Conceptions of kinship

Conceptions of Kinship

Conceptions of Kinship Bernard Farber

Elsevier

Ne\'/ York

• •

Ne\\' York Oxford

Exclusive Distribution throughout the World by Greenwood Press,

Ct.

U.S.A.

Westport,

Elsevier North Holland, Inc.

52 Vanderbilt Avenue, f';ew York, New York 10017 Distributors ootside the United

FJsevier/ North- Holland

States

and Canada:

335 Jan van Galenstraat, P.O. Box 211 Amsterdam, The

© 1 981

Netherlands

by Elsevier North Holland, Inc.

Library of Congress Catal oging in Publication Data

Farber, Bernard.

Conceptions of kinship.

Includes index.

I. Kinship. 2. K.inship-(.;nited States. I. Title.

GN487.F37

306.8'3

80-16712

ISBN 0444-99076-3

Desk Editor

Robert Glasgow

Design Edmee Froment Art Edilors Glen Bunis & Aimee Kudlak J'Jtclw11icalslOpening pages Jose Garcia Prod,crion MaMger Joanne Jay Compositor Crestwood Press Printer Haddon Craftsmen

�-.tanufactured in the l1nited SIDles of America

Contents

XI

Preface

1

Introduction: Popular and Legal Conceptions

of Kinship

Legal Models of Kinship Kinship and Social Structure Popular Counterparts of Legal Models 2 Historical Backgrounds of Kinship Models

Parentela Orders Model: Judaic and Athenian Backgrounds Rome and the Civil Law Model Canon Law Model: Church and State Historical Contexts Compared

3 Kinship Classification and Social Context

Metaphors of Social S pace Kinship Models as Metaphors of Social Structure

Maps: A

Serendipitous Finding

Formal and Popular Concepts of Kinship

An American The

12

19

20 21

27

29

32 33

Summary

4 Cognitive Kinship

2

Strategy

Standard American Model and the Distribution

of Kinship Maps Compon ential Analysis and Standard American Kinship Standard American Kinship: A Summary

34 34

40

45

45

48

56

58

66

VI

5

iatio n s in Kinship Superficial?

V ar

or

Distances: Fundamental 68

Religious Identity and Kinship Orientation

71

Family Income and Minority-Group Status

77

Occupational Status and Kinship Orientation

83

Assimilation and the Standard American Pattern

89

Summary: Variations in Collaterality

6 Family

of Orientation

95

103

Ideological Aspects of Factionalism and Communalism

103

Kinship Orientation and Divorce

106

Kinship Orientation and Religious Endogamy

109 1 12

Kinship Orientation and Age at Marriage The Maternal Role and Fertility

115

Summary: Kinship Ideology and Family of Orientation

7 Family of Procreation

119

121

Models of Family Organization

121

Kinship Orientation and Religious Homogeneity

125

Idealization , Disillusionment, and Marital Instability

129

Kinship Orientation and Family Roles

Kinship Orientation and Fertility

Summary: Kinship Orientation and Family Organization

8 Relatives and Strangers

135

146

154

156

157

Kinship and Social Exchange Location of Residence

162

I nterest in Kinship Ties

166

173

Actual and Desired Contact

186

Summary: Ties with Kin

9 Perspectives

on Kinship: Concluding

Remarks

Summary Statement

187

1 87

The Kinship Grid

187

Kinship Models and Soci al Settings

192

Kinship Model s and Other Perspectives

196

Standard American Versus Middle-Cl ass Kinship

197

Permanent Availabi lity and Kinship Orie ntation

202

Kinship Orientation and Communal Structures

205

Components in Measures of Collaterality

207

Vll

A Brief Epilogue: Models of Collaterality

and Secular

Immortality

Appendix

A.

Multiple Regression

Analysis

Th e Multivariate Procedure

B. Method of

Data

Collection

Morris Axelrod and Ed-.,vard A. Greenberg Phoenix Probability Sample

Supplementary Sample Field Procedures

Jewish

References

Index

214 214

216

Findings

Appendix

211

222 223 229 230

231

245

Perhaps the title of this monograph should be Conceptions of Collaterality. The term collaterality refers to aspects of kin­ ship ties associated with genealogical distances among rela­ tives. George Peter Murdock (1949, p. 1 03) has identified collaterality as a basic dimension in detennining relationships among kin:

Preface

The criterion of co lla terality rests on the biological fact that among consanguin­

eal re lati ves of the

same generation and

sex,

some wi ll

be

more closely akin to

Ego than others. A d irect ancestor, for example, will be more nearly related

than his siblin g or cousin, and a lineal descendant than the descendant of

a

siblin g or cousin. Our ov;n kinship system consistendy recognizes the criterion of collaterality and� �·ith the sole excepti ons of 'cousin', never employs the

same tenn for consanguineal kinsmen related to Ego in different degrees.

This book is based on the premises that (I) any major subsystem of a society (like kinship) can be examined as a means for gaining insight into the charac­ ter of that society; ( 2) since, as far as we know, certain attributes of kinship­ like collaterality-are universal, kinship seems to provide an appropriate means for the study of modern society; (3) unlike other dimensions of kinship. collaterality lends itself to precise formulations that permit measurement of diversity in a population; and (4) inasmuch as collaterality appears to be a basis for mobilization of action by relatives and for significant rules regarding marriage and inheritance, it appears to provide an effective perspective from which to study the relationship between family and society. In particular, collaterality seems to be related to the extent that pluralism is significant in structuring of society. A pluralistic ideology is one that values the principle that the collective welfare of a society is best served by promot­ ing the special interests of its particular components defined by race, religion, economic role, status in the socioeconomic stratification system, and/or eth­ nicity. Kinship may be regarded as a vehicle for perpetuating group identities associated \\'ith these special interests-religious sectarianism, socioeco­ nomic position, and ethnic distinctiveness. Presumably some relatives are more closely identified than other kin with these special interests. Collateral­ ity serves to designate the shading among relatives of their relevance to these concerns. One would then anticipate that populations Ylith a pluralistic ideol­ ogy would develop family and kinship norms that reflect their needs. By way of contrast, in populations that foster universalism and consider the

X

perpetuation of pluralist ideologies as destructive to the common welfare, kinship is used differently. Here the ideology champions either individualism or ·1:otalitarianism. ·'That is, kinship is intended to serve the individual or the nation

as

a whole: it is no longer a vehicle of a special interest. Under these

conditions, the role of kinship (including family) is to provide personnel ex­ clusively for the society and its duty is to keep them as happy, healthy, and generally proficient.as possible. Consequently, this state of affairs demands a different configuration in collaterality. The monograph elaborates upon various implications of the above distinc­ tions and reports the findings of an empirical investigation which, to some extent, tests the hypothetical statements derived. The propositions examined in the Phoenix study are that spatial metaphors symbolized by the components of models of collaterality are isomorphic with communal versus pluralist dis­ tinction in religion, ethnic. and socioeconomic settings in the social structure and that these components are associated with particular kinship nonns and values relevant to the communal-pluralist dichotomy. Many sociologists and anthropologists have \\'ritten about kinship in ways that contribute significantly to the understanding of the relationship between collaterality and social structure. However, this concern with collaterality has generally been peripheral. Focusing on the work from the perspective of col­ laterality in this book may give a description of their positions a strange cast­ a vie\\' of their writings that they had not intended.

Of course, this description

does not touch the core of their work, but hopefully it does reveal implications of their conceptions for collaterality. The research in this monograph was undertaken \\'ith a grant from the Na­ tional Science Foundation

(SOC76-211 10). I am particularly thankful to Pro­

fessor John Atkins, University of Washington, for his extensive suggestions

and criticisms regarding my discussions of models of collaterality. His com­ ments have contributed significantly to the monograph. In addition, I profited

much from the comments on earlier drafts of chapters and introductions to

chapters by Professors G. N. Ramu and Nicholas Tavuchis of the L:niversity

of Manitoba and the late Louis Schneider of the llniversity of Texas at Aus­

tin. Their sharp criticisms were indeed expressions of friendship. J appreciate

also the revie\\' of the final chapter by Professors Joan Aldous of Notre Dame

University and Robert Lewis of Arizona State University. Mrs . Temtlin of

Temple Beth Israel Library was exceptionally kind in providing materials

pertaining to Jewish law and kinship. Finally� J thank my associates Morris Axelrod�

Edward Greenberg� and Kenneth Andersen for participating with

me in this adventure. All three of them have made this an exciting enterprise. But above alii

am

grateful to my mischpokhe and to my dear wife Rosanna

most especially for inspiration. patience, and gentle counsel. Bernard Farber

Conceptions of Kinship

Introduction: Popular and Legal Conceptions of Kinship

Studies of American kinship have been limited in the kinds of insights they have yielded. For the most part, they have been concerned with the extent to Ylhich the nuclear family has become '•structurally isolated' ' in different segments of society and various consequences of this isola­ tion. This concern has stimulated research on patterns of interaction and assistance among relatives, feelings of attach­ ment, and kinds of personal obligations. To a lesser extent, investigators have treated kinship as a cultural phenomenon, dealing with kinship nomen­ clature and the meanings of consanguinity and affinity in American society. In any case, there seems to be an implicit assumption that American kinship is some sort of vestige of an institution that used to be important for main­ taining social structure but now serves mainly as an appendage to the nuclear family or as an ephemeral connection between families. Yet, kinship organization is expressed not only through day-to-day inter­ actions among relatives but also through legal norms that govern such matters as the definition of incestuous marriage, the priorities of succession when intestacy occurs, and rights and obligations pertaining to support and guard­ ianship. Presumably, these legal formulations follow conceptions about kin­ ship ties that are currently in the population. However, family laws often conflict with one another even with regard to such elementary matters as who of a person's relatives are closer and who are more distant. The historical models that provide the basis for computing kinship distances and priorities in modem law differ from each other in significant ways in their ordering of relatives. Comparatively little is understood about these models and their assumptions about the nature of family life. The analysis r epo rted in this monograph is concerned with the extent to which people's ideas about kinship ties conform to the various models for computing kinship distances and the implications of holding these ideas for family life and kinship interaction. In exploring the implications of different modes of handling collaterality, I deal with several questions with regard to kinship as a mechanism for maintaining social continuity. Under what circumstances do people regard kinship as a corporate-like ._reality" by means of which ''the important

2

th ing s in I ife'' are perpetua ted ? What evidence is there that people in A mer­

ican society organize

kinship ties in ways that reflect this ''reality?'' How is this conce ptio n of kinship related to such matters as courts h ip patterns, marital norms, fertility, socialization of children, and so on? What might kno wledge about these matters sugge st about future trends in family life and change s in American legai cod es? This book reports the fmdings of a social survey d esig ned to answer these questions The target group consisted of residents of Phoenix, Arizona who have ever been marri ed and who are between the ages of 1 8 and 45. The 772 cases which were st udied had two components, a pro bab ility samp le of Phoe­ nix and a supplementary sample of Jewish household s The field techniques are described in Appendix B, M e thod of Data Collection. .

.

LEGAL MODELS OF KINSIDP

of kin s hip play an important role i n the analysis, they s hou ld be described first . These models are related to, but are n ot identical with, systems of tenn s by which an individual's relatives are des ignated-parents, uncles, aunts, cou s ins, and so on. A matrix of terminology describing kinship statuses in a particular culture can be regar ded as a t ab le of organization. As such , the table expresses the rights and obligations which, at one time or another in a society ·s history, people have developed to or­ ganize the ir relationships. The biological similari ty of rep roduction for all humans makes it po ssi ble to compare one table or organiz ati on with another. Lewis Morgan, in his analysis of Systems of Consanguinity and Affinity of the Human Family, has su ggested tha t : Inasmuch as legal models

­

In every supposable plan of consanguinity, where marriage between pairs ex­

ists, there must be a lineal and several collateral lines. Each person, also, in

constructing his own table becomes the cen tral point,

or

EGO, from whom

outward is reckoned the degree of relationship of each kinsman, and to whom

the relationship returns. His position is necessarily in the lineal line. In a chart

of relationships this line is vertical. Upon it may be inscribed, above and below

any given person, his several ancestors and descendants in a direct series from father

to son,

and these persons together will constitute his right lineal male

line, which is also called the trunk,

or

common stock of descent. Out of this

trunk line emerge the several collateral lines, males and female, which are

numbered outwardly. . .. The first collateral line, male, consists of my brother and his descend�ts, and the frrst, female, of my sister and her descendants.

The second collateral line, male, on the father's side, consists of my father's

brother and his descendants, and the second, female, of my father's sister and

her descendants;

the

second collateral line, male, on the mother·s side, is

composed of my mother's brother and his descendan t� , and the second, female,

of my mother's si ster, and her descendants. The third collateral line, male9 on

3

father's side , consists of my grandfather 's brother and his descendants. and female, of my father 's sister and her descendants; on the mother9s side, the same line, male, is composed of my grandmother's brother and his descendants, and the same, female, of my grandmother's s i ster and her des­ cendants (Morgan, 1871, pp. 17-18). the

the third,

Morgan continues in this manner through the fifth co l lateral line consisting of the siblings of the great-great-grandparent and their descendants. Analyses of kinship terminology are aimed at determining how the relatives who con­ stitute the person's kindred are grouped by kinship nomenclature to designate the pattern of relationships that one has with relatives. George Peter Murdock has proposed that this grouping can be made on the basis of sex of relatives, generation in relation to Ego, relative age of kin, consanguinity versus affinity, degree of collaterality, and so on (Murdock, 1949 , pp. 136ft). To­ gether, these patterns or groupings constitute tables of kinship organization. Whereas a table of organization based on kinship nomenclature implies a series of rights and obligations pertaining to each relationship, it does not provide a clue to the relative strengths of these obligations to various kin. The table of organization of kinship nomenclature is much like a set of statutes that has evolved over a period of time. There are times when laws conflict and decisions must be made with regard to priorities. Because of these conflicts, nomenclature must be supplemented by other means for des­ ignating priorities. The history of legal codes governing intestacy laws in­ dicates that these laws were fonnulated following periods of conflict (usually between social classes). In complex societies, such as urban civilizations, city states (and dynastic empires, e . g., Feng, 1937), diverse segments of the population differ in ranking which relatives precede others in rights and ob­ ligations. Presumably the legal models express popular conceptions about priorities orHproximities" of relatives. The sections below describe the char­ acteristics of the legal models for determining priorities among relatives. The social contexts in which these models emerged wil l be discussed later. The legal models for determining priorities among consanguineal relatives share some characteristics. Indeed, study of the historical contexts indica tes that, i n temporal sequence, the Parentela Orders procedure seems to have arisen first; the Civil Law model represents a modification of the Parentela Orders; and the Canon Law method revises the Civil Law approach. Because of this relatedness in history, despite differences in organizing principles, the three models express similarities in ways of counting the number of parent­ child links from one relative to another. Like the Genetic model, which is also described below, all three models assume that rights and duties of ''closer'' relatives have a priority over those of more distant ones. They differ only in the way by which they count the parent-child links between relatives. (See Atkins, 1974.)

4

Parentela Orders Model

The Parentela Orders model for computing priorities in the rights and obli­ gations of relatives has had a long history. According to Lewis H. Morgan (187 1 , p. 34), the tenn parentela was applied by the fifth century Helvetians in their domestic laws to refer to "a number of relatives united under the same set of parents as their next common stock (Stamm).'' The human body was used by the Helvetians as a metaphor for a parentela, so that "husband and wife, united in marriage, b elong to the head; the children, born as full brothers and sisters from one man and one wife, to the neck ...[then] chil­ dren of full brothers and sisters occupy that place where the shoulders and arms join. These form the frrst kindred of consanguinity , viz., the children of brother and sister.The others occupy the elbow, the third the hand, etc." As a gen eral model, Parentela Orders organize kinship relations in the following way: I

Each parentela is headed by an ancestor of Ego, the rank of any parentela being a function of the number of generations that this ances­ tor is removed from Ego. The first parentela is headed by Ego and incl ude s all of his descendants; the second parentela is headed by Ego's parents and includes all of their descendants apart from those in the first parentela, and so on . 2. Within each parentela, any relative is located according to the number of generations he is removed from the head of the parentela. By de­ termining the parentela cla ss into which a relative falls and his gen­ eration with in that parentela, one can describe his precise relationship to Ego. .

The principles for determining the order of priorities of relatives in the Parentela Orders model are as follows: I. All

membe rs of a lower-ranked parentela have priority over members of a parentela of a higher rank. All of Ego's descendants (i.e., mem­ bers of the flfSt parentela) have prior rights and obligations over all other relatives. The second class in priority consists of the parents and all of their descendants who are not also Ego's descendants, and so on. In theory , the number of parentelae in an individual's genealogy is infinite. 2. Within each parentela, the head (i.e., a direct -line ancestor of Ego) has priority over all other members and, in tum, each parent has priority over his (or her) children. For example, Ego's children have pri ority over grandchildren and the grandchildre-n over great-grandchildren. In the second parentela, Ego's p aren ts have priority in rights and obli-

5 i

gations over the brothers and sis ters , and the brothers and sis ter s over

the nephew s and nieces. In theory, the progres sio n within the paren t el a continues ad infinitum.

The location of any relative in the Parentela Orders system can be repre­

sented s y mboli cally . Let j be the number of g e neratio nal links between Ego

and any ancestor; from Eg o to Eg o 's parent is one link, to the gran dpare nt is two links, and so on . Sim ilarly , let i by the numbe r of generati onal links between the head of a parent el a and h is (or her) own descendants; for the third p arentela, from the grandparent t o the aunt or uncle of Ego is one link and to Ego's first-cousins, two links. Since the progress ion within pare ntel ae is infinite, the rank ordering of re latives depen ds upon the e ntire set of relatives actually c onsidered in a particular instance. Before one goes on to the ne x t paren te la, one must exhaust the generational depth (represented by i) of kin in the g i ven parentel a. Hence, for a particular set of re lati ve s , the rank order of each relative is influenced by th e deepest generational depth (i.e., the large st i) for that set. (The size of the large st i can be repre se nted by m.) The. r ank order for any relative is then determined by the summation of j (as weig h t ed by (m + 1)) plus i. The formula for the rank ordering of re lati ves in the Parente l a Orders system is thus D = i + (m + 1)j. The ranking of ge ne alogical distances from Ego is prese nte d in Table 1-1. Huebner (1968) suggests that the Parentela Orders were derived by the Germans from traditional Jewish law of intestate succes sion . Morg an (1871), ho w ever, believes that Helvetians borrowed the scheme from the Rom an system of ''numbering c oll at er al or •transverse' lines of co n sangu ines in which the first collateral line consists of the de scen dants of parents (oth er than Ego hi ms el f and his own descendants), while the second line co n sists of the des ce nd an ts of grandparents of Ego (e x cludi ng Eg o ' s parents and their descendants), and so forth'' ( Atkin s, 1974, p. 10). Morgan (1871, p. 35) remarks that uthe German [Parentela Ord ers ] is a very perfect system, but its excellence is due to its fidelity to its Roman model. '' Yet, in the 17th century, John S elden ( 1636; 1640) presented several well-documented analTable 1-1. Ranking of Oktances from Ego Acoording to Parentela Orders Model ·

Parentela orders

(ascending generalions)

Generation within

Parentela (in descending order) 1 2 3

a

Ego (O)a Children •: 1)

Grandchildren (2.,

Ranking of distance from E10.

Parenls (3)

Siblings (4)

0

Nieces and nephe ws (5)

m

Grandparents (6) Aunts and uncles (7) First cou s ins (8)

6

yses to support his view that the Roman system had its roots in earlier Hebraic law-both written and oral. Regard le ss of origins, the Parentela Orders s ystem survives in modem German law governing intestacy (Model, 1964) as well as in contemporary Israeli and U.S. law. In the Ordnung Parentelen of West German law s of succession, those relatives who belong to the seco n d, third, and lower orders (Ordnungen) are called to inherit only if at the time of succession no rel ative of an earlier order is c apable of being an heir (Model, 1964, p. 1 1). Heirs of the frrst order include Ego's children, grandchildren, and other descen­ dants. Heirs of the second order are the deced en t's parents and their des cen ­ dants who are not included in the first parentela. The progression goes on e ndlessl y. Similarly, in Isra�l: The deceased's children or their descendants inherit before his parents, his parents before the parents of his parents .... The Grandparents will not inherit if there are col l ateral relatives who are descendants o f the parents; that is to say, not only the bro thers and sisters of the dece ased , and their children, will oust the grandparents from their right to succeed, but also the grandchildren of the brothers and s isters of the deceased (Tedeschi, 1966, pp. 243-244).

Some states, s uch as Arizona, appl y the principles of the Parentela Orders model in a modified form (Wypyski, 1976, p. 65). Arizona intestacy law (1975, 14-2 103) provides for the follow ing pri ori ties among kin in the as­ signment of estates ( exclu ding the spouse's share): 1. To the issue of the decedent. 2. If there is no surv iving issue, to the dece dent 's parent or parents equally. 3. If there is no surv iving issue or parent, to the issue of the decedent's parents or issu e of either of the parents. 4. If there is no surviving is sue , parent or issue of a parent, but the decedent is survived by on e or more grandparents or is su e of grand­ parents , half of the estate pas ses to the paternal grandparent or grand­ parents equally, or to their issue if both grandparen ts are deceased; the other half passes to maternal grandparents o r their issue in the same m anner . If there is no surviving grandparent or issue on either the paternal or maternal side, the entire estate passes to relatives on the other side in the same manner as the half. Civil Law Model

The Civil Law model for assigning priorities in intestate succession i s asso­ ciated with that body of law developed in ancien t Rome and later codified

7

in about the middle of the sixth century as the Institutes of Justinian. The earliest record of this model appears in the Twelve Tables, which had been devised at the time of the founding of the Roman Republic in the fifth century B.C. The Civil Lav.t became highl y influential as empires spread throughout Western Europe, and rulers so ugh t a ready-made set of laws to govern with. '"The ideological opulence of this Roman Corpus proved irresistible,

because it embodied jurisprudential principles which with some adaptation could be utilized for the service of the Western Roman emperors in their function as universal lords (the domini mWidi)" (LlJlmann, 1975, p. 85). As

in the days of the ancient Roman empire, the claim v.'as made:

That Roman .law and the governmental themes based on it should become the ucommon la\\'" (the ius

commune) of the Western world, its common legal orde r, so that the old thesis that the Roman lex was omnium generalis could become a legal reality (L llmann, 19759 pp. 89-90). ..

The procedure for comput ing genealogical distances in the Civil Law is sim­

ple. In order to compute the distance between Ego and an ancestor or direct­ descendant, one counts the generational links between them. In the ascending

line, a parent is one degree of genealogical distance, a grandparent tv.'o degrees, and so on; in the descending line, a child is one degree of distance,

a grandchild two degrees. For a collateral relative, (a) one counts the number of generat io nal links from Ego to the nearest ancestor \vho is also an ancestor of Ego's collateral relative; then (b) one determines the number of links between the collat eral relative and that ancestor; and fi nally (c) one takes the sum of these (a and b) as the degree of genealogical distance betv.·een Ego and the collateral relative. Symbolically , if j is the number of generational links between Ego and the nearest common ancestor and i is the number of links between the collateral relative and that ancestor, the Civil Law distance

between Ego and the collateral re latives is (i + j) degrees. (Designations i and j are taken from Atkins, 1974). For example, for the genealogical dis­ tance between Ego and an aunt one starts with finding the nearest common

ancestor, who is Ego's grandfather. The number of l inks j between Ego and the grandfather is 2, and the number of links i between the aunt and the grandfather is 1 The degree of genealogical distance bet\�·een Ego and the aunt is thus (i + j) or 3. Table 1-2a presents the configuration of close relatives according to Civil Law degrees. Unlike the arrangement of relatives in the Parentela Orders scheme, the mapping by Civil-Law degree permits sev eral kinds of re latives to be equidistant from Ego. For example, although a niece precedes an aunt in the Parentela Order computation, they are of equal degree in the Civil Law procedure. .

The Civil Law procedure has been utilized in one form or another in most major legal systems in Western civilization. Max Rhe instein ( 1955, p. 11)

Table 1-2.

Degree of Relationship Civil Code Models, by

A. CIVIL CODE MODEL

Generation 2

0

According to Canon Law, Genetic, and Generation of Relative

Distan� from EGO: Degree of relationship

1

2

Grandparents

Aunts and

Parents

0

-1

3

EGO

-2

Children

Siblings

Grandchildren

Great

4

aunts

great

and

uncles

uncles

Nieces and nephews

First cousins

Grandnieces and

grandnephews

( continwdJ

indicates that, HAlthough the [Roman] Civil Law was taught in the conti­ nental universities from the 13th to the 19th century as a common basis, . . . as such Lit] is no longer in force anywhere except, in the form of the so­ called Roman-Dutch law, in the Union of South Africa and in Ceylon." Yet, the Civil Law model is widely applied in the computation of genealogical distances worldwide in legislation dealing with intestate succession and often in laws pertaining to the prohibition of incestuous marriage. During the 19th century, the development of the Napoleonic Code, based explicitly on Roman law, and the adoption of that Code in various legal systems (e.g., Austria, Mexico, Louisiana) did much to institutionalize the Civil Law model in mod­ ern family law. In the United States, the Civil Law procedure is followed in most state laws governing intestacy and incestuous marriage. Canon Law Model The Canon Law computational model of genealogical distances was first

introduced into the legal corpus of the Roman Catholic Church; later it was incorporated into English family law. In general, ''The diverse sources of the Canon Law were collected from the 13th to the 15th centuries in a compilation which is referred to as the Corpus Juris Canonici. In 1917 it was superceded as the official source of the law of the Roman Catholic Church by the Codex Juris Canonici. . . . At the Refonnation the ecclesiastical courts and the Canon Law were continued by the Church of England (Rheinstein, 1955, p. 14)." Since, in England, ecclesiastical courts had jurisdiction over family matters, the Canon Law model was applied particularly in the"elaborate system of marriage law. ''

9

Like the Civil La w model for determining genealogical di stances , the Canon Law procedure also involves the determination of generational l inks. For direct descendants and ancestors, the two models are equi v alent in as sign ing priorities ; one merely counts the number of generatio nal links (i or j) bena,reen Eg o and the lineal relati ve to detennine degree of d istanc e They differ, however, in determining degree of distance of co llatera l relatives from Ego. As in the Civil Law degree, one counts the links j from Ego to the nearest relative which both Ego and t he collateral have in common and t hen the l i nka ge i from that ancestor to the collateral relative. But, instead of adding i to j, a person applyin g the Canon Law procedure determines the degre.e of di stance between Ego and the co l late ral relative to be either i or j, whichever is larger. This sch eme of computation yields the c onfig ur at i on of distances in Table 1 2 b. Ego's aunt, w ho was 3 degrees distant according to Civil Law computation, is only 2 degrees distant by Can on Law rules; in Canon-Lav.· de grees , she now shares her locus not on l y with Ego's nieces and nephews, but also with Ego's grandparents, frrst-cousins, and grand­ ­

.

-

childre.n.

Rheinstein (1955, p. 53) re gards c urrent applications of the Canon Law mode l to the distribution of es t ates (rather than to incestuous marriages) to be a '6Consequence of hist orica l misunderstandings.'' He re fers specifically to Ge org ia lavl (Code 1933, 113-903) (8)) and to an 1878 Georgia decisi on which ruled that:

Under the rules of the Canon

Law.

.

. �·e must count from the intestate up

to

the common ancestor one degree for each generation, thence down the collateral

line to

the contestant; the number of degrees

in the longer of these ty,·o lines

is the degree of kindred between the intestate and the contestant. And by this

rule the grandchildren of an aunt are in the third degree, and are heirs at law

Table 1-2

(continued)

B. CANON LAW MODEL

Genention 2 I

0 -I -2

n

EGO

Distance from Ego: Degree of relationship 2

Grandpare nts Parents

Siblings

Children

Aunts and uncles

First cousins

Nieces and nephews Grandchildren

Great

3

aunts

and

great uncles

Second cousins

Grandnieces and

grandnephews

10

in preference to the great-grandchildren of a brother, who (Rheinstein, 1955, p. 53).

are

in the

fourth

degree

In this instance, according to the Canon Law degree, the aunt's grandchildren inh eri t ahead of the brother's great-g ran dch ildre n ; according to the Civil Law scheme, both are 5 degrees distant from the d ec eased person, and the two

would have sh are d equally; and according to the Parentela Orders procedure,

the brother ' s great-grandchildren would have inheri ted the estate. Had the Georgia court in te rpreted the English Statute of Distribution of 167 1 to refer

to the Civil Law model, as English courts have done, the choice of heirs in Georgia intestac y cases would have been different. Genetk Model

A fourth approach to kin sh ip distances is described by David M. Schne ider in his analysis of A mer ic an culture. Schneider (1968, p. 23-25) suggests that:

the American cultural conception, kinship is defined as biogenetic. This defmition says that kinship is whatever the biogenetic relationship is. If science discovers new facts about biogenetic relationships, then that is what kinship is and was all along, although it may not have been known at the time. . . . Two blood relatives are "related" by the fact that they share in some degree the stuff of a particular heredity. . . . Because blood is a ''thing , , .and because it is subdivided with each reproductive step away from a given ancestor, the precise degree to which two persons share a common heredity can be calculated, and �'distance·· can thus be stated in specific quantitative terms. In

If kinship is reduced to genetic ties, then Schneider's position i mplies that American k inship maps can be drawn in terms of shared chromosomes or

�'degree of parentage'' (Cruz-Coke, 1977). Because of bisexual reproducti on among humans, people derive half of their chromosomes from the mother and half from the father; consequently. the degree of shared chromosomes between two relatives is always some factor of one-half. The degree of par­ entage decreases expontentially with the number of generations betw een two relatives.

Symbolically, the Genetic model can be represented

as

follows:

l. Letj be the number of g enerat ions linking Ego to an ance stor (as in the Civil Law and Canon Law models). 2. Let i be the number of gen erations l inking Ego to a descendant or linking a c ommon ancestor to a co llateral relative of Ego. 3. For lineal relatives of Ego (i.e., ancestors and descendants). the degree

1 1 of parentage p is (lh)H,j•. For parents, j = 1 and p = 'h; for grand­ children, i = 2 and p = 1f.t , and so on. 4. For collateral relatives, the degree of parentage p is the product ( Y2 )J (Ih)t-•. For an aunt, p = (¥.!)2 (¥2)0 = 1"; for a first-cousin, p = (Y.zf (Ih)1 = 1/8, and so forth. (Where half-siblings are involved, p = (Y.z)i

(lh)i .)

The configuration of genealogical distances by degree of shared chromo­ somes is presented in Table l-2c. A comparison between the Genetic model and the Canon Law degree indicates that the two procedures produce similar

results for most close relatives and for children of Ego's direct ancestors. For other collateral relatives, however, the genetic proximity seems to re­ semble the Civil Law degree in assigning priorities. For example, the brother's great-grandchildren and the aunt's grandchildren (in the Georgia case) both are the same degree of parentage from Ego, p = 1/ 16. The parentage degree thus seems to combine elements of both the Canon Law and Civil Law models. Cruz-Coke ( 1977, p. 99) suggests that the Genetic model be applied in laws pertaining to uprohibited marriages, inheritance of property, and even royal succession.·' Unlike the other models of genealogical priorities, the Genetic model is capable of assigning precise locations on a genealogical map to identical twins and to half-siblings and their descendants. For ex­ ample, where Ego is an identical twin, Ego's niece is genetically equivalent to Ego's own child, p = Yl; where Ego has a half-sibling, the daughter of Ego's half-sibling is genetically equivalent to a granddaughter of Ego's full sibling, p = 1/8 . Because of growing divorce and remarriage rates in contemporary society, it may be useful to make a distinction between half­ and full-sibling ties in future legislation regarding incestuous marriage and intestacy. Table 1-2.

(conlinued)

C. GENETIC MODEL

Distance from Ego: Fraction of chromosomes shared 1/2

Generation

2 1

0 -I -2

1/4 Grandparents

EGO

Parents Siblings Children

Aunts and uncles Nieces and

nephews Grandchildren

1/8

Oreal aunts and great uncles

First cousins Grandnieces and grandnephews

12 Comparison of Models The fo ur procedures for determin ing geneal ogical priorities involve variations

in the application of generational l inks to relatives (i and j ) . The Parentela Orders model differs from the others , ho\\·ever, in dete rmini ng the relati ve priorities of ascendants as opposed to descendants . For the Canon La\\' , Civil

Law , and Ge netic models , i and j are symmetrical , and consequently , their

size relative to each other makes no difference in computing the degree of

di stance . A large i and smal l j are equi v alent in degree of di stance to a small

i

and a large j . For Pare ntela Orders , hov,rever, i and j are not sym metrical .

The

j

is w e i ght ed by (m + l) , but the i is not . As a result , in the Parentela

Orders procedure all descendants have priori ty over any ancestors .

The models di ffer pri mari l y in their treatment of collateral relatives of Ego . For collateral re latives , they each deal with the combinations of i and j in a

di fferent way :

Pare ntela O rders : Priority = i + (m + I )j, \\'here m refers to the largest i in the set of relatives considered .

Civil Law : degree of d i stance = (i +

j).

C anon LaYi· : degre e of di stance = i or j , whichever i s larger.

Genetic model : degree of parentage = ( Yl )i - 1 ( 1h )i . (For half-sib lings . p =

( 1fl ) i ( lh )j ) .

KINSHIP A."''D SOCIAL STRUCTURE Laws governing classifications of relatives in marriage and inheritance have persi sted in their general outlines from earlier hi storical eras to contemporary urban society de spite gre at demographic and technological upheav al s .

These

clas sific ation scheme s permit both (a) the e stab l i shment of boundarie s for

determi ning when marri ages are considered to be ince stu ous and (b) the formulation of priorities among relati ves in cases of intestacy . Es senti a l l y ,

the same principles that were appl ied in ancient Israel , Greece , and Rome

and in medieval ecc lesiastical law still endure in the legal codes governing

incestuous marriage and i ntestacy in the contemporary world . Even ne\\l·ly formed nati ons follow the rules for cl assifying relatives drawn from (a) the Parentela Orders mode l derived from ancient Israel and classical Greece , (b) the C i vil Lavv mode l from Roman civil l aw , or (c) the Canon La\\· model developed in the medieval Church . The surv i val of the Canon Law , C i vi l Code , and Parentel a Orders models in modem l aw suggests that they continue to be relev ant to the organ i zation of family and kinship re l ations in the contemporary world . This relevance can perh aps be c l arified by reference. to comparisons \\'ith nonindu stri al so­

cieties . The analyses of nonindu stri al soc ietie s by Eise nstadt ( 1 977), Swan son

13 ( 1 967; 1 969 ) , a nd Paige ( 1 974) seem to converge in their conclusions re­ ardin g the relationship between the mod e of po l itical integration in a society g and the character of family and kinship institutions. Im pl ici tly , Eisenstadt, Swanson , and Paige seem to assume that soc ial order is generated , in pan , by the need to regulate the distribution of scarce things. There e merge s, then , an opposition between a just distribution of goods and property for use by the c urrent generation , that acts to equalize life chances for people throughout the society , and a competitive d i s tri bu t ion of goods and property , th at tends to maximize life chances for some peop le and their desc e ndan ts more than for others . The motive tow ard just distri­ bution serve s to create an extensive network of families, w e avin g the society into a cohe s i v e whole . The motive toY�·ard competitive distribution serves to establ i sh a hierarchical structu re i n vo l vin g c on geri es of re l ated families , di­ viding the society into d i ffe rie nti ated strata . As a

re s u l t of this o pposi t i on, two conflicting kinds of kinship structure may develop--one kind supporting the spec i al interest s of the con g eri es of re l ated fam ilies and the o ther kind promoting the common i nterests that al l famil i es in the soc i ety share . The differe nt kinds of family and ki nship struc ­ tures express in g special versus common intere sts in the soc iety have been conceptualized in various ways by different writers . General ly , writers tend to characterize those fam i l y structures that ex­ press speci al interests of a particul ar group in tenns of some sort of cor­ porate rea l i ty and those fami l y structures which e xpress communal welfare in tenn s of personal inc lin at ions and obligations . In each instance , in de­ scribing famil y life that i s organized to promote a special interest, the writer endows th e ki nsh ip group with po l i tic al, economic and/or religious si g n i fi ­ cance . For example, Jeffrey Paige ( 1974) shows that in s tate l e ss societies , facti onal pol itic al structure i s associ ated w ith patri loc al (or viri local) marital re sidence , Y�·hich permits related men to act as a co he si ve unit in civil affairs . He oppo ses thi s form of kinship organ i zat i on to that fou nd in communal statel ess s ocieties , whereby related men ( e spec i ally brothers) are d ispersed , and their ability to mobilize kinship-ba sed groups is l i m ited. Similarly , Guy E. S w an s on ( 1 969) has shown that a d isti nc tio n between associational and soc i a l - syste m e mph as i s in pol itical regime is related to kinship structure and to re l igi o u s bel ief . In one analy si s , he ha s indicated that fac tio na l societies that empha si ze associational base s for c ol l ecti ve action tend to be patrilineal , wh ereas tho se societies that sttess the common welfare of all members, iden­ ti fying the political re gime as serving the whole social system as a s i ng l e entit y, tend to be matrilineal in kin ship organ ization . (In a second analysis , he h as shovJ n a rel ati onship between factional re gi me s and a propensity for states at the ti me of the R e formati o n to become Protestant as o ppose d to c om m u nal state s , where the i mmanence of God is widely b el i eved , to re m ai n Catholic . ) A t h i rd w riter , Shmuel Eisenstadt h as dealt Ylith the distinction

14

bet\\'een patri monial and imperial regime s . In his conceptual ization , patri­ moni al regi mes are characterized by the c lustering of kinship groups in ways that impinge upon economic , pol itical , and/or rel igious institutions of the society , whereas in imperial regime s , kinship structures are relatively i nde­ pendent of other considerations . In addition , Ei senstadt suggests that while imperial regimes are structured in ways which reveal a dominant core and periphery , patrimonial regimes are general l y decentralized . All three conceptuali zations--i . e. , by Paige , Swanson , and Eisen stadt­ share the following characteristics: (a) There i s a rel ationship between the development of kinship structure and the relati ve emphasis in political struc­ ture upon spec i al interests versus general welfare . (b) Societies which h ave factional (or spec ial-interest) regimes are m arked by a strong dependence of political and economic groups upon k inship ties; societies which have com­ m unal ( or general-interest) regimes are marked by differentiated political and economic structures (i . e . , i solated structurally from kinship) . (c) In societies with factional regimes . kinsh ip structures are organized to act as the primary vehicles for the continuity of the soc i al order; in societies with communal regimes . kinship structures are organized to permit other corporate groups to act as the primary vehicle for the continuity of the social order . The se statements about the relationship between kinsh ip and social struc­ ture are drawn from research on nonindustrial societi es . Yet, in principl e, they see m appropriate t o contemporary societie s a s well . In his analysis of modern social structure , Edward Shils ( 197 5) appl ies an approach similar to that of Paige , S wanson , and Eisenstadt . Shils conceptualizes social order in tenns of center and periphery , which corresponds in some respects to the distinction between communal and factional structures . • •Thi s central ity has , however, nothing to do with geometry and little with geography (Shils , 1975 , p . 3). " Rather, for Shils ( 1975 , p. 3): The center, or the central zone, is a phenomenon of the realm of values and beliefs . It is the center of the order of symbols , of v alues and beliefs which govern the soc iety . It is the center because it is the u ltimate and irreducible; and it is felt to be such by many who cannot give explicit articulation to its irreducibility . The central zone partakes of t he nature of the sacred. Shi l s notes that , as compared with earlier eras , modem techno logical ad vance has elevated the standard of l iving and • •integrated the population into a more unified economy " and , simultaneously , there has been "more widespread participation in the central value system through education , and in the central institutional system through the franchi se and m ass commun ication ( Shil s , 197 5 , p. 14) . " ' One effect of this · ·incorporation of the mass of the population into the central i nstitutional and value systems • , is an emphasis upon a con­ cern with the common interests (and consequent equal treatment) of the mem­ bers of the society .

15

\\'ith an incorporation of the population into this centralized culture , dom­ ns that enable the society to persist as an integrated entity tend in ant i ns titutio national scope . Major decisions affecting personal destinies are to be those of made in l arge-scale bureaucracies and their supporting agencies . In a cen­ tral ized culture , t he i mportant things in life are those which override parochial in terests and local concerns . Ho we ver , Shils regards the i ntegration of institutions in modem society as ex istin g in a state of constant tension bet\\'een the tendencies for the common good and separatist tendencie s deriving from special interests . On the one hand, . . the common good cannot be real i zed in a society consisting only of pri v ate entities (Shil s , 1 975, p . 340) . ' 1 Yet , on the other hand , ' �even the most in tegrated society ever known is riddled \\·ith cleavages and antagonistic actions (Shil s , 1 975 , p . 8 5) , " and "scarcity , ambition , human contrariness , divergent tradi tions , conflicting loyalties , and the great une venness of eco­ logical integration all stand in the way of anything close ly approximating a stable , continuou s , inclusive , all-embracing cultural and authoritative inte­ gration (Shils , 1 97 5 , p . 84) . � ., Shils emphasizes that along side the central ization of culture and its insti­ tutional base , there exist collectivitie s that have been alienated in some man­ ner from the cultural core . These alienated collectiv ities seek to promote special interests that they percei ve to be undermined by the institutional complex supporting incorporation of the mass of the population into the central ized culture . The continual expansion of the core of ' 'public interest , " encroaches upon their special domains of rei ig ious , ethnic , or economic in­ terest . The threat evoked by this encroachment mobilizes these collectivities to counteract the i nfluence of the � 'core . ' ' In their reaction to central ization of culture for the � 'publ ic interest, ' " special-focus collectivities apparently re ly not only upon associational means (e . g . , religio us , welfare , and educational agenc ies) but also upon communal institutions , such as kinship and family , to sustain an independent identity . Eve n i n a bilateral society , kinship seems able to serve as an effective vehicle for the persi stence of (or the enhancement of) the ' �important things of life . ' ' It prov ide s a ready-made base for e stabl i shing generational continuity and personal lo yalty . Shils ( 197 5 , p. 1 22) suggests that kinship forms a primordial foundation for community . The attachment of interaction ,

to another member of ane 's ki nship group is not just a function

as

Professor Homans �·ould h av e it . It is because a certain

ineffable signifteance is attributed to the tie of blood . Even where affecti on [is] not great, the tangibility of the attachment to the other person , by virtue of our perception of h i s membership in the kinship group, i s clearly in evidence .

Th e attachment , in order to be effective , however, seems to me to require a ce ntr al idea around which relatives can rally in mobil izing themselves and

16

their resources . For middlem an minorities , this ce n tral idea unites kin s hip with ethnic and economic survival ; for rel igious sects and minority re l i gio ns , family becomes a vehic le for ac h ie vi ng a set of re l i g iou s ideals . ( Some groups may of course utilize formally-organi zed associations as an alternative to k insh i p ; but when they do , they also pre sum abl y d i se ntan g l e no nn s pert ai ni n g to family and socialization of childre n from the goals i mp l ied in the central idea. By re ly i ng o nl y upon fonnal associations for perpe tuating an i d eal , these grou ps must then cont i nu al l y recruit outsi ders to p arti c i pate in their programmatic effort s . ) Onl y by mobi l izi ng fam i l y and kin shi p structures in support of the central idea can factional ist, separatist , or al ienated groups c re ate an en s u red s upply of adherents ge neration after gene rat io n . The endurance of a central idea that tran scen d s i mmed iat e perso nal need requ i re s parti cu l ar kinds of n onns co n nec ti ng th i s i de a to kinship s truc ture . It is u n do u btedl y true that under most circumstances dire need will fac i l itate the creation of close ties between indiv idu al relatives . But lacking a u nifying s pec i al interest . o nc e personal needs ab a te , the se t ie s m ay eve ntu a l ly dis­ sol ve , and c h ance s are s l i m th at the succeeding g e nerat ion , unless it also faces severe p roble m s , will maintain firm ki nship bonds . Without a tran­ scendant rea son (such as unique comm u n i ty stat us � mi n or ity re l i gion , or et hni c i t y) , there is l i tt l e initial moti vation to conce n trate resource s a m ong people w ith c lo s e ge ne alo gical tie s . The pres ence of a tra ns ce ndan t interest . however, ju stifies the per si ste nc e of k i nsh ip struc tu re s bey on d a s ing le gen­ eration without regard to in d i v id ua l p ri v ati o n . Consequently , th e transcen­ dental interest not only fosters stron g personal c ommi tmen ts , but it also prov id e s a b as i s for the form at i on of cen tripetal k i n s h ip norms. By de ve lo pi n g norms w hich sustain a ce ntripetal foc us , ki n shi p groups encourage member families to maintain an identifiable stru cture over ge nerat i o ns . Concentration of k i ns hip resources in modem soc i ety over an ex te n ded time requ ires a c on stan t battle with the corporate repositories whic h c onstitute the c ore of control ove r s ociety . Indeed , the go vernme ntal , industri al , and educational bureaucracies i n modem soc iety encourage residential and social m oveme nt as a means for maximizing utilizati on of human res ource s . More an d more modem soc i ety d iscourages nonbureaucratic personal c omm itme nts (a) by empha si zi ng personal freedom in non-v.'ork affairs and ( b) by defin i ng traditional fami l y and kinship i nsti tutions as coerc i v e , as i ne ffec ti ve loci of socialization , and as interfering with self-realization i n a pos t- indu strial world (e . g . , note stress upon the • •costs of family life " in Morgan ( 1 975 )) . But de sp i te this atte mpted redefinition of family and kinship structures. there i s some ev ide nce t o sugge st that " primordial " c o l l ectivitie s , based on kin ship ties , do pers i s t as viable structures i n contemporary society . I n h i s study o f urb an soci al s truc ture , Lauman n ( 1 973 , p . 203) decid e s that : \\'bile it would be quite fool ish

scene is

a

for

us.

