burdensome lov e signals either strengthen th e bond or dissolve it; only real lovers accept th e imposition .

The Human Sexual Act as a Test of the Bond Sexual relation s for purposes other tha n procreatio n ar e not uniqu e t o humans . Some other animal s als o copulat e withou t regar d t o th e female' s reproductiv e cycle. Porcupines copulat e several times every night, all year long.9 Many monkeys

Humans 2 1

9

copulate whe n female s ar e no t i n heat . Th e bonob o chimpanzees of the woods of Western Afric a have sexual relation s al l yea r long , i n al l possibl e combina tions—male with female , mal e with male , and female with female. 10 Hoatzins , cooperatin g bird s o f th e American tropics , copulat e durin g border clashes between groups. 11 Babblers , too , sometime s hav e se x outside o f breedin g season . We hav e eve n witnesse d several case s of a male and a female fro m riva l groups meeting and copulating under a bush on the border between their territories, even though ther e was no chanc e thes e Romeos an d Juliets would hav e offspring a s a result o f such trysts. We thin k tha t i n al l these case s copulatio n occur s i n orde r t o tes t th e socia l bond betwee n th e parties . The impositio n involve d i n sexua l relations i s greater than in any other signal of love, and i t increases markedly each party's knowledge of the othe r an d of their relationship. No wonder that animal s who copulat e outside the proces s o f procreation ar e those that maintain long-term partnerships. In reproductiv e terms , a stable pair bond is more important t o a woman than to a man: the numbe r o f children sh e can have is more limited, as is the tim e she has i n which t o bea r them , while a man ha s more chance s to pai r off with othe r women, to the possible detrimen t o f his original mate and her children . Thi s may be why it not uncommonl y takes much more stimulatio n for a woman to becom e sexually satisfied tha n it does fo r a man. Anyone trying to "cure" this tendency is acting against the woman' s unconsciou s mean s of testing the rea l commitment of her mate. What abou t homosexua l relations ? An y trai t tha t i s mor e commo n i n a population tha n a chanc e mutatio n woul d be—an d homosexualit y i s clearl y such a feature—mus t hav e som e benefits, sinc e th e trai t ha s obviousl y survive d natural selection . Thi s i s particularl y s o i f th e trait , lik e homosexuality , seems on it s fac e t o impai r reproduction . Wha t benefit s ca n com e ou t o f homosexua l behavior? As we have seen already , copulation can serve to test the socia l bond betwee n male and female, an d that bond does not necessarily have to do with procreation . For example , some birds, such as stone chats and wagtails, pair up in their winter habitats—not t o breed , bu t rathe r t o defen d wha t i s temporaril y thei r mutua l territory.12 In suc h cases, one often finds that the sam e courtship mechanism that helps tw o individual s for m a partnership i n th e breedin g seaso n serve s them a t other times to for m a partnership solel y for defense. One wh o doesn't know that these partners have no intention of reproducing an d will split up in a few months might think they were going through regula r courtship. 13 Modern Wester n societ y sees homosexuality as an alternativ e to heterosexua l relations. But in some societies, especially ones where men and women ar e strictly segregated, peopl e engag e in two paralle l set s o f sexual relations: thos e betwee n

22O HUMAN

S

men an d women, which produce children ; and thos e between same-se x partners, which hel p sustai n an d enhanc e companionship s o f various sorts . I n bot h cases the sexua l act can serve as a mechanism that lets the partners test the partnershi p and sustai n it. Just a s sex between me n an d women encompasse s more than procreation an d serve s als o to tes t th e bon d betwee n them , se x can provid e infor mation neede d t o maintai n bonds betwee n partner s o f the sam e gender. Again, this is not uniqu e t o humans : Trivers describe s pair s of female wester n gull s behaving a s perfectl y norma l couple s an d togethe r raisin g offsprin g conceive d through extrapai r copulation.14 Of course , the sexual act is not necessarily enjoyable to both participants. Just as consideratio n an d car e for one' s partne r ca n be demonstrate d throug h sex , so can th e lac k o f it . An d tha t lac k o f consideratio n elicit s another messag e in re sponse: one who continues to accept such treatment is telling the other very clearly that, unsatisfactory as the relationshi p is , he or sh e does not wan t to end it . That may b e becaus e o f other perceive d benefit s tha t see m more important tha n care and consideration , or because of a lack of alternatives. Forced sex—rape—i s ofte n use d a s a means o f crud e dominance . Thi s i s as true amon g me n a s between me n an d women . Again , mounting s a s a means of showing dominanc e are no t a t al l uncommon i n th e anima l kingdom, eithe r be tween male s an d female s o r amon g males: this has been show n in numerous species.15 And all human languages, it would seem, have some verb that takes on both the sexua l and the dominativ e meanings of the colloquia l America n English ver b "to scre w (someone)." The sexua l ac t amon g humans show s reliabl y the qualit y o f the relationshi p between those participating in it. Between caring and devoted partners, it expresses that car e and devotion . A n uncaring partner will find it hard to hid e th e lac k of care—which ma y lead th e other member of the pair to leave. The sex act may be used t o prov e th e other' s helplessnes s an d powerlessness , a s in rape . We will let novelists an d poet s detail th e peak s an d abysse s of the huma n sexual act, and all the levels in between. Our poin t here is a global one: that it is precisely the closeness an d invasivenes s of the ac t tha t make it a means of conveying and receivin g detailed an d reliabl e informatio n about th e characte r of its participants an d their relationship with eac h other.

Self-Endangerment in Humans: Suicide as a Cry for Help We hav e als o see n tha t animals—nestlin g birds, fo r example—endange r themselves to extort care. 16 Among humans, too, weaker partners can sometimes extract more from a powerful partner by endangering themselves. A toddler can force its parent t o pick it up an d carr y it by running toward a busy street, or by sitting in a dangerou s place . Som e childre n bea t thei r head s agains t th e wal l unti l thei r parents give in to thei r demands . Youn g childre n wh o clim b ont o furniture an d

Humans 2 2

1

jump at a parent are forcing their parent to catch them. Similarly, women throwing themselves int o thei r lovers ' arm s coul d b e badl y bruise d i f th e lover s di d no t receive them . Al l ar e endangerin g themselves—mos t probabl y unconsciously — and a s a result, they get information the y would no t hav e gotten otherwise ; the y are testin g the socia l bond . The ris k itsel f force s th e othe r part y to com e to th e tester's ai d and thus to express commitment. Suicide i s a n extrem e case . On e coul d eve n redefin e successfu l suicide s a s unsuccessful call s for help. More often tha n not, however, death is not th e result, and the suicid e attemp t cause s friends an d famil y t o give the desperat e risk-taker help tha t was not forthcomin g before. Th e person wh o attempts suicid e may not be awar e of this logic: he o r sh e may honestly prefer deat h t o a n unbearable life . Such a n attemp t a t suicid e i s genuine, an d tha t i n itsel f convince s other s o f th e need t o help. Sometimes the perso n attemptin g suicide may even be half awar e of this logic, a s when someone take s an overdose o f sleeping pills, then call s a friend and says , "I f yo u don' t com e an d sav e me, I' m dead. " Th e ris k i s still genuine , though, an d i t doe s forc e th e frien d t o com e an d help . Th e ris k of death i s what persuades other s o f how desperate the situatio n is—an d indeed, deat h may seem better than the prospec t o f a hopeless an d helpless life .

HUMAN LANGUAGE : COMMUNICATION WITHOUT RELIABILIT Y Human languag e is unique i n th e anima l world i n tha t i t i s a system of communication b y symbols . I t i s possible t o trai n animal s t o understan d an d proces s information passe d b y symbols. Some animals have been taugh t score s of words. Primates an d dolphin s ca n even learn to communicate with people by using symbols, an d birds such a s parrots an d mynas can be trained to produce exac t vocal copies o f human word symbols, and even to use them in a way that makes sense.17 Still, a s we saw in chapte r 6 , there is no anima l that use s a symbolic, word-based language in nature . We believe that animals did not evolve a verbal language because the language of sound s and nonverba l communicatio n serve s them better. Mos t animals live in small groups whose members know each other intimately and spend most of their time together . Suc h individuals shar e the sam e immediate surroundings, witness each other' s actions , an d can discern each other's intentions. 18 For example , whe n a babble r make s a beggin g sound, its comrades do not need to be told tha t it is begging, o r wha t i t i s beggin g for—the y kno w tha t from th e circumstances . Wha t the y can' t tel l i s how badly the beggar wants th e thing it is begging for; the

222 HUMAN

S

intensity of the wish is conveyed reliably by the quality of the beggar's voice. When a babbler threatens , its comrades know without being told who is being threatened, why, and what the threatene r expects the other t o do ; again, this is obvious fro m the circumstances. What the y need t o find out is how reliable and how intense th e threat is. This i s conveyed best an d mos t reliabl y by nonverbal communication . The informatio n tha t nonverba l vocalization s d o conve y i s ver y exact : the y express the degree of feelings much more precisely than words can. 19 For example , the word s I a m angry d o no t conve y how angr y one is ; to conve y the degre e o f anger b y word s alone , on e ha s t o us e mor e words : " I a m ver y angry" ; " I a m somewhat angry. " Even then , words ca n express only a few of the infinit e gradations of anger that are possible; but nonverba l vocalizations reflect such gradations admirably. On th e othe r hand , human beings wh o are not familia r wit h on e another may perceive a give n circumstanc e differently . A perso n listenin g t o a strange r ma y be unabl e t o relat e th e intensit y of vocalization to th e degre e of emotion, a s the stranger's constan t companion s can , from pas t experience . Thi s i s especially true in meeting s betwee n peopl e o f differen t cultures . I n tha t situation , i t i s best t o communicate with words, eve n if the partie s hav e to us e a dictionary to translate from on e language to the other . Verbal language may be a poor, inexact medium for expressin g feeling s an d degree s o f feeling—but i t ca n prevent th e misunderstandings tha t migh t aris e fro m incorrec t interpretatio n o f nonverba l vocaliza tions. We don' t know how symboli c word languag e evolved i n humans. But once it did, it enabled group s of humans who were not eac h other's constan t partners to collaborate temporarily : for protection, fo r war, for hunting. These alliance s were short-term and required cooperatio n amon g partners who did not work togethe r most of the time. Such partners also had to be able to discuss things that were not within view—to say, for example, "There is a saber-toothed tiger on the other side of this hill," or "Yo u com e up the valley, I'll hid e in the canyo n by the spring. " People who ar e accustomed t o working together ac t and communicate differ ently than those who d o not a s a rule work together. A regular crew of movers do not nee d man y words when it comes to carrying a piano up a flight o f stairs. Each of them knows from experienc e where t o grab, how to carry , when t o push, pull, lift higher , or stop . All they need is a signal at the instan t they are to lift th e piec e off th e floor , an d possibl y a few monosyllables a t certain moments alon g the way. A few short grunt s by the leade r may well be sufficient . On th e other hand , if a group of people wh o have not worked togethe r before are calle d upo n t o lif t a heavy item, they will have no choic e but t o discus s each move in detail before attemptin g i t and t o continu e th e discussio n while carrying out the task. Even then, their teamwork will be less perfect than that achieved by the firs t grou p b y means o f a few short grunts . Bu t thi s secon d grou p doe s no t have enough experience to work as a team with few or no words. Verbal language is the only tool that ca n enable them t o labor togethe r successfully .

Humans 2 2

3

The ru b i s that verbal language does no t contai n an y component tha t ensures reliability. It is easy to lie with words. There is no substitute for the reliability and precision o f nonverba l vocalizations , a s we sa w in chapte r 6 . Thi s i s why, even after human s evolved verba l language, no human societ y ever gave up th e us e of nonverbal communications .

DECORATION, ESTHETICS, AND TH E EVOLUTION OF ART As we saw in chapter 4, decorative markings spread in a population when individuals wh o ar e s o decorate d ar e preferre d a s mates ove r one s wh o ar e not . Thi s preference result s when th e markings in question hel p observers reliabl y identif y the better individuals . It follows that those who, in effect, selec t the pattern, have to have some preexisting yardstick to asses s the decorate d individuals. This measuring stick must enable them not only to distinguish the good fro m th e bad, bu t more important , th e bette r fro m th e merel y good. I n humans , this capacit y expresses itself throug h what is generally called sensitivit y to beauty, or esthetics. 20 Obviously, th e markings of other specie s were not selecte d by humans; rather females an d male s of these species chos e as mates individuals decorated wit h th e "right" markings , whic h cause d th e evolutio n o f thi s specifi c decoration . Still , highly decorated animal s impress u s a s being beautiful ; thi s is because the deco ration i s not rando m bu t rathe r tend s t o b e symmetrica l and appropriat e t o th e animal's shape, having evolved to emphasize specific traits by which these species have adapted t o their environment an d way of life. Animals , too, show a sense of esthetics: fo r example, they prefer symmetry and completenes s t o the opposite— whether i n a peacock's tail , or a bowerbird's bowers . A nice example is the pair of long , highly decorative feathers of the king-of-Saxon y bir d o f paradise, which are valued a s decorations bot h by bowerbirds an d by tribespeople: the selectio n process by which thes e feathers evolved was carried out b y female birds o f paradise, not by humans or bowerbirds—but all three species find it attractive.21 Birdsong sounds pleasing and harmonious to us; it may be that birds prefer such songs because these enable them to more easily tell the ability of the performer. Neither th e beaut y o f othe r specie s no r huma n appreciation o f tha t beauty , however , constitute s "art." Ar t i s th e purposefu l decoratio n o f objects , and indee d th e productio n o f object s whos e onl y function i s an estheti c one . Ar t ha s obviousl y been practiced by humans from prehistoric times. Yet how did it evolve? It is highly improbable that art sprang forth full y formed from th e brain of some mysterious prehistoric human who decided t o express an admi-

224 HUMAN

S

ration fo r nature, or who contrive d t o char m animals before a hunt by rendering their likeness in a cave painting such as the one s that surviv e to this day. These cav e paintings sho w a very high leve l o f artisti c accomplishment. Th e craft o f painting is not a trivial skill : it demands a n ability t o see and a n ability t o execute that are both difficul t t o achieve. The skil l of putting a line or a dot in the right place could evolve only if from thei r very first attempts, aspiring artists were rewarded fo r developin g tha t skill . Suc h reward could onl y come from others — but wh y should other s pa y attention t o suc h line s an d dots ? Whateve r human s gain b y appreciating and creating art , its value must hav e an actual an d concret e basis: to claim that its value is "spiritual" or "esthetic" is simply to miss the point, since such "feelings " an d "spiritua l needs" too evolved for specific reasons. How, then, could peopl e have benefited by their ver y first attempts at art? We have seen that wrinkles and eyebrows increase the distance at which facia l expressions ca n be seen . Peopl e ma y have noticed tha t a smudge o f dir t i n th e right places on a face enhanced that effect. The first step in the evolution of human art ma y have been take n when people bega n experimenting with suc h marks and then deliberatel y applyin g them i n the righ t places. The mor e skillfully such lines and dot s wer e placed , th e mor e valuable , and valued, they could be—rewardin g the better o f the budding artists. Once th e productio n o f artifact s evolve d beyon d it s ver y crud e beginnings , people mus t have noticed tha t certai n decorations coul d brin g ou t th e qualit y of the objects the y made. For example , a s we saw, 22 a circle in the center can make it easie r to distinguis h a perfectly round objec t from on e les s perfect. Without a compass, it takes a good bit o f skill to make, say, a perfectly round plate . Artisans who had suc h skill benefited by decorating the cente r with a circle that helped sho w off the perfectio n o f their work . Decorating the objec t itsel f demande d skill , however — if the circl e was even slightly misplaced it would distor t th e perceived shap e of the object . An outline ca n als o show off clearly the quality of a crafted item, by making it impossibl e to hid e imperfection s i n it s edge—bu t onl y i f th e ite m i s outlined perfectly . A circle that i s not exactl y in the cente r or a n outlin e tha t i s crooked make s the produc t see m less good tha n i t is. It ma y have taken humanity a long time t o evolve the talen t t o us e and arrang e shapes an d lines in effective , satisfyin g ways, but a t every stage of that evolutio n th e more talented artisa n could b e recognized as such and rewarded . As the y learned t o mak e new artifacts, then, people als o learned t o decorat e them in a style that accentuate d their quality . Decoration tha t does not brin g out the perfection o f shape, material , o r craft is considered "tasteless " or "kitsch." In other words , w e notice whe n th e styl e is not appropriat e t o th e for m o r th e medium. To tak e a modern example , heavy artwork that bring s ou t beautifull y th e

Humans 225 quality o f materia l an d craf t i n fin e crysta l looks tacky when copie d i n molde d plastic. Plastic , though , look s beautifu l in design s tha t brin g ou t it s own uniqu e qualities. Not all have the skill to create such designs and execute them well. People who d o ar e artist s and talente d craftsperson s who understand an d know how to work wit h th e materia l the y use . Thi s abilit y coul d hav e been—an d apparentl y was—appreciated fro m th e ver y start—fro m th e firs t tim e a smeared-o n do t o f colored mu d wa s applied effectively .

ALTRUISM AND MORA L BEHAVIO R To man y people, th e differenc e betwee n human s and animal s is that animal s act through instinct , t o advance thei r material interests , while humans hav e spiritual and moral drives that are lacking in animals. Yet among babblers and other animals we sa w some behaviors that , i f we found them i n humans, would b e considere d evidence o f high mora l standards . The babbler s sho w a lot o f consideration fo r their companions . The y shar e thei r foo d wit h othe r babblers ; the y come to the rescue of their fellows, endangering themselves; they spend time acting as sentinels for th e group; they feed young that are not their own. They also refrain fro m incest and don' t copulat e i n th e presenc e o f othe r babblers . Thes e "altruistic " an d "moral" acts , a s w e believe , increas e th e overal l succes s o f th e individua l wh o performs them—an d that is precisely why babblers d o perform them. We believe that among humans, too, there is a correlation betwee n actin g morally and ethically on the one hand and success in life on the other. We believe that, other thing s bein g equal , thos e wh o behav e accordin g t o thei r society' s mora l principles—those who can afford to do so—are likelier to succeed than those who do not . What , then , i s there t o distinguis h th e mora l an d ethica l goo d fro m th e material good? We think that th e distinctio n is artificial, based on our limited understandin g of behavior. For example , an altruist is defined a s one who assists another without expecting an y payback. Bu t a gain o r a benefit may com e i n a form other tha n material payback . Altruisti c act s obviousl y demonstrate—an d ar e perceive d a s demonstrating—the abilities of those who perfor m them. No t al l of us can affor d to give away part o f our mone y or other possessions, or to ris k ourselves in order to save another; and among those who do these things, some do them better tha n others. Investing in another's welfare shows of f the altruist's quality , improves his or he r socia l standing , an d increase s hi s o r he r chanc e a t success . True , som e altruists los e more tha n the y gain, particularly whe n the y volunteer to risk them selves for their comrades or their country; but more often, the altruists return from the fron t line s with hono r and laurels—having improved thei r own or thei r children's chances , or both .

226 HUMAN

S

We certainl y don't assert tha t those who volunteer t o risk themselves i n order to save another ar e playacting or operating cynically to further thei r own interests. We have the drive , instilled b y natural selection , t o occasionally risk our live s for another. Bu t o n average , altruists ar e likely to gai n more tha n the y lose. I n fact , self-sacrifice i s no t th e onl y risk y behavio r typica l o f humans . Man y sport s ar e dangerous—auto racin g is a blatant example ; s o ar e suc h pursuit s a s mountai n climbing, voyaging in space, and exploring unknown parts of the planet. But those who succee d gai n fame. Both thos e wh o spen d resource s an d ris k thei r healt h an d lif e fo r fame , an d altruists who invest resources o r risk health and lif e for their comrades, gai n prestige accordin g t o thei r society' s principle s an d needs . A s a rule , w e accor d thi s prestige an d respec t a s a matter o f cours e an d d o no t conside r the m t o b e th e benefits o f altruism. We ar e much more awar e of the cos t of altruism than we are of the benefi t i t brings t o th e altruist , just a s we are much more awar e of cases in which a suicid e attemp t result s i n deat h tha n o f th e fa r mor e frequen t case s in which th e resul t i s that hel p tha t migh t no t hav e been forthcomin g without i t is made available. Patriotism i s a form o f "altruism" that is hallowed i n many cultures. An educational syste m that stresse s lov e fo r one' s country , an d th e rea l need t o defen d that country , can create an atmosphere in which an y attempt t o avoid the dange r involved bring s a loss of social standing. Yet once one reaches the battlefield, th e opinion o f one' s comrade s i s th e immediat e motivation . I f on e ask s officer s i n select combat units what it is that motivates soldiers to risk their lives, they answer that the strongest drivin g force is shame and the risk that one's comrades will think one a coward. Even mercenaries take risks in battle, although they don't have such a stake in the caus e they are fighting for. For them , their comrades ' good opinion is all that motivates them to take risks—and this may be enough t o move them t o deeds o f heroism. Gift-giving i s another altruisti c act that increases th e prestige o f the giver . We are ashamed to give gifts o f lesser value than those we receive. Fo r this reason, we tend t o give our more prosperous acquaintance s more expensive presents tha n we give th e les s well-to-do , eve n thoug h th e latte r migh t hav e greate r needs . An d indeed th e custom of shaming rivals by giving them presents exist s in many human societies. I t wa s highly developed i n the potlatc h o f the Northwes t Indians : the y would hol d extravagan t feasts, a t which rivals were presented with expensive gifts ; a riva l wh o faile d t o retur n i n kin d los t face. 23 I n th e highland s o f Papua Ne w Guinea, too , som e triba l head s lavis h gift s o n thei r rivals , attemptin g t o giv e so much tha t th e othe r wil l b e unabl e t o reciprocate . No t fo r nothin g di d th e old Jewish sage s say that "th e hate r o f gifts shal l live." Fund-raisers are well aware that donations pledge d o r handed over in the presence o f peers ten d t o b e muc h large r tha n donation s give n i n private . Thi s ha s become a cornerstone o f Jewish fund-raisin g i n the Unite d States : charit y events are organized so that donations will be publicized a s much a s possible amon g the

Humans

227

donor's business associates and competitors. Donations are solicited publicly, ofte n at large meetings, by persons of high status, who themselves contribute hefty sums and thus "force" others to do likewise s o as not to lose face . Bakal, who studied charit y in the United States , say s that these methods were first practiced by Josef Willen. 24 Willen mad e constructiv e us e o f the findings of the America n sociologis t Veblen , wh o stresse d th e importanc e o f conspicuou s consumption a s a mean s o f showin g of f economi c status. 25 Afte r all , fro m th e giver's point o f view, conspicuous donatio n i s simply a form o f conspicuous con sumption: bot h prov e tha t on e ha s mone y to thro w away . Showin g of f one's financial well-being i n a reliable manner has a great deal o f purely practical value; for example , it can reassure potential busines s partners and facilitate future deals . The Jewish sage s were wel l awar e tha t donor s gain prestige , just a s they un derstood tha t recipients o f charity lose face; accordin g to Judaism, it is important to giv e in secret, s o as not t o sham e th e recipient . Thi s doe s not contradic t th e notion tha t altruism is a means of gaining acclaim. Donors wan t to increase their prestige no t i n the eye s of those wh o receiv e their charit y but rathe r in th e estimation of their peers, acquaintances, competitors, or mates—who are often aware of th e donatio n eve n when i t i s supposedly secret . "Secret " donation s ar e made to protec t th e self-respec t and reputatio n o f the recipient —whose estee m is quite likely unimportant t o the dono r anyway . When we seem t o pursue ou r own interests, we are considered "selfish" both by ourselves and by others. O n th e other hand, we believe ou r tendency to act for others' benefit , which logicall y does no t see m t o be self-serving , i s an expressio n of "good moral values"; w e feel impelled to follow the path of "altruism," an d we respect other s fo r followin g it . W e ar e vividl y conscious o f the costs , risks, an d dangers entailed—of th e handicap involved—an d that's precisel y why we are impressed b y altruism. Yet we consider i t bad for m t o calculat e the benefit s it may bring us. Still, the sham e we feel whe n w e canno t retur n favor fo r favo r show s that on some leve l w e are well awar e of the prestig e altruism bring s us , and o f the effec t this prestige has on our socia l standing. The wisdom of generations acknowledges that altruis m is indeed rewarded—tha t indeed , a s our grandmothe r use d t o say, "when you do good, you do well."

This page intentionally left blank

EPILOGUE

T

he Handicap Principl e i s a very simple idea: waste can make sense, because by wasting one proves conclusively that one has enough assets to waste and more. The investment—th e waste itself—is jus t what makes the advertisement reliable . Thi s ide a seeme d s o obvious t o u s that w e assumed at first that it must alread y b e widel y accepted , an d s o we searched th e existin g literatur e fo r discussions of it. Our searc h yielde d many previous attempt s to explai n th e waste on e sees i n sexual showing off. Most of these explanations were very complex, and some were aided by mathematical models, but th e Handicap Principl e wa s not amon g them. And t o ou r grea t surprise , this idea, which struc k u s a s self-evident, was bitterly resisted b y the scientifi c establishment . Eve n more to our surprise , this same idea ended u p revolutionizin g ou r understandin g o f communicatio n throughou t th e living world, up to an d including communication within the body. The Handicap Principle states that the receive r o f a signal ha s a stake i n th e signal's reliability, or accuracy, and will not pa y attention t o it unless it is reliable. Thus signals are not arbitrary ; rather, each signal is the one best suite d to reliably convey the specifi c messag e it carries . It follow s that there must be a logical connection betwee n th e message and th e signal . The Handicap Principl e enable s us 229

23O EPILOGU

E

to mak e predictions: it makes it possible t o figure out fro m th e natur e of a signal what message it conveys , and likewis e what might logically serve as a signal for a given message. The Handica p Principl e expand s ou r curren t understandin g o f evolution i n a basic way . Signals, like othe r traits , evolve throug h natura l selection . Bu t wher e those othe r trait s ar e selected becaus e they make an organism mor e efficien t i n a straightforward, utilitaria n way, signals ar e selected becaus e the y handicap th e organism in a way that guarantees tha t th e signa l is reliable. This apparen t parado x masks a basic consistency between signa l selection and the evolution of other traits: in both cases , traits that spread throughout th e population ar e those that improv e an individual's chance s of having offspring tha t in turn will reproduce successfully . This i s just a s true of a signal that entail s a handicap a s it is of a bodily structur e that makes its owner mor e efficient . A theory is valuable only if it can lead to new findings, and in fact, the Handicap Principle ha s alread y opened u p ne w avenue s of research . A decad e ago , many researchers attempte d t o sho w how animal s use signals to mislead ; nowadays an increasing numbe r o f studies sho w tha t signal s reliably reflec t th e intention s an d qualities o f the signaler . Even so , researchers d o no t ye t seek th e specifi c cost — the handicap—tha t make s th e signa l reliable . Sinc e th e handicap—th e invest ment—is what guarantee s that a given signa l is reliable, i t follows tha t by deter mining the specific handicap a signal entails, we can better understand the signaler, the environmental conditions , and the message . Once one accepts the Handicap Principle a s a general rule, one can no longer see signal s i n natur e a s mer e conventions . On e therefor e ha s t o reevaluat e al l signals—including thos e tha t u p t o no w hav e been though t simpl y t o identif y a given species , age , gender, o r an y other grouping—al l the way down t o chemica l signals on cell membranes, which traditionally have been seen as merely identifying the cel l a s belonging t o a specific type. The many examples from humankin d that we have used throughout thi s book show how deeply the Handicap Principle is embedded i n human life. The principle can help u s better understand man y of our natura l tendencies, whic h result fro m the need constantl y to test the bonds linking us with others in various endeavors. The Handica p Principl e als o show s how the nee d t o cooperat e wit h one' s competitors coul d have led to the evolution o f altruism, in humans as in other animals. Human being s hav e lon g marveled a t th e wonder s o f nature. The Handica p Principle is one of those theories that sho w us how nature's intricacies are bound together i n an orderly system— a system that i s logically simple, tha t makes sense, and tha t w e ca n understand . A n awarenes s an d understandin g o f this pervasiv e order ma y demystif y nature , bu t i t detract s no t a t al l from th e wonde r w e fee l when w e look abou t u s at the worl d w e live in. Indeed, nature' s order ma y well be its most awe-inspirin g marvel.

NOTES

Introduction 1. Zahavi , A., 1975 , 1977 . 2. Davie s and O'Donald, 1976 ; Maynard Smith, 1976b; Kirkpatrick, 1986 . 3. Eshel , 1978a ; Pomiankowski, 1987 . 4. Grafen , 1990a, 1990b . 5. Lotem , 1993a ; Maynard Smith, 1991a; Collins, 1993 .

Chapter 1 1. Se e Maynard Smith, 1965. Th e difficult y wit h group selectio n will be discusse d in chapter 2. 2. Zahavi , 1978b. 3. Curio , 1978 . 4. Marler , 1955 . 5. Sordal , 1990 . 6. Morri s (1990 ) describe s ho w ma n ha s fro m ancien t time s use d birds ' mobbin g instincts to hunt them, using decoys of raptors. 231

232 NOTE

S

7. Car o (1994 ) surveys current studie s of slotting as well a s past explanation s of this behavior. Se e also Hasson, 1991a. 8. Zahavi , 1977a , 1987 . 9. Fitzgibbo n an d Fanshawe, 1988 ; Caro , 1994 . 10. Hasso n et al, 1989 . 11. Hasso n (199la ) recently reviewed communication s betwee n pre y and predator by pursuit-deterrent signals. 12. Smythe , 1970 . 13. Rhisiart , 1989 ; Cresswell , 1994 . 14. Wiklun d and Jarvi, 1982 . 15. Ritland , 1991a, 1991b . 16. Eshel , 1988 . 17. Kruuk , 1972 . 18. Eshel , 1978a .

Chapter 2 1. Lorenz , 1966 . 2. Maynar d Smith and Parker, 1976 . 3. Zahavi , 1977a . 4. Ewer , 1968 . 5. Baerend s and Baerends-van Roon, 1950 . 6. Glutton-Broc k et al. , 1982 . 7. Zahavi , 1981b. 8. Morton , 1977 . 9. Zahavi , 1982; see more in chapter 6. 10. Davie s and Halliday , 1978 . 11. Katsir , 1985, 1995 . 12. Se e also chapter 6. 13. Schjelderup-Ebbe , 1992 ; Lorenz , 1966; Marle r and Hamilton, 1966. Se e chapter 12 for furthe r details . 14. Darling , 1937 . 15. Barret e and Vandal , 1990 . 16. Lorenz , 1966 . 17. A detailed discussion of the drawbacks of the model of group selection can be found in Dawkins (1980). See also Maynard Smith, 1964, 1976a . 18. Lorenz , 1966 ; Wynne-Edwards, 1986. William s (1994) uses a different definitio n of "group selection " an d stresse s that h e doe s no t assum e the existenc e o f adaptations that developed "fo r the goo d o f the group." 19. Se e Axelrod, 1986 ; w e will discuss this subject further i n chapter 12 .

Chapter 3 1. Williams , 1966 . Williams' s idea was further develope d by Trivers (1972) who examined its various implications.

Notes 2 3

3

2. Jone s and Hunter, 1993 . 3. Zahavi , A., 1975 , 1977a , 1977b . 4. Nisbet , 1973 , 1977 . 5. Wilhel m et al. , 1980 , 1982 . 6. O'Donald , 1963 . 7. Se e Wynne-Edwards, 1962 . 8. Rya n et al. , 1982 . 9. Glutton-Broc k and Albon, 1978 . 10. Lambrecht s and Dhondt, 1986 . 11. Se e chapter 6. 12. Se e chapter 4 . 13. Thornhill , 1992a ; Moore, 1988 . 14. Eisne r and Meinwald , 1987 , 1995 . 15. Eisne r an d Meinwald, 1987 . 16. McKaye , 1991 . 17. Christy , 1988 . 18. Borgia , 1986; Diamond, 1986a, 1986b . 19. Frit h an d Frith, 1990; Diamond, 1991 . 20. Borgia , 1986 . 21. Borgi a and Collins , 1986 . 22. Petri e et al., 1991 . 23. Moller , 1994 . 24. Smit h and Montgomerie , 1991 . 25. Smit h et al., 1991 . 26. Evan s and Thomas, 1992 . 27. Gibso n et al., 1991 . 28. Petri e e t al., 1991 . 29. Hoglun d et al., 1993 . 30. Gibso n and Hoeglund, 1992 . 31. Snow , 1976 ; Foster, 1981 . 32. McDonal d and Potts, 1994 . 33. VanRhijn , 1973 . 34. Hogan-Warburg , 1966 . 35. Dominey , 1980 . 36. Se e a general discussion of this subject in Taborsky (1994). 37. Darwin , 1859 . 38. Darwin , 1872 . 39. Se e a detailed discussio n in Cronin, 1991 . 40. Fisher , 1930 . 41. Se e Andersson, 1994. 42. Se e Mayr (1972) o n thi s subject (p. 97): "Darwin assume d rather naivel y that 'the best arme d males' were also the strongest and that 'th e more attractive' males were 'at the same time mor e vigorous' . . . there is , however, no demonstratio n of an automatic correlation between th e two characteristics." 43. Thi s kind of solution to the problem of waste in courtship computes the evolutionary

234 NOTE

S

value o f feature s b y thei r frequenc y i n a population. Suc h solution s wer e later calle d by Maynard Smith "evolutionary stable strategy" (ESS) and have been used by him to explain many other social phenomena. See Maynard Smith, 1976c. 44. Se e Fisher, 1930, 2nd edition (Ne w York: Dover Publications, 1958) , p. 155 : "Th e possibility shoul d perhap s be born e i n mind i n such studies tha t the most finel y adorne d males gain some reproductive advantage without the interventio n of female preference, in a manner analogou s t o that in which advantag e is conferred b y special weapons. The establishment of territorial rights involves frequent disputes, but these are by no means all mortal combats; th e mos t numerous , and fro m ou r poin t o f view, therefore, the mos t importan t cases are those in which there is no fight at all, and in which the intruding male is so strongly impressed o r intimidated by the appearanc e of his antagonist as not t o risk the damag e of a conflict. As a propagandist the cock behaves as though he knew that it was as advantageous to impress the males as the females of his species, an d a sprightly bearing with fine feathers and triumphant song are quite as well adapted for war-propaganda as for courtship." Fisher has no explanation for this effect, an d continues (p. 156): "The evolutionar y reaction of war paint upo n thos e who m i t i s intended t o impres s shoul d b e t o mak e the m les s an d less receptive to all impressions save those arising from genuin e prowess." 45. Se e Alcock, 1993. 46. Andersson , 1994. 47. Zahavi , 1981a , 1987 , 1991a . Chapter 4 1. Lorenz , 1966 . 2. Wallace , 1889 . 3. Mayr , 1942 . 4. Smith , 1966 , 1967 . 5. Katzir , 1981a , 1981b . 6. Snow , 1976 . 7. Selander , 1972 . 8. Zahavi , 1978, 1981, 1987 , 1992 . 9. Se e more in chapter 5. 10. Barlow , 1972. 11. Zahavi , 1978a, 1981a, 1987 , 1993 . 12. Hailman , 1977; Morris , 1990 . 13. Tinbergen , 1953 . 14. Se e more in chapter 8 about the benefit s an d drawbacks of two colors rather than one. 15. Hamilto n an d Zuk, 1982. 16. Se e also Hasson, 1991b . 17. Moller , 1990a, 1992 . 18. Watso n and Thornhill, 1994. 19. Thornhill , 1992a , 1992b . 20. Parson , 1990 . 21. Zahavi , 1993 . 22. Se e Zahavi, 1978a.

Notes 2 3

5

23. Petrieetal , 1991 . 24. Ridley , 1981 . 25. Se e for example the article by Gibbs and Grant (1981 ) on changes in beak size that follow climat e chang e an d change s in die t i n finches on Daphn e Islan d i n the Galapagos . See also Weiner, 1994 . 26. Lack , 1968 . 27. Roper , 1986 ; Fugle et al. , 1984 . 28. Jarv i and Bakker, 1984; Norris, 1990 . 29. Rohwe r and Rohwer, 1987 ; Rohwe r and Ewald, 1981 . 30. Se e Saino et al. , 1995 . 31. Se e chapter 7. 32. Hasson , 1991b . 33. Maynar d Smith, 199 Ib. 34. Grafen , 1990a , 1990b . 35. Hendryetal. , 1984 . 36. Thi s may have happened with ducks: the brilliant plumage of drakes in the Northern Hemisphere show s of f their abilit y to avoi d predators despit e th e handica p o f being very visible. Lack (1970 ) pointed ou t that on small, remote islands mature male ducks are often drab. I t ma y be tha t o n smal l islands withou t predators , wher e ever y drake coul d affor d bright plumage , th e brillian t color s los t thei r valu e a s a reliabl e indicato r o f th e drake' s quality, became superfluous, and disappeared . See Zahavi, 1981a . 37. Pond , 1973 . 38. Naam a Zahavi-Ely, personal communication . 39. Se e Zahavi, 1987 . 40. Borgia , 1966 . 41. Hunte r an d Dwyer, in press.

Chapter 5 1. Redond o and Castro , 1992. 2. Eibel-Eibelsfeldt , 1961 . 3. Huxley , 1914 . 4. Cullen , 1966 . 5. Morris , 1957 . 6. Zahavi , 1980, 1987 . 7. Simpson , 1968 . 8. Boake , 1991 . 9. Kreb s and Dawkins , 1984 . 10. Kruuk , 1972 . 11. Th e assumptio n tha t the abilit y to notice signal s precedes th e evolution o f a signal serves t o explai n wh y animals are sometimes attracte d t o signal s that d o not exis t in thei r species (Basolo , 1990; Burley , 1986). Ryan (1990 ) describes man y such case s and suggests that when females are preinclined t o specific sound s o r colors, their sensitivities are utilized by males who adop t thes e attractiv e features. He term s this phenomenon "sensor y exploi tation." But in our opinion , th e fac t tha t observers ar e able to sens e one signal or anothe r does no t caus e the signa l to evolv e i n a particular way. Obviously, n o signa l will evolv e if

236 NOTE

S

the on e signaled cannot perceive it. But neither will it evolve if the receive r finds the infor mation uninterestin g or unreliable . Chapter 6 1. Zahavi , 1982. 2. Katsi r (1985 , 1991 ) foun d that the inversio n frequency of a babbler cal l is related to th e stat e o f the bod y making the call : it is very low when th e babble r i s sitting relaxed on its nest, higher when it is standing in a tree, and higher yet when the babbler i s in flight. On inversio n frequenc y i n birds and its physiological basis , se e Greenewalt (1986) o n the physiological side of birdsong . 3. Darwin , 1872 . 4. Scherer , 1979 , 1985 . 5. See , for example, a review by Murray and Arnott on human vocal emotion (1993) , which was recently brought t o our attention . 6. Rowell , 1962 . 7. Gaion i an d Evans, 1985, 1986a , 1986b. 8. Morto n an d Page, 1992 . 9. Lambrecht s an d Dhondt, 1986 . 10. I t may be that the conflict between th e ability to listen and the ability to concentrate on talkin g is the foundatio n o f lie detectors. Th e person takin g a lie detector tes t may be on the othe r en d o f a phone line , under no fea r o f direct attack . What is it that prevents a liar's voice from soundin g like that of a truthful person ? An inherent differenc e between a liar and one who is telling the truth is that the truthfu l person ha s a true story that does not need t o be changed according to circumstance, while a liar is making u p a story, an d his or her succes s depend s on the abilit y t o convinc e th e listener of something that isn't true. The liar has to expend effor t t o make up the story. And since the liar may not kno w in advance the listener's prior knowledg e o f the subject, he or she needs to pay close attention to the listener's reaction in order to adapt the story, change it slightly—and remember th e changes—in order to be convincing. This close attention t o the listener is very likely to affec t muscle s of the neck and head, and that in turn is likely to affect th e voice to some degree. This differenc e ma y well be the one that lie detectors focu s on (Streete r a t al, 1977) . I f the liar tries to rela x thes e muscles, hi s or her listening abilit y will decrease and with it the abilit y to lie successfully . 11. Anava, 1992. 12. Katsir, 1985, 1991, 1995. 13. Zahavi, 1978b. 14. Payne, 1983. 15. Hultsc h and Todt, 1986 ; Todt an d Hultsch, 1995 . 16. Payne , 1983 . 17. McGregor , 1993 . 18. Pepperberg , 1991 ; Kaufman , 1991 . 19. Ofe r Hochberg, personal communication. 20. Ala n Kemp, personal communication . 21. Hultsc h and Todt, 1986 . 22. Loffred o an d Borgia, 1986 .

Notes 2 3

7

23. Darwin , 1874 . 24. Seyfart h et al. , 1980 . 25. Marler , 1955 . 26. Se e chapter 12 .

Chapter 7 1. Zu k e t al. , 1990 . 2. Holde r and Montgomerie, 1993 . 3. Darwin , 1871 . 4. Sutler , 1994 . 5. Evan s and Thomas, 1992 . 6. Evans , 1991 . 7. Andersson , 1982 . 8. Eibel-Eibelsfeldt , 1970 . 9. Se e also the discussio n o f fishes' fins in chapter 2 . 10. Ale x Kacelnic k pointe d ou t a logical weaknes s i n ou r argumen t that mane s an d other feature s lesse n th e apparen t siz e o f bod y structures . Afte r all , w e asser t tha t erec t feathers an d hair canno t b e mean t t o increase apparen t siz e since watchers ca n detect th e deception. Ye t we ar e suggestin g an opposit e deception—whe n we claim that a frame o f bristling hair or feathers is a handicap that reduces the apparent siz e of the animal. Should watchers no t disregar d thi s deceptio n a s well , eve n i f i t i s base d o n a n optica l illusio n (Ponzo's effect—se e Fujit a e t al, , 1991) ? We accep t tha t th e watche r ma y know tha t th e mane decreases the apparent size of the shape within it. But this effect makes it more difficult for a slightly larger individual to show clearly its superiority in size, a superiority that would have been obviou s i f it were no t fo r th e mane . This is a real handicap: onl y an individua l who i s significantly larger than anothe r ca n afford a decoration tha t decrease s its apparent size without impairin g its ability to show off that in fact i t is larger than others. 11. Richar d Wagner, personal communication . 12. Gior a Ilani , personal communication . 13. Se e also the discussio n o f human beards i n chapters 2 and 18 . 14. Darling , 1937 . 15. Glutton-Broc k et al. , 1982 . 16. Moller , 1991 .

Chapter 8 1. Hill , 1990 . 2. Endle r (1983 , 1987) , Lythgoe (1979), Hailman (1977) , Butcher and Rohwer (1989), and other s suc h as Hamilton, WJ . (1973) , and Kingston (1933 ) have studied th e benefits and drawback s of specific colors . To them it was a question o f balance between advertisin g over distanc e o n th e on e han d an d mergin g with th e backgroun d t o avoi d enemie s an d predators o n the other . They als o studied th e eye's sensitivity to various colors. 3. May r and Stresemann , 1950 . 4. Diamond , 1987 .

238 NOTE

S

5. Maier , 1993 . 6. Anderson , 1996 .

Chapter 9 1. Eisne r an d Meinwald , 1987, 1995 . 2. Schneider , 1992 . 3. Elli s et al, 1980 . 4. A t th e sam e time , though , hig h concentration s o f femal e pheromone impai r th e ability of males to find females. In biologica l pes t control , syntheti c hormones ar e used t o "confuse" males . 5. O n intoxicating beverages and the value of showing off the ability to imbibe without getting drunk, see chapter 13 . 6. Naho n et al., 1995 . 7. Jackso n an d Hartwell, 1990a, 1990b . 8. Ulloa-Aguirre , 1995 . 9. Se e the bibliograph y in Nahon e t al. , 1995 . 10. Se e ibid. 11. Zahavi , 1993. 12. Snyde r and Bredt , 1992 . 13. Se e the bibliograph y in Nahon et al. , 1995 . 14. Zahavi , 1993.

Chapter 10 1. Zahavi , 1979. 2. Zahavi , 197Ib . 3. Morris , 1956 . 4. Se e chapter 12 . 5. Selander , 1972 . 6. Se e chapter 3. 7. Borgia , personal communication. 8. Osztreiher , 1992 . 9. Spiro , 1963 . 10. Va n Lawick-Goodall, 1970. 11. Rasa , 1986 .

Chapter 11 1. Trivers , 1974 . 2. Triver s further complicate d the problem by suggesting that according to kin selection, the chil d also has some interest in its parents' reproduction becaus e of its genetic similarity to its siblings; see Trivers, 1972 . Thi s complication is unnecessary, as we shall see when we deal with the issue of kin selection in chapter 13 . 3. Zahavi , 1977a . 4. Heinroth , 1926 . 5. Feldma n an d Eshel, 1982 .

Notes 2 3

9

6. Se e Redondo and Castro , 1992 . 7. Mock , 1984 . 8. Diamond , 1992 . 9. Sade , 1972 .

Chapter 12 1. Se e articles i n Hebrew by Pozis (1984) , Carmeli (1988) , Katsir (1991), Osztreiher (1992), Anava (1992), Kalishov (1996), Perl (1996) , Zahavi, T. (1975) . 2. Late r in this chapte r we present graph s based o n some of the dat a that hav e bee n collected s o far, much of which is as yet available only in Hebrew . 3. Osztreiher , 1996 . 4. Stache y and Koenig, 1990 ; Rowley and Russel, 1990 . 5. Va n Lawick-Goodall, 1971 . 6. Kruuk , 1972 . 7. Va n Lawick-Goodall, 1970 . 8. Rasa, 1986 . 9. Sherma n et al. , 1991 . 10. Bonner , 1967 . 11. Rosenberg , 1984 . 12. Se e Maynard Smith, 1964 , 1976b . 13. Hamilton , 1964 . 14. Trivers , 1971 . 15. Axelrod , 1986 . 16. Axelro d an d Hamilton, 1981 . 17. Ther e i s extensive literature o n these models. Se e a detailed discussio n an d bibli ography in Dawkins (1989) and a more up-to-date one in Sigmund (1993). 18. Stache y and Koenig , 1990 . 19. Se e chapter 1 , on prey-predator relations . 20. Zahavi , T., 1975 . 21. Carlisl e and Zahavi, 1986 . 22. Kalishov , 1996 . 23. Carlisl e and Zahavi, 1986 . 24. Kalishov , 1996 . 25. Carmeli , 1988 . 26. Slagsvold , 1984 , 1985 . 27. Carlisl e and Zahavi, 1986; Zahavi , 1989 . 28. Schjelderup-Ebbe , 1922 ; see Marler and Hamilton, 1966 . 29. Perl , 1996 . 30. Zahavi , 1988; Perl, 1996 . 31. Perl , 1996 . 32. Gaston , 1978 . Onl y one species o f babblers i n India lives in pairs . 33. Komdeur , 1992 , 1994 . 34. Woolfende n an d Fitzpatrick , 1984, 1990 . 35. Kofor d et al., 1990 . 36. Faabor g an d Bednarz, 1990 . 37. Se e Alexander, 1987 ; De Waal, 1996 .

24O NOTE

S

38. Zahavi , 1995 . 39. Ibid .

Chapter 13 1. Se e a detailed discussio n in Cronin, 1991 . 2. Alexander , 1974 . 3. Se e chapter 14 . 4. Wilson , 1971 . 5. West-Eberhard , in her stud y o f Polistes canadensis (1986) , showed that when on e queen clearl y controls the others, the level of aggression in the nest is low, while when there are only slight differences betwee n th e queens, aggression is high and there may be life-anddeath struggle s among the partners. 6. West , 1969 ; West-Eberhard, 1984 ; Gadagkar, 1991 ; Ito , 1993 . 7. The was p Vespula germanica was brought by European s to Australia . In tropica l regions, i t turne d ou t tha t it s colonie s coul d surviv e the winter , an d ne w colonies o f this species in such regions are formed by coalitions of several queens. Se e Ito, 1993 . 8. Wilson , 1971 ; Heinze e t al. , 1994 . 9. Trivers , 1985 . 10. I t ma y be tha t some workers d o indeed avoi d working, but thi s doe s no t deman d any special explanation. It i s the fact that workers do invest in the colony, as indeed most do, that demands explanation . 11. West , 1969 . 12. Gadagkar , 1991; Heinze et al., 1994 . 13. Marle r an d Hamilto n (1966 ) defin e pheromone s a s chemicals secreted b y one individual in order to elicit a specific reaction i n another individual of the same species. As a rule, pheromone s ar e mixture s o f chemicals, an d i n mos t case s their specifi c component s are not known . Thus , we use the term s queen pheromone or pheromones without gettin g into the specific s of one material or another . 14. Isha y et al. (1967, 1968 ) found that worker orienta l hornets cannot make sugars out of proteins, a process known as gluconeogenesis; they feed the larvae proteins, and eat sugars they get from th e larvae. If such mutual feeding is found between th e larvae and the adul t workers of other socia l insects, it will provide another reason—probably the main one—for workers to feed the larvae and take care of them. However, thi s very interesting area requires more research. 15. Winsto n an d Slessor , 1992. 16. Engel s and Imperatriz-Fonseca, 1990 . 17. Va n der Blom, 1986 . 18. Rosele r an d Honk , 1989 ; se e also Velthuis, 1990 . 19. Diamond , 1990 , 1992 . 20. Veblen , 1899 . 21. Trivers , 1974 , 1985 . Not e tha t by this definition, a host taking care of its parasite's offspring i s an altruist, and indeed Triver s s o defines it (1985). By the same definition, when a helper i s sterile t o begin with , it s reproduction canno t diminis h an y further and thus its help canno t be see n as altruism. 22. Trivers , 1971 . 23. Hamilton , 1964 .

Notes 2 4

1

24. Dawkins , 1989 . 25. Th e calculatio n is: in sexuall y reproducing animals , a random hal f o f one's gene s come from th e mother and the other hal f from th e father. Sometimes the gene of one parent is not identical to but rathe r a n allele of the gene o f the other parent. Since the individual inherits randomly one or the other, there is a 50-percent probability that a given allele found in one individual will also be found in its sibling, and a 25-percent probabilit y that they will also be found in the sibling's childre n ( a probability of 50 percent times another probability of 50 percent). 26. A detailed discussio n of this theory can be found i n any book o n sociobiology. See for exampl e Trivers (1985) ; Dawkins (1989) ; Wilson (1975) ; West-Eberhard (1975); Krebs and Davie s (1993) . 27. Termit e nymphs , like the larvae of ants, wasps, and bees, depend on their caretakers for food . Again , as in hymenoptera , termit e workers ar e not completel y sterile , an d thei r fate depend s o n the treatment they get while growing. To understand what is happening in a termite colony, one would have to find out what it is that prevents individuals from striking out o n their own, and how the workers' servic e to the colon y benefits them individually. 28. Wilson , 1975 ; Kreb s and Davies, 1993 . 29. Gadagkar , 1991 . 30. Holldoble r an d Wilson , 1990 , i n thei r boo k abou t ants , tell o f worker ant s of the species Myrmecocystis mumicus and Solenopsis invuca who move to the biggest nest aroun d and leave their mothers t o die of hunger in the nest where they hatched. Such findings run contrary to the theory of kin selection but ar e in full accord with our suggestion that workers choose the colony where thei r chances o f reproducing are best. 31. Haldane , 1932 , 1955 . 32. Motr o an d Eshel, 1988 ; Eshe l and Motro, 1988 . 33. I n fact , som e researchers who stil l accep t group selection a s a valid theory wonder why i t i s that mos t other s i n th e professio n accep t ki n selectio n a s valid whil e rejectin g group-selection models. Se e Wilson an d Sober , 1994 . 34. Zahavi , 1995. 35. Zahavi , 1974 , 1989 ; Woolfenden an d Fitzpatrick, 1990 ; Komdeur , 1994 . 36. Gadagka r an d Joshi, 1985 . 37. Darwin , 1871 . 38. Holldoble r and Wilson, 1990 . 39. Dawkins , 1989 . 40. West , 1969 . 41. Rosele r an d Honk , 1989 , remar k tha t among bumblebees, th e struggl e for dominance among young queens starts in the first days following their emergence from the pupa stage. 42. Zahavi , 1995.

Chapter 14 1. Trivers , 1972 , discussed i n detail the implications of the conflic t of interest between the genders, an d many have studied th e subject since. We ar e not attemptin g to challeng e his treatment o f the subject but rathe r to add some observations fro m ou r own perspective . 2. Selander , 1965 , 1972 ; Orians, 1969 ; Emle n and Oring, 1977 . 3. Gustafson , 1989 ; Beissinger , 1986; Beissinge r and Snyder , 1987.

242 NOTE

S

4. Dawkin s an d Carlisle, 1977 . 5. Scott , 1988 ; Owe n an d Black, 1989; Forslund, 1990 . 6. Newton , 1989 ; Gustafson , 1989 . 7. Se e Gustafson, 1989 . 8. Montgomerie , 1986; Zahavi, 1986 . 9. Zahavi , 1988a; Perl, 1996 . 10. Moller , 1990 . 11. Morto n e t al. , 1990 ; Wagner et al., 1996 . 12. Zilberman , 1991 . 13. Goldstei n e t al., 1986 . 14. Sugiyama , 1967 . 15. Timna , personal communication; Hani, personal communication. 16. Trivers , 1972 . 17. Reyer , 1990 . 18. Munehar a et al., 1994 . 19. Kraak , 1994. 20. Fishman , 197 7 (Hebrew) . 21. Moren o et al., 1994. 22. Leader , 1996 . 23. Orian s and Beletsky, 1989. 24. Emle n and Oring , 1977 ; Emlen et al., 1989 . 25. Gustafson , 1989 .

Chapter 15 1. Thi s descriptio n is based o n Bonner (1991 ) an d Nanjundia h an d Sara n (1992). 2. Se e Atzmony et al., 1997. 3. cAM P is a chemical with many roles in all living organisms—from bacteria to mammals, including humans . It i s considered a secondary messenger transferring stimuli fro m the cell membrane t o the inside of the cell. 4. Shaulsk i and Loomis (1993) found that some of the amebas that ended up as spores showed sign s of having been prestalk before that. In other words, at the end of the process of migration and stal k formation, some of the prestalk ameba s manage to become spores . 5. Se e Atzmony et al., 1997 . 6. Bruc e Levin, personal communication . 7. Atzmon y and Nanjundiah, personal communication . 8. Se e Rosenberg, 1984 . 9. Se e Zahavi and Rait, 1984 . 10. Shaulsk i and Loomis, 1995 .