. . to conc lude that the

American

urban

mosaic of highly differentiated , self-contained nat ionality -rel igious

17 groups , it �·ould be equal ly fool i sh for

little

or

us

simply t o dismi ss them

as

being of

no importance in accounting for the v ital ity and heterogeneity of con­

temporary urban I ife .

As instruments for mobilizing humans and their resources for attaining special goals-Often millenial-family and kinship i nstitutions cannot be considered

as obsolete ; rather they are seen by some collectiv itie s as being highly rele­

vant to the endurance of group identity over generati ons . Indeed , in the se collectivities , kinship ties are regarded as the very vehicles by \�·hich the important things of life endure . Rather, these collectivities consider indivi­

dunlistic decisions and adaptations made to meet personal needs as resu l ti ng in the long run in a decrease in the abi lity of their fami l ie s to affect ' 'who gets what , vlhe.n , where , how , and why ' '---c ompared with others . As noted previously , the analys i s by Shils about the tension betw·een the centralization of culture in the " public interest , and the plural ism of 'pri­ mordial ' ' col lecti vities someYlhat parallels the studies of non industri al soci­ eties that disti nguish between factional and communal regimes (or bet\\·een patrimon ial and imperi ali stic regimes ) ( Paige , 1974; Swanson , 1969; Eisen­ stadt. 1 977) . The analyse s share in (a) their focus upon the relationship between the w·hol e and the part s of the society as a s ignific ant aspect of social order; (b) their concern with common interests (or public interests) versus special interests (or private interests) as motivation s for generati ng different social structures; (c) their conclusion that special intere sts tend to segmen ­ talize soc ieties and break down boundaries between instituti ons , wh i le com­ mon interests tend to sharpen differentiation of boundaries between institu ­ tions but destroy boundarie s between population segments; and (d) the ir belief that kinship structure is an i mportant element in determi ning the character of ·

the social ord er.

The computational mode l s which fonn the focus of thi s study emerged in social orde rs analogous to the regimes discussed by Eisenstadt, Swanson , and Paige . In the hi story of Western civilization , ( a) the Parentela Orders model of kinship priorities developed in a social setting that fostered an emph asis upon the perpetuation of the · 'house , ' ' along with the development of centripetal kinship norms; (b) the Civil Law model arose in a social settin g in which an effort was made to maintain a balance between ' 'family ' ' and civil soci ety ; and (c) the Canon Law model was developed in a social setting

in wh ic h the perpetuation of the • 'family ' ' was considered as subordinate to the pe rpe tuati on of maj or corporate groups , such as the Church or State . Th e stress between centralization and plurali sm suggests that co ntemporary soci al setti ngs parallel the historical situations from which the kinship models

un der inve stigation emerged . The d iversity of social settings in a highly om pl ex s ociety may stimulate different emphasis upon the role of kins hip

�

1 0 pers istence of social order. Thus , the following spec ulations seem reason­

able:

18

1 . Some famil ies exist in settings that reflect their concern for perpetuating

special interests . The se families are ' 'sectarian ' , in either a religious or sec ­ ular sense . The se groups may be characterized by a minority religion or they may have a special interest in maintai ning a particular socioeconomic position or ethnic identity . These fam i l ie s will appropriate ideas about kinship prior­ ities wh ich follow the Parentela Orders model . 2. Some families exist in settings that reflect concern with the common welfare of the society . Universalistic in outlook, these families affiliate or identify v.'ith religious groups that are less sectarian and that may emphasize the ubiquitous immanence of God in everyday life; they emphasize individ­ ualism as an expression of equal rights and/or the predominance of general corporate structures (such as government or economy) by which the general well-being of the society is perpetuated . The families would appropriate ideas about priorities that approximate the Canon LaYi or Genetic models . 3 . Some families reach a compromise between the special demands of famil y continuity and the more general \\'elfare of the civil society . These families appropriate ideas about kinship priorities which approximate the Civi l Law model . B asic to the plan of analysis in this monograph is an assumption that there are only limited modes of organizing family and kinship structures in relation to the larger social structure . These modes are expressed in legal codes per­ taining to the continuity of family and kinship ' "properties, , � primarily in those l aws governing succession and incestuous marri age . In Western soc iety , three general models have been applied to succession and incestuous mar­ riage-the Parentela Orders , the Civil Code , and the Canon Law . I regard these models as markers for popular conceptions about the extent to which family and kinship are considered to be important for transmitting the im­ portant things about life from generation to generation . Consequently , the analysis focuses upon the degree to which people in different social contexts (and with different ideas about the family and children) al so differ in the extent to wh ich they conform to the configuration of relative s implied in each model . Putatively , the kinship models express orientations to ways of perpetuating the "important things of life ' ': To the extent that people in Hsectarian ' settings regard the family as a vehicle by which their way of life will be perpetuated , they will also: (a) have a larger number of children; (b) give greater weight to family obligations than work obligations; (c) give greater importance to maintaining ties with parents and in-laws and with grandpar­ ents ; (d) regard marri age as a special status rather than as a mere formalization of a man-woman relationship; and (e) consider inducement of self-discipline and instrumental " 'adultlike ' ' orientation to the world as important aims in soci al ization . ·

19

As s i gn i fi can t determinants in organizing family life , the kin ship orienta­ by a model should emerge in specific kinds of social contexts ti o ns e xpre ssed r as ephemeral concerns which predominate at different rather than appea fami ly life cycle or merely as express i on s of interpersonal stage s of the relati on s .

pOPU LA R COU NTERPARTS O F LEGAL MODElS

This introd uctory chapter has discussed legal mode ls for assigning priorities amo ng relati ves . Although their appl icabil ity to inte stacy laws and for defin­ ing m arital prohibitions may be obvious , their i mpl ication with regard to social structure , kinship tie s , family interaction , and socialization of children may not be . The remaining chapters of this book will be concerned w ith an elaboration of these implications . Chapter 2 deals with the historical settings in "''hich these models emerged . Each historical setting seem s to have pre­ sented a different kind of problem with regard to the role of family and kinship in the persi stence of soc i al structure . Chapter 3 turns to modem soci al settings , and i t involves study o f ways in which the place of groups in the larger social structure is related to metaphors of social space . The. conception of soci al space is considered to be a function of group boundedness (e . g . , independent identity as a faction , sect, or some other soc ial c l ass) in relation to the rest of the society . Chapter 4 presents a new kinship model which emerged through an analysis of cases in the residual category of kin ship orientations . This model happens to be y,ridely prevalent among middle-c lass Protestants . The c hapter al so describes a componential analysis of the kinship model s which re veals their relationship to spatial metaphors . Chapter 5 per­ tai ns to diversity in kinship orientation among d ifferent segments of the pop­ ulati on , and it com p ares the efficacy of structural vari ables (such as rel igion , ethn icity , and socioeconomic status) w ith that of personal factors in explain­ ing v ari ation in kinship orientation . Chapter 6 deals with the significance of the family of orientation (i .e . , the parental family) in determining kinship co nceptio ns , and Chapter 7 vlith the relationship between kinship orientation and attributes regarding one 's family of procreation (i . e . , with spouse and chil dren) . Chapter 8, based on theories of social exchange , concerns the co nnectio n betw een kinship orientation and actual ties with rel atives . Finally , Chapte r 9 offers a summary and a set of conclusions dra\\'n from the analysi s . As a check o n spurious interpretations o f the findings , a multivarate analy sis Was u nde rtak en . The results of this analysis are pre.sented in Appendix A . The methods of data collection are discussed in Appendix B .

Historical Backgrounds of Kinship Models

Now I tum to the specific hi storical contexts in which the kinship models under investigation have emerged . The weight of evidence is that these ki n­ ship models were not taken over w i l ly­ nilly in the laws where they have been used . I nstead , indicati ons are that persons who formulated the se laws have found alternative models wanti ng , and neYl models were invented to el im inate deficiencies of those the n in exi stence (See Selden , 1636 , 1 640 . ) . For instance , the Rom ans dispatc h ed th ree com ­ missioners to examine the legal systems of the Greek city- states (especially Athens) when the Twelve Tables were derived at the beginning of the Re­ publ ic (Heitlan d , 1923 , I, 70) . Since the Twel ve Tables do not fol lov-· the Athenian mode of computing kinship distances. presumably the commission­ ers regarded that procedure as inappropriate . Similarl y , one can consider the in vention of the Canon Law scheme for dete rmining genealogical distances as a critique of the Roman system . Orig­ i nal l y , · 'The Christi an l aw took over the Roman sy stem of reckoning rel a­ ti on ship as explai ned by Pau lus [and later incorporated into the Justinian Code , ] that i s , each generation [either ascend ing or descending] constituted a degree (Smith , 1 940 , p . 24) . ' ' But co n s i de rab le controversy ensued among those who formulated canon la\\i . Commentators , such as Pope Alexander II . thought that the Justin ian mode of c ounti ng kinship distances made first cousins appear to be much more distantly related than they considered them to be . u Applied to the church laws of marri age , this method of calculation wou ld have the effect of con tracting by half the severity of the canonical prohibitions of marriage among rel atives (Smith , 1 940, pp . 28- 29) . " Con­ sequently , the canonists sought a procedure that would more adequately re­

flect their conception s about an appropri ate definition of extremely close kin­ ship ties . Significantly , in th e past 800 years , legal codes governing incestuous mar­ riage and intestacy have been able to utilize these earlier kinship models with ­ out significant revi sion . Even the recent emergence of the soci al ist blocs of Eastern Europe has not yet g e nerated new models . (See , for example , John­ son , 1 969 , p. 1 88) . The only contemporary development in computation pro­ cedures pertaining to consanguinity has come from the study of genetics . The

21

Genetic model was invented in the 20th century to acco mmodate knowledge bout the mechanic s of bio logic al kinship through chromosomal reprod uc­ At l east for close relatives , however, the Genetic model generally the same configuration of kin as the Cano n Law procedure . uces The pe rs i stence of ttaditional kinship models in contemporary law s uggests still expres s a mean ingfu l dimension in the life of modem that the se models . The appropriate nes s of a k i nsh ip model apparen tly does not depe nd soc iety pri m arily upon degree of industrialization or modernity; rather it seem s to rely more upon the place of kinship in the particular cultural context. By examin­ in g the historic al conditions in which each ki n sh ip model has e merged , we should be ab le to infer the kinds of functions kinship is expected to pl ay

:;on .

pro­

wherev er th at model is appli ed . The section s below sketch the historical and

re ligious backgrounds out of wh ich Parentela Orders , Civil Law , an d Law models were de v eloped .

Canon

The description of these historical contex t s is n ot meant to i mply a un i l i neal evolution from one model to another; there may be no ne ce ssary evolutionary connec tio n between the m . Rather the inte ntio n is mere ly to that certai n kinds of ki ns hip models have an affinity for p arti c ul ar modes o f social structure .

propose

PARENTELA ORDERS MODEL: JlTDAIC AND ATHENIAN BACKGROUNDS

The Parentel a Orders model i s found both in Judaism and Classical A the ns .

There are several p arallels in Judaic and Athenian social strucblre that seem

bo th societies, (a) ki n ship u nit s pro vided the basic political e ntiti e s ; (b) the continu i ty of soci al structure depe nded upon

to be re l ated to this mod el . In

the abil ity of succeeding g ene rations to c arry on famil i al and religious tra­ dition s ; (c) the concept of contract u nd erl ay the j ustifications of norms of family and kin ship ; (d) kinship endogam y was preferred ; and so on . J ack Goody su ggests that the p aral l el s between Judaic and Athenian marri age no rms derive from similar inheritance sy stem s . This tendency

[ towards kinship endogamy] is particularly marked where \\'omen are heirs , or even res i d u al heirs . to property of inte rest to males, for they may be enco uraged or ob l i ged to marry within a certain range of kin ; this vvas the case ¥lith the daughters of Zelophehad in ancient Is rael as well as in the epi ­ cle rate of classical Athens (Goody, 1 976 , p. 1 4) .

B

ec ause of thi s series o f similarities betv.'een family and kinship norms in c i ent Ju daism and classical Athens , it may be instructive to rev iew rela ­

� �tons hips

a mong social structure , the co ncept of contract , and in te stacy law m the se two societies .

22

Judak Law Historically , Jewish legal scholars have assigned the family a central ro le i n social organization. To Maimonides ( 1963 , 1 967) , for example , the family is the basic element in the maintenance of a just soci al order . In his Guide

of the Perplexed, Maimonides ( 1963, pp . 60 1 - 602) locate s the core of ethi c al conduct within the family in that: Fraternal sentiments and mutual love and mutual help can be fou nd in the ir

perfect form only among those who are rel ated by thei r ancestry . Accordingly

a

si ngle tribe that is united through a common ancestor--even if he is remote­

because of thi s , love one another , hel p one another , and have pi ty on one another; and the attainment of the se things is the gre atest purpose of the Law .

A similarity to Classical Greece is noted in Plato (Law-·s , 790b) : "When th e ri ght regulation of private households within a soc iety is ne g l ected , it is idle to expect the foundations of public Ia\\' to be secure . ' ' Th us both Maimonide s and Plato consider the family to be more fundamental than the state i n the perpetu ation of social order. In both conceptions of society the perpetuation

of "tribe " (mishpokheh) or "house " (oikos ) represent special interests that must be attended to if the social order is to persist . The state itself endures only by giving priority to these domestic interests . The pol itic al regime itse lf is not the core of soc ial structure that dominates the individual lives of the � members of the soc iety . Rather , its existence is contingent on the fundamental obl igations to the family and the "house . . , Obligations deriving from the Conv enan t are basic in Jewish religion and in family continuity . The concept of Covenant appears in Genesis XVII , where God propose s to Abraham that in exchange for an enduring , ' 'whole­ hearted ' ' commitment by the descendants of Abraham ' 'throughout their gen ­ erations , " God will protect them and gi ve to them uthe l and of [their] so­ journs " (Hertz , 1 960 , pp . 5 8- 59 ) . A second statement i s found in Exodus XIX , 5-6. Here , God tells Moses that by undenaking the mi s sio n of bringing other nations ' 'closer to God and Righteousness ' ' by being ' 'a kingdom of , priests and a holy nation , . they would be his "treasure . " The Exodus state­ ment is elaborated in commentary in a way that emphasizes the metaphor of a long-tenn economic arrangement . According to the commentary : ' 'God asked Israel , 'What sureties have you to give that you will keep My Con­ venant? ' They offered the Patriarchs , the Prophets and their righteous rulers as their guarantors . But all of them were rejected . It was on ly when they offered their children as sureties for the pennanence of the Covenant ,. that they were accepted (Hertz, 1 960 , p. 291 ) . " The symbolism in regarding chi ldren as pledged propeny to ensure the carrying out the terms of a contract has several impl ications for fami ly and k i nship. First , grace or salvation is not a functi o n of an indiv idu a1 9s own

23

bu t it i s i nstead dependent upon the conduct of descendants and , c ond uc t, n tl y , one is constrained to e mphasize intergenerational obligation s : conseque '" .& · an d h atnten an ce of the soc i alization of c hildr�n 10r ngh teous ' ' 1 1ves _ S econd , there ts no c losure to the contrac t : 1t I S ·everlasting , . . fili al piety . no completion of the terms of the contract, each gen erati on Si nc e there is passes on its obligations to the next , l ike an infinitely-repe ating decim al •

���

·

sult, the soc i al structu re-in this case , the family and nu m ber . As a re ki ndred-that s ustains these reciprocities endures without ever ending gen­ erational ob ligations . (See Levi-Strauss , 1 969). Henc e , ki nship continui t y is to be real ized even

at heavy costs-and the concerns of succession are fore-

most . The intestacy law s of trad itional Ju daism place much e mphas is upon gen eratio nal obl igation s to perpetuate the family . In his Mishneh Torah , Mai­ mon ides i ndic ate s:

The sage s o f th e Talmud e stablished the following ord er o f le gaJ

heirs : l )

son s

daugh ters and their de sce n dan l s ; 3 ) I h e father, 4) de brothers and the ir scendants ; 5) si sters and their descendanls; 6) the father 's and

their descendants ; 2)

father;

7) the

fat her ' s brothers and their d e sc e n dan t s ; 8) the father 's si slers and

their descendants ; 9) the father 's father 's father ; an d so on . To thi s l i st� imp l i ed in the bi bl ical passages, the s age s added an othe r legal h ei r , lhe husband , \\' hos e

right to the inheritance of his wife 's posse ss ions was inferred from the bi bli c a l

expres sion " 'nearest relative · · (�umbers 27 : 1 1 ) . Each son of the deceased re­ ceives an equal share of the estate , except the firstborn of the father, who

rece iv es a double share . Where there are neither sons nor sons ' children , th e

daughters and their descendants become the rightful he irs . . . . When th ere

are

no heirs in the descending l ine . t he property is tran s mitted to the nearest relati ve

in the ascendi ng l ine . The falher take s priority over the brothers of the deceased

in the absence of e ither sons or dau ghters (Maimonides ,

1967 .

pp .

292- 293) .

The prio rities in succe ssion described by Maimon i des had been establis hed

i n Ju daic law in the Mish11ah , which w a s compiled follow ing the destructio n of the Templ e . Radin ( 1 9 1 5 , p . 69) suggests that the Bcreation of houses of

praye r d emanded local organization , and with local organi zation gradations of me m ber s and the establishment of loc al mag i strate s . . . . The organizatio n o f th e G reek ci ty-state , familiar t o t h e East for many years , became a m odel for th ese corpo rately organized LJev.'ish] communities .. ' in the D iaspora . A�p­ p arently , the necessity for c reating a common social order among the se au­ tonom ou s J ewi sh commun ities had stimulated the codification of Oral LaVw' in the �\fi.�hnah .

Althoug h the order of prioritie s i n succession was similar to Greek l aw· � hav e not been able to determine the hi storical sequence . The ..Mishnah JUsti fied the ordering on the basis of B ibl ic al refe�nces and rabbinic al com­ me ntary . The priorities are di scus sed in the Fourth Di vision of the �Wi.� hnah .

�

24 c alled Ne ziki n , in the section Baba Bathra , subsect io ns 8 and 9 (Danby ,

1 933 , pp . 376- 379) . The justifications appearing in The Babylonian Talmud (Epstei n , 1 93 5 , II , pp . 463- 478 ) are founded on the assumption that , ' '.� tribe [ mu st] not be blotted out from Israel , (Judges , XXI , 1 7 ) . According ly . � 'So shall no inheritance of the children of Israel remove from tribe to trib e (Numbers , XXXVI , 8) '' and , "So shall no inheritance remove from one tribe to another tribe (Numbers , XXXVI , 9) . " To assure that the continu ity of the tribe or family , The Babylonian Talmud (Epstein, 1 93 5 , II , p . 474) propo ses that:

that none of the l ine survi ve s . enquiries the paternal side , and are carri ed on in the case of the descending line ) , unti l

Once it has been definitely established

are instituted in an ascend ing order , on

from father ( i nc l ud ing their heirs ,

the first ancestor of

the

as

tribe is reached .

The �\1ishnah then establishe s the princ i p le for the order of inheritance

an

Judai c law :

This is the general rule: The lineal

succession take precedence .

(Epstein, 1 935 , II,

A

descendants

of

any

one w ith a priority to

father takes precedence over all his

p. 478 ; cf. Danby , 1 93 3 , p. 376) .

descendants

The order of priorities described in the Mishnah has been appl ied in other

historic al settings as well . Cl assical Athenian society of the fourth century

represents such a sett i ng

.

The parallels between Jewish communities of the

Diaspora and Greek city -states s u gge st a similar role to be played by family

and kin ship in maintaining the soc i al stru cture . 1be Athe nian emphas i s on

the continuity of the household , the si gn i fican c e of the concept of contract in social relationships , and the i mportance of intergenerational obligations

expre ss these parall els .

Classical Greek Law

The Athen ian model of assignin g priorities in succession seem s to pres uppose

a soc iety consisti ng of autonomous segments , co - e xi sti ng in a common ter­

ritory , and vying for resource s and s tri v i n g to maintain their individual ident­

ities . Thi s conception i s consi stent with the partitioning of the Greek pen­

insula into city-states . Ehre nberg ( 1 946 , pp . 29- 52) s ug gests that the Greek terrain d i c tated the struc turing of pol itic al units into a matrix of city-states rather than an e mp i re with a sin gle major urban center as a core . Given a land ' ·div ided into numerous small parts and particles , • ' classical Greece was a qu iltwork of auto no m ou s c i ty s tate s converging on one another and com­ peti n g for re source s and soil . Each city-state tended to develop independent political , economic , and cu l tural attributes . Yet , they existed "in such a close -

25 · ity to one another, that no natural and political boundaries c ould ex­ ro le , while at the same time every State w as fully aw are s trife and s trugg rces and �wer of its neighbor (Ehrenberg , 1 94 , p . 42) . " De sp ite re so u

; �pro�the

�

_ o s ingle City-state bec ame the core of an emprre . Ehrenberg In­ n , ts ic nfl co ' 'E ven Athens never attained more than a temporary and tyrandi cate s th at , ny . The Athenian Confederacy was the rule of Athens over a . al h egemo ntC . it never became a true empire , because tt never became one wi de area, but tate (Ehrenberg , 1 946 , p . 45 ) . ' ' Consequently , for the ancient Greeks , the

e ived as an organism with a dominant core exten ding its �orld was not percperiphery . Rather they thought in terms of the pec i al

s in­ influence over a terests of political entities strug glin g to surviv e and prosper vis-a-vis each

other (Ehrenberg , 1 946 , p . 45 ) . The ba sic political unit within classical Athen s w as the oikos, that is , the family as a corporate unity , the "house . , . In fact, Harri son ( 1 968, p. I )

considers the oikos to be ' 'the con s titu ent elemen ts of the Athe ni an city state . ' � O ne cou ld not be regi stered as a citizen of Athens without oikos membership . Because of this pol itical use of the fami l y , it was imperative to ensure the continuity of the oikos . The political s ig n ifican ce of the oikos was complemented by it s relig ious importance . The contin uity of the oikos

depended, in part , upon the fu lfillment of religious obligations in the famil y

cult over successive generations . Indeed , becoming an heir invo l ved an ob­ ligation to participate in the sacred rites of the oikos ( H arri s on , 1 968, p . 1 30) .

In Athens , the concept of ownership (as power over the d i spos iti on of property or person s) �·as more limited than in Rome . The Atheni ans placed a greater emphasi s u pon contractual arrangements , which permitted speci fied uses of property with the stated l i mits of the contract-rather than the unlim­

ited domain by paternal power (patria potestas ) or by the persisting po\\'er of "spirits p (genius ) in Roman society . The limitations of contract see m to have perv aded all relatio ns hips Ylithin the family , includi ng m arri age (Har­

rison , 1 968 ) .

The concept of contract or negotiated reciprocity forms an important aspect of Ari stotle ' s N icomachean Ethics and consequently of his views regarding

fami l y ties . Ari stot le (Book \' . 5 , 1 1 3 2- 1 1 3 3) notes :

It is by proportionate requi t al that the city holds togethe r . Men seek to ret urn either e v il for evil . . . or good for good-and if they cannot do so there is no exchange , but it i s by exchange that the y hold together . . . . If it had not been po ssib le for reciprocity to be . . . effected , there would h ave bee n no associati on of the parties . That demand holds things together as a single unit i s shown by t he fact th at \\'hen men do not need one another , i . e . , when neithe r needs the

othe r or one doe s not need the oth e r , they do not e xchange . . . . Vle do not all ow a man to rule , but rational principle , because a man behaves thus in hi s ow n i n tere st s and becom es a tyrant [ \\'hereas the ru le of rec i p roc ity produces

stable re lationship s based on j ustice . ]

26 Ari stotle cons iders the family as a locus of intensified justice . He reg ards the (Book VIII , 1 2, 1 1 62) . ' ' Identifying family tie s as a c ategory of friendship , he indic ates that ' ' the demands of justice . . . seem to increase with the intensity of the friendship

oikos as . .earl ier and more necessary than the city

(B ook VIII , 9, 1 1 60) . " (Compare with Fortes , 1 969 , pp . 2 1 9-- 249 , on axiom of amity in kinship) . Not that Aristotle equates friendship with egalitari anism� instead , he views the hou sehold as a friendship among persons of unequal status-for the sons , the household constitution is a monarchy ; for the wife an ari stocracy; and for brothers a timocracy . But above al l justice is funda­ mental to household structure : The friendship between a king and his subjects depends on an excess of benefits

con ferred . . . . Such too is the friendship of a father,

though this exceeds

the

other in the greatne ss of the benefits c onferre d ; for he is responsible for the

e xi stence of his children , which is thought the greatest good . and for the ir nurture and u pb rin g in g . These thi ngs are ascribed to ancestors as well . Further ,

by

nature ,

descendant s.

.

a

fat h e r

tends

to

rule

ov er

his

so n s ,

ancestor

o ve r

. . These friendships imply superiority of one party over the

other , wh ich is why ancestors

are

honored (Aristotle , B ook Vlli , 1 1 , 1 1 6 1 ) .

The emphasis upon generational differentiation of obligations in the family � which Aristotle ap plies in his discussion of just fam ili a l relationships , finds expression in the intestacy laws of Athens . These l aws folio\\' the same prin ciples as those of the Mishnnh with regard to priorities . The order of succession in intestacy laws of classical Athens is as follo\\l·s : I . ' 'First in order of succession came the legitimate sons . . . . Rights of succession of descendants did not . . . run out after the third generation , but the line of heirs continued theoretically ad infinitum (Harri son , 1 968 , pp . 1 3� 1 3 1 ) . u 2. ' 'If a man died leaving behind him no sons but only a daughter or daughters . . . . the dau gh ters became heiresses (Harrison , 1 968 , p . 132) . � · 3 . ' 'If a father had a claim upon his son for maintenance in his old age , a fortiori he had a claim upon the sons 's property if the son died first . . . . A man 's heir might be hi s paternal or even his maternal uncle; and it would seem irrational that these relatives should succeed to the exclusion of his father ( Harri son , 1 968 , pp . 1 39- 1 4 1 ) . " 4 . .. 'There were two terms which were used to embrace relatives other than ascendants or descendants . There were anchisteia and syngeneia . The former of these was the narrow er and the more technical . It denoted all those who were rel ated to the deceased , whether on the father 's or the mother'' s side , down to and including sons of cousins (i . e . , first cousins once removed) , or possibly down to and including second cousins . . . . For the l aw of succes­ sion it determined the l imit of relatives of the deceased 's father who were entitled before relatives of his mother could come in ; a father 's frrst cous in

27

�a�

o c

remo \'ed ( possi bly a father 's second cousin) would take befo re an y the other hand had no ve 00 the mother 's side . T he tenn syngeneia on

re h res t ri c t i ve sue

use and would apply to relatives beyond this limit . Th u s all . " e1a , but not aI I syngene1a were anch z�·te1a chisteia were syngen an . • W i thi n the anchisteia there was a fixed order of rel at i v e s entitled to father 's side , and then , i f there was no representative of any · ucceed on the the se c las ses on hi s side , succession passed to re lati v e s on the mother ' s order. Anyone in a nearer group excluded all those more side i n the sam e in the same group shared equally ; if one in a di stantly related , w hile thos e le aving chil dre n , those ch i ldren took his group had predece �sed the de ( ) brothers of the deceased by sh are in equal portton s . . . . Thts 1s t he order: I c the same father and their des endants without limit ; sharing was per stirpes [i . e . , the share of a deceased heir goes to his children] ; (2) sisters of the deceased by the same father and t he ir descendants without limit , also sharing per stirpes; (3) patern al uncle s , thei r children , and grandchildren; (4) paternal aunts , their chi ldren and grandchildren ; there wou ld come next (4a) paternal gre at-unc le s with thei r children and grandchildren and (4b ) paternal great­ aunts with their children and grandch ildren; (5) brothers of the deceased by the same mother; (6) sisters by the same mother, both with descendants ·

·

·

�f

c.u i�

without limit; ( 7) maternal uncles; (8) maternal aunts , in both cases with their

children and grandchildren . . . . This exhaused the

ste ia . If there were

an chi

no relatives within the anchisteia the law simply uses for the next entitled

the vague phrase •the nearest on the father 's side ' (Harri s on , 1 968 , pp . 1 43146) . ' '

ROME AND THE CIVIL LA"T MODEL

The Civil Law model is rooted in the history of ancient Rome . Befo re the fou nding of the Roman Republic , the Patrician s , essenti ally "tribal " heads , pro vi ded the go verni ng elite . The Senate itsel f was a body of these Patricians . The n , as the economic base of Roman society e xpanded , the power of the Plebi an masses increased (Mommsen , 1 905 , I , p. 342- 7 ) . In about 45 0 B . C . � it the cre ation of the Republic , the Twelve Table s were produced , an d the dt st 1nctio n between Patricians and Plebians \vas reduced . The po wer of the Ple b i an s was extended , and five years after the issuance of the T wel ve Tables ,

��

marri ages betwee n Plebian and Patrician families could be validly contracte. d

��ommsen,

1 905 , I , p. 37 1 ; Heitland, 1 923 , I , p . 95 , Sec . 1 2 1 ) . The broad ­ e nm g of t he body politic w eakened those official policie s whi ch required strong centri peta l kinship ties as a basis for sustainin g the social order .

The c ult ural context of the Roman Repub l ic differed in si gnificant \\'ay s from th at of the Greeks . Although the Greek gods were regarded as person s , Ro man deities \\'ere more abstract (Radin , 1 9 1 5 ) . This d ifference became

28

e s pec i a l l y important

v.-ith the g rowth of Roman i mpe-rialism . I n their con­ que sts � the Romans h ad to face problems abo ut the dom ains of the deities i n \\'ays fa r different from those of the Greeks . Since Greek wars i n v o l v e d subjugation of autonomous s tat es, conquest did no t demand a ch an g e i n the jurisdiction of the go d s . Rom an imperial i sm , however , necess it ated an ex­ pan sio n of the pro v inc e of the Roman go ds -a un iversal ization of au tho ri t y­ and the i m po rt ati o n of the gods of the conquered peo p le s into Rome (Rad in � 1 9 1 5 , pp . 44-46) . There "'·as t h u s a clo se connection betv.reen universal i stic re l i gious dom ai ns o f the gods a nd the imperialism of Roman s oc i e t y . The uni versal ism in re- l i gi o n and political s phere extended to the Roma n farnily . PlJ l ria pot£�sras refers not mere ly to the head of a ho usehol d , rathe r it extends to ' "the c h i ld re n of the sons and m o re re mote descen dants t h rou g h males .. withou t any l i mi t other than t hat im po sed by the s pan of hum an l ife (Jol o\\'icz .. 1 967 p . 1 1 8 ) . The pateifalnilills-the o l de s t male ancestor­ h ad compl ete co ntro l ove r the pri v ate. l i ves of h i s une n1 ancipated descendants but l acked j uri sdict ion o ver the i r c i v i l (pub l ic) l i ves . �1 arried females were now go ve rned by another paterJiunilias . Throu gh i t s connec t ion w ith the e xtensiveness of power. the agnatic grou p be can1e the sig n i ficant entity i n Roman ki n ship . hTwo pe opl e are re la t ed agnati cally if they are. in the patria potes ra.l· of the sa me man � or i f t h e re i s some c ommon an ces t or i n \\'hose pov.'er they v.· o u ld both be if he v.·ere alive (Jolowicz , 1 967 , p . 1 22) . " The relevance of this agn ati c group for main­ ta i nin g collateral ties is suggested by laws govern ing i ntes t ac y in the XII Table s , c ompil ed around 450 B . C . (Jolow i c z , 1 967 , pp . 4-- 5 ) . The major di fference betv-.·een the A th e n ia n and R om an models for co m p uti n g kin sh i p di stances l ie s in the d eg ree t o "'· h ich they d i sti n g u i sh betwee-n line al and co l late ra l re l ati ves . The Athenian m odel re gard s direct- l i ne relativ es as pri­ mary l i nk s to Ego and cons iders col l ateral relatives o n l y as seco ndary ex­ t e nsio n s of l i neal kin . The R o m an model , howe.ver, gi ves equ a l \\'ei g ht to both 1 i neal and co l l atera l descent . The agna t i c group , re fe rri ng to persons rel ated to eac h other t h ro u g h males , attac h e s greater importance to col lateral relative s than does the oikos of Athe ns or the " ' hou s e ' ' of the Jews . The inte st ac y laYlS in the XJJ Tables p ro vide for the fo ll o Yling order of �

succes sion:

� '

1 . · �If a m an di e s i ntestate the ftrSt people entitled to succeed to his estate are hi s sui heredes , al l those in his patria pote!itas or manu.-. L i . e . , in hand] ( Jolowic z , 1 967 , p . 1 23 ) . � · A c cordi ng to Watson ( 1 97 1 , p . 1 7 6 ) , dAll sui � male or femal e , \\o·ho were chil dren of the deceased , and his wife in manu . to ok equal shares. ' ' Grandchi l dren , however , were c ap abl e o f inheriti ng o n l y by representation . 2 . hFailing sui heredes the s ucce ss ion goes to the ne are st agnate or ag­ nates , if there are sev e ral in the same degree (Jo lowicz, 1 967, pp . 1 231 24) . "

29

'In default of agnates the succession went to the gentiles [i . e . , members f the gens ] . . . . The rights of the gens were anterior to those of the agWe do not know whether the gens took as a corporation or es o . use of the \\,.ord gentiles suggests , they took as individual s betb er, as the p o 1 24 0 S ee also Watson , 1 97 1 , pp . 1 78-- 1 8 2) . " o l o wi c z , 1 967 . Si nce two indi vidu al s are agnatically related to eac h other to the e xtent that they derive from a common paterfamilias-who may be dead-this fact for determining t he agnatus proximus . The gener­ c an be u sed as the basis 3.

•

:a. �

0

0

at ional distance betvieen the person and the common ancestor determ i ne s the relati ve dilu tion of the ancestor 's potestas as co mpared with the powe.r of mo re im mediate ance stors , who have a greater degree of ' - n a tur al relation­

ship. , .. (See San d ars , 1 8 74 , p . 274) . The degree of " natural re l at ion shi p " implies that an ancestor 's spirit of genius is d i ss i pated at a regu lar rate for

each generation of distance from h i m . Hence , if one i nterprets k ins hip prox ­ imity in terms of the me an degree of distance. from

a

paterfamilias ,

the

common ancestor, the measurement distance i n the X I I Tables seem to flo\\' from the concept of patria pote s tas .

The com p u t ation o f kinsh i p di stances fo und i n the X I I Tables v.'as , Yiith slight modification , taken over i n the Justini an In stitute s at about 530 A . D .

The maj or changes i ntroduced by Ju stinian pertain to the exten sion of the rights and inheritance to cognati , i . e . , person s re lated through fem ale as we.ll

as m a le rel atives . By th i s time , the ge ns had decl ined as a v i abl e kinship

enti ty , and with this dec l ine , the agnatic group lessened in significance . The order of succession described in Book 3 , Title 6 , of

the Institute s , � 'Of The

Degree s of Cog natio n ' " (De Gradibus Cognation is ) are the eq uivalent of

those described in the Napoleonic Code and to

those followed in subsequent

applications of modem c ivil law .

CANOS LAW �IODEL : CHURCH AND ST..t\TE

Unlike the Parentela Orders and Civ il Law models , which emerged through

the explication of laws go verning succession to property , the Canon Law was developed to govern marital relati on ships . Some of the assump­

�del

tion s that gu ided the canonists in their refinement of the model had been stated earl ier in S t . Augustine ' s City of God ( Book , VX , Ch . 1 6) . Here Aug u sti ne described marri age as a basis for e xpandi ng the boundaries of

g roup so l idarity . Augustine ( 1 966 , V . 4 , pp . 503- 505 ) wrote :

For love was accorded its due importance [in choosing marriage partners outside

the family] so that men , for whom hannony \'�'as useful and honourable , might be bound by ties of various relationships . 1be underlying purpose [of marrying nonrelat ives] was that one man should not compri se many relationships in his

o�� self but that these connections should be severall y distributed among in­ dlv td uals and in this way serve to weld social life more securely by coveri ng

30 i n their multiplicity a multiplicity of people . . . . [ I n the beginning , Eve was compelled to

be]

both mother-i n-law and mother to her children of either sex .

But if these relationships had involved two women , one

as

mother and another

as mother-in-law, the bond of social affection would have embraced a wider

circ le . . . . And thus the soci al bond would not be restricted to a small circle ,

but \\'ould extend most v.·idely to embrace a greater number of people through the abundant ties of kinship.

August i n e explained consanguineous marriage as resu lti ng from demographic in suffic i ency , emphasizing the virtues of prohibiting such marri ages . Several centurie s later, in the 1 2th ce ntury , Gratian formulated a connection between consanguineous marriage and the nature of the community . Speci fically , he distinguished between the function of m arriage for Jews and for Christians . Gratian indicated that : God commanded the Hebrev.·s to i nterm arry [ among relatives] bec ause the

salv ation of man was realized in the pure Jew ish race .

.

. . Chri st changed the

nature of God 's people by spreading the faith beyond the bounds of the He­

bre�·s . Faith , not blood , was to be the criterion for me mbersh ip in the chosen

peop le . . . . In h i s discussion of consanguineous m arri ages , [ Gratian] argued

that such marri ages were no longer permitted because purity of b lood no longer served as the foundati on of the community ( Chodorow ,

1 972,

p.

74) .

As Grati an sugge sted , the c onc eption of community in the 1 2th century seems to have involved the idea that mechanisms more inclusive than kinship are req u i red to cement relationshi ps in society . Between the ninth and the 1 3th century , E u rope had un d ergon e profound changes . There was a slow shift in the basis for social structure from exchange of personal service to the les s personal uses of property and money . This period saw the growth of trade centers and c ities as pol iti c al entities; the di ssipation of the vassalage and manorial s ystems; the emergence of the state ' 'as an entity that was auton­ omous, independent, self-sufficient and lived on its own norms (Ullmann , 1 975 , p . 247) "; and the proliferation of universities (Maranda , 1 974 , pp. 25- 4 1) . These movements stimulated efforts toward centralization by both the state an d by the Church . Both of them strove to create una concordia ex diversitate--concord from di ve rs i ty (Ullmann , 1 975 , p . 248) . To create this unity , they had to appeal to n on - fami l ial motivations for community . Both the Church and st ate aimed to unify as large a populace as possible under their jurisdiction , and they were thereby stimulated to s tand ard i ze and codify laws governing social relations . Following Pauline doctrine , the Church v-·as de signated as an organic , concrete union of bel ievers, organized ' 'in accordance with its underly i ng purpose or aim , its •finis ' or 'telos . ' ' ' Canon law was supposed to provide a mechanism for dovetailing the various elements into a coherent corpus . In formulating canon law , ' 'The writers

31

nated by the concept of the universal Church as constituting an were domi d extensi vely (Ullm ann , nic entity " an they app lied organic analogies 3 The Canon Law model emerged in its modem fonn out PP · 442- 44 ) . of th is conte xt. Unli ke o ld Germanic law, v,rhich regarded kinship as analogous to the hu man body , can on law used the analogy of a tree . The line of descent was the trunk of the tree , and distances between tv.'O relati ves cons idered to be like were detennined by the number of generations they w·ere removed from a co mmon line. of descent. By analogy, they were the branches . This analogy had appe ared earlier in different gui ses . Isidore of Seville ' 'held that the son and daughter were to constitute the trunk , and the grandson and granddaugh­ ter were to be the first branch (Smith, 1 940 , p. 25) . " People were to be seen as related when their genealogical connection extended through the sixth degree; in other words , cousinship w as the basis for detennining kinship distances-Ego 's grandchildren (by different children) are first cousins to each other, the great-grandchildren are second cousins , and so on . As John of Orleans later ruled , ' 'Beyond the sixth degree the ties of kin ship [are] conceived as being , in a sense , di ssipated , to be gathered up again by mar­ riage, whereupon the cycle [is] repeated (Smith, 1 940 , p . 27) . ' ' Isidore 's tree analogy was i ncorporated into later versions of the canon la\11· , including the compilation by Gratian in the 1 2th century . During the 1 2th century , the concept of cousinage came into use in France (�laranda , 1 974 , p. 63) and eventually was extended to include " "rel ati ves in general . ' ' By the 1 3th century , the term genealogie had already taken on its modern anthropoliogical meaning (Maranda, 1 974 , p. 67) . The degree of cou si n ship was refined further by Bernard of Pavia in about 1 1 90 in his scheme for computing kinship distance. . · �where the l ine s of descent from a comm on ancestor were unequal , the longer line \\'as to be considered the determinative factor (Smith , 1 940 , p. 3 3 ) . ' · Apparentl y for Bernard , the reasoning w as that , gi ven the common ances­ tor as the core or the trunk in a line of descent, the person farther removed in degree from the ancestor extinguished the relationship to the ancestor and therefore to the other person . For example, let us assume the follo\\'ing situ ati on: one relative A is four degrees removed from an ancestor C (i . e . , A is a great- great-grandchild of C ) . A second relative B is eight degrees re m ove.d fro m this same ancestor C . Since a relationship is considered to be di ssipated after the sixth degree , following Isidore of Seville, and the second relative B is eight degree s distance from the common ancestor C , then the re lationship between the two ' 'relative s , · , A and 8 , should be considered as di ssipated . If both A and B had been within the range of . . kinship � ' to the co mmon ancestor, then they Vw'ould have been kinsmen to one another­ havin g fallen with the cousin ship range despite their belongi ng to different ge ne rations . As it i s , since one of them is no longer regarded as a kinsman

���.