Chapter 16 1. Dawkin s and Krebs , 1979 ; Rothstein, 1990 . 2. Davie s and Brook, 1989 . 3. Lotemetal , 1991 , 1995 . 4. Lote m wrote another article about th e dangers o f evolving the abilit y to recogniz e a parasite's nestling and deser t it or throw it out of the nest. See Lotem, 1993b .

Notes 2 4

3

5. Researc h done b y Yoram Shpirer , Amot z Zahavi, Arnon Lotem , an d Stev e Rothstein. 6. Sole r et al. , 1995 . 7. Zahavi , 1979. 8. Se e chapters 3, 12 , and 14. 9. Newton , 1989b ; Scott, 1988 ; Owen an d Black, 1989 . 10. Yom-Tov , 1989 . 11. Ingle , 1911 ; Wyllie, 1975 ; Witherby et al. , 1949 . 12. Zahavi , 1979. 13. Sole r et al. , 1995 . 14. J . Tengo , University of Uppsala, Sweden, personal communication . 15. Tengo , 1984 . 16. M . Sorensen, personal communication. 17. Smit h et al. , 1984 . 18. Oberski , 1975 ; Kuris , 1974. 19. Schwammberger , 1993 . 20. Salyer s and Whitt, 1994.

Chapter 17 1. Zahavi , 197la. 2. Ward , 1965 . 3. Darling , 1938 . 4. Unfortunately , Ward, th e partne r with who m Amot z developed th e ide a o f information centers , died a t an early age. 5. War d and Zahavi, 1973 . 6. Parker-Rabenold , 1978 ; Broo m et al., 1976 ; Heinrich, 1988 . 7. Brown , 1986 . 8. Heinrich , 1988 . 9. Wagner , 1996 . 10. Zahavi , 1983, 1995 . 11. Ward , 1972 . 12. Fear e e t al. , 1974 . 13. Heredi a e t al. , 1991 ; Hiraldo et al., 1993 . 14. Bucher , 1992. 15. Wynne-Edwards , 1962 . 16. Marx , personal communication. 17. Veblen , 1899 . 18. Ibid . 19. Se e Wynne-Edwards, 1962 . 20. I n ou r origina l articl e on informatio n center s (War d and Zahavi , 1973 ) w e mad e something o f a group-selectio n argumen t ourselves . W e hav e sinc e recognize d tha t early error an d have suggested tha t othe r explanation s shoul d b e sought for the communal displays at roosts (Zahavi , 1985b) . Thi s subjec t has received further treatmen t in our respons e to a n article by Richner and Heeb (Zahavi, 1996). 21. Zahavi , 1983, 1995 . 22. Se e chapter 1 .

244 NOTE

S

Chapter 18 1. Shepher , 1983 . 2. Hess , 1965 . 3. Eibel-Eibelsfeldt , 1971 . 4. Se e chapter 8. 5. 2 Samuel 20:9. 6. Se e chapter 7. 7. Morris , 1967 . 8. Car o and Seller, 1990 . 9. Seve r and Mendelsohn, 1989 . 10. D e Waal , 1995 . 11. Strahl , 1988 . 12. Se e chapter 10. 13. Zahavi , 1971b. 14. Trivers , 1985 . 15. Wagner , 1996 . 16. Se e chapter 11 . 17. Th e gra y parrot Alex; see Pepperberg, 1991 . 18. Th e exception is among the most highly social bees, wasps, ants, and termites, whose communities include thousands of individuals. 19. Th e differences between a language of nonverbal vocalizations and a verbal language is like the differenc e betwee n a n analog speedometer an d a digital speedometer. An analog speedometer's needl e let s us estimate speed prett y precisely, even if there are only two or three numbers o n the face of the speedometer. A digital speedometer, which displays numbers, i s limited t o th e precisio n o f these numbers . I f th e digit s chang e with onl y every 5 additional mile s per hour , the digita l speedomete r wil l not sho w the differenc e between , say, 1 1 mph an d 1 4 mph—while an analog speedometer will . 20. Zahavi , 1980 . 21. Frit h and Frith, 1990 ; Diamond, 1991 . 22. Se e chapter 4. 23. Benedict , 1946 . 24. Bakal , 1979 . 25. Veblen , 1899 .

BIBLIOGRAPHY

Alcock, J. 1975 . 5th ed . 1993 . Animal behaviour; An evolutionary approach, Sunderland , MA: Sinauer Associates. Alexander, R.D. 1974. The evolution o f social behaviour. Ann. Rev. Ecol. Syst. 5 : 325-83. . 1987. The Biology o f Moral Systems. Ne w York: Aldine De Gruyter . Anava, A . 1992 . The valu e o f mobbin g behaviou r fo r th e individua l babble r (Turdoides squamiceps) (i n Hebrew). M.S . thesis , Ben-Gurion Universit y of the Negev . Anderson, S . 1996. Bright UV colou r i n the Asia n whistling thrushes . Proc. R. Soc. Land. B., 263, 843^8. Andersson, M. 1982 . Female choic e select s for extreme tail length in widow birds. Nature 299: 818-20 . . 1994. Sexual selection. Princeton: Princeton Universit y Press. Atzmony, D. , A . Zahav i an d V . Nanjundiah . 1997 . Altruistic behaviou r i n Dictyostelium discoideum explaine d o n the basis of individual selection. Current Science 72: 142-45 . Axelrod, R. 1986. An evolutionary approach to norms. Amer. Politic. Sci. Rev. 80: 10951111. Axelrod, R. , and W.D. Hamilton. 1981 . The evolution o f cooperation. Science 211: 139096. Baerends, G.P. , an d J.M. Baerends-van Roon . 1950 . An introduction t o th e stud y of th e ethology of cichlid fishes. Behaviour (suppl.) 1 : 1-242. 245

246 BIBLIOGRAPH

Y

Bakal, C . 1979 . Charity U.S.A.: An investigation into th e hidden world of th e multi-billion dollar charity industry. New York : Times Books. Barlow, J.W. 1972 . Th e attitud e o f fish eye-lines in relatio n t o bod y shap e an d t o stripe s and bars. Copeia 12: 4-12 . Barrete, C. , an d D . Vandal . 1990 . Sparring , relativ e antle r size , an d assessmen t i n mal e caribou. Behav. Ecol. Sociobiol. 26: 383-87. Basolo, L.A . 1990 . Femal e preferenc e predate s th e evolutio n o f th e swor d i n swordfish . Science 250: 808-10 . Beissinger, S.R . 1986. Demography , environmenta l uncertainty , and th e evolutio n o f mate desertion i n the snai l kite. Ecology 67 : 1445-59. Beissinger, S.R. , and N.F.R . Snyder . 1987 . Mat e desertio n i n th e snai l kite. Anim. Behav. 35: 477-87. Benedict, R. 1946 . Patterns of culture. New York : Penguin Books . Boake, C.R.B. 1991. Coevolutio n o f senders and receivers of sexual signals: genetic coupling and geneti c correlations. Tree 6 : 225-27. Bonner, J.T. 1967 . Th e cellular slime molds. Princeton: Princeton Universit y Press. Borgia, G. 1986 . Sexua l selection i n bowerbirds. Sci. Amer. 254: 70-79. Borgia, G., an d K . Collins. 1986 . Feathe r stealin g in the sati n bowerbird (Ptilonorhynchus violaceus): Mal e competition an d the qualit y of display. Anim. Behav. 34: 727-38. Broom, D.M. , WJ.A . Dick , C.E . Johnson , D.I . Sales , an d A . Zahavi. 1976 . Pie d wagtai l roosting and feeding behaviour. Bird Study 23 : 267-80. Brown, C.R. 1986 . Clif f swallo w colonies a s information centers . Science 234: 83-85. Bucher, E.H. 1992 . Th e cause s of extinction o f the passenge r pigeon . I n Current Ornithology, ed . D.N. Power , vol . 9, 1-26. Ne w York: Plenum Press . Burley, N. 1986 . Sexua l selection fo r aesthetic traits in species wit h biparental care . Amer. Nat. 127 : 415-45. Burley, N. , G. Krantzberg , an d P. Radman. 1982 . Influenc e o f colour-banding on the conspecific preferences of zebra finches. Anim. Behav. 30: 444-55. Buss, L.W. 1987 . The evolution of individuality. Princeton : Princeto n Universit y Press. Butcher, G.S., and S. Rohwer. 1989 . The evolution of conspicuous an d distinctive coloration for communicatio n i n birds. Current Ornithol. 6: 51-108. Carlisle, T.R. , an d A . Zahavi . 1986 . Helpin g a t th e nest , allofeedin g an d socia l statu s in immature Arabian babblers. Behav. Ecol. Sociobiol. 18: 339-51. Carmeli, Z . 1988 . Mobbin g behaviou r i n th e Arabia n babbler (Turdoides squamiceps) (i n Hebrew). M.S. thesis, Hebrew University, Jerusalem. Caro, T.M. 1994 . Ungulate antipredator behaviour : Preliminary and comparative data from African bovids . Behaviour 128: 189-228 . Caro, T.M., an d D.W. Seller. 1990 . The reproductive advantages of fat in women. Ecol. and Sociobiol. 11 : 51-66. Christy, J.H. 1988 . Pilla r function in the fiddler crab Uca beebei. II: Competitive courtship signalling. Ethology 78: 113-28. Glutton-Brock, T.H., an d S.D . Albon. 1978 . Th e roarin g of red-deer an d th e evolution of honest advertising . Behaviour 69: 143-69. Glutton-Brock, T.H., F.E . Guinness , an d S.D . Albon. 1982 . Red Deer: Th e Behaviour and Ecology o f Two Sexes. Chicago: University of Chicago Press. Collins, S. 1993. Is there only one type of male handicap? Proc. R. Soc. Lond. B. 252:193-97.

Bibliography 2 4

7

Cresswell, W. 1994 . Son g as a pursuit-deterrent signal , and it s occurrence relativ e to other anti-predation behaviour s o f skylark (Alauda arvensis) o n attac k by merlins (Falco columbarius). Behav. Ecol. Sociobiol. 34: 217-23. Cronin, H. 1991 . Th e ant and the peacock. Cambridge, UK: Cambridge University Press. Cullen, J.M. 1966 . Reduction of ambiguity through ritualization. Phil. Trans., R. Soc. B. 251: 363-74. Curio, E. 1978 . Th e adaptiv e significance of avian mobbing. Z. Tierpsychol. 48 : 175-83 . Darling, F.F. 1937 . A Herd of Red Deer. Oxford: Oxfor d Universit y Press. . 1938 . Bird flocks and the breeding cycle. Cambridge : Cambridge University Press. Darwin, C . 1859 . O n th e origin o f species b y means of natural selection, or the preservation of favoured races in th e struggle for life. London : John Murray. . 1871 . Th e descent o f ma n and selection in relation to sex. Facsimile ed . Princeton : Princeton University Press , 1981 . . 1872 . Th e expression o f th e emotions in man and animals. London: John Murray. Davies, N.B. , an d M . Brook . 1989 . A n experimenta l stud y o f co-evolutio n betwee n th e cuckoo Cuculus canorus and its hosts. II: Host egg markings, chick discrimination and general discussion. /. Anim. Ecol. 58: 225-36. Davies, N.B. , an d T.R. Halliday. 1978. Dee p croaks and fighting assessment i n toads (Bufo bufo). Nature 274: 683-S5. Davis, G.W.F., and P. O'Donald . 1976 . Sexua l selection fo r a handicap: A critical analysis of Zahavi's model. /. Theor. Biol. 57: 345-54. Dawkins, R. 1989. The selfish gene. 2nd ed . Oxford : Oxford Universit y Press. Dawkins, R. , an d T.R . Carlisle . 1976 . Parenta l investment , mat e desertio n an d a fallacy . Nature 262: 131-33 . Dawkins, R. , and J.R. Krebs . 1979 . Arms races between an d within species. Proc. Roy. Soc. Land. B. 205: 489-511. De Waal, F.B.M. 1995 . Bonob o se x and society. Set. Am. 27 2 (3): 58-65. De Waal , F.B.M. 1996 . Good natured: The origins of right and wrong in humans and other animals. Cambridge, MA : Harvard Universit y Press. Diamond, J. 1986a . Animal art: Variation in bower decoratin g style among male bowerbird s (Amblyornis inornatus). Proc. Nail. Acad. Sci. USA. 83: 3042^46. . 1986b . Biolog y of birds o f paradise and bowerbirds . Ann. Rev. Ecol. Sys. 17: 1737. . 1987 . Flock s of brown an d black new-guinean birds : A bicoloured mixed-species foraging association . Emu 87: 201-11. . 1990 . Kung fu kerosene drinking. Natural History 7 : 20-24. . 1992 . Th e rise and fall o f th e third chimpanzee. London: Radius. Diamond, J.M, 1991 . Borrowe d sexual ornaments. Nature 349: 105 . Dominey, W.J . 1980 . Femal e mimicr y in male bluegill sunfish— a geneti c polymorphism ? Nature 284: 546^8. Eibel-Eibelsfeldt, I. 1961 . Th e fighting behavior of animals. Sci. Amer. 205: 112-28. . 1970 . Ethology, the Biology o f Behavior. New York: Holt, Rinehart, and Winston . . 1971 . Love and Hate. London: Methuen . Eisner, T. , an d J. Meinwald . 1987 . Alkaloid-derive d pheromone s an d sexua l selection i n Lepidoptera. I n Pheromone biochemistry, ed. G.D. Prestwich and G.J. Blomquist, 251— 69. Orlando, FL : Academic Press.

248 BIBLIOGRAPH

Y

. 1995 . Th e chemistr y of sexual selection. Proc. Natl. Acad. Set. US A 92: 50-55. Ellis, P.E., L.C . Brimacombe, LJ. McVeigh, and A. Dignan. 1980. Laboratory experiments on th e disruptio n o f matin g in th e Egyptia n cotto n leafwor m Spodoptera littoralis (Lepidoptera: Noctuidae ) b y excesse s o f femal e pheromones . Bull. Ent. Res. 70 : 673-84. Emlen, S.T., N.J. Demong, an d DJ. Emlen. 1989. Experimenta l induction of infanticide in female wattled jacanas. Auk 106 : 1-7 . Emlen, S.T., and L.W . Oring. 1977 . Ecology, sexual selection an d the evolution of mating systems. Science 197 : 215-23. Endler, J.A. 1983. Natural and sexual selection of color patterns in Poeciliid fishes. Environ. Biol. o f Fishes 9: 173-90. . 1987 . Predation , light intensity, and courtshi p behaviour in Poecilia reticulata (Pisces: Poeciliidae). Anim. Behav. 35: 1376-85. Engels, W., and V.L. Imperatriz-Fonseca. 1990. Caste development, reproductive strategies and control of fertility in honey bees and stingless bees. In Social insects: An evolutionary approach t o caste and reproduction, ed . W. Engels , 167-230. Berlin: Springer-Verlag. Eshel, I, 1978a. A critical defence of the handicap principle. /. Theor. Biol. 70: 245-50. . 1978b . O n a prey-predator nonzero-su m game an d th e evolutio n o f gregarious behavior o f evasive prey. Amer. Nat. 112 : 787-95. Eshel, I., and U. Metro. 1988 . Th e three brothers' problem—kin selection with more than one potential helper . I: The case of immediate help. Amer. Nat. 132 : 550-66. Evans, C.S. 1985 . Displa y vigour and subsequent fight performance i n the Siamese fighting fish Betta splendens. Behavioural Processes 11: 113-21. Evans, M.R. 1991. The siz e of adornments of male scarlet-tufted malachite sunbirds varies with environmenta l conditions, a s predicted b y handicap theories . Anim. Behav. 42: 797-803. Evans, M.R., and A.L.R . Thomas. 1992 . The aerodynami c and mechanica l effects o f elongated tails in the scarlet-tufted malachite sunbird: the cost of a handicap. Anim. Behav. 43: 337-47. Ewer, R.F. 1968 . Ethology o f mammals. London: Logo s Press . Faaborg, J., an d J.C. Bednarz . 1990 . Galapago s an d Harris ' hawks : Divergent cause s of sociality in tw o raptors . I n Cooperative breeding i n birds, ed . P.B . Stace y and W.D . Koenig, 357-84. New York: Cambridge University Press. Feare, C.J., G.M. Dunnet , and LJ. Patterson. 1974 . Ecological studie s of the rook (Corvus frugilegus L. ) in north-east Scotland ; foo d intake an d feeding behavior . /. Appl. Ecol. 11: 867-896. Feldman, M.W., and I. Eshel. 1982 . On the theory of parent-offspring conflict : A two-locus genetic model. Amer. Nat. 119 : 285-92. Fisher, R.A. 1930 . Th e genetical theory o f natural selection. London: Clarendo n Press . Fishman, L . 1977 . Wheatears—liv e warnin g system s agains t snake s (i n Hebrew) , leva Va'aretz 19 : 198 . FitzGibbon, C.D. , an d J.H. Fanshawe . 1988 . Slottin g in Thompson's gazelle : An hones t signal of condition. Behav. Ecol. Sociohiol. 23: 69-74. Forslund, P . 1990 . Mat e chang e reduces reproductio n i n th e barnacl e goose (Branta leucopsis)—the advantag e of having an old partner. Acta XX Con. Inter. Ornithol. 470.

Bibliography 2 4

9

Foster, M.S . 1981. Cooperativ e behavior an d social organization in the swallow-tailed manakin (Chiroxiphia caudata). Behav. Ecol. Sociobiol. 9: 167-77. Frith, C.B. , and D.W . Frith . 1990 . Archbold' s bowerbird , Archboldia papuensis (Ptilono rhynchidae), uses plumes from king bird o f paradise, Pteridophora alberti (Paradisaeidae), as bower decoration . Em u 90 : 136-37. Fugle, G.N., S.I . Rothstein, C.W. Osenberg, an d M.A. McFinley. 1984 . Signal s of status in wintering white-crowne d sparrow s (Zonotrichia leucoptrys gambeli). Anim. Behav. 32: 86-93. Fujita, K. , D.S . Blough , an d P.M . Blough . 1991 . Pigeon s se e the Ponz o illusion . Anim. Learning Behav. 19:283-93. Gadagkar, R. 1991. Belonogaster, Mischocyttarus, Parapolybia, an d independent-founding Ropalidia. In The social biology o f wasps, ed . K.G. Ross and R.W. Matthews, 149-200. Ithaca, NY : Cornell University Press. Gadagkar, R., and N.V . Joshi. 1985 . Colon y fission in a social wasp. Current Sci. 54:57-62. Gaioni, S.J. , an d C.S . Evans . 1985 . Th e rol e o f frequenc y modulation i n controllin g th e response of mallard ducklings (Anas platyrhynchos) to conspecific distress calls. Anim. Behav. 33: 188-200. . 1986a . Mallard ducklin g respons e to distres s calls with reduced variability: Con straint on stereotypy in a "fixed actio n pattern." Ethology 72: 1-14 . 1986b. Perception o f distress call s i n mallard duckling s (Anas platyrhynchos). Be haviour 99: 250-74. Gaston, AJ . 1977 . Socia l behaviour withi n groups o f jungle babblers (Turdoides striatus). Anim. Behav. 25: 828^8. . 1978. The evolution o f group territorial behaviour and cooperative breeding. Amer. Nat. 112 : 1091-1100 . Gibbs, H.L., an d P.R. Grant . 1987 . Oscillatin g selectio n o n Darwin's finches. Nature 327: 511-13. Gibson, R.M. , J.W. Bradbury , an d S.L . Vehrencamp. 1991 . Mat e choic e i n lekking sag e grouse revisited : Th e role s o f vocal display , femal e site fidelit y an d copying . Behav. Ecol. 2: 165-80. Gibson, R.M., and J. Hoglund . 1992 . Copyin g and sexual selection. Tree 7: 229-32. Goldstein, H. , D . Eisikovitz, and Y. Yom-Tov. 1986 . Infanticid e in the Palestine sunbird s Nectarina osea. Condor 88 : 528-29. Grafen, A . 1990a. Biologica l signals as handicaps. / . Theor. Biol. 144 : 517-46. . 1990b . Sexua l selection unhandicappe d b y the Fisher process . /. Theor. Biol. 144: 473-516. Greenewalt, C.H. 1986 . Bird song: Acoustics and physiology. Washington , DC: Smithsonian Institution Press . Gustafsson, L . 1989 . Collare d flycatcher. In Lifetime reproduction i n birds, ed . I. Newton , 75-88. London: Academi c Press. Hailman, J.P. 1977 . Optical signals: Animal communication and light. Bloomington, IN : Indiana Universit y Press. Haldane, J.B.S. 1932. Th e Causes of Evolution. London: Longmans Green. Reprint. Princeton: Princeton Scienc e Library, 1990. . 1955 . Populatio n genetics . New Biology 18 : 34-51.

25O BIBLIOGRAPH

Y

Hamilton, W.D. 1964 . The genetical evolution of social behaviour. /. Theor. Biol 1: 1-52 . Hamilton, W.D. , an d M . Zuk . 1982 . Heritabl e tru e fitnes s an d brigh t birds : A rol e fo r parasites? Science 218: 384-87. Hamilton, WJ. III. 1973 . Life's color code. New York: McGraw-Hill . Hunter, C.P. , an d P.D. Dwyer . The value of objects to satin bowerbirds (Ptilonorhynchu s violaceus). Emu (i n press). Hasson, O . 1991a . Pursuit-deterren t signals : Communication between pre y and predator . Tree 6: 325-29. . 1991b. Sexual displays as amplifiers: Practical examples with an emphasis on feather decorations. Behav. Ecol. 2: 189-97. Hasson, O. , R . Hibbard, an d G . Cebballos . 1989 . Th e pursuit-deterren t function of tail wagging in the zebra-taile d lizard (Callisaurus draconoides). Canad. ]. Zoo/. 67: 1203 09. Heinrich, B. 1988 . Winte r foragin g at carcasse s by the thre e sympatri c corvids, with emphasis on recruitment by the raven , Corvus corax. Behav. Ecol. Sociobiol. 23: 141-56. Heinroth, O.J., an d M. Heinroth. 1926. Die Voget Mitteleuropas. Berlin-Lichterfelde : Hugo Bermuhler-Verlag. Heinze, J., B . Holldobler, and C . Peeters. 1994 . Conflic t an d cooperatio n i n ant societies. Naturwissenschaften 81 : 489-97. Hendry, L.B., E.D. Bransome, M.S. Hutson, and L.K. Campbell. 1984 . A newly discovered stereochemical logic in the structure of DNA suggests that the genetic code is inevitable. Perspectives i n Biology an d Medicine 27 : 623-51. Hendry, L.B., and V.B. Mahesh. 1995 . A putative step i n steroid hormone action involves insertion o f steroi d ligand s int o DN A facilitate d b y recepto r proteins . / . Steroid Biochem. Molec. Biol. 55: 173-83. Heredia, B. , J.C. Alonso , and F. Hiraldo. 1991 . Spac e and habitat use by red kites, Milvus milvus, durin g winte r i n th e Guadalquivi r marshes : A comparison betwee n residen t and wintering populations. Ibis 133: 374-81. Hess, H.E. 1965 . Attitud e and pupil size. Set. Amer. 212: 46-54. Hill, G.E. 1990 . Female house finches prefer colorful mates: Sexual selection for a conditiondependent trait . Anim. Behav. 40: 563-72. Kingston, R.W.G. 1933 . Th e meaning of animal colour and adornment. London: Arnold. Hiraldo, F. , B. Heredia, an d J.C. Alonso . 1993 . Communal roosting of wintering red kites, Milvus milvus (Aves, Accipitridae): Social feeding strategies for the exploitation of food resources. Ethology 93 : 117-24. Hoglund, J., R. Montgomerie, and F. Widemo. 1993 . Costs and consequences of variation in the siz e of ruff leks . Behav. Ecol. Sociobiol. 32: 31-39. Hogan-Warburg, A J. 1966. Social behavior of the ruff, Philomachus pugnax (L). Ardea 54: 109-229. Holder, K. and R. Montgomerie. 1993. Context and consequences of comb displays by male rock ptarmigan. Anim. Behav. 45: 457-70. Holldobler, B. , and E.O. Wilson . 1990 . Th e ants. Cambridge: Harvard Universit y Press. Hultsch, H. , an d D . Todt . 1986 . Signa l matching: Zeichenbildun g durc h mustergleiche s Antworten. Semiotik 8 : 233-44. Huxley, J.S. 1914 . The courtshi p habits of the great crested grebe (Podiceps cristatus) with an addition to the theory of sexual selection. Proc. Zool. Soc. London 35: 491-562.

Bibliography 2 5

1

Ingle, A . 1911. Cuckoos a s nest robbers . Em u 11 : 254-55 . Ishay, ]., H . Bitinsky-Salz , an d A . Shulov . 1967 . Contribution s t o th e bionomic s o f th e oriental horne t (Vespa orientalis Fab.). Israel]. Entomol. 2: 45-106. Ishay, J., an d R. Ikan. 1968 . Foo d exchange between adult s and larva e in Vespa orientalis. Anim. Behav. 16: 298-303. Ito, Y . 1993 . Behaviour and social evolution o f wasps. Oxfor d Serie s in Ecolog y and Evo lution. Oxford: Oxford Universit y Press. Jackson, C.L. , and L.H . Hartwell . 1990a . Courtshi p i n S. cerevisiae: Both cel l types choos e mating partners by responding to the stronges t pheromon e signal . Cell 63: 1039-51. . 1990b . Courtshi p i n Saccharomyces cerevisiae: An early cell-cell interactio n durin g mating. Cell Biol. 10 : 2202-13. Jarvi, T., and M. Baker. 1984 . Th e function of the variation in the breast strip e of the great tit (Parus major). Anim. Behav. 32: 590-96. Jones, I.L. , an d P.M . Hunter . 1993 . Mutua l sexua l selectio n i n a monogamous seabird . Nature 362: 218. Kalishov, A. 1996. Allofeeding among babblers (Turdoides squamiceps) (in Hebrew, with English summary). M.S. thesis, Tel-Aviv University. Katsir, Z. 1985. Vocal signal affected b y body posture and movement in the Arabian babbler (Turdoides squamiceps, Aves : Timaliidae). 19th InternationalEthological Conference, 2 : 398. Toulouse , France : Sabatie r University. . 1991 . Message s i n voca l communication : Investigation s o f th e variation s in th e babbler "shout " (in Hebrew). Ph.D . thesis , Hebrew University , Jerusalem. Katsir, Z . 1995 . Th e meanin g o f th e "variations " i n th e babble r "shout" : A musicalethological approach . Behav. Processes 34: 213-32. Katzir, G. 1981a . Aggressio n by the damselfish , Dascyllus aruanus L., towards conspecific s and heterospecifics. Anim. Behav. 29: 835^41. . 1981b . Visual aspects o f specie s recognitio n i n the damselfish , Dascyllus aruanus. Anim. Behav. 29: 842-49. Kaufman, K . 1991 . Th e subjec t is Alex. Audubon, September-October , 52-58. Kirkpatrick, M . 1986 . Th e handica p mechanis m o f sexua l selectio n doe s no t work . Am . Nat. 127 : 222-40. Koford, R.R., B.S. Bowen, an d S.L. Vehrencamp. 1990. Groove-bille d anis: Joint-nesting in a tropical cuckoo. I n Cooperative breeding i n birds, ed . P.B. Stace y and W.D. Koenig, 333-56. New York: Cambridge Universit y Press. Komdeur, J. 1992 . Importanc e o f habitat saturatio n an d territor y qualit y for evolutio n of cooperative breeding in the Seychelle s warbler. Nature 358: 493-95. . 1994 . Experimenta l evidenc e fo r helpin g an d hinderin g b y previous offsprin g i n the cooperative-breedin g Seychelle s warbler, Acrocephalus seychellensis. Behav. Ecol. Sociobiol. 34: 175-86. Kraak, S.B.M . 1994 . Femal e mat e choice in Aidablennius sphynx, a fish with parenta l care for egg s in a nest. Ph.D. thesis , University of Groningen. Krebs, J.R., an d R . Dawkins . 1984 . Anima l signals : Mind-readin g an d manipulation . I n Behavioral ecology: An evolutionary approach, ed . J.R. Krebs and N.B. Davies, 2nd ed., pp. 380-402. Oxford: Blackwel l Scientific Publications . Krebs, J.R., and N.B . Davies. 1993 . An introduction to behavioural ecology. Oxford : Blackwell Scientifi c Publications .

252 BIBLIOGRAPH

Y

Kruuk, H. 1972 . The spotted hyena. Chicago: University of Chicago Press . Kuris, A.M . 1974 . Trophi c interaction : Similarity of parasitic castrators to parasitoids . Q. Rev. Eiol. 49: 129-48. Lack, D. 1968 . Ecological adaptations for breeding i n birds. London : Chapman and Hall . . 1970. Th e endemic ducks o f remote islands. Wildfowl 21 : 5-10 . Lambrechts, M., and A. Dhondt. 1986 . Mal e quality, reproduction, an d survival in the great tit (Parus major). Behav. Ecol. Sodobiol. 19 : 57-63. Leader, N., 1996 . The function of stone ramparts at the entrance of blackstarts' (Cercomela melanura) nest s (in Hebrew). M . Sc. thesis, Tel-Aviv University. Loffredo, C.A. , and G . Borgia . 1986. Mal e courtship vocalization s a s cues for mate choice in sati n bowerbird (Ptilonorhynchus violaceus). Auk 109 : 189-95 . Lorenz, K. 1952. King Solomon's ring. London: Methuen. . 1962 . Th e function of colour i n coral reef fishes. Proc. R. Inst. Gr . Br. 39: 282-96. . 1966 . O n aggression. London: Methuen . Lotem, A. 1993a. Secondar y sexual ornaments as signals: The handicap approach and three potential problems. Etologia 3: 209-18. . 1993b . Learning to recognize nestlings is maladaptive for cuckoo (Cuculus canorus) hosts. Nature 362: 743^15. Lotem, A., H. Nakamura, and A. Zahavi. 1991 . Rejectio n o f cuckoo egg in relation to host age: A possible evolutionar y equilibrium. Behav. Ecol. 3: 128-32. . 1995 . Constraint s on egg discrimination an d cuckoo-hos t co-evolution . Anim. Behav. 11:43-54 . Lythgoe, J.N. 1979 . Th e ecology o f vision. Oxford: Clarendon Press . Maier, EJ. 1993 . T o deal with the "invisible": On th e biological significanc e of ultraviole t sensitivity in birds. Naturwissenschaften 80: 476-78. Marler, P . 1955 . Characteristic s of some anima l calls. Nature 176: 6 . Marler, P. , an d WJ . Hamilton . 1966 . Mechanisms of animal behavior. New York : Joh n Wiley and Sons . Mathews, C.K., E.M. Kutter, G. Mosig, and P.B. Berget. 1983. Bacteriophage T4. American Soc. for Microbiology, Washington , B.C. Maynard Smith,}. 1964. Group selection an d kin selection. Nature201: 1145—47 . . 1965 . Th e evolutio n of alarm calls. Amer. Nat. 99: 183-88. . 1976a . Evolution and the theory o f game. Am. Sci. 64: 41^t5. . 1976b . Sexua l selection an d the handicap principle . /. Theor. Eiol. 57: 239-42. . 1976c . Grou p selection . Quar. Rev. Eiol. 51: 277-83. . 1982. Do animals convey information about their intentions? /. Theor. Eiol. 97:1-5. . 1991a . Theorie s o f sexual selection. Tree 6 : 146-51. . 1991b. Must reliabl e signals always be costly? Anim. Behav. 47: 1115-20. Maynard Smith, ]., an d G.A. Parker. 1976 . The logic of asymmetric contests. Anim. Behav. 24: 159-75 . Mayr, E. 1942 . Systematic^ an d the origin o f species. New York: Columbia Universit y Press. . 1972 . Sexual selection an d natura l selection. I n Sexual selection and the descent of man 1871-1971, ed. B. Campbell, 97. Los Angeles: University of California Press. Mayr, E. , an d E . Stresemann . 1950 . Polymorphos m i n th e cha t genu s Oenantbe (Aves) . Evolution 4: 291-300. McDonald, D.B. , and W.K. Potts. 1994 . Cooperative displa y and relatedness among males in a lek-mating bird. Science 266: 1030-32 .

Bibliography 2 5

3

McGregor, P.K . 1993 . Signalling in territorial systems : A context fo r individual identifica tion, ranging and eavesdropping. Phil. Trans. R. Soc. Land. B 340: 237-44. McKaye, K.R. 1991. Sexua l selection an d the evolution of the cichlid fishes of Lake Malawi. In Cichlid fishes ' behaviour, ecology an d evolution, ed. M.H.A . Keenleyside , 241-57. London: Chapma n and Hall. Mock, D.W. 1984 . Siblicidal aggression and resource monopolization in birds. Science 225: 731-33. Moller, A.P . 1990a . Fluctuatin g asymmetr y in mal e sexual ornaments may reliably reveal male quality. Anim. Behav. 40: 1185-87. . 1990b . Mal e tail length an d femal e mate choice in the monogamous swallow, Hirundo rustica. Anim. Behav. 39: 458-65. . 1990c . Sexua l behavior is related to badge size in the house sparrow Passer domesticus. Behav. Ecol. Sociobiol. 27: 23-29. . 1991. Influence of wing and morphology on the duration of song flights in skylarks. Behav. Ecol. Sociobiol. 28: 309-14. . 1992 . Femal e swallow preferences for symmetrical male sexual ornaments. Nature 357: 238-240. . 1994 . Sexual selection and the barn-swallow. Oxford: Oxford University Press. Montgomerie, R . (convenor) . 1986 . Symposiu m on mat e guarding . Acta XIX Con. Inter. Ornithol. 408-53. Ottawa . Moore, AJ. 1988. Female preferences, male social status, and sexual selection in Nauphoeta cinerea. Anim. Behav. 36: 303-5. Moreno,}., M. Soler, A.P. Moller, an d M . Linder. 1994 . The function of stone carrying in the black wheatear (Oenanthe leucora). Animal Behav. 47: 1297-1309. Morris, D. 1956 . Th e function an d causatio n of courtship ceremonies . In L'instinct dans le comportement des animaux et de I'homme, ed. P.P. Grasse , 261-86. Fondation SingerPolignae. Paris: Masson. . 1957 . "Typica l intensity " an d it s relationship t o th e proble m o f ritualization. Behaviour 11: 1-12 . . 1967 . The naked ape. Ne w York: McGraw-Hill . . 1990 . Animalwatching. New York: Crown. Morton, E.S . 1977 . O n th e occurrenc e an d significanc e of motivation-structural rule s in some birds an d mammal sound. Amer. Nat. Ill: 855-69. Morton, E.S. , L. Forman, and M. Braun. 1990. Extrapai r fertilizations and the evolution of colonial breeding in purple martins. Auk 107 : 275-83. Morton, E.S. , and J. Page. 1992 . Animal talk. New York: Random House. Motro, U. , and I. Eshel. 1988 . Th e three brothers' problem : Ki n selection wit h more than one potential helper. II: The case of delayed help. Amer. Nat. 132 : 567-75. Munehara, H., A. Takenaka, and O. Takenaka. 1994. Alloparental care in the marine sculpin (Alcichthys alcicornis, Pisces: Cottidae): Copulating in conjunction with parental care. /. Ethol. 12: 115-20. Murray, I.R., and J.L. Arnott. 1993. Towards the simulation of emotion in synthetic speech: A revie w o f th e literatur e o n huma n voca l emotion . / . Acoust. Soc. Am. 92 : 1097 1108. Nahon, E., D. Atzmony, A. Zahavi, and D. Granot. 1995 . Mate selection i n yeast: A reconsideration o f the signal s and the message encoded b y them. /. Theor. Biol. 172 : 31522.

254 BIBLIOGRAPH

Y

Nanjundiah, V. , and S . Saran. 1992. Th e determinatio n o f spatial patter n i n Dictyostelium discoideum. J. Biosci. 17 : 353-93. Newton, I . 1989a . Sparrowhawk . In Lifetime reproduction i n birds, ed . I. Newton, 279-96. London: Academic Press. . 1989b. Synthesis. In Lifetime reproduction i n birds, ed. I. Newton, 279-96. London: Academic Press. Nisbet, I.C.T . 1973 . Courtship feeding , eg g size an d breedin g succes s i n commo n terns . Nature 241: 141-42. . 1977 . Courtship feedin g an d clutc h siz e i n commo n tern s (Sterna hirundo). In Evolutionary ecology, ed . B . Stonehous e an d C.M . Perrins, 101-9 . London : Macmill an. Norris, KJ . 1990 . Female choic e an d th e evolutio n o f conspicious plumag e coloratio n of monogamous male great tits. Behav. Ecol. Sociobiol. 26: 129-38. Oberski, S . 1975. Parasite reproduction strateg y and the evolution of castration of hosts by parasites. Science 188: 1314-16. O'Donald, P. 1963 . Sexua l selection an d territorial behaviour . Heredity 18 : 361-64. Orians, G . 1969 . On th e evolutio n o f mating system s in birds an d mammals . Amer. Nat. 103: 589-603. Orians, G.H., and L.D. Beletsky. 1989. Red-winged blackbird. I n Lifetime reproduction i n birds, ed . I. Newton, 183-98 . London : Academic Press . Osztreiher, R . 1992. The mornin g dance of the Arabia n babbler, Turdoides squamiceps (in Hebrew). M.S . thesis, Tel-Aviv University. . 1996. The competitive relationship amon g Arabian babbler (Turdoides squamiceps) nestlings (i n Hebrew). Ph.D . thesis, Tel-Aviv University. Owen, M., an d J.M. Black. 1989 . Barnacle goose. I n Lifetime reproduction i n birds, ed . I . Newton, 349-62 . London: Academic Press. Parker-Rabenold, P . 1987 . Recruitment t o foo d i n black vultures : Evidenc e fo r followin g from communa l roosts. Anim. Behav. 35: 1775-85 . Parsons, P.A . 1990. Fluctuating asymmetry : An epigeneti c measur e o f stress . Biol. Rev. Cambr. Phil. Soc. 65: 131-45. Payne, R.B . 1983. The socia l contex t o f son g mimicry : Song-matching dialect s i n indig o bunting (Passerina cyanea). Anim. Behav. 31: 788-805. Pepperberg, I.M . 1991 . A communicative approac h t o anima l cognition : A study of conceptual abilitie s o f a n Africa n gre y parrot. I n Cognitive ethology: Th e minds o f other animals, ed . C.A. Ristau , 153-86 . Hillsdale , NJ, and London : Lawrenc e Erlbaum Associates. Perl, J. 1996 . Competition fo r breeding between Arabia n babbler male s (in Hebrew). M.S. thesis, Tel-Aviv University. Petrie, M.,T. Halliday, andC. Sanders. 1991. Peahens prefer peacocks with elaborate trains. Anim. Behav. 41: 323-31. Pliske, T.E. , and T . Eisner . 1969 . Sex pheromone o f the quee n butterfly : Biology. Science 164: 1170-72. Pomiankowski, A . 1987. Sexual selection: Th e handicap principl e does work—sometimes. Proc. R. Soc. Land. B. 231: 123-45. Pond, G . 1973 . An introduction to lace. London: Garnstone Press . Pozis, O . 1984 . Play in babblers (i n Hebrew). M.S. thesis, Tel-Aviv University. Rasa, A. 1986. Mongoose watch. New York : Doubleday .

Bibliography 2 5

5

Redondo, T., and F. Castro. 1992 . Signalling of nutritional need by magpie nestlings. Ethology 92 : 193-204. Reyer, H.U . 1990 . Pie d kingfishers : ecological cause s an d reproductiv e consequence s of cooperative breeding . In Cooperative Breeding in Birds, ed . P.B. Stacey and W.D. Koenig, pp. 527-559. Cambridge, UK: Cambridge Universit y Press. Rhisiart, A.P. 1989 . Communicatio n an d antipredator behaviour. D.Phil, thesis, University of Oxford . Ridley, M. 1981 . Ho w the peacoc k go t his tail. New Sci. 91: 398-401. Ritland, D.B . 1991a . Unpalatabilit y o f vicero y butterflie s (Limenitis archippus) an d their purporte d mimicr y models , Florid a queen s (Danaus gilippus). Oecologia 88 : 102-8. . 1991b . Revising a classic butterfly mimicry scenario: Demonstration o f Muellerian mimicry between Florida viceroys (Limenitis archippus floridensis) and queens (Danaus gilippus berenice). Evolution 45: 918-34. Rohwer, S., and P.W. Ewald. 1981. Th e cos t of dominance and advantage of subordination in a badge signalin g system. Evolution 35: 441-54. Rohwer, S. , an d F.C . Rohwer . 1987 . Statu s signallin g in Harris ' sparrows : Experimenta l deceptions achieved. Anim. Behav. 26: 1012-22. Roper, T. 1986 . Badge s of status in avian societies. Ne w Sci. 109: 38^(0. Roseler, P.F. , an d C.GJ . Honk . 1989 . Caste an d reproductio n i n bumblebees. I n Social insects: An evolutionary approach t o caste an d reproduction, ed . W . Engels, 147-166 . Berlin: Springer-Verlag. Rosenberg, E . 1984 . Myxobacteria. Ne w York: Springer-Verlag. Rothstein, S.I . 1990 . A model syste m for coevolution : avia n brood parasitism . Ann. Rev. Ecol. Syst. 21 : 481-508. Rowel, T.E . 1962 . Agonisti c noise s o f th e rhesu s monke y (Macaca mulata). Proc. R . Soc. London 138 : 91-96. Rowley, I., and E . Russel. 1989 . Splendi d fairy-wren . In Lifetime reproduction i n birds, ed . I. Newton, 233-52. London: Academic Press. . 1990 . Splendi d fairy-wren : Demonstrating th e importanc e o f longevity. In Cooperative breeding i n birds, ed . P.B . Stace y and W.D . Koenig , 1-30 . Cambridge , UK : Cambridge University Press . Ryan, MJ. 1990 . Sexua l selection sensor y systems, and sensor y exploitation. Oxford Surv. Evol. Biol. 7: 157-95. Ryan, MJ., M.D. Tuttle , an d A.S . Rand. 1982 . Bat predation an d sexual advertisement in a neo-tropical anuran. Amer. Nat. 119 : 136—39 . Sade, D.S. 1972 . A longitudinal stud y of social behavior o f Rhesus monkeys. In The functional an d evolutionary biology o f primates, ed. R . Tuttle , 378-98 . Chicago : Aldine Atherton. Saino, N., A.P. Moller, and A.M. Bolzern. 1995 . Testosterone effect s on the immune system and parasit e infestations in the bar n swallo w (Hirundo rustica): A n experimental tes t of the immunocompetenc e hypothesis . Behav. Ecol. 6: 397-404. Salyers, A.A. , and D.D . Whit . 1994 . Bacterial pathogenesis. Washington, DC : American Society for Microbiology . Scherer, R.K . 1979 . Nonlinguisti c voca l indicator s o f emotio n an d psychopathology . I n Emotions in personality an d psychopathology, ed . C.E. Izard, 493-525. New York: Plenum.

256 BIBLIOGRAPH

Y

. 1985 . Voca l affec t signalling : A comparativ e approach . Advances in th e study o f Behav. 15 : 189-244. Schjelderup-Ebbe, T. 1922 . Beitrag e zur Sozialpsychologi e de s Haushuhns. Z. Psychol. 88: 225-52. Schneider, D . 1992 . 10 0 years of pheromone research . Naturwissenschaften 79 : 241-50. Schwammberger, K. 1993. Freilandbeobachtungen zur Nestiibernahme bei Polistes biglumis bimaculatus durc h de n Sozialparasite n Sulcopolistes atrimandibularis (Hymenoptera , Vespidae). Zeitschrift fu r Angewandte Zoologie 79: 291-97. Scott, D.H. 1988 . Reproductiv e succes s in Bewick's swan. In Reproductive success, ed. T.H . Glutton-Brock, 220-36. Chicago: University of Chicago Press . Selander, R.K . 1965 . O n matin g systems and sexua l selection. Amer. Nat. 99: 129-41. . 1972. Sexua l selection an d dimorphism i n birds. In Sexual selection and the descent of ma n 1871-1971, ed . B . Campbell, 180-230 . Los Angeles : Universit y of California Press. Sever, Z. , an d H . Mendelssohn . 1988 . Copulatio n a s a possible mechanis m t o maintai n monogamy in porcupines, Hysterix indica. Anim. Behav. 36: 1541-42. Seyfarth, R.M. , D.L . Cheney , an d P . Marler . 1980 . Verve t monke y alar m calls: Semantic communication i n a free-ranging primate. Anim. Behav. 28: 1070-94. Shaulsky, G. , an d W.F . Loomis . 1993 . Cel l typ e regulatio n i n respons e t o expressio n o f ricin A in Dictyostelium. Dev. Biol. 160 : 85-98. . 1995 . Mitochondria l DN A replicatio n bu t n o nuclea r DN A replicatio n durin g development o f Dictyostelium. Proc. Natnl. Acad. Sci. 92: 5660-63. Shepher, J. 1983 . Incest: A biosocial view. New York: Academic Press . Sherman, P.W. , J.U.M. Jarvis, an d R.D . Alexander . 1991 . Th e biology o f th e naked molerat. Princeton : Princeto n University Press. Sigmund, K. 1993 . Games of life. Oxford : Oxfor d Universit y Press . Simpson, MJ.A . 1968 . Th e displa y o f th e Siames e fightin g fish (Betta splendens). Anim. Behav. Mon. 1 : 1-73. Slagsvold, T. 1984. The mobbing behaviour of the hooded crow, Corvus corone comix: Antipredator defenc e or self-advertisement? fauna Norv. Ser . C 7: 127-31. . 1985. Th e mobbing behaviour o f the hooded crow, Corvus corone comix, in relation to age, sex, size, season, temperature and the kind of enemy. Fauna Norv. Ser. C 8:9-17. Smith, H.G. , an d R. Montgomerie. 1991 . Sexua l selection an d the tail ornaments of North American barn swallows . Behav. Ecol. Sociobiol. 28: 195-201. Smith, H.G. , R . Montgomerie, T . Poldmaa, B.N. White , an d P.T. Boag . 1991 . DN A fingerprinting reveal s relatio n betwee n tai l ornament s an d cuckoldr y in barn swallows , Himndo rustica. Behav. Ecol. 2: 90-98. Smith, J.N.M., P. Arcese, and I.G. McLean. 1984 . Age, experience, and enemy recognitio n by wild son g sparrows . Behav. Ecol. Sociobiol. 14: 101-6. Smith, N.G. 1966 , Evolution o f some arctic gulls (Lams): A n experimental study of isolating mechanisms. Ornith. Monogr. (AOU) 4: 1-99. . 1967. Visua l isolation in gulls. Sci. Am. 217 : 94-102 . Smythe, N. 1970 . On th e existence o f "pursuit invitation " signal s in mammals. Amer. Nat. 104: 491-94. Snow, D.W. 1976 . The we b o f adaptation: Bird studies in the American tropics. Ne w York : New York Times Books .

Bibliography 257 Snyder, H.S., an d D.S. Bredt. 1992 . Biological roles of nitric oxide. Set. Am. 266 : 58-67. Soler, M.,J. Soler, J.G. Martinez , and A.P. Moller. 1995. Magpie host manipulation by great spotted cuckoos : Evidence fo r a n avian mafia? Evolution 49: 770-75. Sordahl, T.A . 1990 . Th e risk s o f avia n mobbin g an d distractio n behavior : A n anecdota l review. Wilson Bull. 102 : 349-352. Spiro, M.E . 1963 . Kibbutz, adventure in Utopia. Ne w York: Schocken Books . Stacey, P.B. , an d W.D . Koenig. 1990 . Cooperative breeding in birds: Long-term studies of ecology an d behaviour. Cambridge, UK: Cambridge Universit y Press. Strahl, S.D . 1988 . Th e social organization and behaviour of the hoatzin, Opisthocomus hoatzin, in central Venezuela. Ibis 130 : 483-501. Streeter, L.A., R.M. Krauss, V. Geller, C . Olson, and W. Apple. 1977 . Pitch chang e during attempted deception. /. o f Personality an d Soc. Psychol. 35: 345-50. Sugiyama, Y. 1967. Socia l organisation of hanuman langurs. In Social communication among primates, ed. S.A . Altman. Chicago: University of Chicago Press . Sutler, E. 1994 . Ar e the peacock' s trai n feather s really upper tai l coverts? /. fur Ornithol. 135: 57 . Taborsky, M. 1994. Parasitism and alternative reproductive strategie s in fish. Adv. i n Behav. Studies 28: 1-100. Tengo, J . 1984 . Territoria l behaviou r o f th e kleptoparasit e reduce s parasiti c pressur e i n communally nesting bees. Abstract volume: XVII International Congress of Entomology 510. Thornhill, R. 199 2 a. Female preference fo r th e pheromon e of males wit h lo w fluctuatin g asymmetry in th e Japanese scorpionfl y (Panorpa japonica, Mecoptera) . Behav. Ecol. 3: 277-83. . 1992b. Fluctuating asymmetry, interspecific aggressio n an d male mating tactic s in two species of Japanese scorpionflies . Behav. Ecol. Sotiobiol. 30: 357-63. Thornhill, R. , and J. Alcock. 1983. The evolution of insect mating systems. Cambridge, MA: Harvard Universit y Press. Tinbergen, N . 1953 . The herring gull's world. London: Collins . . 1965. Some recent studies of the evolution o f sexual behavior. In Sex and behavior, ed. F.A. Beach, 1-33 . Ne w York: Wiley. Todt, D. , and H. Hultsch. 1995 . Acquisitio n and performance of song repertoires: Ways of coping with diversity and versatility. In Ecology and evolution of acoustic communication in birds, ed . D. Kroodsma and E. Miller. Ithaca , NY: Cornell University Press . Trivers, R. 1971. Th e evolutio n o f reciprocal altruism . Quart. Rev. Biol. 46: 35-57. . 1974 . Parent-offspring conflict. Amer. Zoo/. 14 : 249-64. . 1985. Social evolution. Menlo Park, CA: Benjamin/Gumming Publication Company. Trivers, R.L . 1972 . Parenta l investmen t an d sexua l selection . In Sexual selection and the descent o f man 1871-1971, ed . B. Campbell, 136-179 . London: Heinemann . Ulloa-Aguirre, A. , A.R . Midgle y Jr., I.Z . Beitins , an d V . Padmanabhan . 1995 . Follicle stimulating isohormones : characterizatio n an d physiological relevance . Endochrine Reviews 16 : 765-787. Van de r Blom , J. 1986 . Reproductiv e dominanc e within colonie s o f Bombus terrestris. Behaviour 97: 37-49. Van Lawick , H., andj. Va n Lawick-Goodall. 1970 . Innocent killers. London: Collins . Van Lawick-Goodall , J. 1971 . In th e shadow of man. London: Collins .

258 BIBLIOGRAPH

Y

Van Rhijn, J.G. 1973 . Behavioura l dimorphism i n male ruffs (Philomachus pugnax L.). Behaviour 47: 153-229. Veblen, T . 1899 . Th e theory o f th e leisure class. New ed., ed. C.W. Mills. New Brunswick, NJ: Transaction Publishers , 1992 . Velthuis, H.W. 1990. Chemica l signal s and dominance communication in the honeybee Apis mellifera (Hymenoptera : Apidae). Entomol. Gener. 15(2): 83-90. Wagner, R . 1992. Extra-pair copulations i n a lek: The secondar y mating system of monogamous razorbills. Behav. Ecol. Sotiobiol. 31: 63-71. Wagner, R. , M.D . Schug , and E . Morton . 1996 . Confidenc e in paternity , actual paternity and paternal effort b y purple martins. Anim. Behav., in press. Wagner, R . H., 1996 . Wh y d o female birds rejec t copulation s fro m thei r mates? Ethology 102: 465-480. Wallace, A.R . 1889 . Darwinism: An exposition of th e theory o f natural selection with some of it s applications. London : Macmillan . Ward, P. 1965 . Feedin g ecolog y of the black-face d dioch (Quelea quelea) i n Nigeria. Ibis 107: 173-214. . 1972 . The functiona l significance of mass drinking flight s by sandgrouse (Pteroclididae). 7te 114 : 533-36. Ward, P. , an d A. Zahavi. 1973 . Th e importance of certain assemblages o f birds as "infor mation-centers" fo r food-finding. Ibis115: 517-34. Watson, J.P., an d R . Thornhill. 1994 . Fluctuatin g asymmetr y and sexua l selection. Tree 9: 21-25. Weiner, J. 1994 . Th e beak of th e finch. New York: Vintage. West, M.J. 1969 . Th e socia l biology of polistine wasps. Univ. Mich. Mus. Zoo/. Misc. Publ. 140: 1-101. West-Eberhard, M.J. 1975 . Th e evolutio n of social behavior by kin selection. Quar. Rev. Biol. 50: 1-33 . . 1984 . Sexua l selection, competitiv e communicatio n an d species-specifi c signals in insects. In Insect communication, ed. T. Lewis, 283-324. New York: Academic Press. . 1986 . Dominanc e relations i n Polistes canadensis (L.) , a tropical socia l wasp. Monitore Zoo/. Ital. (N.S. ) 20: 263-81. Wiklund, C., and T. Jarvi. 1982. Survival of distasteful insects afte r bein g attacked by naive birds: A reappraisal of the theory of aposematic coloration evolving through individua l selection. Evolution 36: 998-1002. Wilhelm, K. V., H. Comtesse , and W. Pfiumm. 1980 . Zu r Abhangigkeit des Gesangs vom Nahrungsangebot beim Gelbbauchnektarvogel (Nectarinia venusta.)Z, Tierpsychol. 54 : 185-202. . 1982 . Influenc e of sucros e solutio n concentratio n o n th e son g and courtshi p be haviour of the male yellow-bellied sunbir d (Nectarinia venusta). Z. Tierpsychol. 60:27 40. Williams, G.C. 1966 . Adaptation and natural selection: A critique of some current evolutionary thought. Princeton: Princeto n Universit y Press. . 1994. Natural selection: Domains, levels and challenges. Oxford: Oxford University Press. Willson, D.S. , an d E . Sober . 1994 . Reintroducin g grou p selectio n t o huma n behaviora l sciences, behavioral and ^>rain Sciences, 17 : 585-654.