32

of the c ommo n ancestor , the two persons , A and B , cannot be re l at ed to e ach other. In about 1 280 , John de Deo comp leted the model in canon law with his . . Tree of Consanguinity , " whereby the counting of degrees began with , Ego s chi ldren rather than grandchildren: c hildren were one degree di stant , grandc hi ldren two degree s , and so on . By th i s time , p rohib ition of consan­ gu i neou s marri age had been restricted by the Fourth Lateran Council to re­ l ati onshi p s w i th i n the fo urth degree of di stance (S m i th , 1 940, pp . 3 3- 34) . IDSTORICAL CONTEXTS COMPARED

The his tori c al s e tt i ng s in v,rhich t he Civil Law , Parentela Orders , and Canon Lav..- ki nship m ode l s emerged differ co n s i de rab l y from one another. Despite

superficial sim ilarities , the societies in which they d eve l oped d i ve rged in re l i gi o u s c onc e ptions , political structure , and ki nship organization . The Ro­

man Rep u blic , as a c om prom i se between Patricians and Plebians , emphasized d i stin ct ion between private (familial) and publ i c (civ i l ) domains . (See Hueb­ ner , 1 968 , V . 4, p. 719) . In contrast to Roman socie ty , both in an c ie nt Judaic and c lassical A theni an societies , fam il y and kinsh ip units themselves c onsti ­ tuted the basic entities in t he state; there was con seq uently no sharp se p arati on among Hebrews and Athenians between family , gove rnme nt , and re l i gi on : they were all intertwined . Rather, patrimonial segments had to develop s trong reciprocities intern a lly in order to maintain thei r ide ntity as w el l as to ne­ goti ate agreements with other segments for continued coexistence . As for the l aw of the Church , canon law itself emerged in the late Medieval period as a respon se to the growth and centralization of the state . The social con ditio n , which favore d the dev e lopmen t of l arge corporate e nti ties-w ith e ac h at­ tempting to carve out an au t onomo u s domain of l aw s and authority-acted to subdue the power of the fam i ly to affec t the de stiny of its members . Instead , family an d kindred were pawns utili zed by Church and state to central i ze and ho moge nize their domains . The l arg e corporate entities became the basic mechanisms for continuity of social structure . Church writings o ften appl ied org anic analogies i n describing soci al re alit y . Metaphorically , in ho­ mo geni z i n g the societ y , ki nship became a tree with spreading branches , its ' 'fruits ' • s u pply in g the person n el for the l arger and more stable corporate entities . In A t he nian and Judaic social structure s , k i ns h ip prov ided the mech anism

by which patri monial segments were pe rpetu ated , and the ki n sh i p models which were app l i ed in c alc u lati ng ki nsh i p d i stan c e emphas i ze d the di stinc­ t i o ns bet wee n a scendi ng and descending gen e rat io n s . In Rome , the fam i l y fonn ed an i n te gral , but s ubord in ate , part in the persistence of civil s oc i e ty , a nd the model of computing kinship di stances stre ssed the h ierarchical dis­ tances from a common ancestor . In canon law� the fam ily was socially un-

33

.

nt

as c ompared with the Church and state in providing for the co n­ insti tu tional structures , and k inship ties were seen onl y as a 110 off fro m a line of descent. As a m e taph or for social struc tu re , bran c · was · ; at . s e s sentially con tractu a1 &10r the anc aent J ev..-s a nd A thentans IOnship y,ra of hierarchal power rel atio nsh i p s for the Romans; and , for the e pression branches of n i sts . it was an organism , a growing tree th at spread the The differen t models thus imply a range of roles for the family to stence of society-from social con ti nu i ty through fam ily and lay in the persi the irrelevance of corporate family and kinship bonds nsh ip ex ten din g to r the pers i sten c e of social structure .

���aof � ng

·

·

�; :faith . � fo

SUI\1MARY

The dep i c t ion of

the historical roots of the kinshi p computational model s associated with major religious and civil legal systems suggests that these models vary with the social structures they symbol ize . People whose vie\\·s of kinship appro ximat e the Parentela Orders model have tended to see the continuity of social structure as rooted primarily in corporate aspects of family and kinship . They apparently visualize the glue of society as consi sting of a network of co trac t s (or al l iances ) . The obl igat i ons ge nerated in these t co ntrac s serve to weld the special interests of the various segments of the society into a stable network of social relations . In history , Jewi sh l aw seems to express this conceptio n of kinship . People who regard and kinship in ways con sistent with the Civ il Code model of genealogical distance tend to see society as per s isti ng through two separate spheres-the public and the p riv ate . The core of the public sphere is the civil society ; the c ore of the private sphere is fam y and kinship . The persistence of cohesiveness in society rests upon accommodation be­ tween these tYiO spheres of influence and pov.·er. Historically , this conception of social structure emerged with the development of the Roman Republic and, as time went on , the Empire . In the past , people who have shaped family and kinship relations on the basis of the Canon Law model appear to find the basis for the continuity of soci al structure in overarching institutions , notably the C atholi c Church , as coordinating the lives of members of the community . While personal ties amo ng family and ki n may be very strong , these are not seen as the roots of continuity of the social order . Instead , organic analogies are appl ied to ex­ press the mechanisms that coordinate actions of the ind ividuals in the soc iety and give form to networks of social relationships .

n

family

il

Kinship Classification and Social Context

The review of the historical settings in which the kinship models in legal codes emerged lends support to the view that t here are on l y l i mited way s of symbolizing kinship structure s as they re late to the larger soci ety . In t he hi story of Western civilization , (a) the Parentela Orders model of kinship priorities d evelo ped in a soc ial setting that fostere d an em ph as is upon the perpetuati on of the - 'house , ' ' along with the development of centripetal kin ship norms; (b) the Civil Law model arose in a social se tti n g in which an effort wa s made to m aintai n a balance between - 'family · ' and civil s oc iety ; and (c) the Canon Law mode l was developed in a social setting in wh ich the perpe.t uati on of the " fam i ly " was considered as subordinate to the perpetuation of major corporate groups , s u c h as the Church or S tate . But if there is a correspondence between ki n s h i p model and historical origin s , then it seems plausible to spec u late that the underl y i ng dimensions in the models somehow expre ss symbolically the character of the genera l social struc ture . If this i somo rphi s m between kinship model and ge neral so­ cial structure exists , then it is also reasonable to ex pect that in contemporary society people who conceptu alize ge ne al ogic al ties ac c ording to a particular model are integrated into the larger soc iety in ways which parallel that model . This chapter presents a general statement regarding kinsh i p orientations as spati al metaphors of social s truc t ure . First, it e xp l ores pre vi o us applications of s pati al metaphors . Afterwards , it deal s spec ifically with kinship orienta­ tions as spatial met ap hors .

METAPHORS OF SOCIAL SPACE

For some time , p syc hol og i sts have s tud i ed the organ i zati on of sensory impressions . For ex ampl e , Gestalt psychologists (such as Kohler , 1 947 , p . 1 99) co nc lude that ' 'sensory org ani zatio n appears as a primary fact which ari ses from the el e ment ary d y n ami cs of the nervous system . " They al so suggest that ' 'the experiences of the various sense modal ities are loc al ized in a com mon space bec ause we have l earned in early childhood how they

35

y correlated " or because of "dyn amic reasons . " At any rate , s t be spati al l s e ns ory facts do appear i n one space , the space i n which al so t he v i su a l obj ects and the ' isual self are located (Kohler, 1 947 , p . 2 1 5 ) . "

�; 11

:

.

The org anization of sensory facts , however . doe s n ot occur 10 a soctal are given to these sensory i mpre s sio ns and their character­ c uu m . Names .

va

ned by others with Vw'hom a child i n terac t s . As long as the i stic s are defi org ani z ati on of perceptio ns pertains to phys ical things , there is a high p rob ­ abi lity of ac c urate communication with re gard to objects in the e nv i ronment .

A p rob le m in co m mun icatio n with others ( or in th i nki n g about a matter) emerges when the top i c is no longer a physical thing but an idea or some other intan gible event: How should a h ypothe t ic a l individual describe to someone else the c haracteri stic s of a phenomenon that is not a physical thing in a way wh ic h \\'ill maximize the probability of accurate communication ? Ac curacy requires the individual to use words that are u nders tood by the other . Insofar \\'ords describi ng se n so ry have a high prob abi lity of accurate commun ication , their metaphorical appl ication in the case of intan­ gible events (as o pposed to physical thi ngs) would describe these events in terms read i l y understood by the other individuals . We would anticipate then that t ateme nts w h ic h are i ntended to de sc ribe intangible events accurately tend to u se many sen sory metaphors (e specially spatial and visual imagery) .

as

facts

s

There

i s agreem e nt that

social rel ation shi p s consti tute

a

sign ificant are a of

i ntangible e vent s . This being the case , w·e expect sensory metaphors to

abound in the language used in e v e ry day life and in the social science s . Just

as

Kurt Lewin ( 1 93 6) found the

metap hor of psychological

space to be v al u ­

able in the study of ps ych ic al phenomena , the metaphor of social s p ac e has been fruitful in the analysis of social phenomena . The concept ion that phy s­ ical and social space are isomorphic leads Levi-Strauss ( 1 966 , p. 1 68) to suggest that ' 'space is a soc iety of named places , just as pe opl e are l andmarks with in the group . Places and in div i du a l s al ike are d es i gn ated by proper name s , which can be substituted for each o ther in many circumstance s com­ mon to man y societies. ' ' As subject matter for investigation , folk mode l s of soci al s pace may supply vehicles for understanding the deve lopment and main te n anc e of social stru c t ure s (e . g . ) alternati ve proc ed ures for com p uti ng kin ship di st ance) . The concept of soc i al d istance represents a w i despread application of the soci al space me t aph or in the social science s . The.re has been , however, dis­ agre ement over meanings given to the idea of ' ' di st a nce . ' ' For Pitirim So­ rokin ( 1 95 9 , p . 6) , social distance i mp l ie s that: Human beings , who

are

members of the same social groups and who within

each of these groups have the same function ,

are

in

an

identical social position .

Men who differ in t he se respects from eac h other have different social positions .

The

greater the resemblence of the positions of the different me n , the nearer

36

they

are toward

each other

in social space .

The greater and

more numerous

are

their differe nces in these respects , the greate.r is the social distance bet�·een them .

For Emory S . B ogardus ( 1 959) , ho we ver , the e xpre ssion social distance grows out of common-sense usage and draws upon such terms a s intimate re l at i on s , feeling distant, feel i ng near, or feel i ng far away . Bogardus ( 1 95 9 , p. 7) defines soc ial distance as "the degree of sy mp athetic understandin g that function s between person and person, between person and group , and between group and group . Sympathy refers to feeling reactions of a favorably res pons i ve type , and un ders t anding involves that kno wledg e of a person which also leads to favorabl y re spons ive beh avior . ' ' McFarland ( 1 973) refers to S orok in 's co ncepti on of social distance as em­ bodying the notion of c ate go ric al similarity , whereby i nd i v i d ual s or groups who are similar in attributes de s i gna t i n g soc ial po s i t i o n are "close " ; in con­ trast , M cF arl an d refers to Bogardu s ' conception as re l y i ng upon deg ree of probability of s y mpat het ic interaction . The con ceptual distinction between categoric al similarity and degree of pro babili t y has i m ponan t consequence s for the anal y si s of soc ial re l ations . The measurement o f social distance ac­ cording to the Sorokin conception i n vo l ves the c l as sification of i nd iv idu a ls or g rou ps in such nominal c ate go ri e s as occupat i on , income , education , eth­ n i c ity , and so on . Ho we ve r, the measurement of social distanc e s , according to the Bogardu s c on ceptio n, involves a gradien t of interactions accordi ng to i mpl ie d ' ' sy mpat heti c understanding • '-intermarri age , commensal i sm , de­ g rees of residential propinquity , gradations of do ing business togeth er , and the like . The difference in meaning of s oc i al distance between Sorokin and B og ardu s thus l e ads to d ivergent understandings abou t hoy,r social relation­ s hips evolve . For Sorokin , similar social posi tions generate similar economic , pol itic al , and social i ntere sts , and t herefore in a h igh l y diverse society s oc ial nearness promotes the development of facti on s with special intere sts . For B ogard u s , sympat hetic und erstanding transcends group boundaries, and so­ cial ne arne s s promotes the establishment of a cohe s i ve order and good will across po p u latio n seg ment s . The distinction be t\\'een c ate goric al and gradient represen tation s of social space is not re stric ted to the measurement of social distance . A similar dis­ tinction extends throu gh a wide range of social phenomen a and , in eac h case , leads to o pposi n g conce ptio ns about the c haracter o f soc ial existence . For example , in d esc ribin g the growth pattern s of cities , the B urges s concentric­ c irc le h ypot he si s and t he Hoyt sector hy po t hesi s lead to d ifferent i nterpre ­ tations of the proce ss of urban de ve lo pme n t (B urgess , 1 97 3 ; Hoyt , 1 933) . Like the Bo gardu s co nception of social distance , the Burgess concentric­ circle hypothe s i s seems to fall into a cl a ss of s pati al metaphors involvi ng a grad ie nt; l ike the Sorokin co ncepti on , the Hoy t sector hypothesis falls into

37

iff

f

hors that emphasize d ering values or diverse bounded 1 ss of metap an d yiel d a series of nom i nal categories . I will describe these two a regtons s patial metaphors 1n greater detcu·1 below . l asse s o f aJ metaphors , l ike Burgess ' concentric-circle model , c The frr st class of spati i en sion of core versus periphery , intimacy versus remoteness , sts on a d m som e other characterization of a gradient . The use of gradients for de­ i in soc i ology . In their ad i ribi ng soc ial relationsh ps has had a long tr it on mitive classi fi cat i on, Durkhe im and M auss (in Parsons et al . , f an aly sis o pri ! 96 1 ) i ndicate that primit i ve societies based thei r systems of classification

c. a

·

� �

u pon such grad ations:

The frrst c lasses of things v.·ere classes of men , into which these things \\'ere integrated . . . phrateries were the frrst ge nera; the clans , the frrst spec ies . Th ings were supposed to be integral parts of the society , and it was their place in society that determined their place in nature . . . . It is a fact of common observation that the things included in genera

are

generally imagined as located

in a sort of ideal mi lieu , the space dimen sions of which are more or less c learly

defined . It is certainly not Vlithout reason that concepts and their relationships have so often been represented by c ircles--co ncentric or excentric , outside or inside one another, etc . (Durkheim and Mau ss , in Parsons et al . , 196 1 , p . 1066) .

The assumpt i ons made by Edward Shils ( 1 975) about the nature o f society constitute a coherent statement about the ex i stence of a gradient , extend ing

from core to periphery , as expressing a basic dimension in social life . For S hil s , i t i s the length of the gradient that describes the degree of integration of the society : the longer the gradient, the greater is that chance that new centers will emerge , which will , in tum , ac t to increase the size of the gradient further . According to Sh i ls ' exposition :

I . ' ·An human collectivities have a tendencv toward closure into selfcontai nment. They seek th ro ugh their authorities to establ ish and ma i ntain a certai n iden tity , to define their boundaries , and to protect their integrity ( S h il s , 1 975 , p . 45 ) . " '• 2 . The main factors which e stabli sh and maintai n a society are a central authority , c onsensus , and territorial boundedness . . . . Every society , seen macro soc i ologically , may be interpreted as a center and a peri p hery . T he ce nter con sists of those insti tutions (and roles) which exercise authority­ whethe r it be economic , governmental , political , mil itary-and o f those w h i ch cre ate and diffuse cultural symbo ls-rel i g i ou s , literary , etc .-through churches , schools , publishing houses , etc . The periphery consists of those .

strata

or

sectors of the soc iety which are the recipients of commands and of beliefs whic h they do not themselves create or cause to be diffused, and f o those who are lower in the distribution or allocation of re\\·ards , dignities, fac il i tie s , etc . (Shils , 1 975 , p . 37 and 39) . "

38

3 . . .The existence of a central value sy stem re sts , in a fu n d amental w ay , on the need which human bei ngs have for incorporation into someth in g w h ich transce nds an d tran sfi gure s their conc rete individual e xistence . They have a need to be in contact w ith symbol s of an orde r which is larger in i ts dime n­ s i ons than their O\\'D bodies and more central in t he 'ultimate ' structure of re ali ty than is their rout i ne everyday life (Shils , 1 97 5 , p . 7 ) . " Moreover � ' 'This central value system i s the cen tral zone of the society It is cen tral because of its in ti m ate connection with what the soci ety holds to be sacred; it is central because it i s espoused by the rul ing authorities of the society ( Shils , 1 97 5 , p . 4) . " 4 . . .All territorially extensive soc i eti e s tend to have a spati al center as well , which is, or is t hou ght to be, the seat of the central i nstitutional and cu l tural systems (Shi l s , 1 975 , p . 39) . " 5 . ' "Integrated societ ies in Vlhich the authoritative institutional and cultural systems are "·ell established can become civil soc ieties with a wide diffu sion of the virtues requ ired for the effective prac ti ce of c iti zen sh i p . They become so where closure around the ce nter is accompanied by the approximation o f center and periphery . Thi s is th e pat h which h a s been followed over the past century and a half in Western Europe , th e United State s , and Australia , and to a lesser extent by Japan and C anada. In these countries, the mutual ex ­ change between center and periphery , and the he ig htened se nse of affin i ty which attends this i n terc hange , have brought larger proportions of the pop­ ulati on into the center and obliterated to some extent the bo u ndary which has in the past separated center from periphery ( S hi l s , 1 97 5 , pp . 46-47 ) . " .

The second class of metaphors of soc i al spac e , l ike Hoy t s sector model '

in urban growth , refers to qu al i t at i ve differences in the struc ture of soci al relation ships . Like the gradient metaphor, the second metaphor of social space has had a long history of appl ication i n the soc ial sc ience s . It too appears in the analysis by Durkheim and Mauss of pri mit ive class ificatio n scheme s . They write (in Parsons et al . , 1 96 1 , p . 1 067): \\'hat i s here concei ved as perfectly homogeneous is represented elsev.·here as

essentially heterogeneous . For us, space is formed of identical parts, inter­

changeable one v..· ith the other . We have see n , hov.,-ever, that for many peoples

space is profoundly differe ntiated, depend ing upon the region considered . Thi s

is because each region h as its own affective val ue.. Under the influence of

v arious sentiments , a particular region of space is referred back

rel i gious principl e; in consequence , it is endowed Ylith virtues \\'hich distinguish it from any other reg ion .

to

a specific

.. ui

.

generis ,

The dist i n cti on between the gradient and c at ego ri cal met aph or s of social

kin ship organization . Dual ki nship o rg an izat io n i s gene ral l y characterized by the pre sence of moieties that d ivide a society into tw o maj or kin groups . Levi- Strauss ( 1 963 , pp . 1 3 2space appears in Levi-S trau ss ' discu ssi on of dual

39

shows a correspondence between the spatial l ayout of villages and the g an ization of moieties . In societies with moieties , villages tend to be laid

1 63)

C:t that they are divided physically along an axis into two parts , which SS designates as diametric dualism . Sometimes these villages also �vil-S traU mmon with other societies which are not divided into moieties) (in so

co disp ay of concentric circles , u sually with a core and a periphery differen­ a layou t ti ated according to function or residential use . Levi -Strauss emphasizes the i n te rplay of these layouts in structuring social relationships; however .. it may be equ al ly i mportant to stress their distinctiveness as illu strating the two

metapho rs of social space . 1n con centric patterning , the core of the village , according to Levi-Strau ss ,

is ord inari ly reserved for communal functions such as ceremonial acti vities , danci ng , burial , or the living quarters for ch iefs or unmarri ed men . The pe­ riphery holds the individual clans, the homes of marr ied couples , and the

work areas . On the other hand , di ametric structure s signify categorical dis­ tinctions in social rel ationships and responsibilities . Levi-Strau ss suggests

that frequently � 'the opposition between moieties expresses a more subtle di­ alectic " in function . Dualistic structure appears to present not mere-ly a di­ chotomy in social categories but also an opposition: warfare and policing \·ersus the arts of peace and arbitration , continuity versus disconti nuity . In this manner, the diametric structure of dual kinship appears to expres s the barriers , the boundaries , and the conflicting valences described in topological psychology . The diametric structure seems to represent another instance of the c ategorical metaphor of social space , in which a line of demarcation acts as a basis for partitioning the space .

In contrast to users of gradient models , those scholars w ho apply a cate­

gorical conception of social space take as a starting point an initial state of conflict among collectivities--a battle of group against group . In his work on kins hip , for example, Levi -Strauss begins with an original state of hordes

in confl ict and proposes that social structures greater than the nuclear family

e merged as a result of the incest taboo and of the rec iprocities developed and elaborated between groups through marital exchange . From the begi nning , accordi ng to the structural position , community life was based on political

alli ance s of groups real izing their speci al interests in reciprocities established Wit h othe r groups . In the categorical conception of social space , thus , one as su mes as a starting point the predom inance of special intere sts of different groups .

The no min al-category metaphor of social space thus implies that rel ation­ . hi s ps in the family e merge as an entity only insofar as obligations-reci­ proc it i es-within the family coalesce . The categorical representation suggests th t the family affects the collective destiny of its members in competition � W ith simi lar collectiv ities for wealth , power, means of grace , or other scarce gOOds .

40

It may be advisable to summarize at this point . The previous discussion of

soc ial space metaphors and their appare nt implications for the analysis of social relationships may be stated as follows:

l . Soc ial space metaphors derive their util ity from the gre ater effectivene ss

of words pertaining directly to sensory i mpressi on s (especially visual im­

agery) than of more abstract terms to communicate ac c urately the descripti on

of social relationships . 2 . The application of spatial metaphors suggests an isomorphism between physic al and social space , which may exist as explicit ' 'translations ' ' in cl assific ation scheme s of social space . 3 . Two kinds of spatial metaphors appear to be applied generally ; a gra­ dient co nception of social spac e and a nominal-category conception of soc i al space .

4 . The grad ient c onception of soc ial relationsh ips i mplies a soc ial space consisting of

a

core and periphery , with c oncentric circ le s denoting degrees

of remoteness; the categorical conception of social relationsh ips implie s

a

social space con sisting of a set of regions (defined as categorie s of bounded areas) , with a l i ne of demarcation (rather than a single point) as a basi s for generating the model .

5 . These social space metaphors seem to apply to a large variety of social phenomena and , as Durkheim , Mauss , Levi-Strauss , and others have sug­ gested , to express some basic elements in ways soci eties (or social groups) are organ i zed . 6 . The two metaphors funher imply a difference in the way social rela­ tionships develop: the gradient-metaphor rests upon the assumption of tend­ encies toward consensus or common concern as the basis for social structure ;

the categ ori c al metap hor assumes that factionali sm or group conflict over special interests is at the root of social structure .

7 . Given the differences in assumptions about the development of social relationships , the two metaphors presuppose opposing resolutions to disorder in society . The grad ient representation presupposes a strengthening of con ­

sensus with regard to the common w el fare ; the categorical model presupposes

an accommodation by the oppo sing faction s .

KINSHIP MODELS AS METAPHORS OF SOCIAL STRUCTURE Ki nship models seem

to symbolize the relationship between the social whole and its parts in terms of space . In discu s si ng kinshi p ties , people often refer to Hthe proximity of relatives , • ' "di stantly related persons , " "intimate kin, , or ucousins who are far removed . " David Sch neider ( 1 968 , pp . 72- 7 3 ) propo se s that there are three different meanings of di stance in Ameri can ki nship: physical distance , socio-emotional distance , and genealogical dis­

tance . Wh i le these three meanings are not necessarily rel ated , they seem to

41

ted . The co nnec ti on between kinship and the concept of di stance .orrel a be c ests that the ki nsh ip models in leg al code s can indeed be reg arded as . sugg phors of soc tal structure . ta me al ti Hen ry M organ ' s seminal work , co nce ptions of geneasi nce Lewi s l ic al strocrure have been regarded as synonymous with patterns of kinship Using a l inguistic model , anthropologists seek to determine

s�ver C::mencl ature. mantic domains , which are supposed to yield a system of soci al c l assifi­ ��ion based on roles , statuses , and group membershi p . Anal ysis of k i nship n

tenns h as indeed produced n umerou s insights pertaining to fami ly and k inship

i cularly in n on i nd ustri al societies . These insi ghts refer pri m ari ly struc tu re , part to the sem antic dimensions underl ying a tab le of organization of kinship

relationshi ps . The tables of organization vary in the ways they partition rel­ ati ve s accordi ng to co n s a ng uin i ty versus affinity , lineal ity versus c.ollateral ­ i ty , the d i stinction between a scending and descending generations , bifurca­

tion versu s mer ger by relative 's gender, and so on (Murdock , 1 949) . Anthro po logists have uncovered numerous correlations between kinship ter­ minology and table of organ i z ati on . In modern society , the relatively high degree of social stratification , oc­

cupational differentiation , religious and ethnic diversity , freq uent contact with foreigners , and variations in domestic life sty les all i nteract to sustain a

broad spectrum of kinship tables of org a nizatio n . Still , despite the heter­

ogeneity of ki ns hi p tables of organization and their supporting ideologie s , virtually everyone in the society learns a c ommon set o f k i n terms (of the dominant language spoken) . C on seq ue ntl y , even with a com mon kinship terminology , people often do not agree about what constitutes an pprop riate table of kinship org an iz at i o n . For in s tance , some people include in- law s and distant cou si ns in thei r · "k i n universe , ' ' while others do not (Schneider and Cottrell , 1 97 5) . Thi s l ack of association between table of kinship organi zation and term s for de sig nati ng re latives suggests that in contemporary society a supplemen­ t ary c riterio n is needed to handle the tasks which term i n ol ogy ordinarily acco m plishes in noni ndustrial societies . While people may differ in tables of �rganiz ation , they still need a vocabulary to describe priorities of duties and nght s w ith re spect to their · "kin universe . " S i nce contemporary societie s do not gen erally have formall y-organized descent groups in which membership Wou ld define these priorities , another means must be devi sed to serve th i s

a

purpose .

Sc h nei der and Cottre l l (1 975 , p . 92) propose that "one of the fundame ntal fe ature s ' ' of the structure of the Ame ric an kin universe is the di vi sion of

�7l ative s

uin

to those who are 'close ' and those are are 'distant . ' " In fact , th e gre ater the genealogical distance , the more important nongenealogical and n onkin sh i p con siderations become , the more choice Ego has over whether

to

inc lude or e xc lude [the rel ative] from his kin univer se and whether to

42

transmit [the relative] to his children (Schneider and Cottrel l , 1975, p . 96). '' Sheila Klatzky (n . d . , p. 84) also has found that genealogical distance is associated with the structure of kinship relations-in that ·'there i s a greater element of choice in contacts with genealogically more distant relatives.'' Likewise , Litwak ( l960a) has shown that the closer the genealogical tie , the less apt is geographical distance to diminish contact , mutual aid , or se nti­ mental bond . (See also Rosenberg and Anspach, 1 973; Adams , 1968.) The tie between social structure and conception of genealogical space pre­ supposes that the societies in "·hich the kinship models in legal codes "V1'ere first developed exhibited many of the complexities found in modem society. Jewish soci al structure, classical Greece , and the Roman Republic all sho\\' these complexities . Indeed , the presence of laws of succession presumes an elaborate system of property institutions, whereby individuals (or individual families) can accumulate weal th and resources. Each of these societies held urban centers , with a high degree of soci al stratification and with foreign residents interspersed among the population . In each society , there was also conside.rable contact with various other ethnic and/or rel igious groups . Canon law itself was explicitly established to create a unity out of the diversity of European customs and traditions . So there was much heterogeneity in these societies with respect to social and economic hierarchy , diversity of family norms, range of c itizens h ip rights , and so on . The broad ranges of population characteristics in these societies perhaps encouraged the metaphorical use of the concept of social distance (especi ally in the sense in which Sorokin used the term) to describe similarities in property interest by various groupings. Since property interests were generally connected with the family or kin group , it apparently was reasonable to apply the concept of distance to the measurement of genealogical relatedne ss . Despite their considerable complexity , the societies in which the legal models originated in no vt'ay even approximate the size and heterogeneity of modem society . These older societies were still h omogeneous enough for the particular kinship models impl ied in legal codes to serve their special needs for social continuity . The heightened complexity of modem , urban society (as suggested by the discussion in the introductory chapter), however, is capable of accommodating the full range of conceptions of genealogical space found in the legal codes of the Hebrews, ancient Athens , the Roman Twelve Tables, and Catholic canon law. The variations in conception of genealog­ ical distance found in the studies reviewed above, associated as they are with socia1 characteristics, may well be expressions of the models which appear in legal cod es-Parentela Orders, Civil Law , Genetic , and Canon La\v . The discussion of metaphors of social space leads to the speculation that the kinship models applied in legal codes fall on a continuum . At one extreme is the Canon Law model , which expresses the concentric-circ le conception of space , implying consensus or common concern as a basis for social struc-

43 At the other extreme is the. Parentel a Orders model , which yields a set unded cl asses , implying that an accommodation by special interests lies the root of social o rder. Specific i mp l ic ati on s of this sy m bo l i zat i on are

�t�� tha:� people who conceptualize kinship ties in accord ance with the Parentela

Orders model regard their partic-i pation in segmental groups ( perhaps deri v ing ethnic , or some other special interest) as more significant for from re l igi ous , in universalistic associations and social continuity than their involvement move me nts in the civil society . 2. People who conceptualize kins hip tie.s in acc ordance with the Civil Lavl model regard social continuity as deri vi n g most effectiv e l y from a balan ce of parti cipat ion in segmental, special interest groups and in universalistic associations and mov emen t s in civil society . 3. Pe opl e who conceptualize kin sh ip ties in accordance with the Geneti c (or shared chromosome) model or the Canon Law model re gard social con­ tinuity as deriving more effectively from participation in universalistic as­ soc i ati ons , institutions , and movements in civil society than f ro m segmental

groups .

Two investigations h ave been undertaken in which adherence to a particular

kinshi p model is related to place in the co mmunity . In part, both studies dealt with the ex tent to wh i ch participation in a segmental, re l i gi ous c ommu nity that is, Judai sm and Christian sectarianism-is associated with conception of genealogical space in tenns of the prioritie s of the Parentel a Orders model . In the first study , based on a sample of 248 students at Arizona State Uni­ versity, Je�·s and Christian sectarian s showed the greatest conformity to the. Parentela Orders model; denominational Prote stants were next; and Catholics and athei s ts exhibited least conformity to this model (Farbe r, 1977). In the seco nd stud y , based on 182 Jews l iv i ng in the Kansas City met ropol itan area, persons indicating a high degree of conform ity to the Parentela Orders model had the foll owing characteristics: (a) most often Orthodox in upbringing , next Conservative, and least often Refonn; (b) born in rather than m i grated to Kansas City; (c) live in areas of Kansas City with a relatively high concen­ tratio n of Jewish popu lati on ; (d) more acti\'e than the ir own p aren ts in Jewish activities; and (e) di sapprove of intermarri age with non-Jews F ( arber, 1979). The fi ndi ng s of both studies are hence c o nsisten t with the p revio u s discussion regardi ng the relationship betv.'een organization of gene a l og ic al space as ex­ pres sin g an ideolo gy regarding the role of the family in social continuity, Whereby a segmental co mmu nity-i n this case, religious--translates its con­ cerns abo ut c ontinuity into a sy stem of priorities in genealogical space . For ex a ple m , the centripetal family ideo logy of Ju dai s m expresses this concern about c on tin uity through the ve hicle of the Parente I a Orders model, and among Jews. it is those who are most involved in Je�·ish communal life who organ i ze the ir kin universe in terms of this model. ­

44 Since the tY.'O past studies are based on atypical populations (university students and Jews in Kansas City ) , the q ue stion may be asked: To what extent does the relationship between social structure and kinship orientation appear in data generated from a more representative sample? The analysis that fol­ lows is addressed to this question .

Cognitive Kinship Maps:

A Serendipitous Finding

Briefly , the previous chapters in this monograph proposed that the computational measures for deter­ mining priorities in intestacy and for defming certain kinds of con­ sanguineous marri ages as "'inces­ tuous�

9

in I aYi produce orderings

of consanguines that c an be regarded metaphorically as cogn itive kinship map s . Historical analysis indicates that the kind of measure adopted in lav.' is related to the character of the social structure . In his rigorou s analysis of such measure s , John Atkins ( 1974) described in particular the Canon Lav.· Civil Lay; computational models and their homologues . He indicates that, although the homologues may be ' 6h i storical l y distingui shable'' from these measures , they ''create exactly the same ordering of consangu ines {Atkins, 1974, p. 13)." Unexpectedly, this study has unvei led an ordering

and

of consanguines that has not appeared among any of the previou s measures (including those described by Atkins ( 1974)), and the analy sis in this chapter

will focus upon thi s serend ipitous fmding .

FORMAL AND POPULAR CONCEYI'S OF KINSHIP

Althoug h an analysis of the formal measures of genealogical proximities may prov ide numerous insights into the character of kinsh ip in different societies, the questi on recurs : Do these formal models of genealogic al proximities rep­

resen t in any w ay the kinds of kinship maps that people carry around in their own m inds? If these models do represent distillations of popu l ar conceptions of kin ship , then in a population one should be able to find some degree of assoc iat ion betv.'een cognitive confonnity to a model and variou s soci al and cu ltural c harac te.ri stics . Pi lot studie s (Farber , 1977; 1979) on limited popu­ l ati o ns h ave indic ated that there is indeed a relationship between conformity

� ka_ nship models and predictable characteristics in religion , ferti lity , migra­

hon, and edu cation . The w idespread presence of a previously unknown model i n the Phoenix

�pulat io n ha s important implications for the understanding of middle-clas s kinship structure. Consequently, the c hapter also incl udes the resul ts of a compo nenti al analysis that uncovers the cognitive dimens ions shaping the

46

new model-as well as the other models under investigation . This type of analysis clarifies the inte-rrelationships among the various models (Atkins, 1974) and it should expedite. interpretations of the Phoenix study findings in the chapters that follow . In the Phoenix study , individuals were classified by kinship mapping ori­ entation on the basis of their responses to questions penaining to prioriti es in intestacy law . The technique was as follows: I. The respondents were presented with a serie s of paired comparisons of relatives for whom the four models-Parentela Orders , Civil Code , Ge. netics , and Canon Law�iffer in genealogical remoteness . 2. The instruc-tions to the respondents were:

Next , 'Ne have a few questions about inheritance lav.·s . When someone dies without leaving a v.'il l , the la'W·s says that the estate goes to the nearest relatives . The law then tries to follow what most people look upon as their nearest relatives .

Suppose you had to v.·rite a law for Arizona to decide which relati vcs should have a greater claim to the estate when a person dies without leaving a will .

Categories of relatives are listed on this form. These pairs were chosen because they stand for different ways of figuri n g which relatives are closer.

For each pair of relatives , please check the kind of relative who should have a greater claim to the estate-when there is no \\rill .

3. The respondents were then asked to assign priorities in nine pairs

of

consanguineous relatives. The list included parents , children, brothers and sisters , grandchildren , first cousins, aunts and uncles, nieces and neph­ ews , grandnieces and grandnephev,'s, and grandparents . For each pair, the respondents \\'ere given three choices . For example: Comparing brother�· and sisters �·ith grandchildren:

( ) Brothers and sisters should have a greater c.Iaim.

( ) Both should have an equal claim.

( ) Grandchildren should have a greater clai m . In th e comparison between brothers and sisters and grandchildren, the choice of brothers and sisters would be appropriate for the Canon Lav-.' and Genetic models; "equal claim" would be appropriate for the Civ il Lav,r model; and grandchildren would be the appropriate response for the Parentela Orders model .

Nieces and nephews

First cousins versus niece�\ and

versus

Parents versus children

Brothers and sisters versus parents

grandchildren

Brothers and sisters

uncle�

First cousin� vemus aunts and

grandnephew�

First cousins verso�\ grandnieces and

nephews

grandparents

Children

Purents

Grandchildren

Aunts and uncles

grandnephews

Grandnieces and

Nieces and nephews

Nieces and nephews versus

nephews

Niece� and nephews

GrandparentR

Parentela orders

Aunts and uncles versus nieces and

uncles

Grandparents venus aunts and

intestacy question

Kin types compared in

Equal claim Equal claim

Equal claim

Brothers and sisters

Aunts and uncles

Equal claim

Niec� and nephews

F..qua) claim

Equal clain1

Equal claitn

Parents

Equal claim

Aunts and Uncles

Equal clain1

Nieces and nephews

Grandparents

Equal claim

Grandparents

Genetic model

Clasaily Respondents by Kinship Model

Civil law

TABLE 4-1. Patterns of Answen to Intestacy Questions Used to

Equal claim

Equal claim

Brothers and sisters

!:.qual claim

First cousins

Equal claim

Equal claim

Equal claim

Equal claim

--

Canon law

'-J

�

48 4. Each respondent was then classified according to the kinship model to which a majority of answers corresponded . The response patterns used to classify each person by kinship mapping orientation are shown in Table 4- 1 . When there was no majority for any one model, the person was assigned to a residual category. (See Farber, 1977 and 1 979.) AN AMERICAN STRATEGY

In an ideal world, all participants in a social survey would provide responses consistent with the investigators" expectations. In a study of kinship models by which people organize priorities among relatives, all respondents would dutifully provide answers that would permit classification in one of the pre­ detennined categories of kinship orientation. Spec ificall y , each respondent would produce a pattern of responses that could be classified unambiguously as conforming to an identifiable model-Parentela Orders , Civil Law , Ge­ netics , or Canon Law . But, the world is not totally predictable-and not all respondents conduct themsel ves as if they were using the investigator� s re­ search design as a script. In this study , responding to the questions on pre ­ ferred priorities in intestacy law , 17 percent of the sample produced patterns of answers that fell into a residual category . The large number of cases that did not correspond to any of the kinship models under examination evokes the questions: Do sizeable segments of the population conceptualize priorities among relatives in an idiosyncratic fash­ ion? Or has the investigation omitted a widely-applied kinship model from consideration? Supporting the interpretation of residual category cases as idiosyncratic is my experience that many persons say that, in responding to the questions, they are drawing from their own personal experiences with kin. This argument, however, is specious since most people , even when they conform closely to an existing model , are unaware of their rationale . A Serendipitous Model

In orde.r to determine whether many of the residual-category cases actu al ly conform to a coherent pattern , I examined the modal responses to the indi­ vidual questions on priorities in intestacy laYl . The pattern of modal responses in the residual category appears in Table 4-2. A clue to the organizing prin­ ciple of this pattern is suggested by the fact that, with one exception (that of cousins versus nieces and nephews) , the kin types in the modal categories are always in an older generation than are the types with which they are compared. That is, grandparents are given priority over aunts and uncles� parents over children , and so on . The next step in the anal ysis of the residual cateogry is to find a compu-

49 LE 4-l. Modal Responses to Intestacy Questions for Cases Falling into 1AB Residual Category of Kinship Ct.sifications Percentage of cases

types com�ared in Kin O intestacY questiO

Grandparents

falling into modal

Modal kin type

versus aunts and

category a

Grandparents

53.8

Aunts and uncles

78.8

Grandparents

78.8

First

Nieces and nephews

41.7

rust cousins versus grandnieces and

First cousins

75.8

Aunts and uncles

56.1

Brothers and sisters

60.6

uncles uncles versus nieces and • un1s and ·""' nepbews s versus Nieces and nephew grandpareuts . cousins versus rueces and nepbews

grandnephews

First cousins versus aunts and

uncles Brothers and sisters versus grande hi ldren Brothers and sisters "iersus parents Parents versus children a

Number

of cases

classified as

Residual

52.3

Parents

49.2

Parents =

132.

tational formula that y ields the pattern of priorities suggested by the modal re sponse s. The computational model that fits the patte rn is: D = (n + I )i + j, where D refers to priority rank; j as before is the number of generations bet\\·een EGO and the ne arest ance stor in common Vlith a collateral relative; n is th e number of generations betYleen EGO and the m os t re-mote ance stor (i.e., the largestj) to be con sidered in a set o f the computations; and i i s the number of generati ons between the nearest common ancestor and the collat­ eral relative. In com putatio ns involving lineal rather th an collateral rel at iv e s , j alone is used for EGO' s direct ancestors , and (n + 1) i by itself is used for EGo's descendants. Thi s co mputati onal model produces the configuration of kin types in Table 4-3 .

Comparison

There �odel

with Parentela Orders }fodel

is so me similarity between Parentel a Orders and the sere nd i p i to us yi el ded by the residual-category analysis. Both model s depend upon

differenti

ati ng ancestral l ines from lines of desc ent , and both m odel s provide for un iqu e ra nkings of distance for each kin type . D i ffere nces , however ,

a�pear whe n the patte of priorities in Table 4-3 is compared with that Yt�lded by the Parentela Orders model. As noted earl ier, while the compu­ rn

talto nal formula for the se re ndipitou s model is D

=

(n

�

I) i + j, the formul a

50 4-3. Ranking of Distances from Ego According to Model Based on Modal Responses in Residual Category (Standard American

TABLE

Kinship Modelf

Order in

Order in

line of descent {i) 0

3

2

Ego (Q)b

Great-grandchildren (9)

(Ji

+

sisters (4) and nephews

Nieces

Grandnieces and

First cousins (8)

j) .

the appropriate formula ror the Parentela Orde�

distances in the Parenlela Orders model is

(1) Children (2) Grandchildren (3) Great-grandchildren (4) Parents

(7)

grandnephews ( 10)

.. Rmkill@ of distance from Ego.

these kin types,

Grandparents (2) Aunts and uncles (5)

Parents (1)

Brothers and

Grandchildren (6)

" Computational fonnula is Note: For

generations (j)

0

Children (3)

2

asce nding

as

follows:

(5_) Brothers and sisters (6) Nieces and nephews (7) Grandllieces and gnndoepbews

model is: (I + 4j).

Ranking
Grandpareats (9) Aunts and uncles

(8)

( I 0)

FIJ"St

cousins

for the Parentela model is D = i + (m + 1) j, where m in this case refers to the most remote descendant (instead of ancestor), i.e .• the largest i to be condsidered in a set of computations . Whereas the serendipitous measure stretches distances of descendants of common ancestors , the Parentela Orders measure elongates the distances of the ancestois themselves . The serendipi­ tous model thereby emphasizes the closeness ot ancestry and the importance of the I ine of ascent; the Parentela Orders model stresses the proximity of descendants and the significance of the descent line . Symbolically, the ser­ endipitous model sees kinship from the perspective of family history. y..·hile the Parentel a Orders model views kinship from the standpoint of family destiny .

The Standard American Kinship Model

interpretation of the fmding is that the serendipitous model is mere.Iy a version of another, more traditional kinship orientation . However, a more plausible alternative interpretation is that this new model represents a con­ ception of kinship that predominates in contemporary. postindustrial society. There is some empirical support for this position . All 723 cases in the Phoenix probability sample were examined to deter­ mine the extent to \\'hich they conformed to the pattern of modal responses in the residual category. The findings were: (a) responses for almost half the

One

51

Je fi t this pattern; ( b.> most cases for which there had been a tie betv.·een

saJllP t el a Orders and Civil Law scores had an even higher score for this (47 out of 85); (c) l arge numbers of cases which had previously fallen

�:1

�toto the Pare ntel a Orders

. reclassified as conand Ci vil Law categories were .

.

model; (d) the slZe of the restdual category was reduced n rm i g to thi s new to si derabl y (from 17.3 percent to 7.0 percent). In short, the introduction of model into the analysis has greatly modified the distribution of cases .

:�·new

ion . by kinship classificat

Persons conforming to this model seem to be simil ar to those exhibiting ''standard'' American characteristics in general-l anguage patterns , values,

life st yle, and so on. Table 4-4 shows the extent to which certain social charncteri stic s are overrepresented in persons who conform to this model. .Although the overall percentage of case s in th is cl assification �·as 43.5 per­

cent, higher percentages appeared for Protestants in denomi nations originat­ ing in the Reformation era, non-minorities, persons i n professional or tech­ nical occupations, those who have a family income $25,000 or over, and/or people whose fathers are native-born. Conversely, low percentages appeared

for Je.ws, b l acks, and American Indians, semi-skilled and unskilled workers, those v•ith a family income of under $ 1 0,000 , those persons who did not graduate from high school , and i ndividuals with foreign-born fathers . In general, the Standard American kinship model seems to be most prevale nt among middle-class , non-minority Protestants , whose parents are native­ born.

The exte nsiveness of the serendipitous mapping in the sample, along w ith social characteristics of persons conforming to it , suggest that it is not

the

merely a version of another measure in the series examined . Rather, the find­ ings support the view that the serendipitous mapping represents a kinship orientation sui generis. Consequently, pending further analysis, I am naming

this orie ntation the Standard American kinship model.

Both religion and socioeconomic status appear to influence the distribution of the S tandard American model. Table 4-5 compares the percentages of perso ns in the Standard American category when respondents are classified by their relig ion and by the occupation of the male co-head of the hou sehol d. (The Jew ish respo ndents in the supplement ary sample are included here and because of their general occupational homogeneity , all of them are placed in column headed Profession al, Managerial . or Admini strative.) A compariso o ng rel igio us groups reveals that, in the Professional Manager n am , ial, or d . . . Athe mtm s tratJ.ve cate ory, Jews (43 percent) are somewh at less l ikely to hold � (S S tandard Amencan model than are Protestants (54 percent), Catholics th 2 percent), or per sons express ing no rel igious preference (54 percent)­ being l ittl e distinction among the latter three groups f . In Blue Co llar h I es, however, Protestan s (44 percent) are clearly more predom inant in � t S t andard A mencan _ _ rubnc than Cathohcs (28 percent) or those without

�he

e�.