Bibliography 2 5

9

Wilson, E.O . 1971. The insect societies. Cambridge: Harvard Universit y Press. . 1975 . Sociobiology: The ne w synthesis. Cambridge, MA : Belknap Press . Winston, M.L. , and K.N . Slessor. 1992 . The essenc e o f royalty: Honey be e quee n phero mone. Amer. Set. 80: 374-85. Witherby, H.F., F.C.R. Jourdain, N.F. Ticehurst, and B.W. Tucker. 1943 . The handbook of British birds. London : Witherby. Woolfenden, G.E. , and J.W. Fitzpatrick. 1984 . The Florida scrub jay. Princeton: Princeto n University Press. . 1990. Florida scrub jays: A synopsis after 1 8 years of study. In Cooperative Breeding in Birds, ed . P.B. Stacey and W.D. Koenig, 239-66. Cambridge: Cambridge University Press. Wyllie, I. 1975 . Study of cuckoos and ree d warblers. Br. Birds 68 : 369-78. Wynne-Edwards, V.C. 1962. Animal dispersion i n relation to social behaviour. Edinburgh: Oliver an d Boyd. . 1986 . Evolution through group selection. Oxford: Blackwell Scientific Publicatio ns. Yom-Tov, Y. 1989 . Seemingly maladaptive behaviours, individua l recognitio n an d hierarchy. Ornis Scand. 20: 2. Zahavi, A . 1971a . Th e functio n o f pre-roos t gathering s an d communa l roosts . Ibis 113: 106-09. . 1971b . Th e socia l behavior of the white wagtail, Motacilla alba, wintering in Israel. Ibis 113 : 203-11. . 1974 . Communal nestin g by th e Arabia n babbler , a cas e o f individual selection . Ibis 116 : 84-87. . 1975 . Mate selection: A selection for a handicap. /. Theor. Biol. 53: 205-14. . 1976. Cooperative nestin g in Eurasian birds. I n Acta XVI Con. Inter. Ornithol., ed. HJ. Frith, an d J.H. Calaby, 685-93. Canberra , Australia. . 1977a . Reliability in communication system s and th e evolutio n o f altruism. In Evolutionary ecology, ed . B . Stonehouse an d C.M . Perrins, 253-59. London : Macmillan Press. . 1977b . The cos t o f honesty (furthe r remark s on the handica p principle). ], Theor. Biol. 67: 603-5. . 1977c . Th e testin g of a bond. Anim. Behav. 25: 246-47. . 1978a . Decorative pattern s an d th e evolutio n of art. New Sci. 80: 182-84. . 1978b . Wh y shouting. Amer. Nat. 113: 155-56. . 1979 . Parasitism and nes t predation i n parasitic cuckoos. Amer. Nat. 113: 157-59. . 1980 . Ritualization and th e evolutio n of movement signals . Behaviour 72: 77-81. . 1981a . Natura l selection, sexua l selection an d the selection o f signals. In Evolution today, ed . G.G.E. Scudder and J.L. Reveal, 133-38. Pittsburgh: Carnegie-Mello n University Press. . 1981b . Th e latera l displa y o f fishes: Bluf f o r honest y i n signalling ? Behav. Anal. Lett. 1 : 233-35. . 1981c . Some comments on sociobiology. Au k 98 : 412-14. . 1982 . The patter n o f vocal signals and th e informatio n they convey. Behaviour 80: 1-8. . 1983 . This week' s citatio n classic : The importanc e o f certain assemblages of birds as "informatio n centers " for foo d finding . Current Contents 15 : 26.

26O BIBLIOGRAPH

Y

. 1985a. Evolution of group life with special reference to the Arabian babbler (Turdoides squamiceps). I n Acta XVIII Con. Inter. Ornithol., ed . V.D . Ilyche v an d V.M . Gavrilov 2: 1043 . Moscow: Academy of Sciences of the USSR. . 1985b . Som e furthe r comment s o n th e gathering s o f birds . I n Acta XVIII Con. Inter. Ornithol, ed. V.D. Ilychev and V.M. Gavrilov, 2: 919-20. Moscow: Academy of Sciences of the USSR . . 1987 . Th e theor y of signal selection an d som e of its implications. In Proc. Intern. Symp. Biol. Evol, ed. V.P. Delfino , pp . 305-27. Bari, Italy: Adriatica Editrica. . 1988a . Mat e guarding in Arabian babbler, a group-living songbird. Acta XIX Con. Inter. Ornithol., 420-227. Ottawa . . 1988b . Mat e choic e throug h signals . Acta XIX Con. Inter. Ornithol, 956-60. Ottawa. . 1989 . Arabia n babbler . I n Lifetime reproduction i n birds, ed . I . Newton , 253-76. London: Academi c Press . . 1990 . Arabia n babblers : Th e ques t fo r socia l statu s in a cooperativ e breeder . I n Cooperative breeding i n birds: Long-term studies o f ecology an d behaviour, ed. P.B . Stacey and W.D. Koenig, 103-30. Cambridge: Cambridge University Press. . 1991a . O n th e definitio n of sexual selection , Fisher' s model, and the evolutio n of waste and o f signals in general. Anim. Behav. 42: 501-3. . 1991b . Sexua l selection: Badges and signals . Tree 7: 30-31. . 1993 . The fallac y of conventional signalling . Phil. Trans. R. Soc. Land. B 338: 227 30. . 1995 . Altruis m a s a handicap—th e limitation s o f ki n selectio n an d reciprocity . Avian Biol. 26: 1-3 . . 1996a . Th e evolutio n o f communal roosts as information centers an d th e pitfall of group-selection: A rejoinder to Richner and Heeb. Eehav. Ecol. 1: 118-19 . . 1996b . Cooperation among lions: An overlooked theory. Tree 11 : 252. Zahavi, A., and D . Rait. 1984 . Socia l adaptations in myxobacteria. \nMyxobacteria, ed . E. Rosenberg, 215-20. New York: Springer-Verlag. Zahavi, T. 1975 . Socia l behaviour o f the babbler s (i n Hebrew). Hig h school project. Zilberman, R. 1991 . Extra-pair copulation s in a sunbird population (Nectarinia osea osea) in Ramat-Aviv area (in Hebrew). M.S . thesis, Tel-Aviv University. Zuk, M. , K . Johnson, R . Thornhill, an d D . Ligon . 1990 . Mechanism s of female choice i n red jungle fowl. Evolution 44: 477-85.

LIST O F F I G U R E S

Dorcas gazell e xii i Thomson's gazell e slotting xi v Peacock x v A group of babblers xv i Cheetah 3 A sentinel babbler4 Sparrowhawk 5 Babblers mobbin g a n owl 6 Hunting dog s 7 Merlin o n a falconer's arm 8 Caterpillars of the swallowtai l butterfly 9 Tiger 1 0 A falcon an d a flock o f starling s 1 2 Salamander—aposematic coloration 1 3 Ibexes fighting 15 Cichlid fish swimming in parallel 1 6 A bearded man , after Leonard o 1 7 Cichlid fish with stretche d fins , differen t coloring o f fins and bod y 1

8 261

262

LIST O F F I G U R E S

Toads 2 0 A sherif f i n a threatening stance 2 1 Dog 2 2 Terns, courtshi p feedin g 2 5 Crested newt with nuptial cres t 2 6 White pelican wit h nuptial bump 2 8 A stately country house 2 9 Satin bowerbird i n courtship 3 1 Peacock i n ful l displa y 3 2 Barn swallo w 3 3 Manakins dancing 3 5 White-collared ruf f i n display 3 6 Lyrebird i n displa y 3 7 Rooster 3 9 Red deer 4 0 Various species o f ducks in fligh t 4 3 Kangaroo 4 4 Babblers, mal e (above) and femal e (below ) 4 5 A decorate d plat e 4 7 Banded damselfis h 4 8 Danio zebr a fis h 4 9 Lesser black-backe d gul l 5 0 Feathers a t differen t stage s of wear 5 1 A peacock's feather with an "eye" 52 Butterfly 5 3 Convergence—long-claw pipit (Macronyx ) and meadowlark (Sturnella) 5 Red-backed shrik e 5 5 Cream-colored course r 5 6 Great ti t 5 7 "Status badges " 5 8 Lace collar, afte r Ruben s 5 9 Bengal bustard i n display jump 6 1 Bird o f paradise in displa y 6 2 Nestlings begging fo r foo d 6 3 Wolves—a dominant an d a subordinate 6 4 A foot rac e 6 5 Gray heron gettin g ready to fl y 6 6 Gazelle jumping playfully 6 8 Wolves howling 6 9 Shelduck chick 7 1 Crested lar k 7 6 Vervet monkey 7 8

4

List o f Figures 2 6

3

A babbler raising wings in mobbing 7 9 Training a dog 8 0 Rooster 8 1 Male widowbird in fligh t 8 2 Colobus monkey , with a mane and with a shaved head 8 3 Orangutans—adult male and bab y 8 4 California quai l 8 5 Square-lipped rhinoceros 8 6 Elephant seal 8 7 Ibex 8 7 Horns 8 9 Little egre t 9 1 Great creste d greb e 9 1 Red dee r 9 2 Mourning wheatear and deser t wheatea r 9 3 Wheatear tails—Europea n wheatear , black-eare d wheatear , isabellin e wheatear , desert wheatear 9 4 Raven 9 5 Ballerina in performance 9 6 Ballerina in training 9 7 Black swan and pie d goos e 9 8 Humboldt pengui n 9 9 Male moths aroun d a female moth 10 1 Female moth secretin g pheromon e 10 8 Male and femal e bulbuls clumpin g together 11 1 White wagtails , male and female , i n winter territor y 11 3 Peacock turning its back on a female durin g courtshi p 11 4 Babblers clumping an d allopreenin g 11 5 Swallows on a line, spaced a t regula r intervals 11 6 Pioneers dancin g 11 7 Great creste d greb e carrying its young 11 9 A kitten a t the to p o f a tree 12 0 Wild so w with brightly-colored piglets 12 1 A penguin with its chick 12 2 Barn owl nestlings 12 3 Babblers feeding a t the nes t 12 5 A babbler leg with rings 12 6 Babblers a t play 13 0 Dwarf mongooses 13 3 Babbler fledgling s 13 4 Allofeeding amon g adult babblers 13 7 Babblers mobbing a snake 13 9

264

LIST O F FIGURE S

Babbler 14 1 Babblers aslee p on a branch 14 2 Babbler i n fligh t 14 4 Babblers fighting, grappling with each other 14 8 Babbler landscap e 15 0 Mutual feeding between ant s 15 1 Worker bee taking care of cells in a hive 15 3 A nest of Apis dorsata 15 4 Queen cell in a hive 15 6 A queen with her retinu e 15 9 A termite colony 16 2 A nest of Polistes canadensis 16 7 A pair of hawks 16 9 Stilts exchanging places at the nest 17 1 A male Alcichthys alcicoris taking care of eggs 17 2 Jacana 17 4 The lif e cycl e of cellular slime mold, afte r Bonne r 17 7 Dictyostelium, aggregating 18 0 Fruiting bodies o f Dictyostelium discoideum 18 2 A fruiting bod y o f the myxobacteri a Chondromyces crocatus 18 Scrub warbler feeding a European cuckoo nestling 18 5 Eggs of European cuckoo and its hosts 18 6 Cuckoo nestling pushing out a reed warbler's egg 18 7 Great spotte d cuckoo 18 8 An armore d knight 19 0 Cowbird 19 1 Dog eating food provided b y its master 19 3 Bacteriophage T4 19 4 Sandgrouse gathering at a water source 19 7 Breeding colony of black-faced dioch s 19 9 Rooks 20 0 Gulls divin g at food 20 2 Flock o f starlings near a roost 20 5 Hairdo, after Leonard o 20 9 A painted ey e 21 0 Heavy eyebrows, afte r Leonard o 21 1 A baby's face an d a n elderly face 21 2 Masks—a witch an d a clown 21 3 The ideal woman—now 21 4 The ideal woman—then 21 5 A ribbon aroun d a woman's neck 21 6 Lovers' hands 21 8

4

List o f Figures 2 6 Porcupines copulatin g 21 9 Alex the gra y parrot 22 1 Cave painting 22 3 A painted fac e 22 4 Donation box 22 7

5

This page intentionally left blank

INDEX

Aberdeen, Scotland , 200 Acetate, 10 2 Acid, 102 , 160 Actors, 20, 70, 214 Adaptation: an d decorations, 223 ; for group living , 1478; and markings , 54, 57; non-reversible, 148 ; for observing signals, 59; and waste, 38 Advertising; beards, 213 ; color, 237 n 2; disregard of predators, 23 5 n 36; human foo d supply durin g development, 215 ; long neck, 216; males and females, 26; queen pheromone , 159-60 ; roost, 204-5; sexual, 229; size, 237 n 10 ; to one' s mate , 172-3 ; and waste, 229; wealth, 216, 227 Aerial display, 15, 204 Affluence, human , 215 Agamidae, 54 Age: beards , human , 214; and eg g recognition, 186-7 ; identification of , 230; marker s identifying, 43; nose, 213; thicke r eyebrows, human , 212 Aggregate, socia l amebas (slim e mold), 178-8 0 Aggression: altruis m as substitute for, 142; in babblers , 128, 130, 137, 145, 146; in courtship, 113-5 ; human and animal , 15; in insect colony, 155 , 240 n 5;

among kin in social insects , 164-5; to males , 175 ; and rank , 144; threa t as substitute for, 15; young cuckoo begging , 189 Aggressors, and grou p selection , 23 Alarm calls, and babblers , 122 , 134. See also warning calls Alarm system, blackstart, 173 Albon, S.D., 29 Aldchthys alckornis, 172, 172 fig Alcidae, 87 Alcohol, 10 2 Alcoholic beverages , 103, 160 Aldehyde, 102 Alexander, R.D., 15 3 Algae, 26 Alkaloid, 30 Allelopathic chemicals, 181 Allofeeding, babblers , 127 , 137-8, 13 7 fig Allopreening, babblers, 115-6 , 11 5 fig, 127 AlonsoJ.C.,200 Alternative, essential fo r communication , 47 Altruism: XV-XVI; and bright colors , 202; competitio n over, 133-41 , 172 ; definition, 161; evolution of , 125,

267

268

Index

Altruism (continued} 131-3, 149-50, 158, 161 , 230 ; gain from, 167 , 179 ; and grou p selection theory, 161 ; host an d parasite, 240 n 20; in human, 225-7; and kin effect , 166-7 ; and kin selectio n theory , 164; between non-kin , 163 ; among parents, 171 ; and prestige , 144; reciprocal, 132-5, 149, 153 , 163 , 171 ; a s a signal, 150; in social insects, 157-8 , 163 ; stalk formation, social amebas, 183; and suicid e (social amebas), 178-9; theories explaining, 131-33; as threat substitute , 141-3, 149 ; warning calls, 12 Amasa (Bible) , 213 Amboseli, Kenya, 77 Ameba, social. See slime molds Amphibians, 25, 35 Amplifier signals , 56- 8 Amplifying differences , 56 , 58 Analog measuring, and nonverbal vocalization , 244 n 19 Anava, A., 73, 127 , 13 9 Andersson, M., 39 Andrena sabulosa, 190 Animals (general): large groups, 244 n 18; performances, 34; and symbolic language, 77-80, 221; vocabulary, 72, 78 Anis, groove-billed, Crotophaga sulcirostris, 148 Ankle, 217 Ant: 15 1 fig; aggression toward kin, 165 ; evolutio n of social life , 151-4 , 162-3 ; as large social groups, 244 n 18 ; and noxiou s chemicals, 22 ; workers leaving mother's colony , 241 n 30 Antelope, 48-9, 8 6 Anthropomorphic models : alcoholi c beverages , 103; ballet dress, 96; boxers, 10-11; children jeering, 8; contests, 46; crested helmets, 85 ; discussion of , 9; drinking alcohol, 160; and evolution, 60 ; eyeliner, 45; girlfriend leanin g on boyfriend, 116 ; gymnastics, 34 5; humans an d dog s a s partners, 193 ; human information centers , 203; hunting estates, 29; inflation, 59 ; inheritance laws , 123; knights an d serf s as parasite and host , 190-1 ; littering, 132 ; longbow , 82; Mafia mode l o f parasitism, 189; mansions , 29; money (queen pheromone), 159 ; Olympi c games, 65; ornaments that become cheap , 59; plate decorations , 47; skydivers , 62 Anthropomorphism, 9 Antibiotics, 181 , 19 4 Antlers: XIV, 39, 55; evolution of, 89-90; as a handicap, 87-8; i n ritual fighting, 62-3; an d social rank, 14 4 Apes, 132 , 211 Aposematic coloration. See warning colors Appeasement, 19 Arad, D., 183 Araiida, 30, 101 Arena: communal, mating, 199; leks, 34; of peacock, 33; of sage grouse and ruffs , 2 5

Arms race, 185-6, 195 Arnott, J.L., 23 6 n 5 Art, evolution of, 223-5 Artifacts: an d art , 224; in courtship display, 31 Artists, 224-5 Asexual reproduction, 17 7 Assessing, 34, 216, 223 Asymmetry, in relationship, 154 . See also symmetry Atherton, Australia, 97 Athletic contests, 46 , 65 Attack, 144 , 192 , 21 0 Attraction: decorative feathers, 223; and the Handica p Principle, 40; and mal e care of offspring, 172; an d sexual selection, 38; and vigor, 233 n 42; and waste, 39 Atzmony, D., 104 , 179 Auto racing, 226 Axelrod, R., 133 , 232 n 1 9

Babbler, Arabian, Turdoides squamiceps: 125-50 , XVI fig, 12 5 fig, 141 fig , 14 2 fig, 144 fig; aggression, 128, 130, 145 ; allofeeding , 127 , 137-8, 13 7 fig; allopreening and clumping , 116, 127 , 115 fig; altruism, XVI, 133^t , 172 , 179 ; calls , 73, 78-9, 1267; coalitions, 129-30 ; competition ove r altruism, 134 6; competition ove r mating , 129 , 146-7 ; confrontations betwee n groups , 115 ; copulations not for procreation , 219 ; dancing, 116-7, 127 , 143 ; DN A studies, 127 ; eg g color, 121 ; ey e ring, 55; feeding at nest an d rank , 13 8 graph; feeding of nestlings, 127 , 133, 137-8 , 138 graph, 123 fig; female life strategy, 129; fledgling s self-endangering , 120 , 13 4 fig; fights, 15, 115, 143 , 148 fig; food, 126 ; gender marker , 44 45, 45 fig; groups, 127-30,147, 239 n 32; group breeding, 129, 131 ; group-livin g adaptations, 148 ; incest avoidance , 128-9, 210, 225; landscape, 15 0 fig; life expectanc y and strategy , 128-9; long-distanc e calls, 74; loud calls , 74; male aggression towar d females, 113 ; male life strategy, 129; mate guarding, 146, 170 ; mating, 127, 129 , 146 ; mobbing, 73, 127 , 133, 139^11 , 143 , 14 0 graph, 6 fig, 79 fig, 139 fig; mobbing calls , 73, 79; "moral" behavior, 225; need for communication , 221-2; nonterritorials, 127-8, 130; play , 127, 12 6 fig; rank, 128; ran k an d prestige , 143-5; recognizing ability, 158 ; reprimanding , 145 ; reproduction, 129-30 ; research in Hatzeva, XV-XVI; rhythm, 73; sentinel activity , 3, 4 fig, 127, 134-6 , 143, 135- 6 graphs ; shyness ove r copulation, 145-7; subordinate copulating, 147 ; territories, 4, 125 , 127 8; 115-6; threat call s over distance, 21; threats an d audience, 75; and tit for tat, 139 ; trills , 73, 141 ; warning calls, 3-6, 12 , 78; water balance, 127 ; young more colorful, 121 Baboon, yellow, Papio cynocephalus , 121 Baby and adul t face : human, 212 fig; orangutan, 8 4 fig

Index Baby: baby face , 213 ; "blackmail " by, 122 ; crying, 71, 121, 22 0 Bacillus, diphtheria , 19 4 Bacteria: antibiotics , 181 ; blackmail , 194 ; cAMP, 242 n 3; communal, 131 ; virulent an d non-virulent , 194 Bacteriophage T4, 19 4 fig Bakal, C., 227 Balance: assumption of , 195; extortio n an d its cost, 122 ; host/parasite, 185 , 192 ; mobbin g o f cowbirds, 191 ; nestling begging , 189 ; powe r i n a group, 146 Bald head, human , 83 Ballet dress , 96 , 96 fig, 97 fig Bangalore, India, 17 9 Bark: gazelle, XIII; babblers, 5 , 78 Barlow, J.W., 48 Barrette, C., 22 Bars, 16 0 Basolo, L.A. , 235 n 1 1 Bats, 2 9 Beak, 49-50, 53-4, 87, 235 n 25 Beard: human, 17 , 2D-4, 17 fig; ibex an d leopard, 87 Beauty, 46, 52, 214-5, 223, 225 Bedouins, 94-5 , 20 3 Bee, Andrena sabulosa (parasite) , 190 Bee, Apoidea: evolutio n o f social systems, 151-3,162-3; large social group, 24 4 n 18 ; queen cell , 15 6 fig; queen pheromone, 158 ; quee n wit h retinue , 15 9 fig; threats b y noxious chemicals , 22; worker an d cells , 153 fig. See also bumblebe e Bee, honey , Apis mellifera, 153 , 16 0 Bee, Numada marshamella (host) , 190 Bee, stingless, Metiponmae, 15 4 Bee-eater, Merops apiaster, 82 , 9 9 Beech tree , 203 Begging, 63 fig, 189, 22 1 Behavior, human an d animal , 209-10 Behavioral mechanisms an d natural selection , 12-3 , 120, 209 Belladonna, 211 Benefit t o th e group , 23 , 149 , 157 , 179 , 192 , 225 Betel nut, 160 Bible, 129 , 213 Bird o f paradise, king-of-Saxony , Pteridopbora alberti. 31,223 Bird of paradise, superb, Paradisaea raggiana, 61-2, 6 2 % Birds of paradise, Paradiseidae: bristlin g feathers , 83; leks, 35; males no t carin g fo r offspring, 31; tail, 25 Birdsong, 25, 28-9, 72 , 223, 236 n 2 Birdwatching, 67, 205-4 Bizarre signals, 38 Black, use of, 54, 94-7, 20 2 Blackbird, red-winged , Agelaius phoeniceus, 98 Blackbird, Turdus merula, 43, 97 Blackcock. Se e grouse, blac k Blackmail, 120-2 , 194

269

Blackstart, Cercomela melanum, 4, 5, 94, 144 , 172- 3 Blood supply , and color , 51, 93, 214 Blue, glossy, use of, 98; objects, and bowerbirds , 60 ; use of, 96, 212, 21 4 Bluffing, actors , 20; in threats, 16- 7 Boar, wild, Sus scrofa, 121 , 12 1 fig Body fat, human, 215 Body parts as signals, 81-9 2 Bond, testing, 111-7 , 217-23, 230 Bonobo, Pa n paniscus, 219 Borgia, G. , 32 , 60, 97, 114 Bottleneck, reproductio n an d femal e dominance, 17 4 Bowerbird, golden , Prionodura newtoniana, 97 Bowerbird, great , Cblamydera nuchalis, 32 Bowerbird, satin , Ptilonorhynchus violaceus, 31 fig, 60 Bowerbird: 26 , 31-2,144,150; decorations an d stealing, 31-2, 60; female preference, 223 ; showing of f by waste, 157 ; testin g female intentions, 11 4 Bowtie, 217 Boxers, 10-11 , 17 Breast band, 56, 217 Breasts, human, 215 Breeding colonies, 198- 9 Breeding, babblers, 127 , 130 , 14 6 Bright colors , 51, 201- 2 Bristling, 83 Broods: babblers, 128 ; size, including parasite , 19 1 Brook, M. , 18 6 Brother, ki n selection theory , 16 3 Brown, C.R. , 199 Brushrunner, Coryphistera alaudina, 54 Bucher, E.H., 20 3 Buffalo, Syncerus caffer, 8 6 Bulbul, yellow-vented, Pycnonotm barbatus : 4, 5 , 43 , 55,56, 11 1 fig, 116, 14 4 Bulge as a handicap, 84- 5 Bull, Bos taurus, 63 Bullfinch, Pyrrhula pyrrhula, 116 Bumblebee, Bombus terrestris, 160, 24 1 n 41 Bunting, Cretzschmar's , Emberiza caesia, 94 Bunting, Emberiza, 99 Bunting, house, Emberiza striolata, 94 Burley, N., 235 n 11 Bustard, Bengal, Eupodotis bengalensis, 61 fig Bustard, Chlamydotis undulata, 50, 62 , 85 Butcher, G.S., 23 7 n 2 Butterflies, 49, 50, 53, 53 fig, 101, 15 2 Butterflies, arcteid , Arctiida, 30, 101 Butterflies, danaid , Danaus, 30-31, 101 Butterfly, monarch , Danaus plexippus, 3 1 Butterfly, swallowtail , Papilio machaon, 9 fig Buyers, 26, 47

Calls: babblers, 126 ; and language, 77-80; mobbing, 79; over distance , 74; requests and commands , 75;

27O

Index

Calls (continued] threats, 19 ; warning, 74. See also vocalizations, begging Camouflage, 201, 237 n 2 cAMP, 180 , 183-4 , 242 n 3 Canvasback, Ay thy a valisinaria, 191 Carbon monoxide , 10 8 Care of young: forced, 189; fo r prestige, 171- 2 Caregivers: dependence on, 152 , 24 1 n 27; manipulation by , 154 Carlisle, T., 127 , 137- 8 Carmeli, Z., 127 , 13 9 Caro, T.M., 232 n 7 Carotenoids, 93 Cask, 85 Cat, Felts silvestris, 69, 77, 120-1, 12 0 fig Catbird, Australian, Ailuroedus,115 Catch-22, 38-9 Caterpillars, 9 fig Cave paintings, 224 Cavy, Patagonian. Se e mara Cell membranes, 230 , 242 n 3 Cells, 52, 10 7 Cellular slim e mold. See slime molds Charity, 227. See also donation s Cheating; chemical signal, 184 ; clothing , 217; an d conventional signals , 58; cost o f signal, 212; hos t an d parasite, 37; manes, 84; mimics, 9-10; i n morphs, 36 37; prevented b y pattern of signal, 24; quee n pheromone, 159 ; and reductio n i n cost, 59; ritualization, 64; stretching, 18 Cheeks an d lips (human) , red, 214 Cheetah, Acinonyx jubatus, 3 fig, 67 Chemical signals : cAMP, 242 n 3; communication an d handicaps, 30, 101-8 ; differentiation, 178, 181 ; evolution of , 108 ; harmfu l chemical s an d humans , 160; on membranes, 230; in multicellular body, 107 8; queen pheromone , 158-60 ; signal and non-signal , 184 Chick: extortion o f care by, 51; food bottleneck an d female dominance , 174 Chickadee, Pams atricupillus, 148 Child: interes t in parent's reproductio n (ki n selection), 238 n 2; parent-child conflict , 119-21 , 124 Children: benefiting, 225; games, 8, 115; in human families, 217; testing the bond, 111 , 115 , 119-24,21 9 Chimpanzee, Pan troglodytes, 63, 122 , 13 1 Chin, human, 213 Chromosomes, 153 , 162 , 19 4 Cicadas, 25,29 Cichlid fish , 1 8 fig Cichlid fish, Cyrtocara eucmostomus, 31 Cichlid fish , Geophagus surinamemis, 16, 1 6 fig Clothing, 46, 215- 7 Clown, 21 3 Clumping, 115-7 , 143 , 11 1 fig, 115 fig

Clutton-Brock, T.H., 2 9 Coalitions: babblers, 129 ; relatives, 166 ; socia l insects , 154 Cock. See rooster Coexistence, 19 2 Collaboration: evaluatin g interest, 158 ; hos t an d parasite, 37; and human language, 217; humans, 182 ; mating, 114 ; betwee n parents , 169 ; with relatives , 166; socia l amebas, 182 ; socia l insects, 152 ; type s of, 148 Collaborators: human , 209, 217; parasite evolving into, 193-4 Colony: ant workers leaving, 241 n 30; gulls, 198 ; socia l insects, 154-7 , 164-6 ; swallows, 199 ; termites, 241 n 27 Color: an d bloo d supply, 51, 93 ; bright, on eggs , as extortion, 122 ; bright, a s a handicap, 235 n 36; bright, an d predators, 30, 235 n 36; carotenoids, 93; compromise, 98 ; and courtshi p display , 30, 32; cryptic, 201-2; distinct, on body structures, 49-50; and environment , 93-99; in forests, coral reef s and kelp forests , 97; of fin , 19 ; glossy, 98-9; many , 96; and parasites , 51; research on, 237 n 2; two, 50, 97 8, 234 n 14 ; of young, 121 Combination o f signals, 32- 3 Combs, 51, 81, 87, 93 Commands, 75- 6 Common interest , XIV , 5, 43 Communal children' s house, 210 Communal displa y at roosts, 204-5, 244 n 20 Communal nest, babblers, 129 Communal roost, 198-204 , 244 n 20 Communication: babblers, 126 , 222 ; chemical, 101-8, 184; child-paren t relationship , 124 ; differentiation, 178; and th e Handica p Principle , 209 , 229; betwee n host an d parasite, 194 ; human language, 221-3; between parenta l couple , 175 ; prey and predators, XIII-XIV, 3-13, 23 2 n 11 ; and reliability , 70, 195 ; between rivals, 15-24; between strangers , 222; among unicellular organisms , 177; within th e body , 52; wordless, 80 , 221- 3 Communities, as large groups, 244 n 1 8 Companionship, and sex , 220 Competition: over altruism , 133—41 , 149 , 171-2 ; an d genes, 162 ; gymnastics, 34; and th e Handica p Principle, 40; human, 209; over mating and breeding, 127 , 129-30 , 147 ; over mobbing, 141 ; parents over caregiving, 171-2; and prestige, 125 ; over reproduction , 142 ; and set-specifi c signals , 43, 46; socia l insects, 156- 7 Competitors, 204 , 227, 230 Complementary partners, 173 Compromise: in coloration, 98; between hos t and parasite, 37 Comrades: opinions of, 226; a s rivals, 142 Confidence: an d staring , 55; in threats, 18

271

Index Conflict: child-parent , 124 ; and concentration , 21, 236 n 10 ; escalation into fights , 15 ; between groups , babblers, 115 ; between informatio n collectio n between gender s in mating, 26, 242 n 1 ; between parents, 169 , 175; over reproduction, and quee n pheromone, 159; resolution, 142; over territory, red deer, 18-9 ; betwee n workers , 155-6 Conforming i n dress standards, 216; in threats, 23 Confrontations, between babbler groups, 11 5 Congregation, human , 203 Connection, logical , between signa l and message, 22930 Consciousness, 18 9 Conserving resources, through superterritories, 29 Conspicuous consumption , 160 , 227 Constraints: in gathering food for offspring , 156 ; in host-parasite relationship, 188 ; on reproduction , 174; over specific traits, 173 Constructions, animal, 31 Contempt, 21-2 Contests, 46 Conventional signals, fallacy of , 17 , 23, 58, 60, 159 , 183,218,230 Convergence, 53-4, 54 fig Cooperation: with competitors, 230; degree of , 111; in defense o f territory, 113, 125; between do g and man, 193; amon g female babblers, 145 ; between genders , 173; between hos t and parasite , 194; humans, 209, 217, 222, 230; among male babblers, 147 ; between parents, 169; with raptor, 12 ; with rivals, 60; among social insects, 156-7 Cooperative animals , 149 Copulation: o f babblers, choice of male, 145-7, 170 ; o f babblers, new female, 129; of babblers, subordinate , 147; competition over , in babblers, 147 ; and degre e of dominance , 141 ; extrapair, 171; and male caretaker, 172 ; no t for reproduction, 218-9 , 219 fig; shyness about, babblers,145, 225; and wintering pairs, 197. See also sex Coral reefs, 44, 48, 97 Cormorant, Phalacrocorax, 97 Corvids, and cuckoo , 188 Cost-reducing traits, 88 Cost: of altruism, 144 , 150, 226-7; balance o f cost an d gain, 189; importance o f in signa l selection, 59; of low-pitched threat , 19-20; reducing, 88; and reliability of signals, 58, 212, 189; search for, XVI; o f signals and non-signals, 59-60; specific, in threats , 16, 24; in time , 28 Goto Doana, Spain, 200-1 Courtship, 25^0; aggression, 31,113-4; colors, 30, 32; conflict in, 26; constructions, 31,3 1 fig; dancing, 32, 34, 62; ESS, 233-J n 43; feeding, 25 fig; gift-giving, 32; of great crested grebe, 64; Handicap Principle in , 40; and, parenting couple, 169; scents (pheromones), 30; showing ability, 28; testing interest, 114 , 11 4 fig;

vocalization, 29-30,32; in wintering pairs, 219 Cowbird, brown-headed, Molothrus ater, 191-2,191

%

Crabs, 31 Cracidae, 8 5 Creationism, 91 Crests, 84-5 Crickets, 25, 29 Cronin, H., 23 3 n 39 Crow, hooded, Corvus corone, 172, 187-9, 198- 9 Crows, Corvus, 141, 188 Cryptic colors, 11 , 201-2 Cuckoo, European, Cuculus canorus, 185 fig, 186-7,190 Cuckoo, great spotted, Clamator glandarius, 172, 18790, 18 8 fig Curassow. See Cracidae Cursor, cream-colored, Cursorius cursor, 55, 56 fig Cuticle, 9 , 103 Cyrtocara eudnostomus, 31 Damselfish, banded, Dascyllus aruanus, 48 fig Danaidae, pheromones of , 10 1 Dance: babblers , 117, 127, 143; bustards, 85; cost in time, 28; courtship signals, 32, 34; group, 116; group, humans, 117; manakins, 35; sage grouse and ruffs, 2 5 Dancers, ballet, 96 Dancing arena, 25, 114 . See also leks Danger an d fame, among human, 226 Danio zebr a (fish) , 4 9 fig Daphne Island, 235 n 25 Darling, F.F., 22, 87, 198 Darwin, Charles: mimicry and huma n language, 77; sexual display, 37-38; sexual selection, 38, 40, 44, 91, 233 n 42; singing, 70; social insects, 151-2, 164 Darwinian logic, and evolutio n o f altruism, 125 Daughter, inheriting, 124 Davies, N.B., 20, 186 Davis, G.W.F., 241 n 26 Dawkins, R.: arms race, 185; competition betwee n genes, 162 ; evolutionary models, 239 n 17 , 241 n 26; kin selection , 165; "mind reading" , 66 Decorations: as amplifier signals, 57; bowerbirds, 31-32, 114; decreasing apparent size, 237 n 10; emphasizes features, 48; evolution of, 47, 47 fig; feathers, 31,91, 223; function of, 43; human, 211, 215-6; and their message, 54; prove status, 56; selection of, 223-4; and social status, 55; and symmetry, 52 Decoys, 231 Deer, Cervus: antlers, 55, 87-9, 144 ; courtship roars , 29 Deer, red , Cervus elaphus: courtshi p roars exhausting, 29; 40 fig, 92 fig.; fighting between simila r rivals, 22; threats among, 15, 18-1 9 Defense, o f territory, 113, 173-4, 197-8 , 219, 234 n 44 Definitions: o f altruism, 161; of fitness , 162 , 165; of handicap (sports) , XIV; o f prestige, 143^; of queen

272

Index

Definitions (continued] pherormone, 240 n 13 ; of sexual selection, 38; of signal selection an d utilitaria n selection, 40; of signals, 59 Degree: of control, in babblers, 146-7 ; of cooperation , 111; o f dominance, i n babblers, 141 ; o f emotion, 22 2 Dependence: o n caregivers, 152 ; on group mates , 115 Descendants, 162 , 192 . See also offspring Deserting a mating partner, 170 , 219 Deterrence: of pursuit, 232; and the runawa y process, 39; and sexua l selection, 38; and song , 15. See also intimidation; threat s Development: o f human breasts, 215 ; perfection in , 33; symmetry in, 52. See also evolutio n Dexterity, 211 Dhondt, A., 29-30, 72 Dialogue, in singing, 76 Diamond, J.: colo r i n mixed-species flocks , 97 ; humans and harmfu l chemicals, 160 ; menopause , 12 3 DIF, 179 , 181- 3 Differentiation, i n multicellular organism, 178-9 , 181 Digestion, of sugars and protein s i n wasps, 240 n 14 Digital measuring , i n verbal communication , 244 n 19 Dioch, black-faced, Quelea quelea, 198, 19 9 ( fig.) Diphtheria, 19 4 Diploid, 15 3 Direction o f gaze. See gaze Display: of disregard towar d predators , 235 n 36; flights, 34; in flock, 205; leks , 34-5; o f status, 56; of wealth, 60 Diving, larks, 62, DNA analysis , 33, 127 , 130 , 172 Dog, African hunting, Lycaon pictus: 1 fig; groupbreeding, 131 ; group rituals , 117; observing stot ting, 7 Dog, Canis familiaris: friendly , 62; hair raised in threat , 18; partnership with humans, 193, 19 3 fig; scolded, 61; showing contemp t i n threat, 22, 22 fig; testing the bond, 111-2, 117; an d verbal language, 80, 80 fig; vocalization of , 6 9 Dog, commo n red , Cuon alpinus, 9 8 Dolphin, bottlenose , Tursiops: ange r calls and posture , 70; and verbal language , 80, 221 Dominance: bumblebees , 24 1 n 41; displayed by rape, 220; between mates, 173- 5 Dominant babbler : allofeeding , 137-8 ; copulations, 129, 145 ; mate-guarding , 170; mobbing, 139-41 ; sentinel activity, 134-6; sleep, 143 Dominant individual : degree o f control, 146 , 147 ; identified b y scent, 30; larger nose, 213; prestige, 149 ; prestige an d altruism , 144-5; in social hierarchies, 22; threats, 142 ; workers, 155 Donations, human , 160 , 226-7 Dove, turtle, Streptopelia turtur, 34, 62 Doves, communa l roost, 19 8 Drill, 216

Drinking, 160 , 203. See also alcoho l Drones, 154 Drongo, Dicrurus: black color , 95 ; mimicry in calls , 77 Duck, Anas: color, 30 ; constraints on females, 173; duckling vocalization, 71; fledglings , 122 ; identifying, 44, 4 3 fig; intention to fly, 61; losing colo r on islands , 235 n 36; parasitism, 191 ; rape , 170 Duiker, Grimm's , Sylvicapm Grimmia, 89 Dwyer, P.D.,60 Eagle, Aquila-, and verve t monkeys, 77; and warning calls, 7 9 Eagle, black, Aquila verrauxi, 95 Eagle, Wahlberg's , Aquila wahlbergi, 77 Ecological niche : convergence, 54; female dominance , 174-5; and markings, 53; mimics, 9-1 0 Economics, 60, 215, 217 , 227 Edge of structure, decoration , 47, 49 Eel, electric , 15 Efficiency, selectio n for , 40, 23 0 EGF, 10 8 Eggs: 90; of babblers, 129 ; breaking, babblers, 146 ; color, 121-2 ; cuckoo an d host , 186 fig; ejected by cuckoo, 186 ; recognitio n by host, 186-7; Egret, little, Egretta garzetta, 83, 90 fig , 96- 7 Egrets: communal display at roost, 204; parents competing, 17 1 Eibel-Eibelsfeldt, 1,212 Ein Gedi Nature Reserve, Israel, 87 Eisner, T. , 30 Electric pulses , 15 Ellis, P.E., 10 3 Emotion: an d facia l wrinkles and nose, 213 ; human vocal, 236 n 5; an d intensit y of vocalization, 222; and pupil size , 210; Endangering self , 120-2, 220-1, 225-6 EndlerJ.A., 237 n2 Enforcement o f good behavior , 13 3 Ensuring reciprocity, 132 Ensuring reliability: by cost, XVI; by investment, 27 ; by waste, 40 Environment: changin g cost o f traits, 59; an d convergence, 54; and decorations , 223; littering of, 132; and socia l amebas, 179-80; and socia l organization of white wagtail , 198; an d understanding signals, 230 Enzymes, 183 Equilibrium, presumption o f evolutionary, 7, 185 , 187 , 195 Eshel, L, 11 , 12, 122 , 164 ESS, 234 n 43 Esthetics: an d quality , 52; value of, 223-5 Ethics, an d evolution, 225 Ethologists, 218 Eusocial systems, 162 Evans, C.S., 71

Index Evans, M . R., 34, 82 Evolution: altruism , 125, 149 , 150 , 158 , 161 , 230 ; antibiotics an d toxins , 181 ; o f art, 223-5; bizarre signals, 38; "blackmail" systems, 122 ; chemica l signals, 108 ; constraint s on specific traits , 173 ; decoration, 47 ; of decorative markings , 223; definition, 161 ; ethics , human , 225; experiments , social amebas , 178 ; experiments, unicellula r organisms, 60; of feathers, 90-1; grou p living , 147 ; group-specific markers , 43-60; and the Handica p Principle, 230 ; horns an d antlers, 89-90; kin effect , 166-7; and kin selection theory , 163-5; markings, 53; "not ye t complete," 191 , 195 ; parasite-hos t relationship, 185-6 ; of parasitism into collaboration , 193; polymorphism , 53 ; "psychological weapons, " 120; reproduction, 123 ; ritualized signals, 68; runaway process model, 38; signals, 59, social insects, 151-^, 162-3; "spiritual" needs, human , 224; status badges, 56 ; 235-6 n i l, 236 n 36; of style of clothing, 215; symboli c language in humans, 222; unicellular organisms, 177 Evolutionary stable strategy . See ESS Evolutionary systems, and huma n systems , 60 Evolutionary theory, and the proble m o f altruism, 161 Exaggeration: ritualization, 67; runaway process model, 39 Exploitation: o f offspring b y parents, 123 , 153 ; o f one's mate, 169 , 175 ; by parasite of host, 185 ; betwee n partners, group selection , 131 ; amon g social amebas , 182; amon g socia l insects, 124 , 152^1 , 157 ; of weaker partners, 152-4, 157 Expression, huma n facial, 214, 22 4 Extinction, 20 3 Extortion o f care, by self-endangerment: by humans, 220-1; by offspring, 120- 2 Extrapair copulation , 33 , 170- 1 Extravagance, 39, 40 Eyelashes, 212 Eyelid, 212 Eyelike pattern: fins, 50; peacock, 32-33 ; status badge, 55; value and widespread us e of, 52-3, 5 2 fig Eyes: human, 210 fig, 212; movements, 55, 59; ring, 55, 59; sensitivity to color , 237 n 2; stripe an d directio n of gaze , 55, 5 6 Face, human , 212 fig, 213-4, 224, 224 fig Facial markings, 55, 213 Falcon, Falco, 12 fig, 34 Fame, 226 Family business, 165 Fanshawe, J.H., 7 Faran, N., 127 Fat reserves , 103, 215 Fatty acids: cuticle, 103 ; myxobacteria , 182; quee n pheromone, 16 0

273

Fear cry, babblers, 7 8 Feare, C.J. , 200 Feathers: edg e decorations , 49 , 51, 51 fig; evolution of , 90-1 Feeding: ability, shown i n courtship, 27 ; other's offspring, 225 ; subordinat e babblers , 143 , 145 ; young babblers, 127 , 137-8 , 13 8 graph Feldman, M.W. , 122 Female: advertising t o males, 26; aggressiveness toward males, 175 ; babblers , lif e strategy , 129 ; an d beards , 213; constraints , 26, 90, 169 , 173 ; contro l over copulation, 145 ; demanding two morphs as mates, 37; dominance, 174-5 ; evaluating rump , 49; horns and antlers, 90; and human sex , 220; learning from older females, 35 ; pheromones, 102 , 23 8 n 4 Fighting: avoidin g by prestige, 157 ; babblers, 14 8 fig; babblers, and rank , 143 ; comb a s handicap, 81; escalation into , 15 ; gill covers and figh t results , 65 ; handicaps, 87 ; ibex, 1 5 fig; and "propaganda" , 23 4 n 44; ritualistic, 90; similar rivals, 22 Fin: colo r an d eyelike pattern, 1 8 fig, 49-50; stretche d in threat , 1 9 Finch, Melanoderma melanoderma, 99 Finch, yellow-billed, Melanoderma xanthogramma, 99 Finches, 17 5 Finches, Darwin, Geospizinae, 235 n 25 Fireflies, I-jimpyridae, 2 5 Firstborn, an d inheritanc e laws, 123 Fish: decoration s of , 48, 57; female dominance, 175 ; leks, 35; male taking care of others' offspring , 172; morphs and lif e strategy , 36-7; schools , as food, 198 ; sexual display, 25; species identification , 44 Fisher, R.A. : observation o f features, 82; the runaway process model , XVI, 37-40; sexual selection, 38 , 234 n44 Fishman, L., 172 Fitness: an d altruism , 161; definition , 162, 165 ; an d extravagance, 39; inclusive, 162 , 165 ; ki n selectio n theory, 163 Fitzgibbon, C.D. , 7 Flamingo, Phoenicopterus ruber, 93 Fledgling: babbler , 13 4 fig ; babblers, allofeedin g an d rank, 138 ; babblers, commo n nest , 130 ; babblers , incest avoidance , 128; babblers , killed , 147 ; self endangering, 120-2 Flock: black-face d dioch, 198 ; display , 205; findin g food, 203 ; gulls, 201-2; kites, 201; white wagtail, 197-8 Fluctuating asymmetry, 51. See also symmetry Flycatcher, collared, ficedula albicollis, 99 Focus, 210, 212 Food: babblers, 126 ; bottleneck, an d female dominance, 174 ; human, availability, 215; rooks , in severe weather, 200; production, human , 217; an d social organization, 198; storage , human, 193 ; storage, social insects , 152, 156 . Se e also feedin g

274

Index

Forcing: cooperation, 157 ; DIP, 181-3; host and parasite, 189 Fowl, jungle, Callus gallus, 81. See also rooste r Frames, outlining shape, 49 Freedom, an d dominance between genders , 173 Frequency, o f features in population, 233-4 n 43; inversion, in birdsong, 236 n 2 Friendliness: greeting, human, 212, 218; reliabilit y of message, 62 Frigatebird, Fregata aquila, 175 Frogs, courtship call s and predation risk , 29 Fruiting body: myxobacteria, 182 ; social amebas, 178- 9 FSH, 10 5 Fujita, K. , 237 n 1 0 Fund-raisers, 226-7 Fungi, parasites, 192 , 194 Gadagkar, R., 164 Gaioni, S.J., 71 Games: babblers, 127, 12 6 fig; chUdren's, 8, 97, 114- 5 Gaze: handicaps, 85; markings, 55-6; noticed, 59; signals, 59 Gazelle, dorcas , Gazella dorcas, XIII fig Gazelle, Gazella: horns, 86, 89; offspring an d blackmail, 121 ; 6 8 fig; ritual fighting, 62-3; slotting , XIII-XIV, 6-7; and wolf, ritualization, 67 Gazelle, Thomson's , Gazella Thomsoni, XIV fig, 67 Gender: conflic t of interest, 169, 242 n 1 ; cooperatio n between, 173 ; determination, 153 , 162 ; and dominance, 174 ; marker s identifying , 43—5 , 230 Genes: diphtheri a toxin, 194; an d kin selection theory, 162, 164 , 238 n 2; in mate, 26- 7 Genetic code , 58 Genetic: damage , and incest avoidance , 210; differences between morphs, 36; relatedness, and kin selection theory, 163 ; uniformity , unicellula r organisms, 177 Genets, European cuckoo , 186 Genotype: o f cells, 107; non-virulent, 194; an d recognizing cuckoo eggs, 186-7; social amebas, 182 Gibbs, H. L., 235 n 25 Gibson, R.M. , 34- 5 Gift-giving: courtship signals , 26, 32; human, and prestige, 225-7 Gill covers, 65-6 Giraffe, Giraffa camelopardalis, 86 , 8 9 Glossy colors, 98-9 Gluconeogenesis, 240 n 14 Glycogen, 18 2 Glycoprotein: role of, 104-7; sugars of, 105 Gnu, Connochaetes gnou, 11 Goat, Capra aegagros: black or white, 94-5; horns , 87 Gold, value of lace, 60 Goose, pied, Anseranas semipaltnatus, 96 fig Grafen, A. , XV, 58 Granot, D., 104 Grant, P. R., 235 n 25

Grasshoppers, Acrididae, 25, 152 Grebe, great crested, Podiceps cristatus: "blackmail" by fledglings, 11 9 fig, 122; chick color, 51; courting, ritualization, 64; crest, 11 9 fig Green: glossy , use of, 98; use of, 97 Greenewalt, C.H., 23 6 n 2 Greetings, human, 117 , 212 Grooming, human, 214 Group selection . See selection, grou p Group-specific markers, 43-60 Group: adaptation s to living in, 148 ; altruism, 125; benefit to, 157 ; dancing , 116; evolution of , 147 ; incest avoidance, 210; investment in, 167 ; need for language, 221; reproduction and natural selection, 131; rituals , 117 ; singing , 76 Group, babblers, 127-9 : clumping and dancing, 116-7; display of manakins, 35; helper replacin g partner , 172; livin g 131 , 147-8 ; reasons and consequences , 147; rituals , 117; unneeded assistance , 164 Grouse, black, Tetrao tetrix, 25, 199 Grouse, sage , Centrocemcus urophasianus, 25, 34 Grouse, Tetrao, 31, 35,174 Guarantee: by cost, 24, 58; of paternity, 170 Guard. See sentinel Guard, mate. See mate guard Gull, herring, Larus argentatus, 50 Gull, Larus: bright adul t colors , 201; communa l roost, 198; diving at food, 202 fig; eye ring, 55; nuptial feeding, 26 ; parents competing over caregiving, 171 ; species identification, 44 ; tearing grass, 63 Gull, lesse r black-backed, Larus fuscus, 5 0 fig Gull, western, Larus occidentalism pairs of females, 22 0 Gulls: flock , 198 , 201-2; species identification, 44 Gutman, A., 127 Gymnastics, 34, 65 Hailman, J.P., 237 n 2 Hair: raised, 18 , 83; human, 211, 20 8 fig Haldane,J.B.S., 163 Halliday, T., 20 Hamilton, W.D., 51, 162 , 240 n 13 Hamilton, W.J., color, 237 n 2 Hand, human: and long hair , 211; testin g the bond, 218, 218 fig Handicap Principle : acceptanc e of, XVI; evolution of altruism, 125 ; humans, 209; mathematical models, 58; movements and ritualization , 61 ; pattern ensure s threat, 23-4; reducing cos t of signals, 88 ; ritualization, 64, 67; and th e runawa y process model , 37, 40; and set-specifi c markings, 49; summary, 229; and understanding of evolution, 230 Handicap: altruism , 144 , 150 , 227 ; antlers , XIV , 55 , 88; beard, human , 213 ; brigh t colors , 201-2, 23 5 n 36 ; bright color s an d parasites , 51 ; carotenoids , 93; chemica l signals , 101 , 108 ; i n child-paren t communication, 124 ; comb, 81; in courtship, 27-37;

Index definition o f (i n sport) , XIV ; i n fighting , 87 ; i n game o f tag , 8 ; gil l cover s hel d open , 66 ; horns , 88; huma n bab y crying , 121 ; huma n body , 211-7 ; key t o reliabilit y i n threats , 16 ; long tail , 83 ; manes, 84 , 237 n 10 ; necktie, 217 ; ne w bod y members, 91-2 ; no t physical , 57; no t random , 92 ; proboscis, 87 ; pupi l size , human , 210-1; quee n pheromone, 159 ; relaxatio n i n threat , 20 ; ribbo n on neck , 216 ; ritua l fighting , 63 ; selectio n fo r (signal selection) , 230 ; i n al l signals , 58 ; i n singin g (time), 28 ; vision , 84- 5 Handshake, 117 Haplodiploid gende r determination , social insects, 153, 162 Hare, Lepus capensis, 121 Harmful chemicals : DIF, 181 ; an d humans, 160; quee n pheromone, 158-6 0 Harmony, in birdsong, 223 Harriers, Circus, 126, 14 8 Hartwell, L.H., 10 4 Hasson, O. , 7 , 57, 232 n 11 Hatzeva, Israel , XV, 4, 15, 77 Hawaii, 70 Hawk, Galapagos , Buteo galapagoensis, 148 Head: apparen t size , and mane , 83 Health, durin g molt, 33 Heeb, 244 n 20 Heinrich, B., 199 Help: asking for, through suicid e attempts , 220-1, 226; in kin selectio n theory, 161-2; in social amebas , 179 ; unneeded, 164 Helpers: exploited offspring, 153 ; helpin g more than necessary, 164 ; in a hive, 161; kickin g out husband , 172; manakins, 35; partnership with kin, 165 ; relatives, 162 ; sterile , and altruism , 241 n 21 Heredia, B. , 200 Heroism, 226 Heron, Ardea, 91, 123 , 171 Heron, gray, Ardea cinerea, 66 fig Hesitancy, 55- 6 Hess, H.E. , 210 Hierarchies, 22 , 166 Hingston, R.W.G. , 237 n 2 Hiraldo, F., 200 Hive: compared to multicellular organism, 153; rank and prestig e in, 160 Hoatzin, Opistbocomus hoatzin, 219 Hochberg, O., 77 Hoglund, J., 35 Holder, K., 81 Hlldobler, B., 165 , 241 n 30 Homosexuality, 219-20 Honesty, 16 , 24, 212, 230 Honeybee, giant , Apis dorsata, nest, 154 fig Honk, C.G.J. , 160,24 1 n 41 Hoof, 49