�

52 TABLE

4-4. Social Characteristics of Respondents

as


Underrepresented in Standard-American Category Percentage or cases

falling into

Extent of

categOl)·

representation

over- or

Standard american

N

Social characteristics

(k/43.5)

43 5

Total Sample (N=723)

.

Religiona

Protestant (no denomination) Refonnation Era Protestant

denomination

No religious preference Pietistic Protestant denomination

50.0

1 .15

100

49.0

1.13

82 1 39

46.3

1 .06

46.0

47

40.5

1.06 .93

205

40.5

.93

43

39.5

.91

10

30.0

.69

582

45.3

90

38 .9 30.0

1 .04

54

Neo- fundamentalist sect Catholic

Latter Day

(k)

under-

Saint (Mormon)

Jewish Minority Status Non-Minorit}· (i.e., not Black, Oriental, Mexican American,

American

Indian. etc)

Mexican American Black, American Indian,

Oriental''

50

.95 .69

Birthplace of Father

619 51 36

United States Latin America._.

Europe

45 .2 37.3 27.8

1 .04 .86 .64 ( contirtued)

Refonoalion era Protestant deoominations include: Lutberm, Conpeaational, Presb)'terian, EvmtgeJical, Reformed. United Church of Christ, and Episcopalian. Pietialic denominalions include: Methodist,. Baptist (net Primitive or Southern), and Disciples of Christ. Neo-fundamental ist secas BDd deaomiaations include: Cburcb of God. Nazarene, PeiUCOslal, Church of Ouist,. Primitive Baptist, Seventh Day Ad\lentist, and SoulherD Baptist. Other denominalims held too few members in the sample to ooUect into a meanincful category. a

bIn this calegor)' are

31 Blacks, 14 American lodians,

�Includes one person born

in Canada.

4

Orientals, IDd I Black and Me�ican

American

a religious preference (36 percent). As for occupational differences, all

re­

ligious groups show that the Professional, Managerial, Administrative cate­ gory contributes the highest percentages of persons with a Standard American

orientation and the Blue Collar category the lowest. Thus, the data i n Table 4-5 testify that neither socioeconomic status nor religion can be discounted as factors in conformity to the Standard American model.

53 (continued fABLE 4-4

)

----

Percentage

of cases falling into Standard

over- or

american

under-

N

category

representation

{k)

(k143.S)

administrators

2SO

49.6

223

4 1.3

1 .14 .95

set"Vice, laborer)

1.51

35 . 1

.81

45.8 43.7

1.05

ics Social characterist

()ccUpatiooal Status

Professionals, managers. and

Sales, clerical, craftsmen Blue collar workers (i.e., operative, Family Income $2S.OOO or over $10,000 to $24,999 Under $10,000

177

40.5

.53

Some college High school graduate Not a high school graduate

TABLE 4-S.

290

2S9

.89

45.3 48.3

1.11

27.3

.63

1.04 1.04

45.2

121

ReUgion, Occupation of the Male Co-Head of the Household, and Percentages of Persons in the Standard American Category Occupation of

Professional,

managerial, or

�ligion

administrati �e

Jewish(%) N

Prote� (%) N

Catholic (%) No religious

1.01

38 . 7

124

Educational Status Some graduate wort

N

Extent of

43 .48 53h 54.0

139 51.7

male co-head

of household

Clerical, sales,

or craft

Blue collar

44.3

43.9

115

43.9

82

27 . .5

60 66 40 53.6 44.4 3.5.7 N 28 --18 14 ---------·��----------------------------------------bides • few pasons from otber occupatioos . h I ncIudes respondents &om both the probability sample and the supplement.

preference(%)

54 Standard American Kinship: A Speculation about its Sou rces

The unexpected discovery of a new measure of coll aterality stimulates queries abou t the sources out of which it has arisen. The tendency for the Standard Ameri can model to be characteristic of middle-class Protestants in denomi. nations founded during the Refonnation era leads to a speculation that this model may have emerged as family capitalism (and Protestantism) took hold in Western Europe and in Am erican society. With the rise of the bourgeoisie, relationships with kin seem to have been affected markedly both on the Eu­ ropean continent and in England. Maranda (1974) documents chang es in French kinship terminology with the decline of the freresche and the expan­ sion of commerce. He finds that ''the lineal dimension of the French system became accentuated through time by segregati ve terminological refmements and by an increasing stability (p. 137)" and that Hthe system of jural nonns emphasize s the rights of the line al nucleus and minimizes those of collate.rals and affmes (p. 136)." In E ngland , the transformation of ties between kin followed a similar pattern. Stone (1975) notes that by the Eng lis h revolution of the 1640s, cous inage had become "'hope l essly fragmented .'' As trad e in­ crease d in importance, the state became stronger, and its fu nctio ns en­

croac hed more and more upon the kinship domain. Simultaneously, geo­ graphical mobility "'among the lower, lower-middle and middl e -clas s groups .. . was remarkably high (Sto ne , 1975, p. 22) . " Yet, "The growth of pater­ nalism was deliberately enc ouraged by the new Renaissance state (Stone, 1975, p. 54).''lndeed, "The State was as supporti v e of the patriarchal nu­ clear family as it was hostile to the kin-oriented family; one was the buttress and the other a threat to its own increasing power (Stone. 1975, p. 55).,. It is this form of family organiza ti on that is identified with the Protestant refor­ mation (and the emergence of capital i sm) and that se em s to have provided the basi s for contemporary domestic life styles in Western Europe and the United States. The Ameri c an family has had deep roots in the New E n gland tradition that

emerged out of Puritanism. Even in the years fo llowi ng the Revolution, the i ns t itutio nal life of such communities as Salem, Massachusetts, resembled that of the earlier Puritans in various ways. For example, the old Puritan conception of an elitist ru l ing class of mercha n ts still lingered in Sa lem community organ i zati on (Farber, 1 972). Today, the middle- c las s descendants of these early American famili e s still '·associate the m se lve s with e vent s that are of historical significanc e for the community'· through memberships in

hereditary associations (such as the Daughters of the American Revolution) '' and pass these memberships, and their concomitant status, to others in the fam i l y (Rosenfeld, 1974, p. 407)." In a hi ghly industral ized society, the.se hereditary assoc i at i on s ''are enclave s of ascribed s tatus for mobile famil ies,

55 d

[they]

perpetu ate t he symbolic v alue of f amily biographies (Rosenfeld,

�74, P· 40S)." ysis of elite American fam ilies, S avet h (1963) has indicated th at

In an anal historical backgrou n d of the family affects "life style" of individuals; be ily connections provid e a basis for power; and through the individual's

�

with ancestors or in volvemen t n the management of reso urces , "dentification family defi nes one's status in society. More specifically, Beale has sug­ f i tradition of public service and noblesse oblige led Theo­ gested that a am ly dore Roosevelt into public life and induced him into a ''lifetime concern over ubl ic questions (Beale, 1964, p. 166). ''But not all kin are equally important

i

�

p

and power. As Saveth (1963, p. 259) proposes, "It is the achieving i ndivi du als within the achiev ing family who become family history.,' The biographies and genealogi cal location of suc h kinsmen become part of the symbolic property of the mod em American famin establishing a style

of life, stat us,

ily. In discussing his own family hi story, St ewart Al sop ( 1968, p. 40) explicates the role of a symbolic estate in establishing on e ' s personal identity and place in a society which em phasizes achieved status over ascribed s tatus: Knowing something about his ances tors . . . gives a being

part of a

man a

satisfying sense of

continuum, of a process of birth, death, and rebirth that started

long before he was born, a nd \\'ill continue lo ng after he is dead. And as a man ·s

age increases,. this sense of being part of the endless human parade

through history is an oddly comforting sense.

It is reassuring, somehow, to know that all ladies and

rum

sorts

of people-poets and eccentric

importers and reactionaries and politicians and murderers and

\visemen and fools-have gone before In such times as these, it is also com­ � foning to know that, barring some final act of human idiocy. all sorts of people

are

still to come after.

For individual f amily histories to persist as oral traditions, there must be some inte rest and commun icati on among family members. Indeed. F irth ( 1956, p. 28) reports that in London kinship information does not appear to be th e property of an indivi du al only. Instead, t his information �'tends to be drav;n from and c on tri but ed to a household pool.'' M o reove r , ''The contem­ JlOrary f amily has a special archival function as a repository of [memen tos serving as] identity symbols wh ich compose a biographical museum or even [f a am ilial] hal l of fame {Weigert and Has tings, 1977, p. 1174)., The impuls e to communicate inform ati on about kin seems to be strongest . lD the · k. m1 ddie class. By way of contrast, the shallowness of knowledge about 10 and friends among lo\'1· soci oe co n omic individuals h as been indicated in

56

various contexts. In Mexico, Oscar Lewis (1967, p. 12) reports that �.in some villages, peasants can live out their lives. without any deep knov..·ledge or understanding of the people Vlhom they 'know ' in face-to-face relation. ships.,, Similarly, Liebow (1967, pp. 205-206) remarks in his study of street comer men that ' 'a man may have detailed knoYi ledge of his friend's present circumstances and relationships, but little else. '' There is the assumption that '�there's a skeleton in every closet (Farber, 1971 t p. 1 11). '"Studies of kinship among lower-class socioeconomic status individuals evoke such statements as, ''I don't know much about my grandfather and I don't much care,� or, '-All I kno\v is that grandma died when mama \vas little� and she doesn "t te]l me much about her life. She keeps it pretty much a secret (Farber, 1971 � pp. I 1 2- 114). '' Despite an occasional hero� there seems to be little pooling of biographical information in lower-class families in America. This lack of pooled kinship information would inhibit the development of a comprehen­ sive family history by individuals outside the middle class. In kee ping with the middle-class Protestant emph as is upon achievement, however, the Standard American model may emphasize "roots·' in one's 0\\'11 family history. It is plaus ible that� in a society that is more interested in promoting notions of self-actualization and achievement� people remain oriented to\�·ard kinsh ip as pe.rsonal hi story . Instea . d of vie\\'ing kinship in terms of its implications for the future, people with a Standard American kinship orientation, may regard their kin (living and dead) primarily as an orienting mechanism, one pertaining to their ov.'n social placement and achievement-and only secondarily as a determinant of the life chances of their descendants from one generation to the next in an ever-changing world.

THE STANDARD A�IERICAN l\10DEL AND THE DISTRIBUTIOI' OF KISSHIP MAPS

I have reported in this chapter the serendipitous finding of the existence of a Standard American map of collaterality. If its prevalence is compared with that of other maps, one can draw implications about the ability of current

lav..·s that pertain to the family, particularly intestacy law, to represent the views of most people. Table 4-6 describes the distribution of cognitive kinship maps in a prob­ ability sample of 723 Phoenix residents. As the methodolog ic al appendix indicates, this sample is restricted to persons, aged 18--45, who are curre ntly married or who had been divorced, separated, or Vv'idowed. (The table does not include the Jewish supplementary sample.} The percentages in Table 4-6 clearly demonstrate the predominance of the Standard American orientation over other conceptions. In contrast to the 44

57 LE 4-6. Kinship Orientations of Respondents in Phoenix rAB Probability Samplea

---

.

·on

Kinship oneotatt ---

parenaela orde�

staJl(iard amencan Ch'i iiaw

Genetic

c.non law

Residual calegor}" Total 11

r·

in the

Percentage

N

16.0 43.5

116

314 1 10

15.3

56

7.7

76

10.5

51

7.0

72 3

100.0

respondent's confonnity to categories of killShip orienlation generally

occur

between

measures

adjacent in degree of collateral removaJ or tint cousins (e.g., Parentela Orders and � American or Ch·il Law; not Parentela Orders and Genetic or Canon Law measures:•. \\'here sucb ties occur, cases are classified i� the_ caaegory v..ilh a smaller degree of colla� �moval of fi.rsa cousins. For any ca�egory. ties represent a mmonty of cases (e.g .. one percent of cases tn Standard American; 10 percenl in Canon La w).

�

iD Table 1, arc

percent of the cases falling into the Standard American category, Parentela

Orders holds only 16 percent, Civil Law 15 percent, Canon Law 11 percent, and the Genetic orientation only 8 percent. Because of the small number of

persons with a Genetic or Canon Law perspective, these two categories �·ill

be combined in the analyses described in the succeeding chapters. The pat­ terns of collaterality are similar in both Genetic and Canon Law maps among those relatives included in the questions on intestacy law used to classify respondents; for these. kin the two maps differ only with regard to distance of first cousins. (See Table 1-2 in Chapter 1.) The finding that the Standard American model is predominant in popular conceptions of kinship also has serious legal implications. It raises a doubt about the appropriateness of the Canon Law, Civil Law, and Parentela Orders models, now in general use in American law, to express the will of the people regarding priorities in case of intestacy. Specifically, this finding has rele­ e or th vanc f e priority given to children as opposed to spouse in intestacy law. nerally, intestacy laws now in force provide for a sharing of the estate by children and spouse. Table 4-7 describes the answers of the Phoenix respon­ dents to the question whether, in case of intestacy, one's children or one's sho uld have prior claim to the estate. Support for current intestacy spouse law is found only among persons �·ith Genetic and Canon Law orientations cen <52.5 per t); individuals holding other orientations favor the spouse. Per­ sons in the Standard merican category are the most inclined of all to give A the spouse priority over the children. This fmding on priority of the spouse is consistent with investigations on

�

58

TA BLE 4-7 . Children \'ersus Spouse

Kinship Orientationa

in Priority of Claim to Estate, by Kinship orientation Genetic

Priority in claim to estate Children should have pri or

claim (%)

Ratio to total

Parentela

Standard

Ci vil

model or

orders

american

law

canon law

6.0

1 1.3

10.3 1 . 34

5 .7 . 74

.78

Tota] 7.7

1 .47

Children and spouse should 19.8

have equal claim (%)

Ratio

to total

20 . 7

. 70

Spouse should hav e prior

claim (%) Rati o to total

.73

73.7

69. 8

1 32

"' Includes cases in supplementary sample . Cbi

square

1 . 07

63 . 8

1 . 15

1 . 00

362

83

1 . 09

�

30. 2

=

67. 5 16; d . f.

=

6;

52.5

28 . 3

36 . 2

64 . 0

1 . 86

. 57

141

718

p < .00 1

heirs in wills. These s tud i e s show that , overwhelmingly , peopl e leave their estates to their spouse, with the eventual d ispositi on of the property at the spouse 's discretion (Sussman et al . , 1970; Simon et al . , 1 980, Gluck­ man, 1 976) . Hence at least with regard to the choice between children and spouse as heirs , persons w ith a Standard American orientation seem to be clo ser than those with other kinship maps in conformin g to curren t in herit ance norms . Given the prevalenc e of the Standard American orientation (sho\'/fl in Table 4-6) , this fmding may have practical implications . If additi onal research supports the tendency for testation to follow the Standard Ameri can map of priori ti es then legi slators should take this fact into account in future revisions of intestacy law . choice of

,

,

COMPONENTIAL ANALYSIS AND STANDARD AMERICAN KINSIUP

The partitioning of terms applied in American kinship into various compo­ nents has in the past con bib u ted to our understanding of the cognitive di· men si ons that organize this terminology into a coherent semantic pattem .

59

Selby , 1 968, pp . 1 8 1- 1 90 . ) An application of componential B uchler and s of collatemlity may yield significant insights into the ysi s to measure rd ension s structuring the Standa American m odel . their componential analysis, Romney and D 'Andrade ( 1 964) have iden· d three cogn itive dimensions underl ying the classification of kin terms fie !110 Am eric an society. These dimensions are : (a) reciprocity (i. e ., compleentary terms used by two relatives to refer to each other-such as aunts nieces or grandparents and grandchildren); (b) generational distinctions between relati ves (i. e . , the number of generations between Ego and a relative in a superio r or inferi or generation) ; and (c) degree of collateral removal (i . e . � l ateral distance from Ego "s line of descent in any given generation­ brother v ersus cousin , father versus uncle, son versus nephev.' , and so on). The res ults of the Romney and D'Andrade analysis have been verified by Wexler and Romney ( 1 97 2) . Insofar as there i s an isomorphism between tenni nological usage and models of collateral ity , one \\'ould expect these models to reveal comparable dimensions . John Atkins ( 1 974) has demonstrated that the i and j dimensions applied in collaterality measures can be transposed into dimensions derived from componential analysis . Specifically , he has translated i and } into dimensions of generational differences and degree of collateral remo val. Generational difference (G) between Ego and a relative can be expre.ssed as U - i ) , 'N'hile . (

se;: :a 1� :d

degree of collateral removal (R) is given by i or j, Ylhichever is smaller. These transformations pennit the presentation of an orthogonal kinship grid based on G and R instead of i and j. Table 4-8 presents a grid of ki nship relations based on the dimensions of generational differen tiation (G) and collateral removal (R) . In its essential features the grid appears to correspond to diagrammatic representations of American kinship tenninology yielded by compone ntial anal yses (�Tallace and Atkins , 1 960; Romney and D 'Andrade , 1 964; \\lexler and Romney, 1 972) . Presumably , like these analyses, measures of collaterality also vary with differential weights assigned to G and R and �·it h the introduction of additional dimensions . For t�ose models i n which i and j play symmetrical roles , the transfor­ matio ns from i and j into the dimensions of generational difference (G) and co llateral removal (R) are described by Atkins ( 1 974 , pp . 1 82 1 ) as fo llows : Canon Law model

=

G

+

R' ·

Ge neti c model = G + 2R - f (where f = 0 for lineal kin or half siblings, and f = 1 for all other kin) (This model is designated as the Murdock degree by Atkins. ) Civil Law mod el = G + 2R .

60 TABLE

4-8. Grid of Kinship Relationships Based on tbe Dimensions or Generatiooal Ditreretiation (G) and Collateral Removal (R)

Generational differentiation

Collateral removal (R)

First step of collateral

Second step of

collateral remo\1 al

(G)

Direct line descent

+2

Grandparents

+1

Parents

Aunts and uncles

Parent s ' first cousi11

Ego

Siblings

First cousins

-I

Chi ldren

Nieces and Nephews

First cousins ' children

-2

Grandchildren

Grandnieces and

0

removal

Oreal aunts and great unc le s

grandnephews

Grandparents ' first cousins

First cousins '

grandchildren

The diversity of semantic dimensions found through componential an alysis suggests that in discriminating among degrees of relationship , the various measures of collaterality combi�e the G and R components in different ways . An examination of the above measures of collaterality indicates that: 1 . The Canon La"v model treats G and R as equivalent; partitioning of kin is based on gradient distances from Ego . through the nuc lear family , and then outward from there . Thus , immediately outside the nuclear family , no distinction is made among grandparents . aunts and uncles , first cousins , n ieces and nephews , and grandchildren; all of them are two degrees distant from Ego . The next ring of relatives around the nuclear family is three degrees distant , and so on. 2 . The Genetic model apportions to collateral re-moval {R) twice the \\'eig ht of generational differences (G) in assigning degree of distance between Ego and a relati ve . but then i t weakens this weight by f. For close kin (i . e . , where i or j = 0 or 1 ) the degree of relationship computed with the Genetic measure is equi valent to that obtained with the Canon Law measure . But the larg er that i and j become , the closer is the Genetic degree of relationship to that of the Civil Law model . Hence , the Genetic model apparently stands between the Canon Law and Civil Law models in extent of stress on collateral removal and generational differentiation . 3. Assigning a double weight to R as compared with G in the Civil La»' model enhances the role of col l ateral removal in partitioning kin by degree of relationship . The allocation of we-ights in the Civil Law component seem s to offer a compromise between the nuclear family emphasis of the Canon

61

od and the stress on l ine of descent in the Parentela Orders and m el r..aw dard American models . The core of the Civil Law component consists ,

s;a;go lineal links in his family of orientation and family of procreation . l inks place both parents and children immediately adj acent to ;esein ddirect 1 and R 0) . Those kin ju st outside this abbrevi ated i stance (G E th en located one unit away ; thus , Ego 's siblings (G 0; R 1!e al core uiv ale nt to Ego 's grandparents and grandchildren (G 2; R = •5

0

=

=

=

are

= l ) are eq O) in being two degrees distant from Ego . Next in remoteness from this core are t he great grandparents , aunts and uncles , nieces , and nephews , and gre at­ grandchildren , and these are still an additional step away from the abbrev iated descen t line (or three degrees from Ego) . In general , the degree of remoteness of any class of kin from the abbreviated descent l ine is defmed by this ===

component .

The transfonnation of the Standard American and Parentel a Orders models

(G-R) grid is complicated by the fact that in these models i and j are asymmetrical . S inc e j repre sents ascent from Ego into the generational-collateral

and i descent, the translation into generational differentiation

U

-

i)

and

collateral removal (i or j , which ever is smal ler) de pen ds specifically upon whether a particular relative is in a superior generation as opposed to a subordinate generat i on in relation to Ego . The transformation of the se models into G-R terms requires a speci al weighting for th is disparity . In the translatio n of the Parente l a Orders model , three special symbols must be introduced , S, A, and m . Let S refer to the comparative generation of Ego and kin . If the rel ati v e is in an a scending generation . then U >i) ; if the rel ati v e is in a descending generation , then (i >j) . If j is greater, then let S = 1 ; if i is greater , let S = - I ; and if i = j , then S = 0. Furthermore , let the quantity ( I + S)/2 = A . l In the transformation , as in the formula [i + (m + 1 )j ] , l et m refer to the largest i in the set of rel ati ves for whom the comput ations are being performed . The derived formu l a for Parentela Orders is a modificatio n of the Civi l Law transfonnation (G + 2R): G(Am + 1) + R(m + 2) , or GAm + G + Rm + 2R, or (G + 2R) + m( GA + R) . The trans formation of the Parentela Orders fonnula into the orthoganal gene rati onal- collateral form hence produce s a second component [m (GA + R] in additio n to the dimensions in the Civil Lav.' scheme (G + 2R) . The

major elem ent in thi s second component (apart from m ) is the in c orporation �f generation al ascendance versus desce ndancy vis-a-vis Ego into the model

(I. e . ,

A for ascending ge nerati o ns

=

I ; A for descen di n g ge nerat i on s = 0) .

62

The transformation of the Standard American formula [(n + l ) i + j] i s comparable to that of Parentela Orders . The o nly difference pertai ns to the s y mb ol s A and m . For the Standard American translation , let the q ua nti ty ( 1 - S)/2 = D . Then D will = 1 when i is l arger than j , and D will === 0 when i is smaller; in addi t i on , let n refer to the largest } in the set of re l at i ve s for whom the co mpu tation s are being performed . The transformed v e rsion of the Standard American model is as follows :

(G

+

2R)

+

n (GD

+

R) .

The co mponents for the Standard American model thereby resemble those for the Parentela Orders model . Both cons i st of the Civil Law co mpon en t and a unique factor . Distinctions refer primarily to the re vers a l in treatment of ascending and descend in g generations in the fonnulae. The i mpac t of the u niqu e components in the Standard American an d P ar­ entela Orders model s upon c l ass ificati on of kin i s indicated in Table 4- 9 . This table separate s the contributions o f the C iv il La\\' com ponent and the unique Parentela Orders and Standard American com ponents in producing priority rankings . The Parentela Orders c omponents are di s cus se d first , and afte tvlard s those for the Standard American model . A rev iew of the components of the Parentela Orders priorit y scores in Table 4-9 d i scl ose s that , with the effect of the Civi l Law deg ree removed , the remaining compo nent derives directly from a hierarchy of Parentel a Or­ ders (P) . The latter com P.Onen t , m (AG + R), can also be e x pres sed in terms of j. S inc e R is zero for a direct a nce s tor, and since A = 1 in ascending generations , any parente l a class can be identified by (G + 1 ) . But inasmuch as [R = 0, G = j] refers to a direct ancestor , a parentela class can also be identified by U + 1 ) . As the verbal description in introductory chapter attests , the ordering of any p arente l a P thereby depen ds upon the ge nerati ons of Ego 's direct ancestor, who is then ' "head ' ' of that parti c u l ar descent line . In the table, the unique component scores for a de scen t line are the same as that

for i t s ' 'head . ' ' Within a p arente l a clas s , only the Civil Law c ompone n t

varies . A feVto' illustrations from Table 4-9 may clarify the re l ation sh ip betv.'een the. Civil Law and P arentel a Orders components . For great-great grandparents , G = 4 and R = 0. The Civil Law component, G + 2R , produces a sco re of 4, and ( since here m = 3) the Parentel a Orders component , m (AG + R) , y i el ds a sco re of 1 2 . The total priority rank i ng i s (4 + 1 2) o r 1 6 . Rem ov i n g the effect of m , we find that 1 2/3 = 4, and (AG + 1 ) = 5 , which tel ls us th at the great-great g randparent s are at the head of the fifth parentela set . Looking down the unique-component co l u m n , we find that, like the great-

largest j

B�ausc the largest i

c Because the

b

-

=

(G)

2 2

n -

4.

is an absolute number whereas

4 (Great-great-grandparents) .

that

-I

-2 0

1 1

(G• ) =

3

I I

(SG). in which S

3 4 4 5

2

4

2 3 5

3 2 1 I

4

1

0 -I

2R)

(3)

+

2 I

(0

Civil 1aw component

0 1

0 0 0 0 0

0

( 2)

(R)

removal

Collateral

-3 3

-2

3 2 I -1

4

( 1)

3 (Oreat-grand�hildren) . m = 3 .

" (G•) diffen from (G) in descending generations.

First cousin 's child

Grand niece First cousin

Niece

Sibl ing

Great-great -aunt Great aunt Aunt

Parents Children Grandchildren Great-grandchildren

Grandparents

Great-grandparents

-

Great-great-grandparents

Relatives

-- -

Generation (U•)a

==

Unique

+I

S

II

10

9 5 6 7

17 13

2 J

1

==

0 for

0

-

.

R))c

collateral senerations ,

8 12 R 12

4 4

4

12

14

8

4

0 0

0

16 12 8 4

+

(6)

(n(OO

component

Unique

and S

16

=

12 17

II

7 6

8

9

IS

3 2 I s 10

4

-

1 for

(3) + (6) = (7)

ranking

Total priority

Standard american

(3) +(4)= (5)

Total

priority ranking

ror ascendina generations,

6 3 3 3 6 6

12 9

0

0

3 0

6

12 9

(4)

component (m(AG + R)b

.. .

Parentela orders

TABLE 4-9. Priority Kankings for Various Relatives Yielded by Parentela Orders and Standard American Models, with Scores Jlecomposed into Civil Law and Unique Components

0\ w

64

great grandparents , the great-great aunt also has a component score of 1 2 . Her Parentela Orders component score ( 1 2) i s obtained by the figures [ 3(3 + 1)], where m = 3, A = 1, G = 3, and R = 1 . Since the great-great au �t is a descendant from the great-great grandparents , she is indeed a me m ber of the fifth parentela and differs from the grandparents only in the. Civil La\\' component of her priority score .. 5 instead of 4. In Table 4-9 , members of the second parentela, with the pare nts as ' 'heads , " inc lude siblings , nieces, and grandnieces-al l of whom have a 3 in the Parentela Orders component column . For the parents , (AG + 1 ) = 2 , and for th e remaining members o f this set, (AG + R + 1 ) = 2 . Note th at for relatives in descending generations (R + 1 ) is equivalent to parentela membership, 2. Again, variations among total priority rankings within a given parentela class issue only from the Civil Lav.' component . The Parentela Order component thus expl icitly c lassifies kin according to the descent lines from Ego and his ancestors: (a) Ego 's own line of descen­ dants , (b) Ego ' s parents � line of descendants (except for Ego 's ov.'n line) , (c) Ego 's grandparents line of descendants (except for Ego 's parents � line) , (d) and so on . As noted earlier, these classes are based on ) , the remoteness of direct-l ine ance stors from Ego . The Standard American component has a somev.'hat different effect upon the establishment of classes of kin in the measurement of coll ateral ity . Th is component categorizes kin on the basis of i, the remotene ss of descendants from Ego and Ego 's ancestral l ine . For example, in Table 4-9 , the unique­ component wei ght for all of Ego 's direct ancestors is zero ; the weight for the children of Ego and each ancestor is 4 (since n = 4) , and this category includes Ego 's great-grandparents ' child (i .e. , Ego 's great aunt) , his grand­ parents ' child (who is his aunt) , his parents " children ( i . e . , Ego 's own sib­ lings) , and Ego �s own children . For grandchildren of Ego 's ancestral line . the weight i n Table 4-9 i s 8 , and the line 's great-grandchildren are assig n ed a 12. Hence , the Standard American component is defmed by the number o f

generations between Ego 's line af antecedants and c ategories of progeny­ children , grandchildren , gre.at-grandchildren--<:lassified with reference to that line . As in Parentela Orders, the d ifferences in total Standard American priority rankings among the various relati ves within each group accrue from their scores on the Civil Law component . The existence of unique components in the Parentela Orders and Standard American models impl ies that the differences in kin statuses in Civil Law are complemented in these models by categories derived from the distinction between ascending and descending generations (i . e . , S in the componential analysis) . For Parentela Orders , this d istinction results in the partitioning of classes of ancestors and their individual line s of descendants . In the Parentela Orders component, the distance between Ego and any collateral kin is defined

65

ancestor

efore specifically in te rms of how remote their nearest common Ego . For the Standard American model , the se parati on between as­ IS ding and de sce nd i n g generations has a differe nt set of consequences . The S tandard A meri c an component organizes kin solel y on the basis of stance from Ego 's ancestral line, and it defi n es degree of kinship by rethis l i ne . At some point of remov al from the ancestral l i ne , morene ss from i re sumably ki n s hip is ext nguished . I n this respect the S tandard American odel di ffe rs from Paren tel a Orders , which defines line s of descent as the­ oretically infinite . In summary , the c om po ne n ti al analysis of th e models y ields three cumu l ative components in cog nitive kinship m apping:

��ro ce�que � �

1 . A component of g rad i e nt distance of individual relatives from Ego and the n uclear family . This c om po nent is further decomposable into the di­ mension s of (a) ge nerati onal differe nce be tween Ego and a rel at i ve and (b) collateral removal between a rel ative and Ego 's line. of desc ent . It is symbolized in the Canon Law model . 2. A compone nt modifyi g the d i stance component by doubl ing the weight of collateral removal in kinship m app ing . This modification p l ac e s Ego 's immediate lineal ties ( to arents and children) at the core of kinship . This component is symbol ized in the Civil Law model . 3 . A component identifyi ng classes of kin by their rel ationsh i to Ego 's line of descent . This component analyzable into t wo subcomponents . One subcomponent groups kin on the basis of their removal from Ego 's ances­ tors , and it is symbolized in the Standard American model . The other subcomponent groups kin on the basis of their de scent from Ego or from specifi c ancestors of Ego , and it is sy mbol ized in the Parentela Orders model .

n

p

is

p

The three components in the kinship models represent a series in concep­ tio n of s oc ial space . (See Chapter 3 . ) The frrst component organizes this sp ac e into gradients of social distance eman ating from a c ore-Ego . The seco nd c om o nent modifies the first one som ewhat by s tre ssing degree of

p

�l�a�ral re mov al . The third component carrie s this progression further by d• v•dm g soc ial space into categories based on the relationships of kin to E go ' s ance stral li ne . In the t h i rd component, Ego is significant onlv insofar as he too i s an an cestor . The three components therefore repre sen i differences in con�epti ons

of social space-from Egocentric gradients to corporate cate­ gones . The proposition of the Phoenix study i s th at the various components of soc i al space symbol ized in the kinship mode ls are i somorphi c with kinds

of rel i gious , eth n ic , and s ocioec o nomi c settings in the soc ial structure--com-

66 munal versus segmental settings-and that each model i s associated \V i th a

particu lar kinship ideology consi stent with its setting .

ST ��DARD AMERICAN KINSHIP: A SU)fMARY ..

This chapter has reported an unanticipated finding , the prese nce of a w ide­ spread k ins h ip orientation that has not been previously identified . This ori

..

entation is particularly prevalent among Protestants , but it i s also wides pread among others at high educational and income levels . It is underreprese nted

especially among raci al minority groups , Jew i sh respondents , persons with

foreign-born fathers , and poo rly-educated persons at poverty levels . B oth

religion and socioeconomic status seem important in determining its di stri­ bution . Because of the social characteristics associated with the popularity

of this orientation , I have called it the Standard American model . This model

emphasizes the social closene ss of rel ati ves in ascending generations , and it appe ars to have emerged during the Protestant Refonnation . Family and kin­ ship attributes associated '.'l·ith the Standard American model are described

in succeeding chapters . an

The cognitive structure of the S tandard American model is suggested by analysis of its components . Like the Parentel a Order model , it c an be

decomposed into two parts: the Ci vil Lav.' contribution and a unique com­ ponent . The anal ysis then indicated that the Civ il Law component serves to distingui sh among kin ..,vithin categories defmed by the unique component .

In addition , the unique component derives its qual ities from the fact that , like

Parentela Orders , the Standard American model makes a disti nction between relatives in ge nerations above Ego and those in inferior generations . Bec ause

of this di stinction , the unique component in the S tandard American mod el stresses classe s of rel ati ves according to descent from Ego 's own ances tral line . Accordingly , the c hildren of Ego ' s great -grandparents , those of his grandparents , the offspring of his parents , as wel l as Ego 's own progeny al l constitute one c l ass; the grandchi ldren of each of these ancestors make up another class ; and so on . Th i s accentuation of closeness to Ego ' s ancestral

line is consistent with the characterization of the Standard American mo de l as stressing fami ly history . �bile the Parentela measure also exhibits a uni que component that re lies upon a principle of descent , its cognitive structure is different from that in the Standard American model . In Parentel a Orders , the descent-line component refers not to removal of kin classe s from Ego 's an­

cestral l ine but to lines of descendants from Ego and Ego 's own l ine of

ancestors . Unlike the Canon Law , Genetic , and Civil Law components , which generate gradients from Ego in spatial representat ions , the Standar d

American and Parentela Orders compone nts organize genealogical space into

67 regorical s che mes based upon ancestral lines. The cognitive compone n t s kin sh ip model s thus appear to be symbo l ic referents for the spatia l the ed in C h apter 3 , and the se attri butes shou l d be borne i n 0 etaphor s de scrib w hen the fin dings on family and ki ns hip ties are presented in the chap-

c�

:ind

ow . ters that foll

NOTE

This application of the S factor in tbe componential analysis Atkins (personal communication) . 1

was

suggested by John R .

Variations in Kinship Distances : Fundamental or Superficial?

The re i s conse nsus among social sc i ­ entists that col lateral ity i s assoc iated

w ith priorities in ki n sh i p re lat ions . Sti 11 , the concept of genea logical c lose.ness is not syste-matical ly appl ied either by researc hers or� for that m at­ tee by the people stud ied . For example , t he pattern of Kl atsky 's ( n . d . ) data conform s gc nera l ! y to the Civil Lav..· mode l fo r com put i ng kinship pro x i miti e s b ut departs con side rab l y from the C anon Lav-' or Genetic models . Yet K lat sky 4: n . d . . p. 84) w ri t es that Helasticitie s lin contactj differ systematical l y hy ge nealogic a l di stan ce . " " S im i larl y , Sch ne ider and Cottre l l ( 1 97 5 , p. 94) report that . � � \\li th each genealogic al step aYt·ay from Ego , le ss and less substance [ i . e . . . . b l ood · '] i � s hared � yet , the y arc · "carefu l n ot to provide any de finition of l c lose and di stant relati ves] in te nns of ge neration- col lateral re moval ce l l s o r in s uch tenns a s �u ncle- aun C o r ·first cousi n � ( Sch neid er and Co ttre l l , 1 97 5 , p . 97) . The rea�on for avoi ding a precise de finition ' ' is simp ly that the defi nition varies not only for v ari ous popu lations but also from infonnanl to i nforn1 ant w i thin l thcirJ sample . · � In spi te of the ambigu ity of the co ncept of gene al og ical distance in the studies ci ted , findings indicate that vari ation s in its appl ication are not ran­ do mly di stributed . Schneider and Cottrell ( 1 9 75 , p . 97) find that a series of soc ial vari ables, � 'such as age , se x , class , occupati on , eth nic ity , rel igion , i ncome etc . . . te nd to in fluence the inclusion or exclusion decisions of dif­ feren t kinds of Egos about different kinds of [relatives]-particularly d i stant ones . " Excl uding a relative from one 's � 'kin univ erse " obviously impl ie s th at Ego conceives of that relative as be.ing extreme l y distant . B ut even · '

y,rithin the kin un iverse , Sc hne ider and Cottrell ( 1 97 5 , p. 98) hav e found

considerable vari ation- ' �for some informants , the uncle- aunt category i s c lose ; for others , it is distant . ' ' Other studies support their conclusions . Fo r e xam ple , rel igion seems to play a particularly important ro le in conceptions of genealogical structure . Klatsky 's (n . d . ) data sho\v that Cat holic i sm i s re­ l at ed to kin ship contact mai nl y because the Catholics in her sample tended to l iv e c lose by ; Winch ( 1 974 , p. 158 ) , ho�·ever , has concluded that Chicago­ area ' •Jews were not more familistic [ in ties with relative s] because they were less m i gratory , but that they were less migratory because they were more

69

the Christi an s . ' ' The various investigators thus show how il i sti c than urarions of genealogical d i stance vary in pred ictable ways in Americ an g }cinsh i p . . Although s ocial sci entists agree th at there t s const" derabl e van at1on tn k I n-

!�

·

·

·

·

·

·

·

h ip ties in different segments of the American popu l ation , they disagree over

this v ari ati on . For some , th i s variation repre sents a series � signi ficance ssofions of the same basic kin ship structure , in which differing

of dev i ant expre mode s of ex pression derive from the particul ar social cond itions under w hich rate . For others , the se mode s themse l ves convey diverse. princ iples peopl e ope of kins hip structu re fundamentally d i sti nct from one another . In th is chapter 1 shall exam ine the issue of kinship variation and present data re levant to its

resolutio n.

The po siti on of Schneider and Smith ( 1 973) is that the American kin ship is essentially unitary � v..· ith vari ations accru ing \Vith conditions under v,rhich relati ves play the ir roles as persons . Thei r v i ev..· i s as follows: 1 . ' •At the level of p ure cultural conce ptions in the domain of ki nship . . . . Americans of all classes share the same conceptions about the m anner

in which 'blood � is transmitted and the i mperishable bonds it creates bet\\'een people . . . . At thi s leveL each person in h i s capac ity as a relative or ki nsman is equal to each other relative , and undifferenti ated from h i m . (Schneider and Smith, 1 973 , p. 1 03) . " 2 . "At this cultu ral level v.'e may al so say that Americans share a con­

ception of •the nucle ar famil y ' as the basic configuration of kinship elements out of which the whole domain of kinship is defi ned and differentiated

( S chneider and Smith , 1973, p . 1 04) . , ,

3 . Variations \Vhich occu r in family an d ki nsh ip rel ation ships issue from

the fact that people conduct themse.lves in a number of distinct social do­

main s--oc c upational role , family role , associational membership--and par­ ticipation in these domain s molds kinship ties . "We argue that nonnative role structures for famil y members vary precisely because they are conglom­ erate structure s of this kind , and the most important dimension of v ariation

is in the class differences in sex-role (Schneider and Smith , 1 973 , p. 1 05) . " 4 . · 'Over and above these vari ations in the sex-role component of famil ial nonns is . . . a feature. of middle-class nonnati ve structure that the nuclear fam il itself, y as a unit , is singled out for special emphasis (Schneider and S mi th , 1 973 , p . 1 05) . " Indeed Schneider and Smith ( 1 973 , p . 1 07) point o ut, ' 'One of the most i mportant sets of forces affecting fam i l ial behavior is

the rel at ion ship of individual s and famil ies to the economic system . ' · The position opposed to that of Schneider and Smith i s that variations in ki nshi p tie s c an be traced to ideological factors that often fit the socioeco­

n o mic real itie s but which at times override economic consideration s . In Kin­ sh ip and Class ( 197 1 ) , I described these ideologies as American B iblical and Weste rn American kinship . In the American Bibl ical scheme , norms with a

70

Bibl ical source have been adopted to justify middle-c lass family life , and Western American norms fit economic and ecological contingencies ofte n faced in the lower class . HoY.rever, this earlier formulation is too simpl is tic . Indeed , the analysis of kinship maps in the Phoenix study is intended to supplement that statement about sources of kinship conceptions . We. can interpret maps of coll aterality as expressing sets of assumptions about the nature of the connection between domestic institution s and gene ral social structure. The character of these assumptions is suggested by the soci al �tructure in which each map has crystallized historical ly. The previou s chap­ ters have proposed the fol lowing: 1 . The Parentela Orders orientation appears to have emerged in pluralistic settings (e . g . , Biblical Hebrev•s , classical Greece) . The ideology upon whic h this orientation re sts is that kinship is aimed at the perpetuation of the special political and/or religious interests embodied in the mi..fchpokheh . the house , or the oikos . 2 . Indications are that Standard American kinship principles were for­ mulated as Protestantism blossomed and the merchant class dominated eco­ nomic and political life in Western society . The Standard American orien­ tation appears to be associated with high achievement motivation and therefore with the perpetuation of spec ial socioeconomic interests . 3 . The Civil La\v map represents a compromise in the Twelve Tables betY.'een the plural ism of the Patricians and the universalism of the Plebes at the founding of the Roman Republic . Its principles thus seem to express a middle gnJund in the pluralism-universalism continuum . 4 . The Canon Law measure was devised in the 1 2th century to express the aims of the Church toward universal ism in spreading the faith over vast domains-with centralized authority and a common body of law . The main vehicles for the perpetuation of society were organized to promote the com­ mon interests of the entire social organism . Resolution of the issue �·hether American kinship rests upon a single cul­ tural base or upon multiple sources depends , in the final reckoning , u pon three kinds of data: 1 . The first kind of data refers to religious and philosophical ideologies that reveal the structure of group values and their fundamental assumption s about the ·nature of society . If it can be shown that variations in religious belief are associated with conceptions about collateral ity , such findings wou ld lend support to the multiple-source position regarding American kinship . 2. The second kind of data pertains to the predominance of the •nuclear family ' ' as a unit si ngled out for speci al emphasis among middle-class pop­ ulations . If the middle-class chooses a set of priorities among kin that em­ phasi zes the unity of the nuclear family , the data would support Schneider and Smith; however, if they tend to give. priority to lineal relatives outside the n uclear family (e . g . , grandparents or grandchildren) over brothers and 4>

71

the unitary conception of American kinship would be called into si sters , then question . It would be difficult to justify the view that the nuclear family co n stitute s ' 'the basic configuration of kinship elements out of Ylhich the whol e do mai n of kinship is defined ' ' when the putative major protagonist of t hi s po sition-the American middle class�ev iates from it. 3. The third kind of data bears upon familiarity with specific relatives as compared with rel igious identity as factors in kinship mapping . In particular, grandparents are ranked as especially close in the Standard American mode. I , and it Ylould be of interest to determine whether familiarity with grandparents is m ore effecti ve than religion or ethnic origin in explaining conformity of i ndi vidu als to the Standard American model . The materials which follow fall into four parts: (a) rel igious identity and kinship orientation ; (b) occupational status and kinship orientation; (c) income and minority-group status; and (d) assimil ation and the Standard American model . RELIGIOUS IDENTITY AND KINSHIP ORIENTATION

Religious belief can be regarded as a way of symbolizing the structure of society (Durkheim , 19 1 5; S wanson , 1 960) . Briefl y , religious beliefs seem to ' 'arise as symbols of men 's experiences with the basic purposes and decision ­ making procedures of society and of enduring and independent groups with­ in societies ' • (S\\'anson , 1 967 , p . viii) . In preliterate societi es , people \vho aim at the unification of the society tend to promote the existence of a "high ' ' god, that is , a god Y�'ho transcends all speci al interests o f groups and who \\o'elds the divergent factions in ths society into a unified , coherent \\'hole . Societies with multi-dieties , hoY�·ever, re.main factionalized , Y.'ith each group alienated from the central state or other competing collecti vities . In modem soci eties , where tradition dictates the existence of a single diety , different groups may fix upon the special characteri stics of this diety . In effect, these differen ces in defin ition of the nature of the overal l diety are equivalent to the \\'ors hip of spec ial gods . Consequently , modern society itself may be said to be governed by specialized gods while it voices monotheism . God as de fi ned by the Jews differs from that of the Christi ans; the god of the Ref­ onnati o n Protestants differs from that of the Neo-fundamental ists or the Pie­ ti s tic Prote stants ; and the god of the Protestants differs from that of the C athol ics-not in al l w ays , of course , but sufficientl y to generate sentiments for perpe t uating distinctions among the more divergent views . By stres sing theological d istinctiveness , rel igious tracts ordinarily promote t � e perpetuation of a unique identity of rel igious minorities and sects . This hteratu re also stresses the i mportance of the family in maintaining ties and cre ati ng boundaries for the rel igious group. (See , for example, Schlesinger, 1 974 . ) Schneider and Cottrell ( 1 975 , p . 67) suggest that "religious affi l iation

tends to be treated as a b adge or a mark of ide ntity of the entire fami l y , [ and] . . . within the Jewish and

C atho l ic

traditions, there is a

and not count me mbers wh o marry ' ou tside the

faith'

tendency

to 'dro p •

' ' Insofar as fami l y

iden ti ty operates as a vehicle for maintai ning boundaries of distinct rel ig io us

communities , kinship orie ntat i ons that emphasize centripetal tendencies would facilitate

their perpetuation .

that the kin­ would differ from that of

Hence , one w oul d an t i c i pate

ship orie ntatio ns of Jews and Protestant sectarian s de nominational Protestants and tha t these , i n tum , would be dissimilar fro m

C atho l ic s .

Kinship and Religion: Findings

In the a nalysis of data , the re l igiou s iden tity re ported by the respondent was cro ss-t abulated with his (or her ) type of ki nshi p orientation . It was antic i pated that separatist rel ig i ou s gro ups woul d be overrepresented in the Paren te l a Orders category, while groups favo ri ng universalism would tend to do m in ate

the Genetic and Canon Law categori es . It was also expected that intermediate reli gio n s , which aim for a moderate separati sm through associational ob l i­ ga tion s (rather than communal ti es ) , are ov e rre presented in the Standard American and Ci vil Law categorie s . Whether one uses current r elig i ou s iden­ tity or rel i gio n in which the responde nt was rai sed , the res u l ts with regard to kinship orien tati on are s i m i l ar ; c urrent ide ntity is chosen arb i trari l y for the analysi s . Parentela Orders . "a\s

Tab le 5- l

shows , Jews and Neofundamental ist

Protestants tend to fall into the Parentela Orders category disproportionately . The ratio of the per cent age for Jews to the percentage of the to tal sa mpl e in the Parentela Orders category is 1 .47 , and i n

the

c as e of the Neo -f un dam e n­

talists , t he ratio of 1 . 59 . The Catholics are the only others with a ratio g re at er than one in the Parentela Order s category . All three of th ese rel igio us groups rely to a con siderab le extent upon communal ties (as opposed to associational eleme nts) for their coherence as collectivites. Especi ally the Jews and sec­ tarian s strive to pe rs i st as au tonomo us re l i giou s ' 'communities ' '-setti ng themselves apart in identity . Variations among the Jews , Monnons , and Cath­ o l ic s will be di scus sed in succeeding paragraphs . Amon g Jew s , differences i n kinship o rientatio ns are found among the Orthodox , Conservati ve , and Reform branc h es . As Table 5-2 i n dicate s , per­ sons raised in Orthodox homes show a far g re ater pro c li v ity to the Parentela Orders model than do individuals from Conservative an d Reform families . This fi nd i ng appears for both the respondents in the current Phoe nix study and in t he previous K an s as City an al ys is (Farber, 1979) . In contrast, persons b roug ht up in Reform Judaism are most likely to develop Genetic or Canon Law kinship orientations . With t he dec line of Orth odo x Judaism in the Un i ted

73 ABLE S- 1. Current Retigion of Respondent and Kinship Orientation T Kinship orientation

Genetic Parentela

curre nt religion

orders

Jewish ( %)

25 . 9

R atio to total Neefundame ntalist Protestant ( %) Ratio to tOial Pietistic Protestant (%) Ratio to

total

Reform ation Protestant (%) Ratio to total

General Protestant (% ) Ratio to total Roman Catholic (%) Ratio to total Mormon (LDS) (%)

Ratio to tOlal N o Religious Preference

(%)

Ratio t o

TOial •

Excludes

lotal

Standard

american 42 . 6 .91

1 . 47 27 . 9 1 . 59 1 4 .9

44.2

.95

47 . 8 1 . 03

. 85 1 3 .7

5 1 .6 1.11

. 78

17 . 0

50 .9 1 . 09

. 97 19.6

43 . �

.94

1 . 14 10.3

43.6

.94

.59 1 6.0

50. 7

1.09

.9 1 1 7 .6

all religious categories with fewer

46 . 6 dlaa 30

responde.-s .

Ci\'il

model

or

law

canon law

N

1 4. 8

16.7

54

.91 20.9 1 . 29

1 7.9 1 . 11 17.9 1.11

. 85

7 .0

43

1 9 .4

1 34

. 36 .98

16. 8

95

. 85

1 3.2

18. 9

53

14. 3

22 . 2

1 89

.81

. 88 28 . 2 1 . 74 13.3

.82 1 6 .2

. 96

1 . 13

1 7 .9

39

.91 20.0

1 .02

1 9. 7

15 717 a

States , current Jewish identification tends to be restricted to Conse rv at iv e and Reform branche s . But whether current afftliation or branch of upbringing is examined , the results are similar; the more traditional and separatist branches of J udai sm tend toward the Parentela Orders orientation and Reform Judai sm sh ows a greater concentration in Genetic and C anon-Law model s . Alth ough , as a group , Mormons do not exhibit a Parentela Orders kinship orie ntation , the conception of ' 'celestial marriage ' ' in Monnon theology seems co nducive to that perspective . Like Orthodoxy in Judaism, the Mor­ mons who are most conformist in their bel iefs are expected to lead a family l ife in accordance with religious law . In order to qualify for a temple we d ­ ding , Monnons must not only be members in good st an d ing and believe in the ch ari smatic a u thority of the church, but they must also be obed ie nt to churc h law . Because of the rigorous demands on the ' 'orthodox • • Monnon l ife sty le , only a minority of the Mormons have their marriage s sealed in the chu rc h ( Campbell and Campbell , 1 976) . H aving one 's marri age sealed in the temple represents a major step in hievi ng celestial marri age throughout eternity . ln Mormon theology there IS an i so mo rp h i sm between earthly and heavenly family structure: "The or-

�

74 TABLE S-2. Brancb of Judaism and

Ki nshi p orientation

Kinship Orthodo�

Branch during upbringing

Orientation

Branch

of Judaism

Conservative

Refonn

25 .0 4 1 .7

Parentela orders (%)

37 . 5

21 . 1

Standard american (%)

37 . 5

52 . 6

Ci vil law (li )

Genetic or canon law ( %)

N

Current

religious

(%)

Standard american ( %) Ch·il law (%) Genetic or canon law (% ) N Because of

for lhe remaining

8

15.8 10 .5

33 . 3

19

12

3 1 .3

43.8

23. 1 42 . 3

25 . 0

3.8

identity

Parenrela orders

Note:

25 .0

an error c ases .

30 . 8

16

in insblletions to an interviewer, information on brancb

26

of Judaism wa s o ot obtained

ganizational structure of heaven is the extended family-kinship netvw'ork (Campbell and Campbelt 1 976, p. 386) . ' ' According to the Mormon con­ ception , · 'a man must enter into the ' new and everlasting covenant of mar­ riage ' by which he and his wife or wives will be marri ed for all eternity and will have the privilege and duty of procreating spirit children throughout eternity even as God procreated us (Campbell and Campbell , 1 976 , p . 385 ) . " Consequentl y , it is incumbent upon the devout Mormons to have a temple marri age , ' 'which is sealed by the Holy Spirit of promise (Church Educational System, 1976, p . 1 98) . " If one interprets Mormon doctrine as an attempt to justify a distinctive identity of Latter Day Saints , then one would anticipate that conceptions of kinship wou ld be affected . Spec ificall y , devoutness among Monnons , like Orthodoxy among Jews , would be associated with an emphasis upon line of descent in views on kinship organization . This anticipation is borne out in Table 5-3 . Although 20 percent of those persons whose marriages had been sealed in the temple fall into the Parentela Orders category , only 7 percent of the other Mormons are in that class . Conversel y , none of the persons with sealed marri ages conforms to the Genetic or Canon Law model s in responses , but fully 24 percent o f the others do . Hence , despite the small number of cases involved , the findings lend additional support to the putative relation­ ship between religious sectarianism and emphasis upon l ine of descent in conceptions of kinship .

Standard American Model . Table 5- l shows no marked tendency by any major rel igious group to be concentrated i n the Standard American category .

75 ABLE S-3 . Kinship Orientation and Sealing of Monnon Marriage bt Temple T for Time and Eternity (Mormon Respondents) Marriage

tati on Ki nsh ip orien

parentela orders (%) Standard american (%)

Civil law (% ) Ge-netic or canon law ( %) N

sealed in temple

Marriage

not sealed in temple

69

20. 0

..

40. 0

44 . 8

40. 0

24 . 1 24 . 1

10

29

Only denominational Protestants and those with no religious preference reveal

an y s peciaJ procl ivity at all for the Standard American pattern . Presumably , con formity to this pattern depends also upon a range of other variables .

Civil LaM-· Model . Religious groups that have strong associ ational com­ ponents for establishing cohesiveness in promoting their special interests are overrepresented in the Civil Law category of kinship orientation . ( See Farber, 1 979. ) As Table 5- 1 shows , these include many (but perhaps not the most devout) Mormons (with a ratio of 1 . 74 over the expected percentage) and Neofundamentalists (here by a factor of I 29 over the expected perce ntage) . Among Neofundamentalists , data (not reported here.) indicate that members of Pentacostal sects tend to fall into th is category (as opposed to Southern Baptists , who generally hold a Parentela Orders orientation) . Inasmuch as many Monnons and Pentacostal sectarians seem to minimize ethnicity as factors in their lives , the church itself may become the focal point of their activities and their commitments . To a lesser extent, denom inational Prot­ estants organize their existence around the c.hurch and, consequently , one finds a small overrepresentation of Pietistic and Refonnation-Era Protestants (both with ratios of 1 . 11) in this category . .

Genetic and Canon Laltt' Models . In Table 5- l , only the Roman Catholics

are

overrepresented in the Genetic and Canon Law categories (but by only a 1 . 1 3 ratio) . The Roman Catholic Church has always aimed at becoming the univers al c-hurch and , in recent years , has promoted ecumenical move­ ments . The overrepresentation of Catholics both in Parentela Orders and in the Genetic and Canon Law classifications suggests that there are diverse ethn ic and soci oeconomic factors that seem to confound kinship orientation of Catholics .

A supp le mentary comparison of Cathol ics in the Parentela Orders category W it h those in the Genetic and Canon Law class (not presented in tabular fo rm) reveals significant differences in ethnic ity . Of the 49 Catholics in the Pare ntela Orders category , 33 . 3 percent are Mexican-Americans , whereas for

76

the

41

Cathol ics falling into the Genetic and Canon Law c lass , ful l y 4 8 . 7

percent are Mexican American . Thus , the Genetic and Canon Law category seems heavily loaded with Mexican Americans . For the non- Mexican-Amer­ ican Catholics especially , the nativity of the grandparents seems to be

an

important variable in determining kinship orientation . Over 80 percent of the grandparents of the non-Mexican-A merican Catholics �·ith a Parentela Ord ers orientation are European born , while only

69

percent of the grandparents of

those with a Ge netic or Canon Law perspective have been born outside the United States . The difference is espec ially great for paternal grandparents . Among these non-Mexican-A mericans , only 46 percent of the respondents

with a Parentel a Orders orientation are descended from grandparents who were both American born , in contrast to 7 1 percen t for those persons in the

Genetic and Canon Law categories . Hence , Catholics with a Parentela Orders

kinship orientation appear to be closer to their European roots than are those with a Genetic or Canon Law perspective .

Kinship and Religion : Summary The findings pertaining to religious identificati on and kinship orientation indicate qu ite cle arly a rel ationship between separatist , communal tendencies in religious groups and the attractiveness of the Parentel a Orders model in kinship orientation . Jews and Neo-fundamentalists (particularly Southern Baptists) are far overrepresented in the Parentela Orders category . Withi n Judaism , the more separatist and traditional branche s , especially Orthodoxy �

are associ ated with the Parentela Orders model , and the least traditional ,

Reform Judaism , i s tied more often to the Genetic and Canon Law perspec­

tive . Al though as a group Mormons do not hold a Parentela Orders orienta­ tion , devout Mormons whose marriages had bee n sealed in the temple do emphasize Parentela Orders and de-emphasize Genetic and Canon Law con­ ceptions of kinship . By way of contrast, the Standard American mode-l seems to perv ade all religious groups ; only the Reformation era and non-denomi­

national Protestants and those persons with no rel igious preference are to some extent overrepresented . Rel igious groups that seem to h ave strong as ­ sociational components, however , se.e m to dominate the Civil Law category

of kins hip orientation . These groups , e specially the ordinary (as contrasted with very devout) Mormons and the Pentacostal sectari ans , seem to play down ethnicity as a factor in their lives , and apparently the church becomes the focal point of all activities and commitme nts . Finally , rel igious groups that aim at creating unity out of di versity tend to maintain Genetic and Canon Law orientations . Those Roman Catholics without strong European ethnic ties are overrepresented in this category . Thus , viewed as a whole, the find­ ings on rel igious identity support the contenti on that religious pluralism as an ideology is assoc iated with type of kinship orientation .

77

FAMILY

INCOME AND MINORITY-GROUP STATUS

one is sue in an analysis of conceptions that people hold about kinship is the comparative influence of minority-group status and financial resources on their v ie ws. Schneider and Smith ( 1 973 , p. 39) tak e the position that ' 'much 'eth ni c life-style ' is of rural , traditional , lower status ori gin ' ' and that "be­ com in g a middle-class American necessarily involves abandonment of much of the c lass orientation of ethnic life ways . " �1oreover, they propose that , · ·n.e very traditional ism , rural origin , and adjustive functions of these sub­ units provides a certain structural uniformity \\thich coincides with and rein­ forces the general lower-class orientation derived from similar positions in the oc cupati onal , power and pre stige hierarchy of American society . " The opposing position , as expressed by Staples ( 1 97 1 ) in his di scussion of the black family , is that the common experience of oppression and d i scri m in ati o n that pervades all phases of l i fe in a minority group i s more penetrating th an socioeconomi c factors in affecting family and kin ship . ' 'One trait that black families usually share is a history of racial oppress ion , v,rhether under colo­ nialism or slavery , [with the result that] no v,rhite family has faced the his­ torical subjugation and contemporary racism that black families have en­ countered and continue to encounter even within the same s ocioec on om i c stratum ( S tapl e s , 1 97 1 , p . 3) : ' The i ss ue of ethnicity versus socioeconomic facto rs is thus not whether one excl u des the other as an influence upon kinship ; rather the issue i s one of comparative v.'eighting of i mportance (Farber, 1 97 1 , p. 1 1 7 ) . The d ata reported in this section first refer to family income gene rally and then to

minority-group status .

Family Income and

Kinship

Orientation

Findings on the impact of family income upon kinship orie ntation are pre­ sented in Table 5-4. Unlike other indicators of socioeconomic status , family income takes account of multiple breadwinners and the exi stence of inve st­ ments. Because of the influence of religious identity upon kinship orientation , the dat a are sho wn for Jewish , Protestant , and Catholic respondents sepa­ ratel y . Unfortunately , si nce there are fev,r cases in minority Protestant de­ nomin ation s and sects , all Protestant groups must be handled togethe r . A s a wh ole , Table 5-4 can be interpreted a s follows: I The religious groups vary in their concentration of cases in the Parentela Orders category by family income . Only the Protestants produce an overre­ p rese ntat ion of Parentela Orders responses among high income famil ies . Con­ so nant w ith this finding , however, Catholic s with a low family income are Underrepresented in the Parentela Orders orientation . Ne ither Jews nor those Without a rel igious preference reveal much variation in concentration in the Parentela Orders category by income level . ·

78 TABLE

5-4. FamUy Income and Kinship Orientation , with Held Co.wtant

$ 24 , 999

_h

Parentela orders (%)

Ratio of total Standard american (%)

Civil

$ 1 0, 000 10

$ 1 0 ,000

Jewish3

Ratio

Family income in 1 977

Under

Kinship orientation

of total

.

1 8.3

Ratio of total

36 . 7

Ratio of total

t·

.1\o

l' For

Jewish respondents , Chi

cases in

this category .

square

=

15.8

50. 2

48 . 0

47 . 5

1 .0 1

1 . 64

1 . 04

. 54

60

215

19. 1

sisniftcanl .

d. f.

=

6; p =

1 7 .0

1 5 .9

1 . 36

20 . 6 1 13

1 4 . 967;

15. 1

. 93

.99

square not statistically

Protestalll respondents. Chi

1 .0.5

2 1 .6

18. 1

. 82 30. 0

total

15.8

1 2 .6

1 . 06

1 5 .0

Genetic or canon law (%)

1 .03

38

. 77

Civil law (�· ) Ratio o f total

. 97

15

. 79

1 . 15

Standard ameri can (%)

a For

43.4

1 18

Parentela orders (% )

K

44.7

20.0

Protestant c

Ratio of

40 . 0

.88

or canon law (% ) Ratio of total

Total 24 . 5

13.3

Genetic

or over

23. 7

. 92

law (%)

$25 .000

26 .7

1 .09

Ratio of total

N

Religion

18.3

.

9. 8

18.3

1 02

(co1fli�Wed) .

OS .

2. There are inconsi stencies among rel igious groups on the extent to Vlhich low income families are underrepresented in the Standard American and Civ i l Law categories. Family income alone seems insufficient t o explain variation i n conformity to Standard American and Ci vil Law conceptions of kinship .

3. Generally , the re lationsh ip betY.'een family-i ncome and kinship orien­ tation is strongest in the Genetic and Canon Lav.' category . For Protestants . Catholic s , and ' 'no preference ' ' respondents al ike , persons V.'i th a fami ly income of $ 25 , 000 or over are underrepresented in this category , V<�'hile those with a fami ly income under $ 1 0 ,000 are overrepresented . In th is kinsh ip category , family-income level seems to outweigh religious identity as a factor

dirth of financial resource s , families at IO\\' income levels wou ld have to rely upon kin-perhaps on a daily basi s-to provid e in orientation . With

a

the economic assistance and personal services required to avoid disaster and to overcome

cri ses as the.y ari se . As a result , their conceptions of the character

of kinship would undoubtedly be affected . (An alternative interpretation

79 ontinued) fABLE 5·4 (c Family income i n 1 977

J{ioshi p orientation

Under

5 10,000 to

$ 1 0,000

$24 ,999

CatbOiicd

6.3

parentela orders (%) Ratio of total standard american ( %)

. 32

39.1

.90 1 5 .5

1 . 15 9.4

1 . 07

. 65

Ratio of total Genetic or canon law ( %)

22.7

34. 4

Ratio of total

No religious preferencec Parentela orders (%) Ratio of total Standard american ( %) Ratio of total Civil law (%)

1 .52

1 .00

32

1 10

17 6

16.7

.

1 .07 4 1 .2 .81

1 . 02 52 4 1 .03 1 4.3 .

1 1 .8

1 .04

. 86

Ratio o f total Genetic or canon law (%) Ratio of toal N .. For Cllholic rt:spondents, Chi

1 . 17

50.0

Ralio of total Civ il law (% )

N

22 . 7

29. 4

1 6.7

17

Total

20 . 5

1 9. 4

1 06 50 .0

43.5

1 . 15 15.9

1 4.5

1 . 10 1 3.6

22.6

.

.60 44

1 4. 3

.87 57 . 1 1 . 13

14. 3

1 . 04 14.3

. 87

1 . 53 square

$25 ,000 or over

42

1 6.4 50 . 7 1 3 .7 1 9. 2

. 74

14

not statisticall y signifteanl .

• For respondents wilhout a relilious preference.

Oti square DOt

statisticall)' significant.

would be that this orientation toward kinshi p tends to keep them at loy; income levels . ) 4. Among Jews in particular the data indicate that family income has little influence on kinship orientation ; there is l ittle overrepresentation or under­ representation of respondents from different income levels in the various categories of kinship orientation . Th i s fmd ing is consi stent with those of other stu dies , which suggest that socioeconomic distinctions within the Jew­ ish populati on are blunted by at least two factors . First, there. is a growing homogeneity in financial status among Jews as the children and grandchildren of immigrants are absorbed into the American oc.cupational , educational and , economic systems . Second , the distinctions in family and kinship "style of life ' ' among Jews are determined more by religious orthodoxy and "Jewi sh­ ne ss ' ' (Yiddishkeit ) of identity than by economic variables, and consequently those pe rsons who are most assimilated into the Standard American culture si mp ly leav e the fold (Sherman , 1964) . Hence , for Jews, socioeconomic homogeneity and the persistence as a • 'people ' '-religious and ethnic-seem to overshadow family income as an influence upon kinship orientation .

80 TABLE 5-S . Minority-Group Held Constant

Status and Kinship Orientation , witb Mexican

Protestant h Parentela orders (%) Ratio of total

.95 49.6

Ratio o f total Civi l law (%) Ratio of total

Genetic or canon law (%) Ratio of total

Ratio of total Civil law (%)

48 . S

are no

18.2

18 .4

2 .63

12

20 . 5 1 . 03

18.8

45 . 1

40 . 6

33 1 9. 9

.91

43.5

. 93 9.4 . 65 3 1 .3

1 . 19

1 4 .5 22. 0

1 .42 64

1 22

There

47 . 3

3 .0 . 16

17.2 . 78

N

24. 2

1 9 .6 1 .08 15.5 . 84

17.2

Ratio of total

1 6 .0

.Sl

1 .04

Ratio o f total Genetic or canon law (%)

Total

24. 2 1 .5 1

1 .05

341

Catholicl· Parentela orders (% _) Ratio of total Standard american (% )

•

Other minoritites

15.2

Standard american (%)

N

american

Nonminority3

Kinship orientation

Religion

3

Jewish respondents ia minority-group calegories. Of the 75 residents with no religious pref-

erence . only 9 are in minority groups . On lbe other hand, minority groups are prone to adopt relieions in no• Protestanl and non-Catholic categories (e . g �"� For Protestant �apondents.

r

For

Chi

Catholic respondenls� Chi

square =

square

.•

Black

27 .764;

Muslim . Traditional Americaa

d.f. =

not significant.

3;

p< .OOI .

lodiaa, Buddhi st) .

Racial Minority-Group Status and Kinship Orientation Table 5-5 displays the cross-tabulations for racial minority -group status and kinship orientation , by religion . In the table , uOther Minorities " consist mainly of Blacks . The total sample includes 46 persons in this category-2 9 black s , 1 2 American Indians , 4 Orientals , and 1 black and Mexican Amer­ ican . Of the 1 0 minority persons who are neither Protestant nor Catholic ,

most adhere to American Indian , Oriental , or other non-Christian religions; these are exc luded from the analysis . The data for Protestant minoritie s are presented in the upper part of Table

5-5 .

(The 1 2 Mexican- American Protestants are not included . ) Among Prot­ estants , minority-group respondents are concentrated in Parentela Orders (ra­ tio = 1 . 5 1 ) and in the Genetic and Canon Law category (ratio = 2 . 63); they are sparse in the Standard American and Civ il Law orientations (ratio = 0. 5 1 and 0. 1 6, respectively) .

81 fhe data for Cathol i cs appear in the lower part of Table 5-5 . (S ince only n rity-group Cathol ics are not Mexican American , these cases are bfee m i o from this analysis . ) The Mexican Americans are o verre pre sen t ed in a the ae.neti c and Canon L w category (ratio = 1 .42) , an d they are u nde rre ­ se. nted in the Ci vil Lavl classification ( rati o = 0. 65) . In contrast, non­ pre mi n ority group C athol ic s are underrepresented in the Genetic and Ca n on Law = 0 . 7 8) . However, Mexican-American Cathol i c s are si m i l ar category (ratio to Ang l o Catholics in the percentage s fall ing into the Parentela Orders and St and ard Am erican ca te gories . The high c oncen tration of racial minority-group re spondents in the Genetic and Canon Lavl c lassification suggests that family income may pl ay an im­ portant ro le in org ani zin g their k i n s hi p orientati on . An analysis w as under­ taken of the re l at i onsh ip between fam i l y income and kin s hi p orientation for persons in the minority-group category . T he findings of this an al ys i s are pre sented in Tab l e 5-6 . The v arious m inorities are grouped together in Tab le 5-6 on the a ssumpti on that each " �rac ial " ' or e thn ic minority-whether black , Mexican American , or American Indian-has in some way had a history of subjugation and oppression which has g i v en them all a s imil ar kind of ex­ perience The data in T abl e 5-6 signify that some v ari at ion in kinship orientation can

�ropped

i ndeed be attribu ted to di fferen c e s in family i n c ome . The strongest trends of

the data on minority-group income are: (a) the upward skew i ng of cases in the

Parente l a Orders c ategory toward the h i gh - income level s (w ith a ratio of 2 .48 when income i s $25 ,000 or over) and (b) the downward ske w i n g of cases in the Civil Law and in the Genetic and Canon Law classifications toward low in come le.vels . But the role of income in determining kinship orien tati on i s , however . lim­ ited . A multivariate analysis in Appendix A shows that , when a long series of TABLE S..(i .

Family Income and Kinship Orientation Family

Under

Kinship orientation

5 1 0 .000 to

$24 ,999

Parentela orders (% )

1 3.2

10 . 8

Standard american ( %)

34. 2

Civi l

10.5

7 .7

1 . 42 42. 1

1 .04

. 84

Ratio of total

.69

law ( %)

Ratio of total Genetic or canon law ( %) Ratio of total

1 . 19

�

& C hi

square

=

9. 9 1 4;

d.f.

=

6 � p = . 1 0.

38

Minority Groups

income in 1 9778

$ 1 0.000

Rati o of total

in

.69

46 . 1

1 . 16

3 5 .4

1 .00

65

$25 , 000 or over

Total

38 . 9

1 5 .7

38 . 9 . 94

41.3

22 . 2

35.5

2 . 48

. 63 18

7 .4

121

82 social background and family o rga ni zati on variables are taken in to acc ou nt ,

the dire.c t effect of family i nco me on kinship orientation is dissipated . As i n Table 5-5 above , racial minority-group status then re-emerges as an impo rt an t influence on k insh i p orientation . The minority groups ten d to be overrepre­ se nted in the Genetic and C ano n Law categories and u nderrep resen ted in th e Parentela O rders and Civil Lav.' classific ations (\\·here rel igio us factors pre­ dominate ) . Hence , despite the appare nt association between i ncome. and ki n­ s hi p orientation (when m ino ri ty status is held constant in Table 5-6) , the mul­ tivariate analysis re ve als that i t is racial statu s which molds concepti on s of kinship .

The data on racial mi nori t y groups th us c larify the basis for the contrad ic­ tory viev.'s h e ld by S c hne i der and Smith ( 1 973) and by S tapl es ( 1 97 1 ) . When the interaction of m inority-group s tat us and fami l y income are e x am i ne d in i so l at i o n from other background and family variables , the Schne ider and S mi th ( 1 973) interpretation is s upported t namely , that the ec onom i c fac tor is primary in de te rm i n ing kinship perspective . But , \\'he n other competing i nflue nc es are considered , t he i mpact of inc ome is V.'ashed av,ray , and the Staples ( 1 97 1 ) con c lu s ion s urv i ves , namely , that the oppressio n of mi nority groups transcends economics in structuring ideas about k inshi p .

Summary: Kinship, Income , and Minority-Group Status

The rel ations hi ps am ong ki n s hip ori e ntati on , re l igi o n , family i nc om e , and min ority - group status are somev.'hat c om pl ex . The c o nfigu ratio n of results for rel i gio us and rac i al minorities suggests that ethnicity o utweighs income in i nfluencing k i ns hip orientation . Am on g Jew i sh respondenb; , fami ly income i s genera l l y i nco nseq uen ti al for kinship o ri en tat io n , while , among the racial minorities , family income (considered in isolation from other variables) is of su bstantial i mport anc e . But Jewish iden t ity is tightl y bound up \\'ith an ide­ ology ass oci ated with Judaic law , and e ven secu lar Je�·s have not aban do ned this ideo log y altogether . This ideology tends to overwhelm effects of family income on ' 'life style . " Similarly , the an aly s i s of rel i giou s sects in the pre­ viou s section has indicated that Protestant sec t arians , des p i t e their lov.' income and bl ue co llar occ upations , tend to hold a Parentela Orders perspec ti v e and to veer away from Genetic and Canon Law views . B ut as people lose th eir identity as Jews and sectarians , the gu iding fam il y id eol o gy which had pre­ viou s l y overpowered socioeconomic consideration s then Vt'eakens . Mere mem be rs hi p in a rac ial minority , howe ver, is independe nt of ideol­ ogy . R acia l identity is on ly rav1 material for soc ial rel ations . One is identified as a Mexican American , American Indian, black , or Or ien t al by phys ical features and l in gu i stic characte ri stics--that is, by overt bi ol o g ic al and social signs . Consequently , the ideological overtones wh ic h emerge in interacti on

83 vide di rec tion for the development of kinsh ip orientations in these mi­ p ritie s . Oppre ssion of rac ial minorities tends to dri ve persons toward Ego­ , hich i mply a comm unal i stic (rather than factional) ntered k in orientations ':' ncep tio n of the v,ray soctety ought to be . co

: �

ocCUPATIONAL STATUS AND KINSHIP ORIENTATION

�fos t analy ses of k inship ties in modern societies deal with extens ivene ss of

s in te rac tion with discrete ' 'relative ' · rather than \\'ith the gestal t of priorities amo ng ki n . Consequently , statements about k inship general ly refer to vari­

ations in exte ndedness as related to social strati fication . As a resul t , much co ncern about modem kinship has been bound up with the nuc lear-family hou sehold and its re.lationship to the economic structure of society . For ex­

amp le , ec.onomic and technological development of societies has been as­

sociated \\'ith c .the ideology of the conjugal family (Goode , 1 963 , p. 369) : '

and .�dams ( 1 970, p . 577) suggests that � 'this ideology i s buttressed by the personal freedom promised in the urban-industrial economy . , , Similarl y , Smelser ( 1 976 , p. 1 53) proposes that if the family has to move about through the labor market, it cannot transport all its relatives with it or maintai n close ties with them . It would thu s be the economically unsuccessful , immob i le fami lies who deviate from this norm . Accordingly , · -me working classes

. . . express a stronger kin orientation , live c loser to their kin , and i nteract

�·ith them more regularly than the middle classe s (Adams , 1 970 , p. 585) . ' �

Yet one problem w ith studies of extensiveness of ties with relatives is that

they do not take into account the separate effects of migration (Klatzky , n . d . )

and fertility pattern s (Gordon , 1 977) . Since fertil ity and migration themselves affect extensiveness of kinship ties , socioeconomic interpretations may be mi s leading . As a resul t, such studies have limited value in providing insight into lhe rote of kinship in modem social structure . The i ssue remains as to whether ce rtain kinds of orientation toward kin are more conducive than oth­

ers toward the maintenance and/or achievement of high socio-economic lev­ e ls . ( See Litwak , 1 960a; Burr, 1 973 , p. 1 5 6 . ) It is this topic that is the focu s

of analy sis in th i s section . Th e discus sion below i s presented i n three parts : (a) occupatio nal status of male co-head of household . (b) couple.s with the same oc c upational status , and (c) social mobility .

Occupational Status of

D�ta

Male Co-Head

of Household .

re l evant to the relationship bet\\'ee n occupationa l status and kinship n one tation are presented in Table 5-7 . Occupation is used he re as a way of l be li a n g socioe conomic status of the family . Inasmuch as both husband and w 1·fie are regarded as co-heads of the hou sehold , the term male co-head (rather

84

than he ad) i s used in the table . In the case of male respondents, the m al e

co.

head 's occupation refers to the respondent ' s own occupation ; in the case or

female respondents , the male co-head ' s occupation refers to her hu sb a nd '

s

occupation . Occupations \\'·ere cl assified according to the 1 970 U . S . Ce n s u s class ification index . Although the findings by co-head are al ightly less con. elu sive than those for male respondents them selves in describing the i nfl ue nc e of occupational status on kinship orientation , the inclusion of husband� per.

mits a more extensive analysi s . (In the table , religion is that of the respo nde nt (male or female) rather than the religion of the male co-head .

The impact of occupati onal status upon kinship orientation depends in pan upon religion of the respondent . The results , therefore , are presen ted by

rel igious groups separately . (Since virtually all Jewish male co-heads

are

professionals , managers , or administrators , an occupational analy sis i s not feasible , and Jewi sh re spondents are dropped from the table . ) Table 5-7 reveals the fol lowing :

Protestants . Because of the small number of cases i n the various denom­

inations , all Protestants

are

grouped together in the table , and the findings

are presented by occupational group:

1 . Professional and managerial occupations are underrepre sented in the Genetic and Canon Law category (ratio

=

0. 79) .

2 . Persons in clerical , sales, and craft occupational status do not shov.· any marked tendency to occupy any given kinship orientation .

3 . Blue-collar Protestants , however , are sparse in the Parentela Orders class (ratio = 0 . 67) , and they tend to concentrate in the Genetic and Canon LaYl category (ratio = 1 .4 1 ) . Catholics . Occupational status seems to play a different role i n kins hip

ties of Catholics than in those of Protestants . As

a

result, Cathol ic fam il ies

reveal patterns of kinship orientation which contrast with those of Protestant�:

1 . Unlike Protestants , professional-and- managerial-status Catholics are underrepresented in the Parentela Orders category (ratio = 0 . 75 ) ; b ut they are overrepre sented in the Standard American orientation ( rat io

==

1 .2 1 ) .

2 . Catholics i n famil ies where the husband i s i n a clerical , sales , or craft occupation also show a s l ight concentration in the Civi I La\\' category (rati o = I . 20) . 3 . Finally , blue -collar Cathol ics are disproportionatel y represented in the Parentela Orders and the Genetic and Canon Lav..· categories ( ratio

==

1 . 38 and 1 . 4 1 , respectively) but underrepresented in Stand ard Ame ric an ki nship (ratio = 0 . 64) .

85 rAfiLE

5-7. Occupation of Male

Co-Head of Household and Respondent's Kinship Orientation, by Religion of Respondenta Male co-head 's occupation Professional ,

admi nisttati ve

Clerical , sales , craft

Blue c-ollar

Total

17.3

19. 1

1 1 .0

16.4

1 .05 54 .0 1 . 12 15. 1

1 . 16 44 . 3

. 83

1 . 10

.67 43 .9 .9 1 20.7 1 . 14 24 4 1 .4 1 82

managerial , or

kinship orientation Respondent 's

Protestanl

parentela orders ( %) Ratio to total standard american ( %) Ratio to total Ci vil law (%)

Ratio to total

Genetic or canon

Ratio to

total

Law (% )

N

Catholic

Parentela orders (% ) Ratio to total

. 92 20 0 .

1 5 .0

19.7 .99

27 . 5 1 . 38

1 9 .9

27 . 5 .64

42 . 8

1 2. 5

1 3 .9

. 90 32 . 5

24 . 1

.95 1 15

5 1 .1 1 .2 1

43 .9 1 .03

Civil law (%) Ratio to total Genetic or canon law ( %)

1 1 .7

60

16 . 7 1 . 20 19.7 .82 66

25 .0

1 6 .7

N No religious preference Parentela orders (%) Ratio to total Staodard american (%) Ratio to total Civ il law (%) R atio to total

Genetic or canon law (%)

Ratio to total

N

. 84

2 1 .7

.90

1 .25 53 . 6 1 . 15 1 4. 3

. 86 7. 1 .43 28

a For

17.3

16.5

Standard american ( %)

Ratio to total

18.2

13.7 . 79 1 39

. 75

Ratio to total

48 . 2

. 84 44.4 .95 22.2 1 . 33 1 6 .7

1 .00 18

Blue Collu

.

336

1 .35 40

166

14.3

20.0

. 72

35 . 7 . 76 14. 3 . 86 35 . 7 2 14

46 7 .

1 6. 7 1 6.7

.

14

60

Jewish respondents, lhe number of cases in Clerical and allegories was too small for Cases, :rqardless of relisious afflliatiou (includinl Jewish) S.OO; d.f. 6; p = 02

�Dgfu l crosstabulation. Chi squue for all ==

.

=

.

No Religious Preference . The distribution for men in fami l ies V.'here the re spo ndent has no religious preference embodies patterns found in both Prot­ e stant an d Catholic fam ilies: I . Like Protestants , "no preference " people in the profess ional and man­ agerial occupational class tend to concentrate in the Parentela Orders

86

orientation , (ratio = 1 . 25) as oppo sed to a Ge netic or Canon La\\· perspecti ve ( ratio = 0 . 43 ) .

2 . Like Catholics in the clerical , sales , and craft occupations , these i ndi.. viduals are abu ndant in the Civil Law category (ratio = 1 . 3 3) . 3 . U ni qu el y , ' �no pre fere nce ' ' blue-collar famil ies are underrepresent e d i n bot h Parentela Orders and Standard American classes ( rat io = 0. 7 2 and 0 . 76, re spec t i vel y ) and overrepresented in the Gene t i c and C anon L aw category (ratio = 2 . 1 4) .

As a whole , the data i n Table 5-7 indicate that occupational status of the male bre adwinner seems to pro vi de a strong influence on kinship orientation somewhat independently of re l i gious i dentity . Apart from the singular ex­ ception of Catholics with a Parentela Orders orientation , persons in families marked by high oc cupational status (i . e . , professional and managerial) tend to hold Parentela Orders and Standard Ame rican kinship orientations , v;hile those with low socioeconomic status ( i . e . , blue collar) tend tov.·ard a G e netic or Canon Law k i nship orientation . The middl e occupational group (i . e . , cler­ ical , sa les , and craft) clusters in the middle-groun d Civil Law category .

Couples in Same Occupational Status

Prob ably , influence of occupational status on kinshi p orientation is most clearly identified when both husband and Ylife are in the same occupat i onal category . Otherwi se , comparisons might involve (a) households in V.'hich both hu s b and and wife are working as we ll as households in w hich only one is emplo yed or (b) households in which there are diverse occ upation al statuses within the fam ily . Table 5-8 reports on the kinship orientations of persons in families in which the husband and wife are in the same occu pat i on al statu s . E xcept for respondents in the Parentela Orders category , the results in Table 5-8 are si m i l ar to those in the pre vio us section on male co-heads of households . In Standard Americ an ki ns hip , there is a skewi ng of percentages av.·ay from blue collar status; persons in clerical , sales , and craft occupations tend to concen trat e in the Civil Law category (ratio = 1 .23)� in the Genetic and Can on La\v class , the percentage s are skewed away from professional and managerial families to�·ard the blue - col l ar group . As with co-he ad s , the findings in the Parentela Orders c ategory are likely co nfou nded by the d if­ fe renc e s between Catholics and Protestants, inas mu ch as high -s t atus Prote s­ tants (but not Catholics) are overrepresented in that category . Hence , oc cu­ pational status-whether c lassified by m ale co-head 's oc cup ati on , by mal e respondent ·s occupation, or (in th i s case) by joi nt occu pational status of h usb and and w ife-yields a stable series of findings: S t and ard Americ an kinship is associated with pro fes sio nal and m anagerial status , and Gene tic

87 tABLE 5-8. Kinship Orientation of Respondents in Families in Which Husband and Wife Are in Same Occupational Status Couples in occ upational status Professional .

manageriaJ , tati on Ki nship orien

administrative occupation

parentela orders ( %) Ratio to total StaDdard american ( %)

23 . 1

Ch'il law (%)

1 3 .8

Genetic or canon law ( %)

1 2 .3

Ratio to tolal

Ratio to total

Ratio to total

N

or

1 . 13

50. 8

1 . 13 . 85 .68 65

Oerical , sales , or craft

Blue oollar

occupation

occ upation

Total

1 8 .6

1 9. 4

20.4

.9 1 45 . 7

1 .01 20. 0

1 . 23 1 5 .7 . 87

70

. 95 32.3

45 . 1

.72 1 2 .9

16. 3

.79 35 . 5

18. 1

1 . 96 31

and Canon Law perspectives are general ly found in blue-collar families . As the results on religion have indicated , however , in the case of Parentela Orders , socioeconomic influences interact with rel igious variables in a com­ plex manner. Social Mobility

Past research on kinship and social mob il ity has indicated l ittle connection between occupational mob i lity and such variables as o rientatio n toward kin , extent of interaction, mutual aid , and shari ng of households (Adams, 1 970) . The only distin ctio n of note i s that sibling relations are some ti mes affected . Di fferential achievement between brothers " 'usually means affectional dis­ t ance and non-iden ti fication ,' ' but mutual upw ard or downward occ upational mobili ty apparently results in mutual support and strengthened bonds (Ad­ ams , 1 968 , p. 1 72) . The focus upon individual ties among relatives , however, leaves unan swered the q ue stio n of the associ ation between social mobility and gene ral kinship orientation. Table 5 -9 d isp l ays d ata relevant to intergenerational social mob i li ty . The extent of upward an d downvlard social mobil ity of the re spon de n t s is ex­ ami ne d separately for each level of father ' s occupation . The socioeconomi c statu s of the respondent i s class ified by the occupation of the male co-head of the ho useh old ( i .e., by the male re spondents ' own occupation and the female respondents � husband ' s occupation) . The frrst segment of Table 5-9 describes the downward social mobil ity of pers ons whose fathers are cl assified in professional and managerial occupa-

Chi

d Not

c Cb i

b

M For

1 9. 7% 46 . 1 % 28 . 9% 5 . 3% 76

=

==

3; p

3; p

-=

d. f. 8 . 68 1 ; d.f. -

14. 239;

statistically significant.

square

square

=

-

.OS.

.01.

Occupation ' '

33

is

30 . 3% 30. 3%

12. 1 % 27 . 3%

collar

to blue

cJericaJ

sales , <..Tafts

male respondents, • •Male Co-Head 's

Parentela orders Standard american Civil law Genetic or canon law N

orientati on

Kinship

F.xtreme professional

13 .3%

.95

of respondent;

for

83 female

13 . 3 %

53 .0%

1 . 69 . 18

20 .5%

or managerial

respondents,

1 7 .0% 29 . 8% 47

46 . 8%

6 .4%

collar

To blue

of' child :

mobi l ity

Dow nward

professional

48

1 2 . 5% 1 6 . 7%

1 2 . 5% 58 . 3%

•

sales ,

73

occupation.

23 . 3 %

17 .8%

42 . 5 %

1 6 .4%

- --

crafts

clerical ,

spuuse ·s

or m anageri al

Extrerne: To

· - --

M ild : To

. 72

.70

1 . 37

. 76

mobility

upward

To mild

extreme

Ratio of

Compared with

8 lue collar"

as

Upward mobility of child

• •Male Co-Head ·s Occu pation ' is

.1M . 45

3 .20 1. 13

mobility

downward

upward to

Ratio of

Clerical , sales, craft c

�

Father 's occupation

of c hild ; To

mobility

Upward

1 .63

mobility

downward

mild to ex tre me

Ratio of

occ upation

of child

Mild: To

Downward mobility

-- ·

Professional or managerialb

Male Co-Head's Occupation••

TABLE 5-9. Kinship Orientation and lntergeneratiunal Social Mobility: Respondent's Father's Occupation

00 00

89

here is generally more m i ld downward mobility (to the clerical , sales , tions . T nd crafts c ategory) than extreme doVw'nward movement (to the blue- collar 11ateg ory ) . However, as the ratios of mild to extreme downward m obi l it y te. considerable d ifferences exi st among the various kinship orienta­ nd ica is relati vely less extreme downward mobility among persons in ions . There the parentela Orders and Standard American categories than in the Civil LaY-�' and the Ge netic and Cannon Lay; groupings . In fact, the chance of extreme do\\'llward mobility in the Genetic and Canon LaY-�· category tends to be high (rati o == 0. 18) . The middl e segment of Table 5-9 refers to the intergeneration al social mob ility of children whose fathers are classified as c lerical . sales , or craft occupat ion s . As a whole , there is more upward mobility of the children (to professional and managerial statu s) than there is downward mobility (to blue­ collar occupations ) . But the differences among kinship orientations are pro­ found. As before , persons with Parentela Orders and Standard American perspectives exhibit less downward mobil ity than do those in the Civil Law and Genetic and Canon Law categories . Persons with a Parentela Orders orientation are particularly more prone to upward than to downward mobility (ratio = 3 . 20) , and those individuals with a Genetic or Canon La\\' orientation are especially more often doVw·nwardly mobile (ratio = 0 . 45) . The third segment of Table 5-9 deals v.'ith the upVw'ard social mobility of children whose fathers are (or had been) blue-coll ar workers . ��s expected , the Standard American orientation is found more often among respondents \\ilh extreme upward mobil ity (to professional and managerial status) than among persons with mild upward movement (to clerical , sales , or craft status) (ratio = 1 . 37) . All other orientations , however, are associated w ith m i ld upward mobility from a blue-collar status of the father . Since this finding is contrary to expectations regarding the Parentela Orders orientation. one. sus­ pec ts that other influences-such as sectarianism , Cathol icism. or minority grou p statu s-may have confounded the findings . In gene ral , Tab le 5-9 suggests that Parentel a Orders and Standard Amer­ ic an k ins hip orientations seem to sign ify the existence of some sort of brake on do wn \\'ard social mobility in certain families and i ndeed the probable

�

:

presence of pressures toward upVw'ard mobility . In contrast, a Genetic or a C anon Law orie.ntation appears to express the absence of such a brake .