275

Hook, 50 Hoopoe, Upapa epops, 8 4 Hopping, 34 , 62 Hormones, 107 , 215 Hornbill, Buceros, 85 Hornet, German . Se e Vespula germanica Hornet, oriental . See Vespa orientalis, Horns: color distinct , 49; evolution of, 86-91, 89 fig; female, 90 ; male gazelle, 62- 3 Hospital, 16 0 Hosts and parasites : 185-95; acceptance by host of parasite, 191 ; altruism , 240 n 21; bees, 190 ; constraints, 188 ; Europea n cucko o genets, 186 ; hosts showing themselves to parasite , 192; Mafia model, 189-92; neutering, 192 ; partnership, 37, 193-4; prestige, 188 ; reproduction , 185 , 19 5 Hugs, 21 8 Hultsch, H., 76 Humans: 209-27; actor s an d credibl e threats , 20; aggressiveness an d resolvin g conflicts by threats, 15 6; altruism , 150, 226 ; ar t an d decorations , 223-4, 223 fig, 224 fig ; assessmen t o f anima l signals , 34; alcoholic beverages , 103 , 160 ; baby , 121-2 ; bal d head, 83 ; beard, 17 , 213; body , 210-6, 21 3 fig, 214 fig, 216 fig; boxers, 17 ; breasts, 215; cAMP , 242 n 3; conspicuou s consumption , 227 ; cultura l an d economic developmen t an d evolutionar y systems, 60; differentiatin g morphs, 36 ; diphtheria , 194 ; dogs a s partners , 193 ; eyeliner , 45 ; eyes , 210 fig, 211-2, 21 1 fig; face, 212-3 , 21 2 fig, 213 fig; food storage, 193 ; gende r an d reproductiv e constraints , 219; grou p singing , 76 ; greeting, 117 ; hai r an d dexterity, 20 8 fig, 211; Handica p Principle , 209 , 230; huma n mal e features , 212, 216-7; human female figure , 21 3 fig , 21 4 fig ; inces t avoidance , 210; informatio n centers , 203-4 ; inheritanc e laws , 123; li p color , 51 , 93 ; long neck , 216 ; lou d calls , 74; lov e signals , 218; mal e aggressive stance, 17 , 2 1 fig; media attentio n t o aggression , 16 ; moral behavior an d altruism , 225-7; menopause , 123 ; nose, 86 , 213; observin g babblers , 133 ; pai d enforcers o f rules , 23 ; pitc h o f threat , 20 ; postur e and vocalization , 20; rap e an d dominance , 220 ; reciprocal altruism , 132 ; rhythm , 73; self endangerment an d suicid e attempts , 220-1 ; se x an d testing th e bond , 218-20; signa l selection , 40 ; symmetry, 51 ; testin g th e socia l bond , 217-23 , 21 8 fig; threat b y voice , 21 ; threat s and audience , 75 ; verbal language , 58, 80 , 221-3; vocal emotion, 236 n 5 ; waist , 217. Se e also anthropomorphi c model s Hummel, 8 8 Hummingbirds, 9 9 Hunter, P.M. , 60 Hunting estates , 29 Hunting, 175,231 Huxley, J.S., 64

276

Index

Hyena, spotted , Crocufa crocuta: choosing prey, 11; group-breeding, 131; observing slotting, 7 Hyena, striped , Hyaena hyaena, bristling hair, 83

Ibex, Capra ibex: beard, 87 , 87 fig; black color , 94; horns, 89; ritual fighting , 1 5 fig, 63 Identification o f species, 43^18 , 57, 230 Iguanas, desert, Iguanidae, 54 Hani, G., 87 , 171 Imagination, 211 Imperfection, 47 , 224 Imposition, 112-7 , 217- 9 Impostor, 56 Inbreeding, 210 Incest avoidance : babblers , 128-9 ; human, 210, 225 Inclusive fitness. See fitness Indians, Northwest , 22 6 Indirect reciprocity, 149 Individual selection . See selection, individua l Individual space , 116 Inflation, 59-60 Information centers , 197-205 ; 202; function, 202-3; group selectio n argument , 244 n 20; human gatherings, 203^ Information: collecte d unconsciousl y by humans, 210; conveyed by altruistic behavior , 143 ; conveyed b y beard, human, 214; conveye d by human non-verbal communication, 222; conveyed b y symbols, 221; in emergency, 200; about foo d sources, 199-200; human eyes , 210-2; provided by the sexual act , 218 20; by self-endangerment, 221; signal definition, 59; social bond , 112 ; solving conflicts , parental couple , 175; uninteresting o r unreliable, 23 6 n il Inheritance, 35 , 38-9, 123 , 135 Innate behavior, 21 0 Insects, social . See social insects Insects: an d babblers, 126 ; an d food stored by humans, 193; leks, 35; morphs, 37 ; sexual display, 25; swarms, as food, 198 . Se e also social insects . Insulting, 218 Intelligence, 211 Intensity of vocalization, 222 Intent, showing , 61, 210- 2 Intimidation, 39 , 234 n 44 Inversion frequency , 236 n 2 Investment: in advertising the roost , 204; in altruism , 132, 150 , 167 , 225 ; in "blackmailing" offspring, 122 ; and conventiona l signals, 58; ensuring reliability , 27; of human collaborator s i n partnership, 217; in offspring, 120 , 123 , 131 , 169 ; b y male to persuad e female, 171 ; i n partnership, 112 ; an d reliability , 59 ; and signa l selection, 40 ; in all signals, 58; of time, 28; waste, in signals, 229; by workers i n the colony , 240 n 10 Iris (o f eye), 44^15, 212

Iron, 194 Ishay,J.,240nl4 Jacana, Jacana jacana, 174, 17 4 fig Jackson, C.L. , 104 Jay, Florida scrub , Aphelocoma coerulescens,\48 Jay, Garrulus glandarius, 77, 14 8 Jeering, 8 Jerusalem, cit y of, 46 Jewish sages, 226-7 Joab (Bible) , 213 Joshi, N. V. , 164 Judaism, 227 Judges, contests, 2 3 Kacelnick, A., 237 n 1 0 Kalishov, A., 127 , 13 8 Kangaroo, Macropus, 44 , 44 fig Katsir, Z., 74, 127 Katzir, G., 44 Kelp forests, 97 Kemp, A., 77 Kenya, 77 Keto group, 16 0 Kibbutz, 117,21 0 Killing offspring , 147, 17 1 Kin effect , 166- 7 Kin selection. See selection, ki n Kin: altruism and parasites, 163 ; aggression towar d kin , 164; effec t o f altruism o n reproductio n of , 167 ; human, 210, 211 ; offsprin g as question of , 165 ; unneeded assistance , 164. See also relative s Kissing, 117,21 8 Kite, red, Milvus milvus, 20-1 Kitsch, 224 Knights, as parasites, 19 0 fig , 190- 1 Krebs, J. R. , 66, 185 , 24 1 n 26 Kruuk,H., 11,6 7 Kudu, Tragelaphus, 8 9 Labor, and alcoho l tolerance, 16 0 Lace, 60, 60 fig Lack, D., 54, 235 n 36 Lambrechts, M., 29-30, 72 Language-, human, 80, 220-3; verbal, and animals, 77 80 Langur, Presbytis entellus, 84, 171 Lark, crested, Galerida cristata: and convergence , 54 ; 76 fig; crest, 85 ; mimicry in calls , 77 Lark, desert, Ammomanes deserti, black color , 94 Larkman, T., 127 Larks: singing in flight displays, 25; dives, 62 Larvae, of kymenoptera: 154-5 , 158 , 193 , 24 0 n 14 , 241 n27 Law, human: incest, 210; inheritance, 123 Leader, N., 173

Index Leaning, girlfriend on boyfriend, 11 2 Leks, 34-35, 199 Leopard, Panther a pardus : an d ibex, 87; killing of offspring, 171 ; spots , 48; spots a s signal to prey , 11; and vervet monkeys, 77 Lie detectors, 23 6 n 10 Life cycle, slime molds, 178, 18 3 Life expectancy , babblers, 12 8 Life strategy, and morphs, 36 Light, ultraviolet, 99 Lines. See stripes Lion, Panthera leo: courtship roars , 29; cryptic coloring, 11 ; group-breeding, 131 ; mane , 84 Lips, human, 51, 93,214 Listening, conflic t wit h concentrating , 236 n 10 Littering, an d reciproca l altruism , 132 Lizard, zebra-tailed, Calsaums draconoides, 7 Lizards, 126 Logic: in nature, and the Handicap Principle , 229-30 ; in prey-predator interaction , 6-7; relationship s in insect colonies , 156 , 16 1 Longbow, 82 , 88 Loomis, W.F., 183 , 242 n 4 Loren, Sophia, 213 Lorenz, K. : coral fish, 44; human aggression , 15 Losing prestige, at human fundraisers, 22 7 Lotem, A., 127 , 186-7, 190 , 24 3 n 4 Loudness: begging, 189 ; calls , 74 Lovers, human, 218, 221 Low-pitched threats , 19-2 0 Lundy, K., 127 Lying, and verbal language , 223. See also cheatin g Lyrebird, Menura, 37 fig Lythgoe,J.N.,237n2 Macaque, Macaca, 12 3 Mafia mode l o f parasitism, 189-90 Magpie, Pica pica, 55, 188 , 19 0 Makeup, 214 Male: advertising, 26; aggressiveness toward females , 113; babble r life strategy, 129 , 145,147 ; gazelles, 62 3; hymenoptera, 153 ; in leks, 35; moths, 238 n 4; no t caring for offspring, 31; parenting constraints, 169 , 219; rivals, 234 n 44; and sexua l act, human, 220; taking car e of offspring, 170- 2 Mallard, Anas platyrhynchos, 43, 50 Mammals: adaptations fo r sea , 148; altruism , 150 ; cAMP, 242 n 3; gender roles, 169 ; leks, 35; sexual display, 25; scents, 26, 30; weaning, 119-2 1 Manakins, Pipridae,3l, 35 , 62, 35 fig Mane, 83-4, 83 fig, 213, 237 n 10 Manipulation, of larvae by caregivers , 154 Manor house, 29, 29 fig Mantids, Mantodea, 5 3 Mara, Dolichotis patagonica, 7 Markings: as amplifier signals, 57; and directio n o f

277

gaze, 55; evolution of , 45^18, 53^1, 223; feathers, 51; function of , 45; implications o f selection b y competition, 49; individuals in large gatherings, 205; lines an d stripes , 48; mutation in, 45; polymorphic species, 54 ; queen pheromone, 159; showin g differences, 47 ; of species, gender, age, 43-^, 215 ; and status, 55- 7 Marler, P., 24 0 n 13 Martin, purple, Progne subis, 171 Marx, I., 203 Masks, 213 , 21 3 fig Mate: bright-colored adults , 202; choice, 47; decorations, 223; deserting, 170, 219 ; ducks , 173 ; genes, 26; guarding, babblers, 146 , 170 ; menstruation, 214 ; prehistori c humans , 211; providing for offspring , 27 ; selection of , 25-40; selection a t roost, 199 ; showing off, human, 227 Mathematical models: th e Handicap Principle , XV, 58; kin selection , 163-4 ; reciprocal altruism, 133 ; runaway process, 38-9; o f sexual showing-off, 229 ; versus anthropomorphic models , 9 Mating arenas, 199 Mating flight , 15 5 Mating: babblers, competition over, 127; inces t avoidance, 129 , 210 Maynard Smith,]., 16 , 58, 231 n 1 Mayr, E., 44, 94, 233 n 42 McGregor, P.K., 77 Meadowlark, Sturnetta, 54 fig Media, 15-1 6 Meeting afte r separation , 218 Meinwald, J., 30 Membrane, cell . See cell membran e Menopause, 12 3 Menstruation, 214- 5 Mercenaries, 226 Merlin, Falco columbarius, 8, 8 fig Message: chemical, 184 ; and contents , 57; and cos t of signal, 16 , 229-30 Messenger, secondary, 242 n 3 Metabolism, 158 Microbiologists, 17 7 Microorganisms 177 , 181 . Se e also unicellular organisms Migration, 31, 201 Millionaires, 160 Mimicry: of females (polymorphism), 36-7; vocalization, 76-7; warning colors and cheating, 9-1 0 Mind-reading, 66, 103 Misleading signals, research hypotheses, 230 Mitochondria, 18 3 Mobbing: by babblers, 5, 6 fig, 79,79 fig, 127,133,13941, 13 9 fig, 140 graph, 143; calls, babblers, 73 , 79; of cowbirds by song sparrows, 191-2; and rank, crows, 141; raptors, 5; risks in, 6; used by humans, 231 n 6 Modality, 32

278

Index

Models (discussion) , 9. See also mathematical models; anthropomorphic models Modesty, 225 Mold, cellula r slime. See slime molds Mole-rat, naked, Heterocephalus glaber, 131 Molecules. See chemical signals Moller, A.P., 33, 51, 88, 170 Molt, 33, 51, 186 Money, 159, 225. See also wealth Mongoose, dwarf , Helogale undulata, 117, 133 fig Mongoose, Helogale, 131 Monkey, vervet, Cercopithecus aethiops, 77-80, 78 fig Monkey: 71, 8 3 fig, 86, 218 Monogamous species, 45, 199 Montgomerie, R., 33-35, 81 Mood, and posture, 7 1 Moose, Alces alces, 87 Moral behavior: animals , 225; human, 210, 225—7 Moreno, J., 173 Morphs, 36-7 Morris, D.: mobbing, 231 n 6; pitch of threats, 19 ; red lips, 214; ritualization, 64; Morton, E.S., 171 Moth, 26, 101 , 103, 160, 101 fig, 108 fig Mother: and child , 119-21 ; daughter inheriting from , 124; social insects, 153 Motivation: parent, 172; show n reliabl y by threats, 24; of soldiers, 226; in threats, 18-9; varies during conflict, 22 ; Motro, U., 164 Movement: courtship signals, 34; exaggeration, 67; eyelid, 212; of eyes, 55; and gloss y color, 98-9; ritualization, 61-2; and white color, 96 Multicellular body: chemical communication , 107-8; communication within, 52; comparison to hive , 153 ; primitive differentiation, 178 ; reliability in communication, 58 Murray, I.R., 236 n 5 Musical contests , 46 Mutation: chance, 219; DIF, 183; European cuckoo and ree d warbler, 186 ; and group selection, 23 ; of markings, 45; for selfishness , 161; social amebas, 178; unicellular organisms, 177 Myna, Gracula, 77, 22 1 Myrmecocystis mumicus, 241 n 30 Myxobacteria, 182 , 184 fig Nahon, E., 104 Nail, finger, 49 Nakamura, H., 18 6 Nanjundiah, V., 17 9 Naphtha, 160 Nadv, N., 20 Natural selection. See selection, natural Neatness, 216 Neck, 50, 216, 216 fig

Necklaces, 216 Necktie, 217 Negev, 94-5 Nestling: cuckoo, 186-7, 189 ; feeding of, babblers, 137; gaping, 62, 63 fig; recognizing parasites, 243 n 4; self endangerment, 220 Neuropeptides, 10 8 Neutering, 192 Newt, crested, Triturus vitatus, 26, 26 fig Niche, ecological . See ecological niche Night roost , 202. See also communal roost Nightingale, Luscinia megarhynchos, 15, 76 Nipples, 215 Nisbet, I.C.T., 27 Nitric oxide, 108 Nonmonogamous species, 11 4 Nonverbal communication, 80, 221-3, 244 n 19 Nose, 6 , 86, 213, 213 fig Notices (conventiona l signals), 58 Noxious chemicals , 22 Numada marshamella, 190 Nuptial gifts, 2 6 Nursing, 119-2 1 Nutrients, prespore and prestalk social amebas , 179-80 Nymphs, 241 n 27 O'Donald, P., 28 Observation: of babblers, 133; of babblers' fights to th e death, 15 ; of features, 82; of prey behavior, 66; of signals, 59; o f stotting, 7 Offspring: an d "blackmail, " 122 ; constraint s on gathering food for, 156; dominant female, 175; early in season, 191 ; and evolutio n o f altrusim, 125; existing, 123 ; exploited b y parents, 123 ; extrapai r copulation, 171 ; host care, 189; human, 209-11, 219, 230; as kin, 162; not crying , gazelle, 121; not one' s own, investment in, 131-3; and parents, 119-24 ; relations betwee n parents , 170 ; runawa y process model, 38-9; selectio n o f father, 49 ; sexual reproduction, 169 ; social insects, 152; widows, 172; of workers, 155 Olympic games, 65 One-celled organisms . See unicellular organisms Open spaces, 95 Opium, 160 Orange color, 98 Orangutan, Pongo pygmaeus, 84, 84 fig Ornamental feathers , 62 Ornaments: becoming cheap, 59; marked, 32 Ostrich, Struthio camelus, 148 Osztreiher, R., 116 , 127 Outlining shape , 49 Ovaries, 15 9 Oviduct, 15 3 Owl, 6 fig, 86, 123 Owl, barn , Tyto alba, nestlings, 12 3 fig

Index Paintings: 223-4, 223 fig, 224 fig Pairbond, 116 Parasites: 185-95; accepting, to minimize damage, 191; altruism of host, 240 n 21; arms race, 185-6, 195 ; balance with host , 185 ; an d bright colors , 51; care for, proving ability, 172 ; a s collaborators, 190, 193 ; European cucko o populations, 186 ; grea t spotte d cuckoo an d crows , 187 ; amon g kin, 163; less an d more virulent, 194; neutering host, 192 ; recognizing as nestling, 243 n 4; risk to nest , 152 . See also host and parasite Parasitism, social. See social parasitism Parent-child conflict , 119-20, 124 , 238 n 2 Parental manipulation , 123 , 153 ^ Parents: babbler s an d adul t offspring , 128 ; competition ove r caregiving , 171-2 ; exploitin g offspring, 123 , 153 ; gende r constraints , 169 ; genuine warnin g calls , 12 ; human bab y crying , 71 , 121, 220 ; n o parenta l involvement , 27 ; an d offspring, 119-24 ; parenta l couple , collaboratio n and conflict , 169 , 175 , 18 9 Parker, G.A., 1 6 Parker, P., 127 Parrots, Psittacidae, 77, 80, 116 , 123 , 221, 22 1 fig, 244 n 17 Partner: aggression among, 240 n 5; babbler, dependance on , 115 ; convincin g of quality, 171; evaluating, 160 ; gai n by partnering with relatives , 165; guardin g territory, 188; i n a hive, 157-8 ; human, and bab y crying, 121 ; language and cooperation, 222 ; mate and territory , 174 ; parasite against other parasites, 190; prestige, 149 ; prestige, babblers, 146 ; reciprocity, 134; showing off wealth, 227; strength of bond, 111 ; testin g the bond, 117 ; weaker, and threa t o f self-injury, 120 , 220 Partnership: an d altruism , 172; babblers, ne w group, 130; breakup, crows, 188 ; complementar y partners, 173; group selectio n an d exploitation , 131 ; host an d non-virulent parasite, 194; humans, 217-220; humans and dogs , 193 ; insects, 152-4; with kin, 165 ; new pair, 188 ; strengt h of, 112 ; testing the bond , same-gender, 220 Partridge, chukar , Alectoris chukar, 55 Paternity, guaranteeing, 170 Patriotism, 226 Pattern: emphasize s features , 48 ; eyelike, 50; human beard, 213; selection of , 223; of threat, prevents cheating, 23-4. See also markings Payne, R.B., 77 Peacock, Pavo cristalus: color, 30; combination o f signals, 32-33, 32 fig; courtship, 114 , 11 4 fig; escaping predators, 33; eyelike patterns, 32-3 , 52-3 , 52 fig; hen's preference , 33, 223; males not caring for offspring , 31 ; tail XIV, XVI, 25, 34, 52, 55, 82, 144; waste, 150 , 15 7 Pecking order, 22 , 144

279

Peek-a-boo, 115 Peers, 204, 227 Pelican, Pelicanus, 50, 84 Pelican, white , Pelecanus onocrotalus, sexual display (bump), 26, 28, 28 fig Penguin with chick, 122 fig Penguin, Humboldt , Spheniscus humboldti, 99 fig Pepperberg, I.M., 244 n 17 Peptide, 104- 5 Peregrine, Falco peregrinus, 126 Perl, Y., 127 , 146 Personal space , 218 Pest control , 23 8 n 4 Pest, 189 Petri dish, 178, 180 Petrie, M., 33, 35, 52 Petting, 218 Pheasant, Pbasianus, 25 Phenotype, 107 , 182 , 186-7, 194 Pheromones, 30-1 , 101-2 , 104-7, 152 , 159-60, 238 n 4. See also queen pheromon e Physical contest, 63 Physiology, human and animal , 209 Pigeon, passenger , Ectopistes migratorius, 203 Pigeon, rock , Columba livia, 174 Piglets, wild , Sus scrofa, 121 , 12 1 fig Pilots, 216 Pioneers, 117 , 11 7 fig Pipit, long-claw , Macronyx, 5 4 fig Pitch, vocal , 19-2 0 Plasmid, 19 4 Plastic, 225 Plate, decoratio n of , 47, 47 fig Playing. See games Plovers, 85 , 171 , 17 5 Poison: i n butterfly pheromones, 30 ; DIP, 179 , 181-3 ; queen pheromone , 158 ; warning colors, 9-1 0 Police, 23 Polistes biglumis, 193 Polistes canadensis, yellowjacket, 165-6 , 167 fig, 240 n 5 Polyandry, 170 Polygamy, 45, 16 9 Polymorphism, 36-37, 53-54 Ponzo's effect, 23 7 n 10 Population: crow , 188 ; huma n hair, 211; passenge r pigeon, 203; reed warbler, 186 ; runawa y process model, 38; social amebas, 178, 180 ; spread o f marking in, 53^&; spread o f traits in, 230; traits spreading through, 189; unicellula r organisms , 177 Porcupine, Hystrix indica, 83, 218, 21 9 fig Posture: an d clothing , 46; and color , 30; and mood, 71; and voice, 20, 69 Potlatch, 226 Pozis, O. , 127 Prayer, communal , 203 Precision, 29-30 , 216

28O

Index

Predation: DI F as protection, 181 ; forcing cooperation, 157; youn g cuckoo, 18 9 Predator: XHI-XIV, 3-13; adaptation s to avoid, 59; avoidance of , peacock, 33 ; and babblers, 133-4 ; and breeding colonies , 199 ; and communa l roost, 198; and different warnin g calls, 78; and ducks, 23 5 n 36; and foo d storage , 193; individual space , 116; mobbing of , babblers, 141 ; nests, 122 , 152, 173; and nonterritorial babblers, 127; observing behavior , 66; observing rump muscles, 49; and prey, equilibrium, 186; showin g self t o prey , 11 , 232; signalin g to prey , 10-1; an d spores, 178 ; warning calls, 4-6, 74; and warning colors, 9 , 237 n 2; and weaning , 121; and young cuckoos, 189 Prediction, 23 0 Preening, 115-7 , 143 Preference, 3 8 Pregnancy, human, 214-5, 217 Prehistoric humans , 211, 223 Preparatory movements , 66 Presents, 26, 32, 226. Se e also gift-givin g Prespore, socia l amebas, 178-9 , 181- 2 Prestalk, social amebas, 178-82 , 242 n 4 Prestige model , host-parasit e relationship , 18 8 Prestige: accorde d handicaps , 55; and altruism , human, 226-7; altruism, social insects, 157 ; avoiding rights , 157; and bab y crying, 121; commands, 76; competition, 125 ; and copulation , babblers , 147; crows mobbing, 141 ; definition, U3-4; an d gain in feeding cuckoo , 188-9 ; gai n from altruism , 167; in a hive, 160 ; and quee n pheromone , 158-9; an d rank, babblers, 143-5 ; showing off in large gatherings, 205; and shyness over sex, babblers, 145 ; status badge, 56; and takin g care of young, 171; warning calls, 12; zerosum game, 149 Prey: communicating with predators, XIII-XIV , 3-13, 232 n i l; cooperation wit h predators , 12 ; and female dominance, 174 ; and predator, equilibrium , 186; selection of , 67; signaling to, by predators, 10-11; warning calls, 5 Primates, 22 1 Princesses, 15 6 Printing pres s workers, 203-4 Proboscis, 8 7 Procreation, an d sex, 218-20 Propaganda, 23 4 n 44 Propeptide, 104- 8 Property, 216 Propheromone, 104- 7 Protection: allelopathi c chemicals, 181 ; and extrapai r copulation, 171 ; humans, 217; of money, 159 ; of nest, insects, 152 , 155-6; by parasite, 190, 193; of spores, 178 ; of territory, 125, 131; by toxin, 194 Protein, breakdown , 182-3 , 24 0 n 14 Prowess, genuine , in Fisher, 23 4 n 44 Psychological weapon , 119—2 0

Ptarmigan, Lagopus mutus, 81 Pubs, 160, 203 Puffin, Fratercula arctica, 87 Punishing: hos t an d parasit e (Mafi a model) , 189-90; of non-conformists (threats) , 23; in reciproca l altruism , 133 Pupae, stealin g of , 193 Pupil (eye), and impression o f human character, 210-2 Pursuit-deterrent signals , 232 n il Quail, California, Lophorfyx californicus, 85 , 8 5 fig Quail, European , Coturnix coturnix, 85 Quality: an d beauty , 52; in a mate, 27; of offspring, 124; in raptors , 175; and rar e objects, 60; of voice, 222 Queen pheromone, 158-61 , 240 n 13 Queens, social insects, 152-61 , 15 6 fig, 159 fig, 241 n 41 Race, foot, 65 fig Railroad timetable s (a s conventional signals) , 58 Rank: babblers, 128 ; and copulation , babblers , 147, 170; and mating , babblers, 146 ; and prestige , 143-5 ; among relatives , 166; and sentinel activity, babblers, 134-6; social insects, 156-6 0 Rape: house sparrows and ducks , 170; humans, 220 Raptors: and babblers, 126; babblers mobbing , 73 , 133, 139; chance s of catching babblers, 5; femal e dominance, 174-5 ; flight display s by prey, 204; humans hunting mobbing birds , 231; prestige, 144; prey cooperating with, 12 ; warning calls, 3-6 Rattlesnake, Crotalinae, 54 Raven, Corvus corax, 95, 97 , 9 5 fig Razorbill, Alca torda, 87, 199 Receiver: of chemical signal , 184 ; of signal, 59; of signal, and reliability , 229 Recipient, o f secret charity , 227 Reciprocal altruism , 132-5, 149 , 153, 163, 171 Recognition: o f parasite, 186-7, 243 n 4; of prestige , 158 Red, use of, 96, 98, 214 Redhead, Aythya americana,¥)\ Reducing cost of signals, 59, 8 8 Reflection, 95 , 98, 212 Reindeer, Rangifer tarandus: evolutio n o f antlers, 90; rivals, 22 Rejection, o f cuckoo's eggs, 186 Relatedness, 163. See also ki n Relationship: in hive, 154; human sexual act, 220; between parents , 170 , 173; Relatives, aggression toward , 164 ; incest avoidance, 210; kin effect , 165 ; kin selection theory , 163^t ; partnership with, 165-6; reproductio n of , 162. See also ki n Reliability: 195; ability in a hive, 157-60; altruism , 144, 150; chemica l signals , 184; in child-paren t communication, 124; clothing, 217; in color, 30 , 235