ASSIMIL ATION AND THE STANDARD A�IERICAN PATTERN

The fi ndi ngs in Chapter 4 on on the social characteristics of persons con ­ form ing to the Standard American model suggest that adoption of this model c an be appl ied as an index of assimilation into middle-class American culture . S i m ulta neously , the study of assimilation seems to provide an opportunity to

90

investigate the. comparative infl uence on conceptions of col l ateral i ty deri vin g from i deo l ogical factors as opposed to viewing relatives as persons . This se c ti on compares re l ig iou s factors and fam il iarity with grandp aren t s as they e ac h impinge upon conformity to the Standard American model . Because the vast majority of grandparents of respondents are. no longer alive , this analysi s focuses upon fam ili ari ty with them at the time v.·hen the responde n t was gro wing up . According to the position that , above all , variations in c oll aterality depe nd upon personal rel ati o n ship s rather than ideo l og y , one would expect that an individual \\'ould assign a higher prio ri ty to those kinds of rel at i ves he kn o\\.·s well than to those he does not know . For e xample , if an individual has known a grandparent well , under the pri nciple of relatives-as-persons , he v.·ou ld probably think of ' 'grand pare nts " ' as close , whereas if he has never kno\\' n any of his g randparents , he would likely cons ider grandpare nts to be some­ Y.'hat distant re l ative s . Moreover, if du ri ng childhood , the person has kno�·n his grandparents wel l , ch an ces are high that he also is quite familiar Vw'ith au nts or uncles . Compared with the various kinship model s under investigation , the Stand­ ard American pattern gives a higher priority to grandparents ( and others in ascending generations) over other relatives . In this pattern, grandparents (fol­ lowed by u nc le s and aunts) are p l aced second only to parents in priority . By way of contrast , in Parentela Orders , the grandparents come after all descen­ dants of Ego and Ego 's parents . In Civil Law , the grandparents follov.· both parents and chi ldren , and they are equ al in closeness to Ego 's grand­ children and siblings . In Genetic and Canon Law model s , all nuclear family members are given priority over grandparents , and grandparents are on a par with aunts and uncles , grandchildren , and niece s and nephews (and , in Canon Law , with first cousins as wel l ) . Hence , with grandparents as reference point. an analy si s of co nformity to the Standard American model se.ems appropriate for dete rmining w hether acquaintanceship with a grandparent is associated \\l'ilh assignment of a high priority to grandparents in c on c e ptions of collat­ era lity . One of the findings in Chapter 4 is that those respondents with a European ­ born father are far less like ly than chi ldren of an Ameri can-father to conform to the Standard American model . Obviously , fewer A merican -born perso ns

with a fore ign -born father would knov-' their grandparents and aunts and unc les than w·ould those whose father had been born in the United State s . Thi s contingency evokes the qu e sti on : I s a person \\l'ith European born grand­ p arents less apt to confonn to the Standard American model bec ause he has not known them or is it becau se of a diffe ren ce in more general conceptions of k i n shi p ties ? The remainder of this sec tio n is concerned '-"'ith the resolution of this issue .

91

Religion and Familiarity with Grandparents

able 5 - l 0 displays the percentages of persons vlith a Standard American T ori entatio n , sorted by their own religion and the birthplace of the ir grand­ parents . The table also indicate s , for each re.ligiou s g roup and grandparental bi rthpl ace , the percentage of persons who had known their grandparents well pri or to ad ulthood. (In order to incorporate the findings for Jew ish responTABLE 5-10. Percentage of Persons in Standard American Category and the Respondent's Familiaril)' with Grandparents Prior to Adulthood, by Respondent's Religion and Birthplace of

Grandparents

Grandpare nt born in

Specific

Percent in standard

grandparent, by respondent ' s

american category

religion

Father's father

4 1 .9 4 2.2

Jewi sh

Protestant Catholic

Grandparent born in United States11

Europe3

Percent

Percent io

Percent

standard

fami liar with

grandparentsc

N

41.9

43

familiar w it h

american

category

grandparents c

50.0

60 .0

49 . 8

s 8

23 4

45

52 . 2

48 . 8

42 .9

41

35 1

39 . 0

56 1

41

60.0

80 . 0

10

44

52. )

61 .8

2 36

36

40.0

54 .0

55

7

.

.

38 . 2

230 57

Falber 's mother

Jewish Protestant Catholic

Jewish Protestant Catholic

56. 1 40. 1 38 .5

41 49 36

42 9 52 5 40. 9

85.7 6 1 .0 56 . 8

66.7 53 . 5 57 .6

42 43

62 . 5 52.4

100. 0 71 .0

42 1

69 . 1

.

37 . 2

41 9

Catholic

.

4 3 .2 40. 3 44 . 2

For

.

31

.

55 . 2

53 . 9

8 1 .4

38 .7

52 . 3

60 . 9

46.6

39. 5

54 . 5

persons with European-born grandparents, Rho betwee n percent i n Standani with grandparent m 0. 02 .

J)ereeru familiar b

.

persoas with American-born graodparents, familiar wilh grandparent is 0.62.

perce..

Rho bet\\l·een

our father 's father "'�I')· we l and fairly

H� fShr)

8

248 76

wm not

American allegory aDd

percent i n Staadard Americaa cateJory

�Based o n responses �

236 66

grandparents

Protestant Catholic

a For

48 . 8

42 9

Protestant

Jewish

37 .8 47 . 4

40.8 47 .2

Mother's Mother Jewish

Means for all

40. 9

3 8 .9

Mother's father

.

and

t o question, ·'When you w e�e growi ns up. how well d id you know these relatives?(Grandfather on your father 's side, and so on . ,.amil iaru includes response catqories: �II. · · unfamiliar ' ' response categories .-e: NOt Jo �ll. I did 1t0t blow him (her), and

alilie .

•·

.

92 dents, 1 had to compute percentages with bases less than 1 0; but even this small group shows much consistency with others in the analysis.)

In the interview, the respondents were asked, "When you were growing

up, how well did you know these relatives?" The four grandparents were included in a I ist of consanguines: your father's father (grandfather

on

your

father's side), and so on. In the analysis, respondents are classified in t\�io

with the grandparent in­ cludes the response categories; (a) knev.' him very \\'ell and (b) kne\\1· him groups-familiar and unfamiliar. Being

fairly well. Being

wifami/iar

familiar

with the grandparent refers to the categories:

(a) not so well; (b) I did not know him; and (c) he was not alive.

Presumably, if the relative-as-a-person hypothesis is operative, there should be a linear relationship between percentages of persons knowing their gl1lJ1dparents well and percentages of persons with a Standard American ori­ entation. Persons who have known their grandparents well should conform

to the Standard American pattern of collaterality-regardless of religious category. Hov.·ever, if the ideological hypothesis is true, differences in per­

centages of persons in the Standard American category should vary according

to the religious characteristics found in Chapter 4--rather than by familiarity with grandparents. In order to test the relative-as-a-person hypothesis, Spearman rank corre­

lation coefficients v.'ere computed to determine whether the percentage of persons who are familiar with grandparents varies directly with the percentage in the Standard American category (over all religious groups and all four grandparents together). For persons whose grandparents �·ere born in Europe

the correlation coefficient,

rho,

is

zero

(rs

=

.02; N = 12),

which indicates

that there is no relationship between familiarity with grandparents and Stand­ ard American kinship orientation. For persons \\'hose grandparents were born

in the United States, however, the results are ambiguous. Although the rlw is high (rs =

.

62;

N = 12), an inspection of the data casts doubt on its mean­

ingfulness. First, among individuals v.rith American-born grandparents, there

is a constant percentage of Protestants in the Standard American category

(about 52 percent despite a considerable variation in familiarity with different

grandparents (50 to 71 percent)). Second, within religious groups there is

little relationship between percent in the Standard American category and

familiarity with grandparents. Third, Jews and Protestants both tend to adopt

the Standard American orientation and to know grandparents more often than do Catholics. The

rho

is thus strongly influenced by religious factors.

Further comparisons among Jews, Protestants, and Catholics rein force

(rather than dispel) doubts about the relative-as-a-person hypothesis. \\'hen

comparison is made bet"'·een persons whose grandparents have been born in

Europe and those whose grandparents

are

American-born, there is consid­

erable diversity among religious groups in familiarity \Vith grandparents. For

Jews, the average difference in familiarity with grandparents, by birthplace of the grandparents, is 26.2 percent (or

55.2%

versus 81.4%); for Protestants,

93 the average difference is 22.2 percent (or 38.7% versus 60.9o/o); and for atholics the difference is only 7.9 percent (or 46.6% versus 54.5%). These

C figures suggest that the amount of contact between the respondent's parents and his grandparents during the respondent's childhood is itself, in part, a tunction of religion. Insofar as contact with grandparents is influenced by

i the data imply that religious ideologies themseJ. ves dif­ rel gious community, fer in the salience they attach to relationships with grandparents--with a higher salience among Jews and Protestants than among Catholics. Familiarity with grandparents itself seems unimportant in holding a Stand­ ard American orientation. Rather, the extent of assimilation into American

culture appears to be a key element. Assimilation, in turn, seems to depend upon religious codes. When comparison is made simply between persons whose grandparents are European and those whose grandparents are Ameri­

can-born, religious grou ps vary systematically in conformity to the Standard American model. As Table 5- 1 0 indicates: 1. For Jews, the average difference between persons with European as contrasted with American-born grandparents is II percent (or 43% versus 54%). But among Je\\'ish respondents, the nativity of grandmothers seems

to be of particular importance-much greater than that of grandfathers-in inducing assimilation into the Standard American system. In traditional Jew­

ish law, it is the task of males to maintain ties with ascending generations

and of females to care for children ( and grandchildren). Sixty-one percent of Jewish persons with a American-born grandmother (N = 18) are in the Stand­ ard American category as compared with only 4 1 percent of those Yl'ith Eu­ ropean-born grandmothers (N = 83). This finding of a 20 percent difference implies that the respondents with American-born grandmothers have parents who themselves have been integrated into American middle-class culture.

l

2. Among Protestants, consistently 0 to 15 per cent more respondents with American-born grandparents hold a Standard American orientation than do those whose grandparents were born in Europe (with an average difference

of 12.0 percent). 3. Catholics, however, are least affected by birthplace of grandparents in their conformity to the Standard American pattern. If anything, having an American-born grandfather decreases the probability of their being in the

Standard American category (an average difference of -4.7 percent). In general, over generations Catholics seem to be less assimilated than Jev.,·s or Protestants into the Standard American system.

American and European Grandparents: Summary

This section has dealt with an issue regarding the basis for differences in conformity to the Standard American model between persons whose recent

ancestors were born in Europe and those whose close antecedants were born

94 in the United States. Ulti mately, the issue rests upon the kinds of factors

which influence kinship orientation generally.

One position is that, since the Standard American model gives a high

priority to grandparents, aunts, and uncles, people �·ho are quite familiar �·ith their relatives in ascending generations will tend to confonn to this

pattern. (Presumably, those individ u als who in childhood had known their grandparents well also were familiar with aunts and uncles .) From this per­

spective, which depends upon considering relatives as persons, differences between respondents with European-born grandparents and those with Amer­

ican-born grandparents in conforming to the Standard American model stem

directly from the fact that fewer individuals with European grandparents would have known them (andlor aunts and uncles). ..

The other position is that ideological factors provide a more powerful

influence than personal ties in the organ izati on of collaterality. The ideolog­

ical posi t ion is that certain religiou s norms and justifications facilitate accep­

tance of the Standard American model and encourage close ties with

grand­

parents. Hence, from this perspective, distinctions benveen persons with European-born and American-born grandparents in conforming to the Stand­

ard American model are greater among different religions than by degree of

fami I iarity.

In brief, the findings on assimilation of European families indicate that

(a) knowing a grandparent well per se seem s to play only a minimal role in

determining kinship orientation, (b) by the third American generation, the

effects of foreign origins (like those of minority-group status) are sustained mainly through the intervention of religious or other ideological factors; and (c) at least in an American social context, the various religions seem to differ in

the salience they attach to relationships with grandparents-with Jews and

Protestants assigning a greater import ance to them than do Catholics. (This

proclivity apparently prov ides Jewish and denominational Protestant ideolo­ gies with a good ''fit'' with the Standard American kinship model.)

At frrst glance, the finding that Jewish respondents with American-born

grandmothers have an affinity to the Standard American model seems to

contradict the data in Table 5- 1 that show that Jews are underrespresented in that model. But two other facts are also relevant, namely, that orthodoxy is

as sociated with the Parentela Orders model, and other investigations have

revealed that the number of generations in the U.S. is associated with languishing of orthodoxy (Farber et al.,

1977).

a

Loss of orthodoxy decreases the traditional emphasis upon lines of descendants and apparently permits the high salience of ties v.tith grandparents to become manifest. Consequently, the connection between having an American-born grandmother and holding a Standard American kinship orientation is actually not inconsistent with

previous findings. Rather it confirms them in describing assimilation to the American middle class from a kinship perspectiv e

.

(These conclusions

supported by the multivariate analysis in Appendix A.)

are

95 VI\tMARY: VARIATIONS IN COLLATERALITY S The opening section of this chapter posed the question: Are variations in

y in American kinship merely an elaboration of themes on a single coUateralt el, or do these variations reflect a diversity of ideological sources cultural mod In that section, it was proposed that three kinds of data of kinship norms? migh t shed light on the resolution of this issue: data on religion, data on specific socioeconomic variations in kinship mapping, and data on familiarity

with specific relatives. The information gathered in the Phoenix survey was analyzed to determine how religion, socioeconomic factors, and familiarity are rel ated to kinship mapping.

Religion With regard to religion, the findings are that (a) persons leaning toward orthodoxy in sectarian-like religious groups--neofundamentalist Protestants, �1onnons, and Jews-tend to hold a Parentela Orders kinship orientation to a greater extent than others do; (b) Protestants in the traditional denominations

are

overrepresented in the Standard American category; and (c) persons with

religious orientations esJX>using universalism-Catholics and those with no religious preference-are overrepresented in the Genetic and Canon Law classification. These findings are gene rally similar to those of other studies involving university students (Farber, 1977) or a Jewish community (Farber,

1 979). My interpretation is that the findings do confmn the proposition that

degree of pluralism in religious ideology is associated with type of kinship orientation.

It might be argued that it is the religious label, rather than ideology, that

is responsible for these variations. However, such an argument fails to deal with the fact that Jews tend to be overrepresented in the pattern of collaterality

found in Jewish sources (e. g .• the ML�hnah) and Catholics appear more often than do other religionists in the mode of collaterality described in the Canon Law of the Church. Nor does this argument consider the detai l that Western religions in general are explicit in their judgments regarding family life and

that, in particular, sectarian-like groups (such as the Monnons) organize appropriate family relationships in terms of the model appearing in the Pen­ tat euch. Hence, although these findings may not convince those persons who are ske ptical about the influence of religious ideology on kinship mapping,

neither would they sway anyone from the conclusion that ideological content has an effect on these maps.

Religi on appears to be augmented by minority-group status in structuring kinship maps. The analysis indicates that, when other variables are taken into ac count, income is less powerful than minority-group status (as Mexican­

�merican.

black, or American Indian) in the patterning of kinship orienta­ lions. This find ing suggests that membership in a racial minority group carries

96

with it experi en ce s of oppression which may generate a communalist tion-and a Ge netic or Canon Law kinship p erspec ti ve .

posj ..

Socioeconomic Status

As for socioeconomic characteristics the-mselves, the pattern of overrepre­ sentation in the findings is that (a) there is a tenden cy for persons in profes­ sional and managerial families to fall into the Standard American classifi­ cation; (b) persons in the midd le socioeconomic category (i.e. , clerical, sales, and craft occupations) are more often found in the Civil Law grouping; and (c) those in blue-collar families cluster more than anticipated in the Genetic and Canon Law pattern. The re sul ts on the Parentela Orders pattern sho·w the confounding influence of religion and ethnicity. (When the respondents are cl assifie d by family income or by educational level, the resul ts are similar to those based on occupation .) This configuration of find ings for socioeconomic characteristics contradicts the expectation in the Schneider and Smith ( 1973) formulation , name ly, that the middle class emphasizes the unity of the nuclear fami ly to a greate r extent than does the lowe r class. Instead, the middle-class respondents-profes­ sionals, managers, and administrators-stress line of descent at the expense of nuclear family ties, partic ul arly the ancestral line in the Standard American mode l . It is clearly the blue-collar respondents who, in h ol di ng Genetic or Canon Law persPective s , give priority to nuclear fami ly members over other kin. One mi ght argue that the blue -collar respondents place much value on the coherence of the nuc lear family because of the fragility of such ties in the lowe r class. However, other findings in this study, such as those pertaining to the Srole i ndex of anomia ( S role , 1956; Miller, 1977, pp . 375-377), fail to support thi s argument. 1 Ap pare ntl y aside from economic ad versity , people who regard social ties as fragile also displ ay other attributes of an alienat ed Weltanschauung-a sense of po werless ness , pessimism about the future state of soc ie ty , and a lack of trust in others . These elem en ts are tapped in the Srole index. As Table 5-11 indi cate s , the anomia attributes are more prevalent among people in lower than at higher socioeconomic levels, minority groups, low educational level , inc ome less than $10,000, and persons in blu e-col lar families. The fmdings on socioeconomic char acteristi c s of persons with high ano mia scores support the Schneider position; but other findings seem even more important for the resolution of the issue . Table 5- 11 al so shows that ( e x cept for persons Ylith a grad uate school education, with no religious pref­ erence, and with incomes over $25,000) people with Genetic and Canon La,�· mappings have a higher than average score on the anomia i ndex regardless of rel igion, minori ty-group status, educational level, family income, or oc­ cupation of the male co-head of the household . Indeed, for all social char-

97 anomia score in (with the few exceptions noted above), the mean cteristics of any of the Genetic and Canon Law category is actuall y the highest ions. It is the consistent tendency for persons with a Canon kinship orientat ardless LaW o r Genetic orientation to have a high score on the Srole scale, reg in the interpretation of ial char acteristics , that seems to be significant

�e

of soc

the find ings. and the Canon Lav.' (or Ge ne ti c ) orientation Both the Srole Anomia Sc.ale al vehicle s for social continuity are t he major eco­ i mply that (a) the princip the society rathe r than ic, poli tical, and religio us corporate structures in

nom

TABLE 5-11. Mean Scores on Srole ''Anomia'' Scale for Respondents in Genetic and Canon Law Category of Kimhip Orientation, by Social Characteristicsa Mean sc ores on srole ··anomia·· scale

Persons

in

genetic and canon law categor}·

Total sample

Jewish

2.11

1.30

�otestant

2.30

Catholic

2.02

1.15 1 . 77

Social characteristics Religion

No religious preference

Minority-group status Nonminority

Mexican Americans

Other (Black, American Indian, Orienta])

and

N 54

387

189

1.60

1.63

1.69 3.00

1.49

590

2.64

80

3.00

2 . 89

46

3.18

2.64

104

2.09

1.92

245

75

Educational level �01 a high school graduate

High School graduate Some College

1.73 0.56

Some graduate work

Family income in 1977

UDder $10,000 $10,000 to $24.999 S25,()()() or over

2.61 2.08

1.16

1 . 25

65

2.25

1 14

1.82 1.16

Occupat ional status of male co-head of household Professional, managerial, and

1.44

administrative Clerical, sales, and crafts

2.05

Blue collar a A

bigh

score denotes

2.89 a

high degree of

.. anomia.··

303

1.08

1.25 1.87 2.34

387 200

268

213

139

98

family and kinship and (b) everyone ought to have ample chance to participate

in these structures. An individual with a high score has a view of the world such that (a) life chances generally are diminishing; (b) the dimness of the future suggests that one should not bring additional children into the world· '

(c) the larger social forces are so overwhelming that one cannot pl an ahead effectively; (d) one cannot place trust in others; and (e) government 'NiH not

be forthcoming with realistic solutions. By implication, though, one ought to be able to do these things; the larger forces and the major corporate structures ought to facilitate life chances, provide an optimistic setting for a fulfilling family life, and be truly adaptive in the face of social problems and crises. The perception of failure by the larger social structure to deal effec­

tively with problems permeating the society leaves high scorer on the '·anomia'� index with only- a·-fragile institution to deal v.'i�ne 's nuclear family.

Social �lobility

The connection between kinship orientation and striving for achieveme nt is revealed in the analysis pertaining to intergenerational social mobility. These

findings indicate that Parentela Orders and, in particular, the Standard Amer­

ican model are associated with upward mobility and restraint on downward movement. At this point in the analysis, one cannot determine whether the kinship pattern resulted from the intergenerational mobility or whether some­ thing in the individual ,s family of orientation \\'as responsible for this move­ ment. The premise that kinship mappings represent more extensive ideolo­ gies, that may be transmitted from one generation to the next, leads me to the latter interpretation. The significance of the Standard American model for upward social mo­ bility and restraint on downward mobility suggests a tantilizing interpretation of the relationship between kinship orientation and achievement motivation. Since the Standard American map emphasizes a sense of closeness to relatives in ascending generations, one can interpret this finding (in a fanciful way)

to mean that achieving individuals regard their place. in society as reflecting

the "inherent" worthiness of their family line-surely their ancestors and hopefully their progeny. Such an orientation would also lead people who are currently of loYl socioeconomic status to ''prove'' the worthiness of their line by inducing a high degree of achievement motivation in their children. Car­ rying this interpretation further, one might even consider the Standard Amer­ ican model as a �'survival'' of the Puritan doctrine of the predestined Elect� whose Earthly success signaled their fated Eternal Salvation and set them apart from the mass of the society. But, of course, such an interpretation cannot be tested empirically.

99

Grandparents and Kinship �lapping the conclusion that European heritage The findings on grandparents confirm . (but only in conjunction \vith religion) has a bearing on kinship mapp i ng Distance from European roots seems to have most affected assimilation of Jev.'s in to the Standard American system and touched Catholics least. Still, Catholics with European grandparents tend to provide Parentela Orders map­

pings, while those with non- European grandparents are more concentrated in the Genetic and Canon-LaYl category (and not in the Standard American

pattern). Together these findings make a strong case for regarding European tradition, functioning together with religion and minority-group status, as an

element which shapes kinship orientation. Viewed collectively, the fmdings on familiarity with grandparents and conformity to the Standard American model do not support the hypothesis that kinship mapping derives primarily from factors pertaining to relatives­ as-persons. Rather, the findings indicate that birthplace and childhood fa­ miliarity with grandparents are dependent upon ideological factors to deter­ mine the ways they affect kinship mapping.

\'ariations:

Fundamental

or Superficial?

The thrust of the findings is that variations in patterns of collaterality in American kinship appear to represent ideological differences associated �·ith

position in the social structure rather than superficial elaborations of differ­

ences in personal attributes. But, the issue is not whether the configurations

of attributes that define relatives as persons are alone important in influencing

interaction between kin. Certainly, interaction with s pecific kin is influenced

by personal characteristics apart from relatedness-gender, age, common interests, personal resources, religious labels, pleasantness, and so on. But

a formulation that emphasizes these attributes as the primary basis for col­

laterality does not explain why individuals whose relatives have similar con­ figurations of attributes exhibit diff erent patterns of collaterality (or vice vers a). Rather the issue is whether patterns of collaterality imply, in addition to personal proclivities, a series of ideologies that serves to organize rela­

tionships with family and kin. True, one's perceptions of relativ es as persons may incorporate whatever else these relatives are (apart from their being

biologically related), but the grounds for these perceptions, it seems, depend upon th e kinds of assumptions about the nature of kinship and its relationship the g eneral social structure. My view is that these grounds are expressed

�0

In

terms of k i nship models that have emerged in the historical process . Schneider and I reach different conclusions with regard to the significance of vari ations in collaterality. Why? Sch ne i der ' s position is that these varia-

100 tions are superficial man ifestations stemming primarily from the fact that people i n modem societies participate in several social domains , "'·hile my position is that these variations in collateral ity themselves involve funda­ mental ideological distinctions . Perhaps a major source of this difference l ies in the methods we use to analyze kinship distances. Schneider relies heavily upon the genealogical method of collecting data on kinship ties. Because of the huge financial i nvestment, laboriousness , and time expenditure required to construct reliable and ''complete'' genealog ies, Schneider·s analyses are necessarily restricted to a relatively small number of genealogies. Much of Schneider's work is based upon 43 genealogies gathered in 1961- 1 963, supplemented by extensive i nterviews with 47 cou­

ples, several women , and 99 children and their mothers (Schneider , 1968� Schneider and Cottrell , 1975) . In addition , the analysis of lower-c lass kinship by Schneider and Smith (1973) is based on "59 Afro-American , Southern White , and Spani sh-American families liv ing in Chicago (of which 40 pro­ vided relatively full information) . '' The interviews covered a wide range of topics dealing with kinshi p , occupation, and status domains , but in addition they included a rdJlge of factual informati on of a genealogical nature. The amou nt of detailed information in these interviews and genealogies is im­ mense .

The richness of materials in Schneider's studies must reveal an extensive amount of variation (within the same genealogy) in the perception of social distance among relatives of equal genetic relatedness . The extent of this variation within genealogies undoubtedly has evoked questions by Schnei der regarding the reasons why an individual would make these distinctions. A reasonable response to the se questions (which Schneider provides) is that the individual perceives relatives as persons, whose genetic relatedness is but one facet of their configuration of personal and social attributes. From this point, it is only a short step to conclude that , since social relationships are reciprocal , individual respondents differ (i . e., variations bern·een genealogies occur) because of the di versity of individuals' own role configurations . For Schneider, the respondent too is a person whose non-kinship attributes influ­ ence his interaction v.·ith relatives. Thus , Schneider"s conceptions regarding the character of collaterality in American society seems to be tied to a par­ ticular mode of data collection which permi ts ample observation of within­ case vari ations , but \\'hich is limited in the extent to wh ich it allo"'·s the observation of regularities between groups in a highly heterogeneous society. l\.1y own method is intended to discount variations within individual ge­ nealogies and to focus instead upon general patterns of collaterality in a large diverse population. For example, my technique for classifying respondents by kinship orientation requires only that a majority of responses correspond to a given pattern . The procedure thereby rests on the assumption that vari-

101

gene alo gies are irrel evant in a ss es si n g an overal l pattern for an ations within d individual. Consequently , by i gno ring details of geneal og ie s , I can etermine

whether

reg u lari ties

occur in a differentiated population, and I then ascribe l ogic al factors. Whereas the focus o f Sch neider·s these reg u laritie s to ideo anal ys is is upon "the relative" (or, to use his term, Alter), the focus of my anal ys i s is upon the ind ivid ual (i.e., Ego) and the organization of ties w ith

relativ e s .

The di fference between S ch neider s and my conceptio n of collaterality in American k insh i p appears to be analogous to a controversy occuring in the '

field of l ingu istic s . Some l i nguists believe that the various regional, socio­ economic, and ethnic variatio ns in the use of English represent d i a lectal deviations from Standard E ng l ish ; others regard some of this diversity as

repre sentin g distinct systems , each \\ ith its o\\l·n rationale and set of usage rules, so that Standard English is but one of sev eral suc h systems. In looking at collaterality, S chneid er seems to take the "dialect" position, while I hold '

the "system s'' position .

NOTE

1Carr ( 1971) reports a study sh ow ing that the Srole index is affected by tendencies of blac ks

of low socioeconomic status to ac quiesce to any statement in an inter\'iew. Carr's criticism, however, is not appl ic ab le in t his investigation.

1. First, unl ike his sample, Phoeni x respondents of lower occ upation , education, and income

levels do not overwhelmingly tend to agree wi th the Srole it em s . In the Carr sample. the rat io of responses in the high catego ri es (4,5) as compared with those in the low cate gori es (0, 1) is

19.3 for blacks at the lower end of socioeconomic scale. The Phoenix investigati on . however� shows little such skewness in responses toward acquiescence. In the Phoenix study, the ratio of

high to low score s for persons with less than four years of high school is I. 04; t he ratio for

persons with family incomes under $10,000 i s .67; and the ratio of respondents from blue-collar

families is 0.8 1 . Blacks and Mexican American s indicate a slight skewness toward agreement.

(For blacks, the ratio

IS

I. 79 and for Mexican Americans 1.25.) But this degree of skewness

cannot be interpreted as indicating that the S ro le index measures acquiescence rather than a fonn of disillusionment with society.

2 . A sec ond reason for rejecting the acqui e scence vi ew is that the findings on the re lationship

between anomia and ki n sh ip orientation persist e ven when such "'ariables a� re li gion , minority­ group status, educational level, family income. or occupational statu s are hel d constant statis­ tically· Similarly, when Lenski and Leggett ( 1960) remo ve effects of acquiescence from their sample, the rank order of percentages of persons in various socioeconomic catego ri e s who agree with a Srole item rem ains unchanged.

3. A third basis for rejecting the acquiescence position is that in a ater chapter ( Table 7-4) tbe associati on between anomia and kinship orientation is sustained when the respondent's own score on anomi a is compared with his estimate of his spouse's answers lo the Srole questions.

l

This comparison is made by the individual, and many distinctions be t wee n self and spouse occur � all socioeconomic levels. It thus appears that individual s in the Phoenix study have little diffic ulty in discri minating betw ee n anomie reaclions of self and other.

102 In summary, three reasons for rejecting the Carr position have been put forth: (a) the lack of skewness toward acquiscence among low socioeconomic status respondents; (b) the persistence of statistical relationships e,·en when socioeconomic variables are held constant; and (c) abilitv of individuals at all socioeconomic levels to distinguish between anomia in one's self as co �­ pared to spouse. The findings pertaining to the Srole index thereby argue against the Carr interpretation that it represents an acquiescence scale rather than a sense of disillusionment or anomia.

The findings in the previous chapter indicate that kinship orientations are as sociated with religious tenets and socioeconomic status. Insofar as conceptions of kinship represent ideo­ log ical po sitions, one would antici­ pate that they tend to be transmitted from one generation to the next in family lines, from parents to chil dren . Accordingly, families of persons with dif­ ferent kinds of kinshi p orientation should differ in important ways from one another. Diverse kin ship orientations should thus reveal particular character­ istics not only about the res pondents themselves but also about their parents and siblings. Here I shall di scu ss the relationship between kinship orientatio n and attri­ butes of the respo ndent s ' parental family. The frrst section sets forth a series of basic propositions regarding this relations hip ; it refers to the ideologic al aspects of factionalism and communalism. Su cceedin g sections deal with fmdings pertaining to respondents' parents and si bli ng s , and the y involve such characteristic s as preva lence of divorce, religious en dog amy , age at marriage, maternal e mployment , and fertility patterns.

Family

of Orientation

IDEOLOGICAL ASPECTS OF FACTIONALISM

AND COMMUNALISM

As the earlier chapters have proposed, charac teris tic s as sociated with kinship orientation s seem to be derived from the social positions of population seg­

ments and, ultimately, from the general social structure itself. This section deals with the ideo logical content of factionalist versus communal structures in society , particularly as this content refers to marriage and family ties.

Earl ier chapters distingui sh betYleen s oci al structures giving priority to special interests of a single group-factional i sm-and those giving priority to c om mon interests of all grou ps in a society-communalism. Factionalism implie s a situati on in which special interest groups are vying for su periority over other groups for econom ic, politic al , or religiou s " good s "-that is, ac cess to wealth, to power, or to eternal or secular salvation. Since faction­ alist polity re fers to organization aimed at procuring or m ain t aining superi­ ority, it s pre.sence may be regar ded as a mechanism for stratifying a societ y .

1 04

Factions are, after all, a means for the gathering of forces and of mobilizing members for conflict or competition. Insofar as factionalism in kinsh ip or­ ganization involves this pulling inward and engaging the obligation and/or loyalty of members, it seem appropriate to call this type of kinship organi­ zation centripetal. Communalism, on the other hand, implies a situation in which special interests are subordinated to common concerns. In stateless societies these common concerns may well emerge from economic interdependence or the presence of a common enemy. In societies with more highly developed gov­ ernmental institutions, other common interests may exist as well-the pres­ ence of a universal religion (as opposed to sects and denominations), nation­ alism (as opposed to subnational "ethnic" identities), centralized political bureaucracy (Weber, 196 1, p. 5 1), values of political, social and economic -equality, and so on. For example, Max Weber repons that in medieval Eu­ rope, the Church �'strove to abolish the rights of the clan in inheritance so that it might retain land willed to it. The common concerns would best be served if members of kin groups were to be d ispersed throughout the society. In this manner, the maximum number of social networks would be created to scatter the loyalties and obligations of any individual as widely as possible. Under the conditions of communalism, major forces would exist to minimize property-based familial ties and to expel family members outward; it thus seems appropriate to regard this kind of kinship associated with communalism as centrifugal. As a consequence, those kinship systems that are oriented toward the dispersal of human property throughout the society (as a matter of common interest to all kinship units) stimulate the homogenization of society and bolster the development of communal regimes. Because of an orientation toward status maintenance in an indefinite future, kinship systems that tend to emphasize saving (i.e., centripetal systems) involve long-range perspec­ tives, establishment of mechanisms for stabilizing obligations, and a redun­ dancy of family roles. Hov•ever, because of an orientation toward immediate psychological and physical comfort of the mass of the population, other systems which stress distribution of people and their property throughout the society (i.e., centrifugal systems) involve short-range perspectives, use of mechanisms for maximizing cross-pressures, and a minimum of jural kinship obligations. Many observers have noted the ''fit'' bet\\·een the decline of the signifi­ cance of the extended family and the rise of modem urbanized industrial society. This fit is attributed to ease of mobility and the reduction of obli­ gations which might impede individual destiny (Goode, 1 963: Burgess, Locke, and Thomes, 1963). Still, even in societies that ideological ly uphold communal norms in family organization, there may be strong motivations among those groups vying for power to adopt centripetal norms. Although '·

1 05

elites i n the United States around 1800 were not constituted as competing formal c.orporate entitie s , they organized themselves i nto collections of fam­ f ing political factions, pooling resources, intermarrying with one il ies, onn another, and creating almost separate soci al worlds (Farber, 1 972). Indeed, in the abse nce of constraints on familial accumulation of resources , one \vou ld anticipate that rival elites v.'ould favor centripetal norms of kinship f or them­ selves (but perhaps not for others). Cen tripetal and centrifugal tendencies in kinship seem to generate kinship no rms that are related to the ki nsh ip models under investigation. In the case of ce ntripe tal kinship this re lationship appears to occur in the following way:

1. When members of a speci al-interest group, at some time in its history,

reach a consensus that its continued existence (or position in society) as an ident ifiab le entity (or as '-successful'') i s threatened by the incursion of other groups- who wou ld 'ioraid its members or " .. reduce'� its pos ition in soc i­ ety-the members v.'ill resort to the strategy of mobilizing the group to resist '·

this incursion. 2. If the threat persists, this mobil ization stimulates the group to organize itself

in

a centripetal manner, that is, to create sharply defined boundaries

between it and other groups and to draw its members inward.

3. The mobilization fostered by threats to the long-run integrity of the

special-interest group also requires a high degree of compliance (both coer­ cive and voluntary) by the members to promote its en ds .

4. As a fundamental institution that enables the. group to persist as an

identifiable entity (or to perpetuate its position in society) over generations, kinship too is governed by rules emphasizing centripetal tendencies.

S. The mobil ization of centripetal ki nship groups to draw me mbers inward \l;ould encourage strong collective efforts to mai ntain finn group boundaries and to sustain famil y stabi l ity.

6. Consequently , groups characterized by socioeconomic , religious, or

other speci al interests that depend for their perpetuation upon centripetality

in kinship would, over a series of generations , adopt fam ily and marriage norms that foster thi s inward movement. Such n orms would include endo­ gamy , reluctance to resort to divorce, high priority given to the maternal ro le , and speci al meani ngs appl ied to fertil ity.

7. Since Standard American and particularly Parentela Orders kinship mode l s appear to represent centripetal approaches to kinship organization, the norms specified above should be more pronounced among persons hold ing the se orientations than they are among people with C i vii Lav.' , Genetic, or Cano n Law orientat ions. Co mplementing this set of statements on centripetal kinship norms is an­ other set showing how communal ism as an ideology is associ ated with cen­ trifu gal kinship nonns: When people regard any special-i nterest group as functi onal ly equiva-

l.

1 06

lent to other groups in the soc iety ., they have little stake in the perpetuation

of any given soc ial entity . (But where any particular group is cons id e red indispe nsible for everyone 's benefit, e . g . , the Catholic comm unity , the a i m

is gene rally to u niversalize that commun ity , to identify it \\'ith the com m on

good . )

2 . I ns t e ad , people. consider it to be imponant that spec i al i ntere st s o f

d i ffere n t groups be subordinated to the com mon good and that the po \\-· e r of

the se speci al - i n terest g rou ps be elimi nated .

3 . Th e po\\'Cr of special-interest grou ps is mi n i m i zed if i n d i vi d u al s

are.

p l aced u nder cross-pre ssure s as mu c h as pos s i b l e . \Vi t h confl ic t i ng loyalti e s ,

peop le are. le ss l i kely to b e mobilized to g i ve priority to a s pe c i al i nterest . 4 . C ro ss-pressure s are

maxin1ized \l.t'he n bou ndaries bet\\'ee n g rou p s

vague and there arc stro n g centrifu gal te nde ncies in grou p organ i zation .

are

5 . i\.s a fundame ntal i n s t i t u t i on in a soc ie ty � kin shi p al s o would be go v ­ e rn e d by ce ntrifugal ru le� that stress th e creation o f a s man y ne t works a n1ong

fant i l ie s as possible .

the perpetu al i on of spec i al -interest groups , w h i le at t h e s a me time , it \-..· ouJd maxin1ize the n u n1 ber of i nd i v id ual rel ati v e s \\'ith w hom personal tie s ( and 6 . The diffu se ne s s of ce ntri fug al kin ship net\vorks Vlould min in1 i ze

pe rs on a l demands ) c a n b e sustai ned . 7 . Consequently ,

families c h aracte r i z ed by c e n t ri fu g al ki nship n o nn s

,-,.·o u l d emphasize personal ties in do me s t i c

re-lation sh ips . I n the i r atte nti ve ness

to per s on a l q ualities in d o m e stic relatio ns , these famil ies V.'ould be inc l ine-d to be tol erant about departures from e ndog amy , d ec is i o n s to divorce , the

pe.rfonn ance of the maternal role , and rate s of fertility .

8 . S i nce Geneti c and especially Canon L,av.' kinsh ip models appear to

represent c e ntrifugal perspecti ve s in k i n s hi p organ ization , the norms l isted above s hould be prevalent among pe rsons \\'ith t he se kinship orientations .

The succeeding sections of this c ha p ter are devoted to te sti n g the rel ati on ­

ship be t ween kinship orientat ions and norms associated v.'ith cen tripetal ver­ sus ce ntrifug al tendencie s .

KINSHIP ORIENT ATION AND DIVORCE

Soc ial scientists have noted that prevalence of divorce is related to the Y�· ays

soc ietie s are organized . Using the Human Relations Area Files , Ackennan

( 1963) has shown that divorce rates are consistently low in those bil ateral

soc ietie s where there is community endogamy and frrst-cousin marri a ge and in those unilineal societies in which the lev irate is present. In addition , Mu r­

reports that in societies permitting frrst-cousin marriag e , rela­ tionship s between cousins are seldom fonnal; similarly , in unilineal socie tie s

dock ( 1970)

with levirate or with sororal marri age , brothers- in-law and sisters-in -law te nd

1 07

to

be i nformal in interaction . This tendency toward infonnal it y (and often

sexual l ice nse)

im p l i es a sense. of closeness amon g these rel atives . N o ti n g

that societies display ing a high degree of factionalism also tend to be char­ acte rized by community endogamy , fi rst -cousi n m arri age , or the lev irate , arber ( 1 975) found th at soc ieties Vlith fac t i o nal re gi mes al so t e nd to have F to w d i v orce rates , while societies �·ith communal regimes have h igh divorce

rates . Fin ally , cross-cultural research on indu stri al societies i n d ic ates that div orce rates are inverse ly related to soc ioec on om i c status (Goode , 1 962 ) . Given this array of fi n d i n g s on the rel ati o ns hip between soc i al structure and

divorce in societies with v a ry i n g de g re es of complex ity , it seems reasonable

to e xpec t that kinship orientation-wh ich al so appears to be assoc iated with locus in the social structu re-i s related to divorce . Insofar as ki nship orien tation has an i deo l og i c a l basi s , its i m p act should be felt throughout the re spo n de nts · fam i l ie s-n ot only in the p aren t a l gen­ eration but al so amo ng brothers and sisters . Table 6- 1 present s data on the. prevalenc e of divorce among the respondents · parents , by kins h ip orientati on and by t he re l igion in wh ic h t h e respondent s \Vere rai sed . The table indicates that for the diffe re n t rel igiou s groups , p rev al e nce of d i vorce i s lowest among parents of Jewi sh respond en ts and next lowest among Protestants and Cath­ olics , and divorce i s mo st v.'idespread among parents of persons raised in

secular home s . When kinship orientation is taken into account for each re­

ligious group , the results are. as foll ows:

1 . W h i l e Je w i sh re sponde nts with divorced p are n t s are u nderrepresented

in the Pare n tel a Orders c lassification , they are fairly h igh in the remaining categ ories of ki nship o ri en t at ion , al th ou gh sti l l lo\\' in compari son w i t h the ·

other rei igious groups .

2 . Protestants w i th di fferen t k i n s h i p orientations sho w little variati o n i n

pre v ale nce of p are n t al divorce : there i s , hov.'ever , a s l i g h t tendency for di­ vorce amon g parents to be o verrep re se nted in the Gen e t i c and C anon Law c at egory .

3 . For C athol ic s , d i v orce occ urs l east ofte n among fat he rs and mothers of perso ns in the Parentela Order s c lass and mo s t frequently amon g those in the Civ il Law , Ge netic , and C anon Law categori e s . Table 6- 1 al so pre sents information on prev al ence of d i vorce among re spo ndents " parents for different socioeconomic groups . Jn the tab le, persons are cl as s ifi ed by t he i r father 's occupation prior to the adulthood of t he respon­ de n ts. Co ns iste nt y,rith other investigations , the data sho � a clearcut rela­ ·

tion sh i p between socioeconomic status and divorce : fathers and mothers in the professi onal , managerial , and admini strative c lass are least l ikely to di­ vorce , wh ile the blue-collar parents are the most i nc lined. But �·ithin soci o ­ econom ic s trata , kinship orientations still differ in prevalence of marital di sruption. For all three strata, divorce is w i de spread in the Genetic and

1 08

TABLE

6-l.