Index n 36; communication, babblers, 222; communication between parental couple, 175 ; communicatio n within the body , 52, 58; conventional signals , 58; in courtship signals , 26; demanded b y observers, 59; eyes, human, 212; friendliness, 62 ; genetic code, 58; in hormones, 107 ; host an d parasite, 194-5 ; and human language, 58, 221; human nonverba l communication, 223; intent an d pupil size , human, 210; need for food, 122 ; red lips, human, 214; selection for , 230; showing off wealth, 227; signals, XIV-XVI; signal selection, 40; status badges, 56, 170; testing fo r committment, 218; testing the bond , 112, 117 ; i n threats, 16 , 19; threats an d threa t substitute, 142 ; vocalizations in chase, 9; vocal signals, 69-70; and waste, 229 Religion, 217, 203 , 210 Repression, o f larvae, 152 Represser protein, diphtheria toxin, 194 Reprimanding, 14 5 Reproduction: an d altruism , 161; babblers, 129 ; bottleneck an d femal e dominance, 174 ; chances of social insects, 157; competitio n over , 142 , 159 ; deserting a partner, 170; favor s "selfish " mutants, 23; gender constraints, 169 , 219; of host an d parasite , 185, 195 ; huma n gender constraints, 219; ki n selection, 162 ; natural selection, 122-4 , 123 , 131 , 151; i n other nests , ants, 241 n 30; parenting couple , 169; parent/child an d kin selection, 238 n 2; parentoffspring conflict , 120 ; partnership wit h kin (socia l insects), 165 ; and prestige , 145 , 158 ; ree d warble r and Europea n cuckoo , 187 ; rhythmi c song, 30; sexual selection, 38 ; species identification, 45; and spread o f traits in populations, 230 ; and ston e collecting, 173 ; and takin g care of others' offspring, 188-9; of workers, social insects, 153 Reproductive organs , attacked by parasites, 192 Reptiles; adaptations for sea , 148; morphs, 37; sexual display, 25 Requests, 75 Reserves: of fat, 103; of glycogen, 182 Resolution o f conflicts, 15-6 , 142 Resources: of neutered host , 192 ; for surviving DIP, 182-3 Revolt, 137 Reward: for altruism , 139; for art , 224 Rhinoceros, Asiatic, Rhinoceros unicornis, 86 Rhinoceros, square-lipped, Ceratotherium simum, 86, 86 fig Rhinoceroses, Ceratotherium, 63 Rhythm, 21, 30, 72^t Ribbon, 216 , 21 6 fig Richner, 244 n 20 Ring, around eye , 55 Risk: altruism and risk y activities among humans, 225 7; ensuring credible threat, 24 , 68; of fighting, 234 n 44; to kin, in kin selection theory , 164 ; to nest, 152 ;

281

mobbing, 6 ; of predation, youn g cuckoo, 189 ; suicide attempts a s a request fo r help, 221; warning colors, 9 Ritual fighting, 62-3, 9 0 Ritualization, 61-8; descriptio n of , 64; natural selection , 68; reliability, 64-5; signals , 66-7 Rivals: communication between, 15-24 ; comrades as, 142; cooperation with, 60; evaluating, through quee n pheromone, 160 ; gift-giving to , human, 226; Handicap Principle , 40 ; hierarchies, 22; "propaganda," 39, 234 n 44; rump, 49; song, 15; stretching displays, 19 Roar: peacock, 32; red deer , 15 Robbery: forcing cooperation, 157 ; of money, as mode l for quee n pheromone, 159 ; parasit e as protector , 190, 193 ; risk to nest , 152 Rodents, an d food store d b y humans, 193 Rohwer, S., 56, 237 n 2 Rollers, Coracidae, 99 Roman centurions ' helmets, 85 Rook, Corvus frugilegus, 20 0 Roost, communal. See communal roos t Rooster, Callus gallus, 39 fig, 51, 81, 81 fig, 87, 93, 234 n44 Roseler, P.F., 160 , 24 1 n 41 Rowell, T.E., 7 1 Royal jelly, 153 Ruff, Philomachus pugnax: courtship dances , 62; dancing in arenas, 25; dark and light morphs, 36 , 36 fig; leks, 35, males not carin g for offspring , 31 ; 199 ; polymorphic markings, 54 Rump, 48-9, 5 3 Runaway process model , XVI, 37^0 Ryan, M.J., 29, 235 n il Sagi, Y., XVI Sahara, 94- 5 Salamander, Salamandra, 1 3 fig Sambar, Cervus unicolor, 98 Same-gender partnership, 220 Sand castles , 31 Sandgrouse, Ptemclididae,\<)l fig , 19 9 Sandpiper, spotted, Actitis macularia, 174 Scale (reptile), 90 Scale insects, 162 Scents, 26, 30, 51- 2 Scherer, R.K., 70 Schools of fish, as food, 19 8 Schwammberger, K., 193 Scorpionfly, Panorpa, 5 1 Scorpions, 12 6 Seal, elephant, Mirounga leonina, 87, 87 fig Secret charity , 227 Selander, R.K., 45 Selection, group : altruism , 131 , 161 ; information centers, 204, 244 n 20; and kin selection , 24 1 n 33;

282

Index

Selection, grou p (continued] Mafia model , 190 ; neutering hosts , 192 ; queen , 161 ; social amebas, 178 , 181 ; an d social parasitism, 131 ; superterritories, 28 ; theory, 23, 163-4 , 232 n 17 , 231 n 1 , 232 n 18 ; threats, 23; unicellular organisms, 177 Selection, individual : altruism , 158; communa l displa y at roost, 205; compared wit h kin selection theory , 164; socia l amebas , 179-81 ; unicellular organisms, 177, 182 Selection, kin: and altruism , 132, 149 ; ants leaving colony, 241 n 30; calculation, 241 n 26; description , 161-3; and grou p selection , 164 , 241 n 33; kin effect , 166; an d parent/offsprin g conflict, 23 8 n 2; problem s with, 163-5; reception o f theory, 163 ; selfish kin , 164 ; social insects, 153 ; unneeded assistance, 164 Selection, natural : alternating signa l and utilitarian, 90 1; for altruism , 158 , 226 ; and behaviora l mechanisms, 12-13, 219; "blackmail" systems , 122; group breeding, 131 ; grou p selection , 23 ; and th e Handica p Principle, 230 ; incest avoidance , 210; an d kin, 165-6 ; kin selectio n theory , 163 , 166 ; merging of morphs, 36; psychological factors, 120 ; reproduction , 123 ; ritualized movements, 68; set-specific markings, 45; sexual display, 37; sexual selection, 38 ; social insects , 151; testin g th e bond, humans, 218; utilitarian selection an d signa l selection, 40; warning calls, 12 Selection, sexual : 38, 40, 44, 91, 233 n 42 Selection, signal: altruism , 125 , 158 ; chemica l signals, 103; contrar y to utilitarian selection, 92 ; cost, 58-60; defined, 40 ; and the Handicap Principle , 230; horn s and antlers , 90-1 Selection, utilitarian , XV, 40, 59, 90-1, 23 0 Selection: advantage in, 53; altruism, 167; decorations , 223; definition, 161; an d the Handicap Principle , 230; of markings , 45; ritualizatio n and reliability , 64-5 ; selecting mechanism, 47; selecting prey, 67; selectiv e factor, ki n selectio n versu s kin effect , 166 ; o f mate in information centers , 199 ; parenting partner (choic e and number) , 169; selectin g brightl y colored mates, 202; throug h quee n o r through workers, 153 ; utilitarian and signal (defined), 40. See also selection, group; selection, individual ; selection, kin ; selection , natural; selection, sexual ; selection, signal ; selection , utilitarian Self-confidence, toads , 20 Self-control, human , 215 Self-defense, warbler s agains t cuckoo, 186- 7 Self-endangerment: animals , 225; humans , 220-1; offspring, animals , 120- 2 Self-injury, 12 0 Self-sacrifice: amebas , 178; humans, 226; kin selectio n theory, 163 "Selfish gene, " 162 Selfishness: an d altruism, 161 ; human , 227; kin selection theory, 164 ; litterers, 132 ; young camouflaged individuals , 202

Sensitivity to color, 23 7 n 2 Sensory cells, 10 2 Sensory exploitation, 235-6 n il Sentinel activity : babblers, 3 , 127 , 133-6 , 144-5 , 225, 4 fig, 135 graph, 13 6 graph; cost i n time, 28; Serfs, 190- 1 Set-specific markers, 43-49 Sexual act, 117 ; forced, 220; not fo r procreation , animals, 218-9; same-gender partnerships, 220 ; testing the bond, humans, 218-20; Sexual display, 25-6, 37-8 , 22 9 Sexual partners, selection an d number, 169 Sexual reproduction, parenting couple, 16 9 Sexual selection. Se e selection, sexua l Sexual showing-off, mathematica l models, 22 9 Shame, 226 Sharing: assets, human partners, 217 ; a territory, ma n and dog , 193 ; territory, pair , 11 3 Shaulski, G., 183 , 24 2 n 4 Sheath, of beak, 87 Sheep, bighorn, Qvis canadensis, 63 Shelduck chick, Tadorna tadorna, 71 fig Sheriff, 2 1 fig Shouting, 74-5, 12 7 Showing off : ability as provider, to one's mate, 189 ; ability, by waste, 157 ; abilit y and commitment , by altruism, 158 , 172 ; abilit y to feed, 27-8; adaptation , by handicap, 57 ; by altruism, humans, 227; art and imperfection, 224 ; by clothing, 216-7 ; by color, 93; confidence, 56; confidence, by threat, 18 ; contempt , 22; differences , 47; by donations, 160 ; lac k of parasites, by color, 51; in large gatherings, 205; by long neck, 216 ; motivation , b y pitch o f voice an d stretching, 19 ; motivation, b y threat, 24; neatness, by uniform, 216; perfectio n o f edge, by line, 47; posture, b y color, 30; sexual, 229; status, to peer s and competitors , 204 ; symmetry, by decoration, 48 , 52; wealth, through mansions , 29; wealth, 227. See also advertisin g Shpirer, Y. , 187 Shrike, great gray, Lanius excubitor, 98 Shrike, red-backed, Lanius cristatus, 55 fig Shrikes, Lanius, 4, 26-7, 5 5 Shushlui Nature Reserve, South Africa , 49 , 86 Shyness, babblers, 145-7 , 225 Siamese fighting fish, Eetta splendens, 6 5 Sibling: helping, 162 , 164 ; parent/child conflic t and ki n selection, 23 8 n 2; preference, 123 . Se e also selection , kin Side effects: antibiotic s an d DIP, 181 ; benefi t t o group , 149; kin effect , 166 ; specie s identification, 57 Sigmund, K., 239 n 1 7 Signal selection. See selection, signa l Signals; acetate , acid , alcohol , aldehyde , 102 ; altruism, 149-50 ; betwee n experience d co-workers , 222; chemicals , signa l an d non-signal , 184 ; clarit y

Index of, 64 ; combination s of , 32 ; conventions , 58 , 218; correlation t o quality , runawa y proces s model , 39 ; cost an d message , 229-30 ; cres t an d mane , 84 ; definition, 59 ; deterren t t o pursuit , 23 2 n 11 ; DIF not a signal , 182-3 ; evolutio n of , 23 5 n 36 , 235- 6 n 11 ; ey e movemen t an d markings , 59 ; an d handicaps, 57 ; huma n bod y fa t an d breasts , 215 ; human bod y parts , 212-7 ; huma n clothing , 216; human hair , 211 ; markings an d arbitrariness , 47 , 53; o n membranes , 230 ; menstruation, 214; neuropeptides, carbo n monoxide , nitri c oxide , toxic chemicals , 108 ; observation of , 59 ; b y predators t o prey , 8-11 ; b y pre y t o predators , 3 10; quee n pheromone , 159-60 ; receive r an d reliability, 229 ; reliability, XIV , 229; ritualized movements, 67-8 ; set-specifi c markings , 43 , 58 ; a signal o r no t a signal , 82 ; signa l selection , 40 ; status badges , 56 ; threats , 17 , 24 Simpson, M.J.A. , 65 Sinai, 94-5, 172 , 203 Singing: and abilit y to feed , 28; courtship signals , 32; Darwin, C. , 70; sexual display, 25; showing abilit y to escape, 8 ; threats, 15 ; as waste, 39 Size, apparent, 213, 237 n 10 Skill: and art, 224; an d long hair , 211 Skin, exposed, and heat loss, 214 Sky divers, 62 Skylark, Alauda arvensis: singing in escape, 8; wing load, 88; white color , 94 Slagsvold, T., 14 1 Slime molds (socia l amebas), Dictyostelmm discoideum, 131, 177-84 , 177 fig, 242 n 4, 18 0 fig, 182 fig Slug (aggregat e of social amebas), 178-9, 182- 3 Smith, H.G., barn swallows , 33-34 Smith, J.N.M., 191 Smith, N.G., gull species identification , 44 Smythe, N., 7 Snakes, mobbed by babblers: 79 , 126 , 133, 139, 139 fig Sober, E., 241 n 33 Social amebas . See slime mold s Social behavior: and environment , whit e wagtail, 198; of humans, and the Handicap Principle , 209 Social bonds: XVI , 111, 113, 117, 217-23 Social hierarchies , 22 Social insects : 150-67 ; aggressio n amon g kin, 164; altruism, 179; basic problem, 152 ; colony formation, 154-5; exploitation , 124 ; food gatherin g constraints , 152, 156 ; gender determination , 153 ; group selectio n and altruism , 161; kin selection an d haplo-diploid mechanism, 162 ; small colonies , 158; social parasites, 157 Social parasitism, 131 , 157, 163 Social rank. See rank Social standing: humans, 216, 226-227'. See also prestig e Social status, an d markings , 55. See also prestig e Social structure , evolution of , social insects , 154

283

Social system: assumption o f equilibrium, 195 ; humans, 209, 223; social insects, 153 , 162-3; unicellular organisms, 177 Sociobiology, 24 1 n 26 Soldiers, 226 Solenopsis invuca, 241 n 30 Soler.M. and]., 190 Soliciting donations , humans , 227 Song, threat and/o r courtship, 39, 234 n 44 Sordal, T. A., mobbing, 6 Sorensen, M. , 191 Space exploration, 22 6 Space, personal, 21 8 Sparrow, Harris's, Zonotrichia querula, 55-6 Sparrow, house, Passer domesticus, 170, 198, 204 Sparrow, song , Melospiza melodia, 191- 2 Sparrowhawk, Accipiter. 5 fig, 126, 169 fig Species-specific markers , 44-5, 53, 68, 215, 230 Species: coalitions with different (symbiosis) , 166, 190; identification, 45 , 57; identification of , 215, 223, 230; markers identifying, 43; markings showin g differences within , 47; signals marking, 58 . See also markings; decoration s Sperm, 90 , 153 Spikes, 50 Spines, 49 Spiritual values , and evolution, human , 224-5 Spores, socia l amebas , 178-80 , 242 n 4 Sports, 34-5, 46, 65, 203, 214, 226 Spots, 11,4 8 Spread o f traits in population, 53, 230 Squirrel, Sciurus, 98 Standardization, in gymnastics, 34 Standards, conforming to, o f dress, 21 6 Standing guard . See sentinel activit y Staring: humans, 212; and markings , 55; as threat, 21 Starling, Sturnus vulgaris: communal roost , 198 , 202, 204, 205 fig; cooperating wit h raptor , 12 , 12 fig; individual space , 116 Starlings, African, 14 8 Status badges, 55-7, 5 8 fig, 170 Status, 29, 55-7, 149 , 170, 204, 227. Se e also prestig e Stealing: bowerbirds, 32 , 60; social insects, 156 , 159, 193 Sterile individuals, socia l insects, 152 , 162, 241 n 21, 241 n 27 Sterility, caused by parasites, 192 Stilt, Himantopus, 17 1 fig Stone collecting, 172- 3 Stone curlew , Burhinus oedicnemus, 50 Stonechat, Saxicola torquata^ 44 , 55, 21 9 Storing, o f food: by humans, 193 ; by insects, 152 Stork, white , Ciconia ciconia, 50 Stotting, XIII , XIV fig, XV, 6-7, 232, 232 n 7 Strangers, language and the limit s of nonverba l communication between , 22 2

284

Index

Stresemann, E., 94 Stretching, 18- 9 Stripes, 48, 55, 121 ; through eye , 55 Struggle: among kin, social insects, 165; among workers in a hive, 156-7; minimizing damage, 192 Style, 46-7,215, 217,225 Subordinate, 56, 143 , 149, 212 Sugars: digestion, in wasps, 240 n 14 ; in glycoproteins , 105 Suicide, 178, 220-1, 226 Sulcopolistes artimandibularis, 193 Sunbird, Palestine, Nectarinia osea, 171 Sunbird, scarlet-tufted malachite , Nectarinta johnstoni, 34,82 Sunbird, yellow-bellied, Nectarina venusta, 28 Sunbirds, Nectarina: color , 30; glossy color, 99; mobbing raptors , 5 Sunfish, bluegill , Lepomis macrochims, 36 Superterritories, 28- 9 Surgeonfish, Zebmsoma, 50 Surrogate mother, parasite and host, 191 Survival: lon g hai r a s a handicap , 211; nee d for cooperation , socia l insects , 157 ; and DIF , 180-3; socia l amebas , 179-80 ; tin y rate , o f young , 151 Swallow, barn, Hirundo rustica, 33, 33 fig Swallow, cliff , Petrocbelidon pyrrhonota, 199 Swallows, Hirundo, communal display at roost, 204; individual space, 116, 116 fig; long tail, 82; symmetry, 51 Swan, black, Cygnus atmtus, 98 fig Swan, Cygnus olor, 96-7 Swarm, 156, 198 Symbiosis: compared t o collaboration wit h relatives, 166; and th e Mafi a model , 193 Symbolic language, human, 221-2 Symmetry, 48,51-3,223 Synchronization and eyelike pattern, 53 Synchronization, of development, 52-3 Taborsky, M., 233 n 36 Tag game, 56 Tail, 25, 33^, 49, 53, 55, 82, 144 Taste, clothing , 216 Tastelessness, 22 4 Tel-Aviv University zoo, 85 Termites, hoptem, 151 , 153, 162, 162 fig, 241 n 27, 244 n18 Tern, Sterna hirundo, 25 fig, 26-7, 82 Territory: babblers , 4 , 125 , 127-9, 131 , 147-8 ; crows, pai r cooperation , 188 ; dark ruffs , 54 ; defending b y threats , 23 4 n 44 ; defens e of , an d dominance betwee n genders , 173-5 ; givin g up , 16 ; miniterritories, 34-6 ; morphs , 37 ; pair , 113; parasite, 190 , 192; prehistoric humans , 211; readiness t o protect , 15 ; shared b y human s an d

wolf-dogs, 193 ; status badge , 56 ; Superterritories , 28-9; territoria l versu s migrator y kites, 201 ; white wagtail, 197- 8 Testing: b y children , 111 , 114-5 ; commitmen t o f groupmates, 115-7 ; o f commitmen t o f male , 175; by dogs , 112 , 117 ; female intentions, 114 ; humans, 117; humans , 111 , 114-5 , 117 , 217-8, 220-3 ; of partnership, throug h sex , 219-20; shynes s a s tes t of prestig e 145 ; of socia l bond , 111-7 , 217-23 , 230; o f th e theor y o f signa l selection , b y inflation , 59 Testosterone, 5 6 Theory: evolutio n an d altruism , 131, 161-3; value of (general), 230 Thomas, A.L.R., 34, 82 Thornhill, R., 51 Threats, 15-24 ; aeria l display , 15 ; altruis m a s a substitute, 141-3 ; altruis m a s surrogat e threat , 149 , 157-8, 160 ; by approaching , 17; and audience , 75; bluffing, 17 ; bristling, 83 ; calls , 19-21 , 115 , 222; contempt, 21-2 ; grou p selection , 23 ; t o inferior , 116; no t conventions , 23—4 ; noxiou s chemicals , 22 ; by parasite , 190 ; preferable t o fighting , 16 ; raise d hair, 18 ; rank , 144 ; reliability of , 16 ; ris k in , 68 ; ritual fighting , 63 ; afte r a roun d o f fighting , 22 ; o f self-injury, 120 ; sexual selection , 38 ; singing , 15 ; staring, 21 ; stretching , 18 ; substitut e fo r aggression, 142 ; vocalization, 19-21 ; wit h witnesses, 14 2 Thrush, laughing, Garrulax, 76 Ticks, 49 Tiger, Panthem tigris, 29, 1 0 fig Timaliidae, 12 5 Time, wasting, XIV, 28 Timna, Y. , 171 Tinbergen, N., 50 Tit, great, Pams major, breas t band compare d with human necktie , 217; courtship song , 29-30; eyestripe, 55, 57 fig; feeding ability, 27; rhythm, 72; status badge, 55-6 Tit, long-tailed, Aegithalos caudatus,148 Tit-for-tat, 133^, 139 Toad, Bufo bufo, 20 , 2 0 fig Tobacco, 16 0 Toddler, 220 Todt, D. , 76, 127 Toxic chemicals, a s signals, 108 Toxins: diphtheria , 194 ; evolution of , 181- 2 Traits: assessing, 216; cost, 59; brought out by decoration, 223; human social, 209; and ki n selectio n theory, 165 ; spreading, 189 ; surviving, 219 Tricking, host b y parasite, 189 . See also cheatin g Trills, babblers , 5, 73, 78, 137 , 141 Trivers, R.; altruism and parenting, 171; gender conflict, 23 2 n 1 , 242 n 1 ; kin selection , 241 n 26 , 238 n 2; pairs of female gulls, 220; Parent-child

Index relationship, 119-20; reciprocal altruism theory, 132, 163 Tropicbird, Pkaethon aethereus, 34, 82 Tufts, 8 6 Turkey, Meleagris gallopavo, 51 Ultraviolet light, 99 Unconscious, behavior mechanisms, human, 210 Uneven cost, 58 Unicellular organisms, 60, 150 , 177-8, 184 Uniforms, 216 Unpalatable prey, 9 Unsaturated fatt y acids , 101 , 160 Utilitarian selection. See selection, utilitarian Values, spiritual, 224 Vandal, D., 22 Veblen, T., 160 , 203-4, 227 Verbal language : animals , 77-80, 221; humans, 221-3 , 244 n 1 9 Vespa orientalh, oriental hornets: 155 , 240 n 14 Vespula germanica, German hornet , 155 , 240 n 7 Vidua, Vidua, 2 5 Vigor, related t o attractiveness , 233 n 42 Violence, 16 0 Viper, Vipera, 5 4 Virulence, 194 Viruses, 194 Visibility as a handicap, 235 n 3 6 Vision: handicaps to, 84-5, 211; monocular, 55 Vocabulary, 72 Vocal emotion, human, 236 n 5 Vocalization: 69-80; babbler research , 127 ; in a chase, 9; courtship, 29-30; distance , 20-21; and emotion, 222; message, 71; mimickry, 77; pattern an d it s message, 72; reliability, 69-70; threats 19-20 ; and verbal language, 244 n 19 . See also call s Voice: an d posture, 69 ; quality of, 222; concentration, 236 n 1 0 Vulture, Egyptian, Neophron percnopterus, 202 Wagner, R., 87, 199 Wagtails, Motadlla alba: roost, 202, 204; social organization, 197-8; wintering pairs 113 , 113 fig, 219 Waist, 217 Wallace, A.R., 44 War paint , 39 Warbler, great reed, Acrocephalus arundinaceus, 186-7 Warbler, scrub , Scotocerca inquieta, 185 fig Warbler, Seychelles , Acrocephalus sechelensis, group living, 14 8 Warbler, Sylvia, 5 Ward, P., 198 , 243 n 4 Warmth, 199 Warning calls : babblers, 3 ; genuine, 12 ; loudness, 74 ;

285

and prestige , 12 ; questions pose d by, 4-6; various, 78-9 Warning colors, 9-10, 1 3 fig Wasps: colony , aggression, 164 ; colony formation, 154 , 156, 166 ; evolution o f eusocial systems, 162; helpless offspring, 152 ; host an d parasite as collaborators, 193; large social groups, 244 n 18 ; larvae, 241 n 27; larvae, digestion o f proteins fo r workers, 240 n 14; reproduction b y workers, 153 ; temperate an d tropical regions, 240 n 5 Waste: 28; altruism , 150 ; attractin g mate s an d deterring rivals , 39 ; i n breasts , 215 ; ensurin g reliability, 40 ; ESS , 233-4 n 43 ; menstruation , 214; natural selection , .38; prestige, 149 ; role i n th e Handicap Principl e (summary) , 229; runaway process model , 37 ; socia l rank , 144 ; of time , XIV, of warmth , 21 4 Water balance , 127 Waterbuck, Kobus ellipsipryninus, 4 9 Weaker partner : threat of self-injury, 120 ; exploitation, 152, 15 7 Wealth, 29, 60, 216 Weaning, 119-2 1 Weapon: horn s and antlers , 89; parasite and host, 185; psychological, in weaning, 119-2 0 West, M.J., 165-6 West-Eberhard, M.J., 240 n 5, 241 n 26 Wheatear, black-eared, Oenanthe hispanica, 173 Wheatear, desert , Oenanthe deserti, 94, 93 fig Wheatear, isabellina , Oenanthe isabellina, 94 Wheatear, mourning, Oenanthe lugetts, 93 fig Wheatear, white-crowned black, Oenanthe leucopyga, 172 Wheatears, Oenanthe: marks, 44; mobbing raptors, 5; stone collecting, 172 ; tails, 94 fig; warning calls, 4 White color , 54, 94-7, 201-2 Widowbird, Euplectes pronge, 82, 83 Wilhelm, K. v., 28 Willen, J., 227 Williams, G.C., 26 Wilson, D . S. , 232 n 18 , 241 n 33 Wilson, E.O, 165, 241 n 26, 241 n 30 Wing, 49-50, 88, 97 Wingload, 88 Winter gatherings , rooks, 200 Wintering pair, 113, 197, 219 Witches, 213 Witnesses, an d threats, 142 Wolf, Cants lupus: observin g stoning , 6-7; as partner (dog), 193; 64 fig, 69 fig; ritualization, 67; wolfgazelle interaction, XIII-XV Words: animals , 78; humans, 80, 222-3 Workers, socia l insects: 151-3 , 155-7 , 159-63, 240 n 10,241 n 27, 241 n 30 Worms, a s parasites, 19 2 Wright, J., 127

2 8 6 Index Wrinkles, facial, human , 213, 22 4 of Wynne-Edwards, V.C. , 203^ 1 askin

, for prestige, 171 ; coloring , 121 , 12 1 fig; cuckoo, g for food, 18 9

Yaeger, Stercorarius, 82 Zaharas Yeast, Saccharomyces cerevisiae, 26, 104- 7 Zahavi Yellow color, 96 Zebra Yellowjacket. See Polistes canadensis Zero-su Yom-Tov, Y. , 18 8 Zilberman Young: babblers, feeding of , 127 ; camouflage , 202 ; car e Zuk

, 203 , T., 127 , 13 4 , Equus, 48, 98 m game , 14 9 , R. , 171 , M. , 51