Prevalence of Divorce among Respondents' Parents , by Kinship Orientation, Religion in Which Respondent ��as Raised , and Father's Occupation Parentela

Standard

orders

american

Genetic

Civil law

or

canon law

Total

Religion in which respo ndent was raised No rei igious

preference (%)

N

Jewish (%) N

30 .0

33.3

20

15

6 . 71'

15

Protestant (%)

22 . 4

Catholic (%) N

1 S . 8b

N

67

22 1 8 .8

20. 0h

K

Entire sample (% ) N .. Ratio of cell h

1 97

21 .3

89

percentage

26 . 2a

72

84

26. 73

420

25 .0

2 1. 9

44

20 1

1 6 .7H

1 5 .4a

1 3.4

39

2 46

30

23 . 1

42 23.7

35

1 08

27. 5

37

91

37

1 9 .0

1 9. 5

1 26

3 33

2 1 .6

to total percentage is 1 . 2 is .8

38

24.3b

1 16

or

or

54

20.8

29. 7

Ratio of cell percentage to total percentage

11. 1

17

Father 's occ upation while respondent was growing up Professional, managerial , or administrative (%) 1 2 .4 1 1 .5 N 52 1 13 Clerical , sales , or craft (%) N B lue collar (%)

35

1 1 .8

1 3 .68

38

3 1 .4

28 . 3 " 46 3 5 . 38 51

27 . 7� 141

2 1 .2

23 . 8 227

29. 2 2 16 2 1. 4

7 16

over.

under.

Canon Law c ate go ry ,

and except for the blue-collar group, it is le a st prevalent i n the Parentela Orders category . The overall p attern of pre valence of divorce among the respondents par­ ents is generally repe ated in the next generat i on . Table 6-2 reports the per­ centage of respondents ' siblings \\'ith previous marri age s . (Some of the sc s ibl in gs have rem arried; others have re mai ned as divorced persons.) The percentage of previously married brothers and s iste rs is lo west in the Parentel a Orders category ( 1 9 . 8 percent as compared with 1 9 . 0 perce nt for thei r par­ ents ) , and it i s highest for s i bl i n gs of persons fall ing into the Gene t i c and Canon Law c lass ( 27 . 2 percent as compared w ith 27 . 7 percent for parents) . The per ce ntages for sibl ings of persons with S t and ard American and Civil La\\' orientations are intermed i ate . Thu s , as anticipated , both parents and sib I ings of pers on s with a Parentela Orders orie n tation show the highest degree of marital stabil ity , and families of individuals in the Genetic and '

1 09 TABLE 6-2. Kinship Orientation and Percentage of Respondents' SlbUngs with Previous Marriages (Whether or not Currently Married) Respondent s '

s iblings Pc.rcent with previous marri ages

Total number of siblings ii

Parentela orders

Standard american

Civil law

Genetic or canon law

1 9. 8

2 1 .9

24 . 7

308

'1:1 . 2

67 5

247

298

Inc lude s siblings born before 1 96 1 for siblin gs groups of si x or fe\\'er brothers and sisters . For each size of group up to the point, there are at least 20 families; tbe N decreases rapidly for larger sized sibling . ups gro •

tbe sibling

Can on LaY.' classification have the greatest propen sity toward marital break up . These tendencies still persist even when the considerable effects of religion and socioeconomic status are taken into account .

KINSHIP ORIENTATION AND RELIGIOUS ENDOG Af\iiY Rules govern ing religious and ethnic endogamy can be regarded as extensions of those covering kinship endogamy . Of particular interest in European his­ tory are the Jewish stet/ and German kinship arrangements , in that both systems applied the Parentela Orders procedure in detennining collaterality . Both Jewish stetl and the old German kinship structure involved a highly

integrated bilateral kindred-the Jewish mishpokheh and the Gennan maegth or magschaft . (See Huebner , 1 968 , v . 4 , p . 587 . ) Signi ficantly , both Jewish and old German systems had rules of preferential marriage with close rela­

tives . Early German law permitted marri ages between any kindred outside the nuclear family , and evidently , 'so-called endogamous marri ages seem to have been the rule; that is , marri ages between members of the same sib (Hu ebner� v . 4 , p . 594 and 604) . ' ' In Jewish stet/ kinship , cou sin marriage \\'as also widespread and stron g bonds were maintai ned between co- pare nts­ in -l aw , w ho were often times siblings . There is , hence , some basis for spec­ ul at ing that Parentela Orders kinship orientation is associated with endoga­ mous rules of marri age . 4>

Tab le 6-3 presents data regarding the extent to which the parents of persons With different ki nship orientations married acro ss boundaries of maj or reli­ giou s group s-Judaism . Protes tantism , Catholicism, or secularism . The table efers to the religions in v,rhich the parents had been raised : it does not take Into ac cou nt premari tal or postmarital conversions . Even though the data may

�

�nceal some conversions , they still reveal considerable variation in religious

Intermarri age by kinship orientation . Si nce the prevalence of i ntermarri ages among Jewish parents is lov..· (below ten percent) , Table 6-3 describes only the extent of intermarri age for women

1 10 reared as Prote stan ts or Catholics . For respondents ' mothers raised as Pro. testants, intermarriage is con s iderab l y less widespread among parents of per.. sons in the Parentela Orders

and Standard American categories than

of t hose

in the Civil Law , Genetic , or Canon Law classes . For those whose mothe rs h ad been reared as Catholi c s , however , only persons with a Parentela Orde rs

orientation consiste ntl y come from hom es in which both parent� h ad been raised as C atho l ic . Similarly , for the entire s ampl e , the Parentel a Orders c ategory stands out as e xh ibiting re li g ious e ndog amy among parents of respo nd ent s . The re l ations h i p between kinship orientation and rel igious

ho mogene ity of fam i l ie s is particularly c le ar when the relig i on of the respo nd en t s brothers .and si sters is examined. T ab l e 6-4 shows t h e percentages of respondents w hose siblings are all in the same re lig i on The table excludes persons who are only children or who have ju st one sibl ing . The table doe s not take i n to account , however, shifts in rel igious id enti ty that have occurred among sib­ lings during the ir adulthood. Still, if re l i g ious ideology is a strong detenninant of kinship orientation , then religious homogeneity among sibl ings does sug­ gest the stubborn pers iste nce of ki ns hi p orientation as a fami l y attribute even after th e married children have established the i r own homes . For th i s analysis, the d i versit y among Protestant denominations is seen as representing some­ what divergent k i ns hip i de ol og i e s , and when siblings appear in two or more Protestant ca te gori es ba si c al ly different in doctrine (i . e . , Re fonnat ion Era , Pieti stic , Neofundamentalist , or Mormon) , they are classified as be i ng het­ e rge neous in relig i on Table 6-4 i nd icate s that rel ig i ous homogeneity amon g sibl ings is highe st '

.

.

among Jewish respondents , next among Cath olics , then Prote stants , and low­

est among sec u laris t s . The fin d ings on kinship orientation is as follows: I . For persons with at l east one sibling with no rel ig iou s preference , hom-

TABLE 6-3. Percent of Respondents' Parents Who Intermarried Across Religious Groups, By Kinship Orientationc

Religion in which

Parentela

orders

american

Total sample (41 )

14. 8h

26 . 0

1 08

27 3

mother was

raised

N

Mother raised as

N

Prote stant (% )

Mother raised as Catholic (%)

N

a

Ratio of cell percentllge to

b Ratio

of

cell

percentage to

20.011

24. 1

Genetic

or

canon law

Total

27 .7

27 . 3

24 .6

101

1 32

614

3 1 .�

32.43

26 . 4

law

158

64

71

348

9.7"

26 . 3 H

25 . 9-'

22.7

22 . 5

31

76

27

percentaae

is 1 . 2

is .8

or

or

44

178

o"·er.

under.

lnlennmiagc � refers to marriage between persons raised Protestant, Catholic , or sec u lar (i.e. � n o religious preference) .

r.

Civil

55

Total percentaae Total

Standard

in differeat major relirious groups--Je wi sh .

11 1 ABLE 6-4. Kinship Orientation aud Percentage of Respondents Whose T Siblings Are all in the Same Religion Percentage of fami lies in which respondents ' siblings are all in same religionc Religion of at least one siblin g No

preference (%)

Parentela orders

Standard american

Civil l aw

Genetic or canon law

Total

so .oa

S l .S•

22 . 2h

2l . lb

40. 0

14

N

Jewish ( % ) �

52 . 9 51

N Cathol ic (-I ) N

Ratio or cel l pcrcentaJe to

of cell percentage

:1

Ratio

Eu:ludes

d

to

1 04

total percen&age total percen&age

7 1 .4

68 .2

72 . 9

59. 1

252

is 1 . 2

is .8

ss

44

28

75 61 .5

30

65 . 4

28

54. 6

51

70 .7

75

85.7

58 8 .

1 38

19

d

45 .511

57.2

86. 7"

N All religious groups (% )

c

9

d

23

Protestanf! (%)

a

33 91 .3

88

295

1 77

46 . 5 11

58.9

1 13

5 57

or over.

or

under.

families in wbicb respondenl is only child or bas onl)' one sibling.

Fewer than S cases. locluded in computatioas for tolal perceatages .

" Categories of Prolestant denominalions \\!"hen siblings as

beiD!

appe• in two or more

in different

religions .

arc

Reformation

Era., Pietistic,

Protestant cate:Bories (or

as

Neofuodameotalist, and Mormon .

Jews, noae , or Catholics) ., they

arc

class ified

ogeneity among brothers and sisters i s greater among Parentela Orders and Standard American respondents than among those in the Civil Law , Genetic , or Canon Law categories . 2 . Among Jews , religious homogeneity among sibl ings is high regardless of kinship orientation . Yet , the percentage i s slightly higher among persons in the combined Parentela Orders and Standard American category than among the fe\\' individuals in the remaining classifications . 3. Protestants reveal most sibl ing homogeneity with re spect to religion in those kinship categories most often associated with Protestantism-the Stand­ ard American and Civil Law models . Like the other religious groups , though, Pro testants show least homogeneity in the Genetic and Canon Law classifi­ c ation . 4 . Unl ike Protestants , Cathol ics in the Parentela Orders category are most homogen eous but, like Protestants , Catholics in the Genetic and Canon Law cl ass ind icate greatest diversity in siblings � religion . "Wnen all religious groups are combined , the results are quite clear: Par­ entela Orders category i s highe st in percentage of homogeneous sibling grou ps , and the Genetic and Canon Law class the lowest . Therefore , the fin din gs on the religious homogeneity among siblings , like those on parental e ndogam y , are consistent wi th the characterization of Parentela Orders kin -

1 12

s hip orientation as ce n tripetal and Genetic and Canon La\\' perspecti v e s centrifug al .

as

KINSHIP ORIENTATION AND AGE AT MARRIAGE which mark transitions in life cyc l e appe ars to reflec t modes of insitutional p artici p ation . Traditionally , marriage has rep resen ted

The timing

of events

one of the most · significant transitions in a person 's life , and the signi fic ance

depends upon its relationship to other status changes that occur in the l ife co u rse Historical analysis suggests that age at marri age is rel ated to social c on straint s on the dec i sion to marry . People have be come of age at marriage

.

relati vely more free to marry

accordi ng

' '

to pre ference

[ and

t i mi ng ]

because

structure conditions impose fewer obstacles to matrimony than was once the ca se (t\1odell , Furstenberg , and S trong , 1978 , p . S 1 29) . " In recen t decade s , social co nd i tions indeed , seem

to

' ·pose fev.'er i mpedime nts t o arranging the

l ife course ad l i b ( Model l , Furstenberg ,

re su l t , ' 'The basis for

'

and S trong 1 978 , .

p . S l 47) . , . .�s a

de ci s i on · about marri age ti mi ng has ch an g e d

from in v o luntary to preferent i al , from a structurally

vidually determined basis

S l 33) . ' �

(Model l , Furstenberg ,

c o ns trai ned and S trong

to

,

.

.

.

an indi­

1978 , p .

Given the n orm of free mari tal choice , in the absence of social c on strain ts ,

there i s

little reason to del ay marri age once the right spouse has bee n found .

I ns ofar as ce ntripeta l ki n sh ip sys tems general l y pl ace gre ater c on stra ints upon

life-course transitions th a n d o ce ntri f ug a l s ystems , one w ou l d anti c ipate that

t hese constraints would opera te in marri age as we l l . Prev ious analyses in this c hap ter h a ve alre ad y i ndicated a lo \ve r prev al ence of divorce and greater degree of rel ig i o u s homogamy among parents and sibl ings of persons in the Parentela Orders category . One would expect then the app l ic at ion of re straint also in e nteri ng marri age On the other hand , given the empha s is in C an on Law ori e ntations to create new t i e s to other kinship groups , Canon La\�· persons would be encouraged to marry early . Conseq uently , the Phoenix d at a .

shou ld reveal that parents of persons in the Pare ntel a Orders category are hi ghe st in age at marri age and parents of those in the Genetic and Canon La\�· c las s are the lowest .

Table 6-5 describe s the rel ati on sh i p betvieen kin s hi p orientations and me­ dian age of mothers at the ti me they married the re spon dent s ' fathers . In the table , two control variab les are i ntroduced : rel igion in \Vhich

the

re spon dents

v..·ere raised an d their fathers ' occ upation at that time . C onsistent with the findi ngs of other studies , the data indicate that Je\vish respondents ' pare nt� tend to marry late and Protestants e arly and that more persons in professio na1 . m a nagerial and admini strative positi ons g ene ra l l y del ay marriage t h a n do ,

those i n bl ue collar j ob s .

1 13 4BLE 6-5. Kinship Orientation and �tedian Age at Marriage of T.. Respondent's M other , by Religion i n Whi£h Respondent \\'as Raised and Father's Occupation Parentela

Standard

orders

american

Civil law

Genetic or canon law

Total

R eligion in which respondent was rai � No religiou s preference (yrs . )

N Jewish (yrs . )

N Protestant (yrs . ) N

Catholic (yrs .) N

20. 8

20 . 0

19

14

22 .0

14

23 .7

25 . 5

20. 5 33 2 1 .3

9

22

8

20.5

1 9. 9

18.9

1 9.6

62

1 82

19. 3

67

375

22 .6

20 . 5

20. 3

20. 1

20. 8

35

78

26

35

1 74

2 1 .4

Father �s occupation at time when re spondent Professional , managerial or 22.2 administrative (yrs . )

was

64

2 1 .3

2 1 .0

20. 6

Clerical, sales or craft (yrs.)

111

2 1 .5

2 1 .0

40

36

Blue collar (�Ts .)

20.0

20. 1

34

83

Entire sam pie (_yrs . )

2 1 .4

1 17

N N

N

33

53

growing up

49

N

23 . 3

10 1

20. 1

20. 7

35

42

1 9. 9

18.9

20 . 9

20.4

20. 3

305

1 03

1 25

27

43

236

20. 8

21 1

1 9. 8

1 87

The ge neral pattern of re su lts in Table 6-5 is that , de sp i te minor variations

in different rel ig io u s an d soc ioeconomic groups , mothers in the Pare ntela

O rders category have the h ig h est median age at m arri ag e , and tho se in the Genetic and Canon La�· c lassification have the lo west m edi an age . Among religious groups , thi s tendency is clearest for Pro tes t ants and Cathol ic s �

among socioeconom ic groups , this tenden cy is s harpe st for both t he top

(professional , manageri al , and administrative) and the bottom (b l ue -co ll ar)

categories . The Jewi sh mo ther s , who marry late anyway, show l itt l e syste.m­

at ic v ariation ki n s hip orientation on age at m arriage . But, de sp i te this de via­ tion , as a v.'hole , the data for m others support the c onte ntio n that persons With a Pare ntela Orders o ri ent at ion will te nd to be c o ns trai ned to marry late , While per sons with a G e net ic or Canon Law orientation Y�·ill be free (and perh a ps encouraged ) to marry early . The dat a for fat he r s are somev.'hat more ambiguous than those for mothers . Table 6-6 pre sents the median age at marri age for the respo ndents , fathers .

Li ke th e prev ious t ab le , this one al so appl ies re l igio n and father ' s occupation as c ontrol variable s . As in the ca se of mothers , Jewish fathers are ol de st at

1 14

marri age and Protestant fathers the youngest, bu t whereas the profess i onal � managerial , and admini strative fathers have the highest med ian age , it i s the middle-range wh ite-collar workers (rather than blue-collar workers) w ho hav e

the lowest median . This shift in pattern of of findings results from the l ar ge age discrepancy betv.•een hu sband and wife in the blue-collar class ( i . e . , a median difference of over 5 years) .

A review of the general configuration of the data in Table 6-6 reveal s th at ,

whereas the median age for fathers of Parentela Orders respondents ten ds to be high ( as expected) , no definiti ve pattern emerges for the remaining kinsh ip

orientations . The complexity of the findings suggests that numero us co n­ strai nts have impinged upon the marital decision of the fathers of the re spon­

dents to affect age at marri age . Since many of these men had been married in the years between 1935 and 1 95 5 , their l ife course seems to hav e been compl icated by famil ial , occupational , and educational adaptations ari sing

from World War II and the Kore an conflict . (Unfortunately , data were not

collected for age at marriage of respondents ' siblings . )

TABLE 6-6. Kinship Orientation and Median

Parentela orders

Standard

american

Marriage of which Respondent Was

Age at

Respondent' s Father, by Religion in Raised and Father's Occupation

Civil

Jaw

Genetic or canon law

Total

Religion in which respondent was raised No religi ous

23 . 0

preference (yrs . )

I'

25 .0

Jewish (yrs . )

N

Protestant (yrs.) N

Catholic (yrs . ) N

16

27 .7

27 . 0

26.0 14

26 . 5

24.0 30 27 . 0

8

9

52

23 . 3

23. 1

23.0

23.4

171

63

56

347

14

21

24 . 9 57 25 . 8

24. 5

23 .0

24. 7

24 . 7

31

74

24

30

1 59

23 . 8

23.9

24 . 9

Father 's occupation at time when respondent was growing up Professional , managerial administrative (yrs . ) N

Clerical , sales or

craft (yrs . )

s Blue collar (yrs .) N Entire sample ( yrs� N

or

25 . 9

24 . 8

49

1 08

39

36

23 . 8

22 . 9

24. 0

23 . 2

31

96

34

39

24 . 3

24 .0

23.3

24 . 7

27

73

2SJ)

26

23.9

'2 3 . 7

1 07

286

100

38

24 .0 115

232

23. 1 200

24.0

1 64

1 15

1}1E l\1A TERNAL ROLE AND

FERTILIT\p

fhe. mobi lization of family ties to promote the i nteres ts of the fam ily sugges ts that the m atern al role in centripetal kin groups differs considerably from th at al groups. Insofar as family members can be regarded as "asse ts , " in centrifug ticipate that with centripe tal organ ization , the maternal role one wou ld an wou ld be oriented to w ard m ax im i zing the � ' worth ' ' of each member . The inves tment in human capital among fam i lies and kin o rg anized centripetally m ay lead to hi g h fe rtil i ty rates and certai nly to speci al efforts by mothers to

instil in the ir c hi ldren values and sk ills that would facilitate ,; 'success ' ' in later vears. S i nce a strong c o m mitme nt to the maternal role is time consuming, it �ems l ikely that mothers in centripetal sy stem s would be less apt to work outside the home than mothers in c ent rifug al systems . (See H il l 1 977) . I n terms of kinship models , one would expect that in families organ i zed on the basis of the Parente l a Orders model , mothers would be employed less fre­ quently than would mothers in families organi zed on the basis of the Ge neti c ,

or Canon Law model . This section is d i v id ed into tv.'o parts . The first part pertains to employ me n t of the respo nde nt s ' mothers during the child rearing years; the second deal s with level s of fertil i t y .

Maternal Employment

Table 6-7 displays pe.rcentages of moth ers who had been employed while the respondents were growing up . The working arran gements had been quite di ­ ve rse: some of th e s e mothers had \\'Orked full-time , others part-time ; some had held a job co ntinuously , others sporadic al ly . Because it would be difficult to ac commodate the large vari ation in pat tern s of Ylo rk , the table refers merely to emp lo me nt at any time prior to t he re sponden fs adulthood . Perusal of y Tab le 6- 7 re vea ls that the major determinant in maternal e mployme nt i s the nu mber of chil dren in th e family; the larger the number of c h i ld ren , the less l ikely is the mother to work . Whereas over half of the mothers w i th only­ ch ildren had been emp loye d outside the home , only a fourth worked when there were four or more chi l dren . Ho wev er , when the number of children is taken into account. Table 6-7 indic ates that fewer mothers of persons in the Parentela Or ders c atego ry we re empl o y e d while the respondent was gro w i ng up than were mothers in the o ther c atego ri es . By way of contrast, in families with fou r or more children , mothe rs of persons in the Standard American classification tended to be emp lo yed m ore of te .n than did other mothers . In general , though , the data

reve al only that mothe rs of respond ent s in the Parente l a Orders c atego ry are less i nc line d than others to work wh ile their children are growing up. At least

116 TABLE 6-7. Kinship Orientation, Mothers Who Had Been Emp loyed W h U e Respondents Were Growing Up, and Number or Respondent's Siblings Percent of mothers employed while respondents were growing up

Number of respondent 's siblings

Parentela

Standard

Civil

orders

american

law

51.5 33 38.7

63 . 7 11

Responde nt only child (%)

45 . Sb

One to

37. 1

N N

three ( %)

70

Foor or more (%) All families (%)

b

Ratio of cell

percentJlle

to

30. 0"

83 37 . 8

1 26

N Ratio of cell

217

1 3 . 3b 45 29.Jh

N

;t

11

tolal

percc nlage is 1 . 2

perce.-age ro tocal percentqe is .8

33 3

40.8 76

27.6

29 39. 5 1 16

Genetic or canon law

Tota]

63 . 7 II 37 .0

57 .6

73

28 .6 56 36 . 4 1 40

66 38 . .5 436

25 . 8

213 36 . 5 715

or more .

or

under .

in this respect, the find ings support the spec u l at i on that comm itme nt to the mate rn a l role is associated with c entripetal tendenc ies in kinship orientation .

Patterns of

Fertility

Table 6-8 describes the fertility of t he respondents ' parents . This t ab le takes

into account the re l igion in v,rhich the responden t was raised , the occupation of the father at the time that the respondent was growing up, and the respon­ dent 's kin sh i p orientation . In order to avoid fertility histories complicated by d i vo rce and widowhood , the table includes only ful l siblings of the res pon­ dent.

Both re l i g ion and paternal occupation influence fert i l ity rates . The ra nk

ordering by rel ig ion shows Catholics with the highe st me an (3 .42 sib l in gs o f the respondent) , Protestants next ( 3 .09) , seculari sts third (2 . 69) , and Je.vt's last ( 1 . 59) . This ranking is con si stent w ith the larg e number of ferti l ity studies over the past generat ion (e . g . , Freedman et al. , 1 959; Rainwater, 1 960) . Fi nd ings by father 's oc c upati on also follow a famil iar p atte rn , with b lue­ col lar parents highest (3 . 40) , clerical , sa l e s , and craft wo rkers next ( 2 . 76) . and pro fe ss i on al , managerial, and ad min i strat i v e workers lowest (2 . 5 6). Thus , the findings ind ic at e that the parents o f the respondents are not atyp ic al in the ir ferti l ity h i s tory .

117 'fABLE 6-8. Kinship Orientation and Mean Number of FuU SibHngs of the Respondent , by Religion in Wbich Respondent Was Raised and by Father's Ocrupation8 Mean nu mber of full siblings

which . Religi on in

�ot was nu� or

Parentela

faJ)ler 's occupatton

orders

Religi on No religious prefere nce

Standard

Ci vil

american

law

2 . 5 5 ('

Jewish

N Protestant

1 . 36

15

22

3 . 25

2 . 63

2 . 79

3 . 26

1 96

72

3 . 19

3 . 50

38

90

30

2 . 73

2 . 35

N

35

15

2 .20h 67

N Catholic

Total

2 . 69

2 . 87r:

20

N

Genetic or canon law

1 . 59

1 . 35C

S4

17

4 .43

3 . 09

3 . 98

3 . 42

44

202

2 . 79

2 . 74

2 . 56

409

74

Father's occu pation

Professional , managerial or administrative

Clerical, sales or craft

52

l l6

42

38

248

3 .06

2 . 39

2 . 94

2 . 76

1 08

38

46

227

Blue coDar

35

3 .26

3 . 86

3.1 1

3 .03

3 . 40

37

35

Enlire sam ple N

3 13

99

3 . 92

2 . 69

2 . 87

326

1 14

N

N

N

• Falher's occu pation at tbe time during respoadeat 's cbildhood.

.

1 22

that respondent was "growilll

up.

hilf\est mean for each rei igious or socioecoaomic group is i o

b

The

r

Categories

combi.aed because N i n either

category is under 1 0 .

"

51

222

3 . 97

1 38

Excludes c.ses where

fatber

had died

ilalics .

Whether one scans the fertil ity data in Table 6-8 by religion or by socio­ economic level , three regularities are apparent:

1 . Parents of persons in the Standard American category generally have the fe we st chi ldren regardless of religious or socioeconomic status . 2 . Ex ce pt for Catho lics , "Ylithin each religiou s o r socioeconomic grouping , fathe rs and mothers of individuals in the Parental Orders category have a higher mean fertility than does that grouping as a whole.

3 · Parent s of people classified as Genetic or Canon Lav.' generally have the larg est number of chi ldren of any kinship orientati on .

The tendencies in fertil ity clearly indicate that. at least in the parental generatio n. kinship orientation is associ ated with c hild-beari ng . The overall Pattern for the enti re sample reveals the proclivity of the Standard American

1 18 parents to limit the number of c h ildren ; the Parentela Orders fath e rs a d

mothers to be above average in chi ldbearing; and the pare n ts of perso n s a

Genetic

or Canon Law to be h i ghe st in child p rodu cti on

.

w�

When the fe rti l ity of the re spondents ' sib l in g s are e xamined i n T ab le 6. 9 some difference s by gener at i on emerge. Whereas the parents of Ge neti c

and

Cano n Law re spondents are highly prolific , the fertility of their childre n

(other than the respondent) shows a marked dec line relative to the oth er ki nship orientations . This drop is most dramatic among sibl ings of respo n

..

as

Catholics and among those whose parents are in th e l ower socioeconomic levels . This shift mirrors the findings of rec e nt feni lity stud-

dents raised

TABLE 6-9. Kinship Orientation and Mean

Number of SibHngs' and by

Religion in Which Responden t Was Raised Fatber's Occupationa

Children by

Mean number of siblings ' children

Religion in which

respondent was raised

Parentela

or father 's occupation

Standard

orders

american

3 .28b

2.53 51

Geneti c or

Civi l law

canon law

Religion

No rel igious preference Nt

7d

2 . /3 29

Jewish

N<

2. 10

Protestant

:!'ll�

232

N•�

2 . 15 409

Total N

95

37(' 1 29 21

1 . 44

.

1 .92 464

2 26 208

28 1 1 .91 824

1 07

.

1 . 89

2.19 1 41

Catholic

2 . 20

1 . 54

.82 28

1 .60

2. 1 3 237

1 . 67 215

2 .0 1

1 . 87 490

2.04 1 17

2 . 08

2.31 1 13 1 . 99 34 1

1 . 92

336

Total

78

2.06 1 141 1 . 84 744

Father's occupation

Professional , managerial or admi nistrative

Nt:

Clerical , sales. or craft

2 . 04

1 . 74

1 57

263

N( B lue collar

2 . 44

2. 1 6

NC

1 52

Total

2 .09

336 1 .92

413

822

N

r.

Father's occupation at lhe

!)

1be

time

duriag .respondelll 's childhood.

e

104

1 . 74 223

1 . 66

highest mean for

lncluda

both

thai responde•

each religioliS

full and half siblings .

or

•

Althougb

mean

Categories

combined because

data

N in either category is

1 . 59 1 48 237 1 . 86 493

1 . 93 645 1 . 68 5 86

2. 1 7 838

• •growing up. · · Cases omitted wbere father had died

sociceconomic group

is based on a ..,.ery small N. the P.-entela Orders and Standard American fertility.

d

\\'85

1 . 68 111

1 08

is in italics .

are presenled in order

under 1 0.

to show

the contrast

berweet'l

1 19

ay a pictu re of shrink i ng differences in the prod ucti o n of chil ­ ,v bic h portr among the various religious and socioeconomic groupings (Westoff and

:n .

Ryder , 1 977 ) . part from the shift in fert il ity patterns among s ibl ings o f perso ns i n the A ti c and C ano n Law c ate go ry , however , the re l at i onsh ip betwe,en kinship

G e ne ber o f chil dren i s sub stanti all y the same as th at for the rie n tatio n and num 0 n nts ' parent s . For the re m ai n in g ki nsh ip orientations , the broth ers and s po de of persons i n the Paren te la Orders category have the l argest fam i l ies and the sibl i ngs of those with a S tandard A me ri can orientation have the

:srers

findings suggest that fam i l ie s with th e se orientation s are n ot smallest . The se ns itiv e to ch ange s in the soc ial structure as those with Genetic or Canon

as

.

Lav.; orien tat ion Particularly , the re sults indicate a c h angi ng pattern of adh e renc e to Catholic

.

fam ily ide olog y As in prev iou s years , Westoff and R yder ( 1 97 7 , p. 338) report that , ' 'The nu mbe r [of chi ldren] wanted among C atho l ic s . . . [sti l l]

varies direc tly and strongl y with simple measures of rel ig i ous co mm it ment . � ' That bein g the ca se , recent events--perhaps pertaining to the changing ro le of women in modem soc i et y-appear to hav e affected the Cathol ic family more than they ha ve the fam i l ie s of Jev-·s or Prote stan ts . Mo re ov e r , the find ing s imply t hat this chan g e has oc curre d most profoundly am o ng those Cath­ ol ic fami l ie s with a Ge ne tic or C a no n Law orientation-precisely the orien­ tation associated traditionally with the C hurch- leav i n g the other ori en t at io ns rel ati\ e ly un affected .

­

SUMMAR Y: KINSHIP IDEOLOGY AND FAMILY OF ORIENTATION

This chapter has dealt with the re lat io n sh ip between kinship orientati on and attribute s of the respondents ' parental family . The analysis has util ized the

co ncepts of centripetal and ce ntrifugal kin ship organization . Theoretical l y , cen tripetal kinship emerges (or is s u stained ) in a situation in which a spec ial

interest

group is vying w i th other groups for economic . pol itic al , or re l i gio us

' 'goods ' ' . Centrifugal kinship , however , is fo stered when special interests are su bo rdin ated to the broader concerns which extend bey ond the particular

�0�P-

As a general rule , centri fugal sy stems are organized in ways that in­ dt va duais are p l ace d u nder as many cro ss- pre s sures as possible and kin g roup bound ari es are vague . These modes of organ izatio n yield distinct nonn s re­ garding marri age , di vo rce , and the mate rna l role .

!h

e theoretical discussion also proposed that the models of kinship orien­ tation under investigati on can be pl ac ed upon a continuum i n the extent to � hi c h they repre sent centripetal versus centrifugal properties . Moreover , the km sh ip models repre se n t ideological stanc es that are tran smitted from parents to ch i ldren.

1 20

To support the above pro pos ition s I have presented data to show that th norms i mp u ted to the models are held by the parents as w el l as by the sibli n e gs o f respo ndent s with given kin s hip orientations. These data reveal d i fferenc acco rd ing to kinship mode l in s uch characteristics as p rev al e nce of in the family ; rel ig io u s endogamy between parent s ; religi ous hom o gene i among si b l i n g s ; age at wh i c h the parents married . the maternal rol e ; and fert il ity (O bv io u sly the data would hav e been more defin i t iv e h ad each p arent and sibl ing been i n tervi ewed personally. But using the respon dent a s family informant on obj ect i ve matters-such a s divorce, age at marri a ge , an d fertility-maternal employ men t does lend plausibility to the assumption that the data are fairly reliable.) This chapter focused upon the family of orientation as the source of the respondent 's kinship ideo logy The fol lo w i ng ch apter shifts the attentio n to the respondent as the transmitter of the i d eo logy to the next gene ration . ,

divor�s ;

.

,

.

The previous chapter on parents and siblings o f re sponden ts outlined

FamilY

of Procreation

characteristics of family relation­ s hips associated w ith the appropri ation of particular

i

kinship orienta­

tion s by i n d i v i du al s . These charac­

teristics included such d ve rs e

p hen o me n a

as

age at m arri ag e , rel igious homogeneity in the family , the maternal role, and fertility . The interpretation offered �·as that the configurations of these attributes expre ss family ideo­ logies associated w i th the kinship ori entatio ns under investigation . Here I shall continue the an al ys i s of the content o f family nonns and values connected with kinship orientations--this time in the re spon den t s ' own fam­ ilies of procre ation . I shall discuss topic s that are similar to those handled in the p re vious chapter II except that greater e mphas i s is gi ven to marital ties between the respondents and their spou se s . Moreover, whereas in the pre­ vious c hapter I u ti l ized a typo l og y of ce n tripetal and centri fu gal k in sh i p organization to interpret the data , I now translate that typo logy into compe tin g models of family org ani z ati on in contemporary soc i ety . MODELS OF FAMILY ORGANIZATION

The dicho tomy of ce ntripetal and centrifugal kins h ip organization appears to be re flected in the di stinctio n betYieen "natural " and "legal " mode l s of

i

fam i ly org an zatio n . The unatural " family mod e l rests upon a c l e ar discrim­ i nation between natural and human creation s . Con si ste nt with th is model , hi storical ly family l aw in Weste rn Europe has d i s tingu is hed between "arti­

ficial ' II and ' 'natural , . nuc lear- famil y relations hips . Accord in g to the "nat­ u ral ' · mode l , the nuc lear family derives its au tho ri zati on or charter for exis­ tence from a source outside the state-in religion , in the mystique of blood

maintenance of estates , or in other sources of traditional Values ( Schneider, 1 968 ) . Artific ial family ties , in this co ntex t , are me re l y a creation of secular legislatures and derive their legitimation from the state. re lati onships , in the

The n atural fami l y model provided the basi s for American family law from Its ve ry be i n n n . In c o lon ial New England , .. 'positive l a w , that the laY/ .

is, g i g �ade by man , bel ieved to have to comport with a h i ghe r divine l aw for al id ity Thi s i de a of higher Ia�· , al though l ater sec ularized ., was to become Its V

.

Vlas

,

1 22 an e n d uri ng legacy of the colon ial period (H aski n s , 1 %0 , p . 1 40) . ,

s0111 aspects of Mosaic l aw were thought of as n atural and immutabl e-ot h e aspects as mod i fiab l e and posi t i v e . The assu mp t i ons that underla y Pu ri t soc i al theory in uthe conception of the social convenant and the e mp h as · upon a G od- gi ve n fundamental law were the g ene si s of id e als which in the e i gh tee nth century and shaped po l i tic al and l eg al th in k in g in th e e ar)v days of t he Republic (Haskin s , 1 960 , p . 229) . ' ' Becau se the natu ral or ­ criptive fami ly d e ri ves its authorization for existence out s ide th e st ate , the household may be regarded as extraterritorial go vernme n t com i ng u nd e r state j urisd iction only i nso far as it i m ped es or interferes with the nonnal l ife of the c om mu nity . This assum pt ion lay behind John Locke ' s state ment in his ' ·second Treati se of Civil Gove rnment ' ' that: ·

: flowe� �

A ch ild is born

a

subject

of no country or government . He is under his father · s

tuition and authority unti l he comes to age o f discretion , and then h e i s a freeman , at l iberty w h at government he \\' i l l put hi mse l f under, what pol iti c he wi l l unite himself to . is the

same

.

. The power that

a

father hath naturally over his chil dren

of natural obligations and commonwealths .

w herever they be born , and the ties

bounded by the positi ve limits of kingdoms

are

not

An alternative cultural mode l emerg i ng to replace the .. .. natural ' ' fam i ly i s the con ce pt io n of the family as a mere legal enti ty . In this modet since fami l y relationships exist only as a category comparab le to bu sine ss enter­ pri ses , the differen c e between natural and artificial family ties d i sappe ars . The sta te is the n the only chartering agency for the leg i t imation of family re l at ion sh i ps (Farber, 1 97 3 ) . In con trast to the Puritan con c e pti o n that Election to Grace produces a natural h i erarc hy among men , the En l ightment doctrines (tha t all men are c reated without differential status ascribed by birth , and all are endovled with th e same civil rights) assume a l ac k of "natural " i n s titu tion a l structure . A natural structure would i mp I y th at there be d iffere n ti al treatment bet wee n

some men and others and that those who s upport the natural s tructures should

be rewarded , while those who do not should be punished . Presuppositions

of the lack of natural structure imply that any structure is c reated by m an­

made rule s , which are lhen codifi ed in law . Hence , in this conce-ption of la�· !

one need not posit ' 'natural , ' ends or purpose s of i n stituti o ns in pro viding

for the perpetual con ti nu ity of soci al institutions . Inasmuch as in stitutions are

seen as continuing to exist for reasons other than tho se that brou gh t them int o

be ing , p art i c u lar .. . natural " ends become irrelevant .

Whereas traditional conceptions of the fami ly have pre suppo sed a fixed se t

of functions deri ved from ' 'nature , � ' l aw i n contemporary society is faced with the prob le m of provid in g conti nu ity in social structures w ithou t presup­

posing fixed ends for these structure s . To so l ve this problem, fam ily lav.l has

had to a ss ume that the perpetuation of soc ial structures (such as the family)

1 23

. des irable only insofar as people have reason to continue them . Any given mi ght have to accommodate a variety of values and goals . Conse­ s uendY , governments have had to modify laws in ways that admit a multi ­ q s and that assume an equi valence of d ifferent value systems . !ic it)' of end e ide a that all domestic groups have equal rights and stand ing in the soc iety thereby oppo ses the notion that congeries of families should be encouraged homogeneous clusters (or facti ons) organized on the basis of to perpetuate soc ial c l a ss , religion , ethnicity , or some other ascripti ve criterion denoting "natu ral ly ' ' derived values . a set of , T he. disti n ction between " natural , and "legal " family models impl ies also a difference between corporate and individual privacy . The concept of privacy itself rests upon the degree of autonomy from the general com munity . To the. extent that conduct is not subject to publ ic survei l lance and accountability , it remains private . In corporate privacy , it is the Hnatural " fami ly v.·hose autonomy is to be protected ; in pe rsonal privacy . it is individu al au tonomy

1�crore

�

that is paramount . The coexistence of both "natural ' ' and ' 'legal ' ' family models i s indicated by the justifications that have been suppl ied in landmark dec i sions of the U . S . Supreme Court . For example , the Court decision in Gris'rvold v. Con­ necticut (38 1 U . S . 47 9) relies heavily upon corporate privacy in marri age . This decision propose.s that: We deal

with

a ri ght of privacy older than the Bill of Ri ghts--ol der than our

political parties , older than our school system . M arri age is a

com i

ng

together

for better or for worse , hopefully end urin g , and inti mate to the degree of being

sacred. It is an association

that promotes a way

of life, not causes; a h armon y

in li ving , not political faiths ; a bilateral loyalty , not

commercial

or soc ial

projects . Yet it is an associ ation for as noble a purpose as any involved in our

prior decisions .

Concurring 'N·ith the Court dec i sion in Grist,t:o/d , Justice Goldberg cites ' 'the traditional relation of the family " as being . . a rel ation as old and as fundamental as our entire civilization � ' and su ggests that: The entire fabric of the

Constitution and the

pu rpo ses that c l e arly underl ie its

speci fic guarantees demonstrate that the rights to marital pri v acy and raise a family

are

speci fical ly prot ec ted .

and

to marry

of similar orde r and magnitu de as the fundamental rights

Hold ing that the Ninth Amendment to the U . S . Const itution impl ies a recognition of additional basic and fundamental rights omitted from the B i l l of Rights , Justice Goldberg concludes that: To hol d th at as

a

right

so

basic and fundamental and so deep-rooted in our society

the righl of privacy i n marr iage may be i nfringed because that right is not

1 24 guaranteed in so m any y,rords by the ftrst eight amendment s to the Cons t it u tio n is to i gno re the Ninth Amendment and to give it no effect whatsoever.

Yet the interpretation of marri age as a fundamental relatio nshi p acts counterpoint to the interpretation put forth in other c ases that ' 'ou tside areas of plai nly harmful conduct , every American i s left to sh ape hi s o�·n life as he thinks best , do what he pleases , go where he pleases (Cited in Roe "' · Wade , 35 L Ed 2d 147). " Indeed , in Eisenstadt v. Baird ( 1 05 U . S . 438) , the Court contends that:

:

It is true that in Griswold the ri gh t of pri vacy in question inhered in the maritaJ

relationship. Yet the marital couple is not an independent entity with a mind and heart of its own , but

an

assoc iation of two ind ividuals eac h ¥lith a separate

intellectual and emotional makeup . If the righ t of privacy mean s anything � i t

is the ri ght of the individual , m arried or single , to be free from unwarranted

go v ernme nt intrusion i nto matters so fundamentall y affecting a person as the

deci sion whether to bear or begat a child .

In Planned Parenthood of Missouri v. Danforth , the Court expl icitly rejects the following argument of corporate pri vacy , namely , that: Recognizing that the consent of both parties is generally necessary . . . to begin

a family , the legislature has determi ned that a change in the family structure set in motion by mutual consent should be terminated only by mutual conse nt ( B rief for Appellee Danforth

38).

That rejection would also discount the v iews th at the decision to complete the birth process involves obligation s by both husband and V.'ife and that , in case of disagreement , the wife already has veto power over her husband ·s desire for abortion . Instead , the Plm1ned Parenthood of Missouri dec i sion takes the pos ition that the husband 's rights with regard to abortion are not considered to be fundamental to the structure of the family , but rathe-r, that they are dele gated by the State: We are not unav.·are of the deep and proper concern and interest that a de..,· oted

and protective husband has in his wife 's pregnancy and in the growth and

development of the fetus she is carry i ng . Neither has this Court failed to ap­ preciate the importance of the marital relationshi p in our society

.

over , we recognized that the decis io n whether to u ndergo or to forgo

. . . More­

an

abort ion

may have profound effects on the future of any marri age , effects that are both

physical and mental , and pos sibly deleterious . Notwithstanding the se fac tors � we cannot hold that the State has the constitutional authority

to

gi ve the spouse

unilaterally the ability to p rohi bi t the wife from term inating her pregnancy

.

when the State itself l acks the right . . . E ven if the S tate had the ability to

1 25 the husband a power it itself cou ld not exercise , it is not at all l ikely deleg ate to ac tion v.·ould further , as the District Court majority phrased it , the that such state in protecting the m u tua lity o f dec isions vi tal to the marri age • interest of the ' relationship .

Th e di stin ction betv-'een corporate and individual privacy suggests the exis­ tionship between modes of kinship orientation and nonns rel ­ tence of a rela ev an t to fam ily privacy . The. emphasis on line of descent in the Standard

meric an and (particularly) the Parentel a Orders models impli es that holders of tbe.se models tend to bracket ascendants and/or descendants (in addition A

to parents and children) within the ir conceptions of especially close kin . Since grandparents and grandchildren are not normally included Vlithin the house­

bold , this bracketing seems to denote a corporate conception of kinship and

family ties and , consequently , a conception of family organization in terms

of corporate privacy and autonomy . At the other extreme � the Genetic and

Canon Law models do not reveal any particular structure on the basis of generational di stinctions or any particular bracketing of kin; as a resul t , the

Ego-centeredness of these model s appears to reflect a stress upon indi vidual

pri vacy and autonomy in family rel ationships . (The Civil Law model seems to

be intermediate between the two extremes) .

Insofar as corporate. privacy is associated w ith centripetal kinship nonns

and individual privacy with centrifugal nonns , the fmdings in the chapter on Family of Orientation should be replicated in the respondents own families

of procreation . The data that are presented in the succeed i ng sections of this

chapter are expected to show further support of the rel ationship between family ideology and kinship orientation with regard to such s ignificant matters

as

(a) religious homogeneity within the family; (b) disillusionment v..- i th the

spouse in the marital relationship ; (c) proclivities toward early marri age and di vo rce; (d) norms governing family roles ; and (e) ideas on fertil ity .

KINSHIP ORIENT ATION AND RELIGIOUS HO�IOGENEITY One of the marks of a coherent corporate structure is the unity of underlying Princi ples by which the collectivity operates . Basic values in the fami ly are

generally grounded in religious ideals and interests . To the extent that kinship orientations represent a range in corporate characteristics , one would expect them to vary also according to religiosity and religious homogeneity in marri �e . ·

Religiosity Among Protestants and Catholics , one of the chief indicators of rel igiosity is the frequency of church attendance (G l ock and Stark , 1965) . Tab le 7- l sho w s the num ber of time s per month that Protestants and Catholics report

1 26 attending church services. For both Catholics and Protestants , more person

n8 are weekly church goers . Among Protestants , those in the Genetic and Cano s Law classifications are least like l y to attend frequently , while am on g

in the Parentela Orders category than in the remaining kinsh ip orientatio

Cath�

olics , the lowest attendance appears among those person s with Stan d ard

American and Civil Law orientations . Hence , insofar as church atten d an ce reflects religiosity , both Catholics and Protestants in the Parentel a Orde rs

category reveal a greater religious interest than do persons with o ther k in ship . . perspectives , while among Protestants , the data clearly show least i nterest among those in the Genetic and Canon Law grouping .

Religious Intermarriage and Homogenization Given current courtsh ip practices , whereby religion ordinarily

plays a

mini­

mal role in selection of serious-dating partners , chances of religious inter­ marri age are con siderable . Under these circumstances , the religious unity between man and woman-as an expression of corporate ex istence-may operate less frequently in marital selection than might be achieved later in

the marri age itself. Table 7-2 displays the percentages of intermarri ages and mixed marri ages by the kinship orientation of the respondents . In this study ,

intermarri age refers to those couples in which husband and wife have been

raised in different rel igions (whether or not there has been a conversion prior to the marri age itself. ) Whereas intermarriages ordinari ly are defmed as marri ages in which bride and groom differ i n rel igion at the time of the wedding , mixed marriages , however, are those marriages in which the hus­ band and wife continue to maintain their separate religious identities after­ wards .

TABLE 7-1. Kinship Orientation and Frequency or Attending by Protestants and Catholics Frequency of atte nding church ser..- ices Protestants

At least once a week (%)

I to 3 times pe r month (%) Less than once a month (%)

N

Catholics At least once

a

week (% )

1 to 3 times per month (%) Less than once a month (%) N

Parentela orders 37. 1 17 7 .

Standard

american

30.4

16.3

45 . 2

.53 . 3 1 84

45. 9 24.3 29 .7 37

32. 5

62

21 .7 45 . 8 83

Church Services

Ci vil law

Genetic or

31 5

28 . 2

.

2 1. 4 47 . 1

70 33.3 18.5 48 . 1 27

canon law

1 2.7 59. 2 71 35 . 1

28.6 38 . 1 42

1 27 n T able 7-2 Jewish re spondent s show the least amount of i n termarri age I d rn i xed marri age , and Catholics the mo st . As for k i nsh ip orientation , anrson s with a Parentela O rders gene ral l y are l e ast apt to i n termarry regard­ of the cou p le ' s religion . Among Protestants, both Parentela Orders and dard Am erican perspectives are lower in percentage of intermarriage than Stan are Civil Law , Genetics, or Canon Law ; but among Jews and Catholics , on ly parentela Orders and Civil Law c ate go ri e s are rel atively low in perce ntag e of intermarriage . The findings on mixed marri age s are generally s imil ar to those on i n term arri age (except that Catholics in the Standard American and ci �· il Law c ategori es tend more often to st ay in mixed marriages than do those in the Parentela Orders, Genetic, or Canon Law classification) . Th e sim ilarity in find i n gs between intermarriages and mixed m arri age s informs us only that the greater the orig i nal he te roge neity i n rel ig i on , the more often it will persist. B ut perhaps the distinctive mark of a corporate entity is special procl ivity toward homo genization of values that had been differe nt initially . Table 7-3 shovls the extent to which religious groups d iffer in homogenization . This table presents percentages of intenn ani ages that remain as mixed m arriage s among the different re l igi o us groups and ki nshi p categories . Even given the smallest number of i n term arriag es among the

J: s

TABLE

7-2. Percentages of lntennarriages and Mixed Marri ages, by Religion and Kinship Orientation

Religion of one of lhe spouses Religion in which raised Protestant (%) N Jewish (%) N

Catholic (%)

N

All cases8 ( %)

N

Parentela orders

Standard american

Jewi sh {%) N

Catholic (% ) N

All c ases �'� (%) N

Genetic or canon law

Total

50.6

55 . 1

49 .6

law

Intermarriages

49. 4 81

47 . 3

85

98

520

33.3

50 .0

22.2

42 . 9

40 . 3

18

26

14

67

60 . 4

70 . 1

9

66. 2

1 37

60 . 9

66 . 2

53

65

30 1

52 . 1

62 . 0

59 . 7

60 .4

1 17

32 1

1 09

1 34

2 5 .8

27 . 1

25 6

Religion now Protestant (%) N

C i vil

66

46

Mixed marriages 42 . 3

40.0

32. 1

203

78

80

1 2.5

29. 2

22 . 2

27 . 3

23 .2

16

24

1 1

38.6

50. 5

44 . 7

38 . 5

60

44. 9

44

11 1

38

52

245

42 . 2

4 1 .4

1 16

1 40

30.9

1 40

3 1 .9 332

9

427

• For uall cases · · onJ)' , intermarriage s and mixed nwria1es include marrillges between major ProlestaDt classifications: Reformation era denminations, Pietistic Protestanas. Neofundamcnlalisl sects. and Latter Day S.ints.

128 TABLE 7-3. Percentage of Marriages (in Wbicb Persons Had Been Raised in Different Religions) Which Remain as Mixed P.tarriages Religion of one of the spouses Protestant Jewish Catholic

All casesa

Parentela

Standard

Civil

Genetic or

orders

american

law

canon law

Tota]

52.2

57.3

72 . 6

37 . 5

1 00.0

63 . 6

63 . 9

58 . 4

83 6

72. 0

73 . 4

58.2

64 . 7 57 . 8

59.3

5 1 .5

70. 7

68 . 5

.

67 . 8

all cases " only. intemwriaaes and miJled marriages include marriages between major Protes tant • For class i ficalions: Reformation era deoomioations, Pietistic Protestants, Neofundameatalist sects , and Latter Da) Saints. . .

religious groups , Jewish respondents are most likely to homogenize the re­ l igions of husband and "'·ife; that is, among Jews the percentage of mix ed marri ages remaini ng (5 8 percent) is lower than it is for Protestants ( 65 per­ cent) or Catholics (68 percent) . Overall , fewer persons in the Parentela Orders and Standard A merican classifications remain in mixed marriage s than do those in the Civ i l Law , Genetic , or Canon Law categories . Among religious groups , this tendency toward religious homogenization holds in particular for Jews and Prote stants; for Catholics , homogenization i s greater in the Parentela Orders , Genetic , and Canon Law classes than it is for persons with a Standard American or Civil Law perspective . Thus , except for Cathol ics , the data on religious homogenization of intermarriages indicates that marri ages of Parentel a Orders and Standard American persons show greater evidence of corporateness than do marri ages of person s with C ivil Law , Genetic , and Canon Law viewpoints . The data for Catholics, however , suggest that for them, the Standard Amer­ ican and Civil Law orientations represent a falling away from the norms associated with the Church and perhaps a Protestantization of vie ws on the family . Kinship Orientation and Religious Homogeneity: Summary

Collectively , the data pertaining to kinship orientation and religious homo­ geneity confmn the expectation that Parentel a Orders (in panicular) and Standard American kinship (to a lesser extent) are related to greater religiou s intere st and homogeneity than are the Civil Law , Genetic , and Canon La\\' approaches to kinship . Persons with a Parentela Orders orientation are the most constant churchgoers and they (along with those in the Standard Amer­ ican category) generally homogenize the religion of the couple in intennar­ riages . Insofar as the coalescence of a common value system is a mark of

1 29 rat s tructure, the fmding s on homog en i zation of religion supports the c.orpo e andard American and particularly Parentel a Orders models ex­ vieW th at St res s corporate con c eptions of family organization .

p

JDEALIZATION, DISILLUSIONMENT, AND MARITAL INSTABILITY Before midcentury , when Freudian psy cholog y was popular in sociological analysis , the roots of marital disruption were sough t in the irrationalities of marital s election. At that time , much research was devoted to the prediction of marital adjustment (e. g., Burgess and Cottrell , 1 939; Burgess and Wallin , 1 953 ; Horst, 1 94 1 ; Locke , 1 95 1 ; and Terman , 1 938) . O f particular interest was the phenomenon of idealization Y�'hereby an individual proj ec t s unreal virtues upon a beloved . According to Willard Waller 's ( 1 95 1 , p. 1 20) de­ scription of idealization, • 'In romantic love one builds up an almost com­ pletely unreal p i c ture of a person which he calls by the s ame name as a real person , and vainly imagines to be like that person , but in fact the only authentic thing in the picture is the emotion Y�'hich one feels toward it . ' ' According to this theory , when the couple settles down to a routine of living together afte r marriage , realism is bared , disillusionment ensues , and the

marriage flounders . But the view that ideal ization during courtship is at fault in p roduc ing marital maladjustments has not survived the test of evidence . Studies of cou rtsh ip and marital pred ic t ion have shoY�·n that most relationships even­ tuatin g in marri age are not intense , romantici zed affairs; ins tead the findings reveal that courtships ordinarily consist of compan ionate relationships that become progressively more intimate v.;ith the passage of time ( Burge ss and Wallin, 1953 ) . On the whole , marital-prediction research shows that couples have a realistic view about their partners ' faults and virtues . But despite the rea l i sm that dominates during courtship " over the years couples do exhibit signs of disenchantment with the ir spouses . The B u rgess longitudinal stu dy has indicated that after fifteen or so years of marriage , ' ' Regardless of ini t i al level of Lmarital] adj ust m ent , the majority of husbands and wives suffered decline in level of adj ust ment . Clearly , this is not si m ply a change built into the mate selection process ; nor does i t seem that husbands and w ives are s i m ply regressing to a theoretical mean level of adju s tment (D i z zard , 1 968 , p . 1 3) . " It seem s that, rather than a di ssipat i o n of ideali­ zati on , there is instead a prog res s i vel y increas i ng disappointment in the spou se . Th is shift in pe rcepti on of the spouse evokes questions about the processes by Y�'hich di si llu sionment occurs and marri ages may eventually di s i nteg rate .

1 30

Irreligiosity of the Spouse When marri age is regarded as a mechanism for unifying basic values an d attaining collective ends , then it achieves a corporate identity that e mbod ies the individual personal ities of hu sband and wife; however, without a c on ­ sensus on val ues , marriage tends to signify only a rel ationship betvl een tw o indiv iduals in a household arena. For example , earlier I indicated how ki nshi p

orientations vary in the extent to which homogenization of initially differe nt religions occurs (with Parentela Orders being highe st in homogenization , and Genetic and Canon Law the lowest) . Instead , if people regard their marri ages

as relationships bet\\'een individuals , then with disillu sionment and inc rea sed maladjustment , they would project upon their spouses those attributes that conflict v,rith their own . As a result, they wou ld likely view their spouse s

as

differing i n values and aims . As a means of self protection , the se indiv iduals , faced v,rith d i sillusionment in marriage , would tend to see thei r spouses as

holding more profane ideas than they themselves do . This profanity can be expressed in respondents ' perceptions of their spou ses as general ly maintain ­ ing an irreligious posture relative to their own . Putatively , the k in ship models under investigation represent a conti nu um of c orporateness. Pare ntel a Orders and , to a lesser e xtent , Standard American

model s , in sofar as they give much weight to line of descent , imply a cor­

porate , enduri ng existence of the family (or "house " ) . While the Civ il Law model is in an intermediate position , the Genetic and Canon Law models are Ego-based . Given this continuum , one would expect that persons with

a

Genetic or Canon Law orientation would be more likely to be v ulnerable to disi ll usionment in marriage and to proj ect profane ideas upon the spouse than wou ld those with a Parentela Orders or Standard American perspective .

The relationship betv.,·een general disill usion ment and kinship orientation

is indicated in Table 7-4 . The data in this table refer to Srole anomia scale ,

which deals with general disi llusionment-loss of faith in the future , dis­ courage ment of having chi ldren , and lack of trust in others and in the pol itical process . (A h i gh score on the anom ia scale implies extensive disillusion ­ ment . ) According to the table , persons with a Genetic or Canon Lav.,' orien­ tation tend more often than others to ascribe to themselves and to their spouses a high level of disi llusionment , while most of those with a Parentela Orders perspective vi ew themselves and their spou se s as low in disillusionment . Respondents v.'ith a Standard American or Civil La�· conception of kinsh ip

are

intennediate , but those individuals with a Ci vil Law viewpoint tend some­

what more often to assign both themselves and their spouses a low anomia

score . The findings in Table 7-4 hence clearly indicate a variation by kinship

orientation in extent of general disillusionment . A connection between kinsh ip orientation and view of the spouse as irre­ l igious is suggested in Table 7-5 . This table presents ratios of ratings for self and spouse on the extent to which various religious beliefs are held . A ratio

131 tABLE 7-4. Kinship Orientation and Percentage of Married Persons with Low Anom1a Score for Self and Spouse on

soore

srole scale anomia

Standard

Parentela

orders

american

Both self and spou se assigned low score (%)H Responde nt has low anomia score

53.7

39.9

58 .6

spouse has ]ow anomia score (% )"

56.6

47 . 3

C:%)a

N a

A lo'-'' anomia score is 2. 3, 4 , or 5,

score

greater than

is 0

or

and il

one

106

256

Genetic or canon law

47 . 8

35 . 6

58 . 3

38.8

49 . 0

40 . 8

96

98

1 , and it denotes a )0\\o· degree of disillusionment with lhe society; denotes much disiUusionment.

a

hish anomia

i ndicates that the respondents tend to reg ard themsel ves as

more re l i g i ou s (or to hold spouses; a ratio less than

as

44.2

Civil law

more fervent .

one

a

more fervent religious position) than their

i nd i c at e s that the individuals see their spou ses

Three statements about rel igious be l ief are presented . The most ge neral

statement merely asks , " 'How important would you say that re l ig i on is to you

(your spouse)? ' ' The second s t atement refers to th e role of God in everyday

life, and the third , which is most specific in c onte nt , pertains t o the belief

that ' 'the institution of m arri age and

TABLE 7-S.

fami ly was estab lished by God , '-and

Kinship Orientation and Ratio of Ratings for Self and Spouse Extent to \\bich Religious

Religious beliefs

Beliefs Are

Helda

on

Parentela orders

Standard american

Civil law

Genetic or canon law

1 .40

2.40

2.25

2 . 87

. 93

1 . 50

1 .64

2 . 20

. 53

.60

.15

1 . 67

How important would you say that religion is to you (your spouse)? Would you say it is very

important, fairly importanl, fairly unimponanl, or not important at

all?

Extent of agreement with •

•1

am

statement,

convinced that God plays

a strong part in what happens in

our daily lives. ' ' b

Enent of agreement with statement, · 'The institution of marriage and family was establ ished by God . nh

aA

ratio

Spouse;

b

a

greater

than

ratio less

lJM

than

indicates

nne

T be res ponse c ategories

lhal respondents

tend to view themsel\.·es

indicates that tbe individuals

are :

st,.ongly a � . a,r�� .

110 1

sec

lheiT spouse as

Ju,.e, duagree,

and

as more

more

religious

rel igious.

strongly disagree.

lhaa their

1 32

therefore , by i mpl ication , more fundamental than secu lar law . For all th r ee statements , respondents in the Parentela Orders category have th e lo \ves t ratios , thereby revealing the least tendency toward attributing profane vie w s to the spouse , whereas the persons in the Genetic and Canon La w catego ry uniformly have the largest ratios , and consequently appear most like ly to impute such views to the spou se . Ratios for the Standard American an d Ci v ii Law respondents l ie in the middle of the range . Hence, the data support the contention that Parentela Orders , as representing a corporate concepti on of kinship , permits least latitude in seeing the spouse as irreligious , and that Genetic and Canon Law views , as representing individualistic c onc e pti on s of kinship, are most conducive to disparaging the spou se ' s rel igious ideas . Like the findings on general disillusionment, the evidence regarding the spouse as irre-1 igious lends credence to the position that kinship orientati ons are associated with modes of family organization reflecting the centripetal ­ centrifugal kinship conti nuum .

Kinship Orientation and 1\-larital Status

One would expect that , where corporate conceptions of kinship predominate, kinship ideologies would militate against marital instability . Specifically , one would expect to find less divorce and a later age for marri age amon g persons with Parentela Orders and Standard American perspectives than among per­ sons with other kinship orientations . Indeed , the data in the chapter on respon­ dents ' parents and siblings show such tendencies . For the respondents themselves , infonnation on marital status appears i n Tab le 7-6 . A mong men , chances are greater for divorced than for marri ed persons (either in frrst or subsequent marri ages) to fall into the Standard American grouping , and the probability is especially high for the divorced to be in the Genetic and Canon Law categories . Among women , however , there is little relationship between kinship orientation and marital status . The data on marital statu s (by se x of respondents) are thus equivocal in that they indicate a relationship between kinship orientation and divorce among me n (albeit a complex one) but not among women . Hov,rever, when data on marital stabil ity are examined by rel i gion of respondents (in Table 7-7) , a somewhat different picture emerges . Compar­ isons between those in the Genetic and Canon Law category (where women predominate) and the entire sample reveal that, except for Catholics , persons with Genetic and Canon Law orientations tend less often to be in their frrs t marri age than the average for the total sample . For Catholic s , the Genetic and Canon Lav.' category (presumably reflecting Catholic doctri ne) shows about the same percentage of first -marri ages as the total sample does . Hence , when rel igion is taken into account , the data indicate that Ego-based kinshi p orientation s do generally hold fev.'er first-marri a ges than the average . Table 7-7 al so displays the percentage of first m arri ag e s for both self and

133 tABLE 7-6. Respondent' s Marital Status and Kinship Orientation Now

married

First

marriage

Kinship orientation Men• Parentela orders (%) Standard american (41;)

Civil law ( %) Genetic and canon law (%) N

Remarri age

22. 6

22 .4

46 .6

44 . 9

1 8 .3

1 4. 3

1 2.5

1 8 .4

208

49

15.2

1 8 .4 35 . 5

not

Divorced, now marri ed 7.7

56. 4 0 .0

35 .9 39

Womeo h

Parentela o rders ( %)

Standard american (%) Civil law (%) Genetic and canon Iaw (%) N

• b

For For

men, Chi squ.-e

= 19. 24.5;

d . f.

=

6; p

=

d . f.

=

45 .7

48 . 9 1 5.9

21 . 1

1 6 .0

20. 1

25.0

24. 7

264

16

81

.01.

womea, Cbi square nol sigoificaat. For total distribution (men and

1 4 . S97;

1 3.6

6; p = .02.

women

combined)� Chi

sqlml'e

=

spouse in the Genetic and Canon Law c ate gory Whereas about 67 percent .

of the married persons in the entire sample report that both are currently in

their frrst marri age , only 57 percent of the persons in the Ge n e t i c and Canon

Law category reveal that neither husband nor wife has been marri ed before .

TABLE 7-7. Kinship Orientation and Percentage of Respondents Now in Their First Marriage for Persons in the Genetics and Canon

La"·

Classification as Compared with the Entire Sample, by Religion

and Marital Status of Spouse Religion and

spouse 's marital status

Genetic and canon law

Total

classification

samp le

46 . 7

54 . 7

·�

75

Religion No religious preference (%) N

Jewish (%)

55.6 9'

74 . 1

Prote stant (%)

50. 7

64 .9

71 69 . 0

387 69 . 3

42

1 89

57.0

67 .4

N

N

Catholic (% ) N Marilal statu§ of spouse Both respondenl and spouse in flrsl maniage ( %) Number of marri ed couples

107

54

5 96

1 34

These data therefore bolster the findings on religion that the Genetic and Canon Law kinship orientations ( e xcep t as an e xpre ss ion of Cathol ic id eol­ ogy) are associated with marital instability-past or present . Marital instability is particularly preval ent among persons marri ed in the teen age years , especially among those women marryi ng under the age of eighteen (U . S . Health Resources Administration , 1 974) . Table 7-8 reports the perc entages of women married before the age of eighteen for the v ari ou s kinship orientations of the respondents . 1be table includes not only the women respon dents them­ selves but also the wives of the male respondents Because wives may not al way s hold the same kin sh ip orientations as their husbands , greater weight should be given in interpreting the data to the distributions for the women respondents themselves than to those for wives of respondents . ­

.

·

I n Table 7-8 , except for persons now divorced , the data for women respon­

dents indicate a fairly uniform progre ss i on among the kinship orientations with regard to the percentage married before reach i ng the age of e i g hteen Married \'/ O me n (now either i n thei r first or sub sequ e nt marri age) in the Paren te la Orders class are least likely to marry before the age of eighteen , .

'

TABLE 7-8.

Kinship Orientation of Respondent and Percentage of Women Married Before Reacblng the Age of Eighteen Civil

Genetic or

law

canon law

Total

1 4. 8

1 1 .6

23 . 8

16. 3

1 35

43

1 0. 8

1 1 .9

1 9. 8

20. 9

14. 9

65

1 93

71

91

420

1 1. 1

1 3 .5

11.1

16.0

1 2 .9

36

25

202

5 .0

9.3

1 4.3

15.3

1 0. 7

40

1 29

42

52

263

Wives of men respondents

00. 0

1 4. 3

23 . 1

16.2

3

21

13

37

Women respondents

00 . 0

8. 1

7 .7

20 .0

9.8

11

37

13

20

81

38 . 5

22 . 3

14. 3

40.0

30.4

13

18

35 .6

29.6

14

27

Marital

characteristics

Parentela

Standard

orders

american

1 6. 3 61

All persons

Wives of men respondents N Women respondents

N

Persons now in flrst marri age Wives of men respondents N

Women respondents N

Persons now remarried <.!

N

N

Persons now d i\.·orced

Wives of men respondents

N

Ylomen respondents N

11

Ale

at previous marriage.

45

96

7

43 8 .

16

46

285

8

46

36. 8

35.6

19

76

1 35 while those with a Genetic or Canon Law perspective are most apt to do so . Standard Of those women now in their first marri age , fev;er of those in the American category marry before eighteen than do those in the Civil Law

category , but this difference disappears among the remarri ed . Cu rio us ly ,

h owe ver, among divorced women , those with a Standard American orien­ tati on least often enter marri age prior to eighteen , while those in the Civil Law group most often do so . Still , the overall fmdings do indicate a rela­

tion sh ip between emphasis on line of descent in kinship orientation and age at m arri age . The findings i n Table 7-8 for wives of respondent� do not, however, permit

as c lear an interpretation as those for the women respondents themse lves . The dat a for wives of respondents suggest only that men in the Genetic and Canon Lav; classific ation s are more prone to marry teenage women than are me n in the other kin ship c atego rie s .

I n summary , the data on age at marri age , like those pertaining to marital

s t atus , support the position that emphasis on line of de scent in ki nship ori­

entation expresses a corporate conception of family and inhibits the disillusion­

ment with the spouse . These fmdings are consistent with those pertaining to the

respondents ' pare nts and siblings , which reveal a similar connection between kinship orientation and such characteristics as marital status and age at marri age .

The weight of evidence thus seems to imply a qualitative difference in family

sbUcture among the various kinship orientations . The following section on family roles and statuses e l aborates further on this difference .

K�SIUP ORIENT ATION AND FAMILY ROLES The existence of a distinction between ' 'natural ' ' and ' 'legal ' ' cultural models of family organization assumes a corresponding di s tinctio n in defin itions of appropriate family roles . Insofar as the "natural ' ' family presupposes a con­

ception of corporate priv acy , one would expect that family roles suitable for this model would emphasize the need for conduct that \\lould del imit family boundaries in order to defend the family domain against external encroach­ ment. Co nsequently , families operating with ' 'natural ' ' model assumptions wou ld be inc l ined :

1 . Not to permit the immediate situations in everyday life to dictate con­ duct gove rning family relationships , but i ns tead to be gu ided by general

Princip les aimed at protecting the privacy of the family as a corporate unit . 2 . Not to allow extrafamily obl igations to subvert household demands , but to g ive pri ority to home and fami ly on case of conflict .

3 . Not to emphasize the personal aspects of social relationships over those

qualiti es pertaining to legitimate status , but inste ad to demarcate family boun­ daries as di stinct in qual ity from interper sonal relation s .

1 36 In contrast to t ho se familie s organ i zed on the basis of the ' ·natural ' ' model families w o rk i ng with a " l e gal " model v,ro uld hold oppos i te v i e\\·s . To gre ater extent, the members of these families wou l d tend : 1 . To adapt their conduct to accommodate s pec i fic situations and th e reb:y reveal greater role flexibility . 2 . To weigh each case of c onfl ict between home and extrafami l y ob li g a­ t io ns on the basis of personal d es ire s and wishes . 3 . To emphasize interpersonal tie s rather than famil ial status in organi zin g

�

conduct re l ated to marri age , the home , and fertil ity . This se ct ion deals with the relati on s hip between c oncep t ion s of roles ,

which are associ ated theoretically with the Hnatural ' ' versus "legal " family mod el s , and types of kinship orientation , which are putative l y expressions of vary i n g degrees of corporateness of kinship s tructu re . Thi s anal y sis rests on the as s u mp tion t hat famil y ·is an i n heren t s ubstructure of the ki n sh ip un it . The connecti ons between family rol es and k i n sh ip orientation are ex ami ned in t\\l·o segments : (a) age , sex , : and household composition as reveal i ng the extent to which family roles are s it u ati ona Uy detennined and (b) the sal ience of marital roles as i nd ic atin g social demarcations of family boundaries .

Age, Sex, and Household Composition

A focus on age , sex , and household composition provides i n fonnati on rele­ vant for investigating the i nfl uen ce of ideology versus s ituat ional factors in kinship orientation . Age and se x constitute distinctions for conduct in almost every society . B ut although all societies make use of age and sex in as crib i ng roles, the specific c o ntent of the se roles varies from group to group . The first question to be addre. ssed , however, is whether age and sex influence kinship orientation a t all . Tab le 7-9 di sp l ay s the kinship ori entat ion s of re sponden t s accordi n g to

the ir age and sex . For men , Parentela Orders and Geneti c and C anon La�· responses vary by age , while Standard American and Civil Law categorie s are not affected . Al though 27 pe rcen t of the men born from 1 93 2 to 1 939 fa l l into the P are n te l a Orders category , only 17 perce nt of the men born in the pe rio d 1 950 to 1 959 do so . Conversely , whereas ju st 9 pe rce nt of the men in the earlier cohort ( 1 932- 1 939) are c las s i fied as Genetic or Canon La\\' , fu ll y 23 percent of the later cohort ( 1 950-- 1 959) are . However, as noted above , age seems to have littl e bearing on the Standard American and Civil La�· distributions among men . Ac c ord ing to Table 7-9 , ag e influences kinship orientations of ""·omen differently from those of men . For ""·omen , Civil Law , Genet i c , and Canon

Law orien t at io n s appear to be unaffected by age . The major influences of age are on the Parentela Orders and Standard A merican perspecti ve s . As for men , the probabil ity of falling into the Parentela Orders c ate go ry is greater for

1 37

T

ABLE 7 -9. Kinship Orientation and Respondents ' Year of Birth , by Sex of Respondent Year of birth Men respondents�

Kinship orientation

\\'omen respondents"

1932-1 939

1940- 1949

1950-1959

1 932- 1939

1 940- 1949

27 .0

1 8 .2

1 7 .2

2 1 .5

16.5

5.8

47 . 2

49 .0

45 . 3

38 . 5

46 . 2

54.8

16.9

1 4 .7

14. 1

1 8 .5

1 5 .4

17 .3

9.0

18.2

23 .4

2 1 .5

22 .0

22. 1

N

89

1 43

64

1 35

182

1 04

• For men, Chi square n01

signifJCanl .

1 950-1961

parentela

orders (%)

standard american ( %) Civil law (%)

Genetic or

canon law (%)

•

For

women,

Chi square

=

1 3 .573� d. f.

=

6; p

=

.05.

older than younger '-"'Omen-22 percent in the 1 93 2- 1 939 age cohort as compared with 6 percent in the 1 950-- 1 96 1 cohort . But the chances of having a Standard American orientati on decrease with age-5 5 percent for younger \\'Omen (born 1 950-- 1 96 1 ) in contrast to 39 percent for older women (born 1932- 1 939) . In brief, the influence of age on kinship orientations shows these tenden­ cies : (a) there i s a greater probability among both men and women to hold a Parentela Orders orientation at older (than younger) age levels ; (b) whereas younger men tend to fall more often into the Genetic and Canon Law cate­ gory , younger women more frequently have. a Standard American orientation . The presence of age and sex differences in kinship orientation evokes a fu rther question as to whether these variations can be attributed to stage in the family life course or to the differences in the historic al contexts in which the individuals grew up . Research on family life cycles i s concerned with the effect s of transitions in role and status on the lives of the family members . A s Rei ss ( 1 960) has found , interaction with relatives varies at different stages of the family development . If kinship orientation is situationally determined , then one wou ld expect that persons shift in orientation as their children gro\\' and eventually leave home . However, it is also possible that age and sex differences derive not so much from life course transitions , but from the diverse hi storical experi ences of different birth cohorts . The economic depres­ s ion of the 1 930s may have infl uenced the kinship ideologies of the older respo ndents ; World War II and the Korean conflict may have affected kinship ide as of the middle-range respondents; and movements toward equal rights may have S\\'ayed the younger respondents . Or , since kinsh ip ideologies

1 38 appear to be tran s m i tted in families , the particular historical experi e nces of

respondents � parents and grandparents may have been infl uential i n shap ­ i ng the re spon den t s · own idea s abo ut ki ns h ip . While it may not be poss i b le to delve into the precise impact of historical periods on kin sh i p ori e n tat io n ' it may be pos si b l e to determ i ne whether age differences can be at tri b u ted to birth cohorts or, conve rsel y , to family life cyc l e e vents . At first g lan ce , the fmdin gs on the concordance of age and sex on k ins hip ori e ntat i on suggest that as people get olde r, the i r i n stitut ional i n volv e men ts stabilize and their interest in l i ne of de sce nt increases . But there are p uzzl i ng elements in the fi ndi ngs . Es peci al ly baffl i n g is th at , among young people ( ge nera l ly those below thirty) , men and '-"·omen appear to differ in the ir perc e pti ons of the role of k i n s h i p in soci al endurance . \\'hereas younger m en te nd to re g ard kinship s t ru c ture from an Ego-cen tri c perspe c ti ve ( i . e . , G e ne tic a nd C an on L a w vie wpo i nt) , younger women more often see kinship as a means for social p l ac e men t . Possi b l y , the women born after 1 950 h a ve been influenced s i gn i fi c antl y by the recent feminist movements, affirmative action programs , and c areer concerns . One would then expect that young V.'omen would d i spl a y considerable interest in upw ard social mobility and V�·ou l d , with dec l i ning birth rate s , g i ve greater attention to asce n din g ge n e rati o n s in their struct uri ng of kin sh i p t i e s . Pe rhaps on l y later do familial i n fl u e nc e s ope rate . Insofar as age d i ffe rences i n kinship orientati on am o ng women may rep­ re se n t cohort shifts in oc c u p ati o na l orientation and the role of women in soc i ety instead of life cycle transitions , a further analysis by educational level may shed l ight on th i s i ssue . Table 7- 1 0 shows k i ns hip orientation and ed ­ u c at i on al level by se x of re spon de nt . For me n , educational l eve l affects kin­ sh ip orie ntation in V.'ay s similar to age . Those men with a hi g h e d uc a tion le v e l ( i . e . , grad uate \\l·ork or professio nal school) tend to fall into the Parentel a the

Orders category, while men at low educational levels (i. e . , less than a high school graduate) are overrepresented in

the Genetic and C ano n Law grou pi ng .

Since the find i ngs for men do not discriminate between age and education

as i nflu en ce s in kinship

orientation , these fmdings are equivocal as far as

the

life c yc le versus coh ort i ssue is concerned . Among w om e n , however, age and educati on seem to differ in the ir effects on kin ship orientation . Unlike the age data, ed u c ati onal level of women is not related to ho ld i ng a Parentela Orders orientation. Instead , while those women at low educati ona l levels (i .e. , less th an a hig h schoo l gradu ate ) are o v e rre pre se n ted in the Genetic and Canon Law class, women at high level s ( i . e . , g raduate or profes s io na l school) , as well as being young , are more often in the Standard American category . Hence , there appear to be tv.ro se parat e i nflue nce s affec ting the re lat ion sh ip between age and k i n ship ori­ e ntati o n amo ng w ome n - one of them related to Pare n te la Orders , derived fro m maturation , and the other related to Standard Ame ric an kinsh i p , derived from the hi storical events affecting the birth cohort .

139 TAB LE 7-10 . Kinship Orientation and Educational Level ..-\chieved, by Se¥ of Respondent Educational level Graduate

Kinship orientation Men respondents•1

Parentela orders ( %) Standard american (%) Civil law { %)

Genetic o r canon law ( %)

s

Women respondents h Parentela orde rs (%) Civil

la w (% )

Genetic o r canon

N

a b

Chi

Chi

=

law ( 4i )

20.478 ; d . f.

college

graduate

29 .0

20 . 8

45 . 2

49. 4

1 6. 1

1 3.0

9. 7

1 6 .9

31

1 7 .6

square DOt signifiCant .

square

work

school

=

9;

1 54

Less than

high school graduate

Total

1 9. 2

15.2

20 . 6

46 . 2

45 . 5

47 . 6

19.2

1 5 .2

15.2

1 5 .4

24. 2

1 6. 6

78

33

296

1 4.8

1 3.2

21 . 1

15.4

49 . 0

49 . 1

25 . 4

45 . 8

5 .9

16.8 1 9. 5

1 8.3

1 6. 9

17.6

1 8 .6

34

149

58 . 8

Standard american (4l )

High Some

19.2

1 67

35 . 2

2 1 .9

71

42 1

p = .01 5 .

The in terp re tati o n that the data on age and se x may refl ec t both cohort and

maturation influences on k inship orientation suggests still another analysis

based on household compos ition . Data re le van t to the family life cycle are presented in T ab l e 7- 1 1 which de scri bes characteristics of the respondents ,

children and households . This t able includes reports o n whether the children

had been born to the. re sponden t , ages of children , res idence of children i n

the h ouse ho ld , and marital statu s of adult children . The fin dings i n Table 7 1 1 are :

1 . Recruitment of children . The qu e st ion as to whether children in the responden t s ' famil ie-s have bee n born to the re spondent points to the distinc ­ tion between the "natural' � family and those domestic rel ationships formed through adoption , marri age , or foster pl ac e men t. The data indicate that per­ sons with a Genetic or Canon La\\1· o ri en t at i on more often than others have children in their families who had not been born to them . This finding is of co urse c o n s i ste n t Ylith that on marital status , which has indicated that rel a­ tively few er couples in the Genetic- Canon Law c ate go ry than others in the sample are in their first marriages. (See Table 7-7 . ) 2 . Ages of children . The dat a on ag e s of children are classified in Table 7- 1 1 , generall y according to level of schoo l ing-5 or under corresponding to preschool , 6- 1 2 to elementary sc hool , 1 3- 1 7 to high school , and 1 8 or

1 40 over to c o llege or independent residence . The o v e. rdl l data i ndicate that ch i} .. dren of parents in the Parentela Orders , Genetic . and Canon Law cate go ri e s are slightly older than children with Standard American or Civ i l La \\' p are n ts .

The differen ces among kinship orientations , hoYlever. are generally le ss t h an five percent ( and alV�·ays less than ten p ercen t ) for any age group . Vari ati o ns in

ki ns hi p

orientation by age of respondent far exceed those according to age

of chi ld re n . (See Table 7-9 . ) Hence . fev.' findings regarding respon dents '

ages appear to derive from differences in family l ife cycle stage a mo ng

kinsh ip orientations .

3 . Residence of children . i�S i n the case of children 's ages , the data on re sidence do not prov ide explanatory po\\'·er for di ffe re nc e among ki nship (Jrientatio ns . Th e re is a sligh t tendency for re spo n den ts in the Genetic and Canon La\v category to be overreprese nted in the . . cn1pty ncsf · stage ( i . e . , \\'·here all c h i l dre n have left home) a nd to be underrepresen ted in households in '-"''hich a l l childre n are still in the h o me . B ut the se findings do not seem sufficient to prov ide a reasonable expl anat ion for v ari at io n s in re sponden ts ' age among kin sh i p orientat ions . 4 . A1arital :J·tatus c�f adult children . The percentages of respon d e nt s ,-.,· ith at least one marr i ed chi l d do sho\lt' a considerable di ffe ren c e by kin sh ip orientation . �1ore respo ndents w i th a Genetic or Canon Lav.,· o rie ntatio n have at le ast one chi ld who has marr i ed than do respondents with other orienta­

tion s . Certai nly , some of this tendency can be explai ned by t he relative ages of c h i ldren whose parent s are in the Genetic and Canon Law grouping . S till , by way of contrast, al thou g h children of parents \\l'ith a

Paren te l a

Orders

pe rs pec t i v e are also general ly older than are those of Standard American or

Ci vil Lav.,· parents , they appear to leave hom e late

and

to marry late . The se

data are consi stent with the findings that the parents of respondents in the Parentel a Orders category ,

as well as the respon dents themselves , tend to

marry at a late age , Vt'h ile those in the Genetic more often marry at an early age .

and Canon

Law classific ation

(See Tables 6-5 , 6-6, and 7- 8 . ) Hence , although marri age of adult children i s the only family life cycle vari able that is clearly as soc iat ed with variation in ki n ship orientation , it appears to be

influenced by kinship orientation rather than the reverse .

The overall findings on age , sex , and household com pos iti on suggest that

whereas age and sex differences in kinship orientation

do

occur, these dif­

fe rence s se.em to deri ve more from maturation and cohort experiences-for example , economic depression , �'omen 's movement . or war-rdther than

from transi tions in the family l i fe cyc le. This set of results appe ars to prov i de

fu rther support for the contentions th at kin s h i p orientati on is more a produ ct

of ideology than personal adap t at i on to family s ituations .

14 1 ABLE 7-11 . Kinship Orientation and Characteristics of ChUdreo T and Household

Characteristics of children and housebold

Parentela orders

Recruitment of children Percent of respondents whose children are all � 'natural � (i. e none i s adopted , step, or

Standard

american

Civil

law

Genetic and canon law

•.

.•

foster)

Ages of children 1 8 or over (born before 1 960)

( %)

1 3- 17 (born 1 96 1 - 1 965) ( � )

(r l 2 (born 1 966- 1 972) ( %) S or under (born 1 973- 1 978) {%) Total ( %)

Number

Residence of children Percent of househol ds in whic h all children have left home Percent of households i n which

N

all children are still in home

Marital staws of adult children Percent of respondents with

least one married chil d

69 . 6

62 . 9

67 .6

28.7

25 . 2 22 . 4 3 1 .6

26 . 5 22 . 2

24. 0

22 . 2 1 00. 0

100.0

25 . 6

29 . 0 1 6. 6

20. 7

355

745

99 . 9

30. I

14.0 3 29

1 3 .8

10.2

9.5

68 . 8

68 . 8

74 . 3

285

1 05

44 . 2

51 .2

45 . 5

62 . 2

43

84

33

3i

1 12

Number of responde nts with al

The Salience of Marital Roles

302

31.9

8.0

at

least one child born 1 960 or earlier

99 . 9

29 . 1

55.2

63 . 8 1 16

In contemporary society , the emergence of popular movements pertai ni ng to equal ri g hts and equal treatment of all individuals (reg ardless of ascribed c haracteristics or of pe-rsonal beliefs) have had significant effects o n the family . While critics of famil y life for c-enturies have urged changes in marital' relati ons hip s (e . g . , John Milton and Frederick Engels) , apparently only re­ centl y has the state of tec hn olo gy , literacy, and standard of l iving been suf­ ficient to permit th e enactment of such chan ges . The women 's movements in particular have encou raged freedom of c-hoice re gardin g domestic versus occupation role and, in some instances, legal-marri age versus living-together. Insofar a s kin sh i p orientations seem to reflect i deol og i cal stances , one wou ld expect that labor force commitment as well as w i ll in gnes s to l ive together prior to marri age would be related to c onc e pti ons of kinship . This section

142 deals with respondent's views on the issues of work versus domestic com­

mibnent by women and of the propriety of cohabitation. These issues bear on the maintenance of family boundaries in social space and time.

Table 7- 12 shows the percentages of women in the labor force (or in

college), classified by kinship orientation and year of birth. The data for women respondents show clearly that among younger women (i.e., born

1940 or later), those in the Parentela Orders category are underrepresented

in the labor force, while Standard American women are overrepresented.

Among wives of men respondents, the percentage of younger women in the Parentela Orders category is also low; however, there is little difference in labor force participation among women in the other kinship categories. (The

data for older women, i.e., born in the 1930s, are less significan t than those for younger women in that most of the former women do not have young

children in the home.)

Table 7-12 indicates further that most women in the study are employed.

To determine re l ati ve commitment to household versus occupational demands (at least in the abstract), the respondents were asked the extent to which they

agreed (or disagreed) with the statement, "In a c.onflict between demands of home and j ob , a woman's frrst loyalty should be to her home.'' Since the

number of children living in the household may affect the degree to which people give priority to the home, responden ts with no or one child in the home are compared separately from those with two or m ore children. The f indi ng s appear in Table 7-13. In Table 7- 1 3, the data for women reveal that in small fami lies , those in the Parentela Orders category are most prone to favor familial demands over job demands, whereas women in the Genetic and Canon Law group are least apt to do so (80 versus 59 percent). For larger families, the trend is in the

opposite direction: Parentela Orders women are least willing to describe their TABLE 7-12.

Kinship Orientation and Percentage of Women Force or In College, by Year of Birth Parentela

Year of birth Born 1940

or later

(%)

N Born in 1930's

N Wives of men respondents Born 1940 or later N

Born in 1930's

N

Genetic or

american

Standard

Civil la\\o·

75 . 2

66.7

66.7

79.3

63 . 5

76 .0

7 5 .9

29

52

25

29

50.0

57.7

57.0

58.1

38

97 68.2 22

30

31

orders

Women rc:spondents

Either in Labor

58 . 3 36

61.1 18

141

canon law

63

48

50.0 16

143

ABLE 7-13. Kinship Orientation and Percentage of Respondents Wbo Agree T with Statement: ''In a conflict between the demands of home and job, a woman 's first loyahy should be to her home," by Sex of Respondent aad Number of Children Living Home Women respondents None or one

child in home Parentela orders ( %) N

Standard american (%) N

Civil law (%)

N

Genetic N

or

canon law (%)

Two or

more

children in

Men respondents None or one

home

child in home

80.0

78.3

15

46

84.2

76.3

86.0

76

107

19 72.7 55

Two or more children in home

91.9 37

94.0

67

71.4

88.6

59.4

92.7

77.3

93.8

55

22

16

21

32

44

78.6

14

81.5

27

frrst loyalty as being to the home, and the Genetic and Canon Law "·omen

are

most will ing (78 percent versus 93 pe rcent) . Note however that for women

in the Parentela Orders category, size of family appears to make littl e dif­

ference in the prec i se percentage ag reeing with the s tatement (80 versus 78

percent); I in terpret this s im il arity in percentages to mean that, for Parentela

Orders women, degre e of loyalty to home versus job is a matter of prin c i p le and not circumstance . At the other e xtreme, with a great discrepancy in

percentage by family size among Genetic and Canon Law women (59 versus

93 percent), presumably situational factors p lay a more significant role in determining one 's first loyalty.

The d at a for men in Table 7-13 are less conclus ive than those for women. Aside from the slightly larger perce ntage of Parentela Orders men in small families agreeing v.·ith the statement, there is littl e si milarity between findings

for men and for women. For most cate go ri es , the men show a gre ater pro­ cl ivi ty to agree with the state m ent than do women-regardless of kinship orientation.

The salience of marital role over occupational role amo ng Parentela Orders

women is parelleled by views of these women on restricting female-male

cohabitation to married coup l es . The percentages of respond ents who agree

with the statement that, ''It is a good idea for a man and a woman to live together for a while before they marry," are pre sen ted in Tab l e 7-14. For

men, there is little distinction among kinship orientations with re gard to ag reem ent with this statement. A mong women, however, there is a wide variation by kins hip orientation. Whereas o nl y 15 perce nt of the women in

th e Parentela Orders category approve of premarital cohabitation, fully 46

perce nt of those in the Genetic and Canon Law classes

are

in agreement with

144

TABLE 7-14.

Kinship

Orientation and

Percentage of Respondents \Vho Agr

with the Statement: "It is a good idea for a man and live together for a while before they marry.''

Kinship orientation Parentela orders (%) N

Men respondents who agree

Women respondents

44.3

15.3

61

Standard american(%)

51.0

Civil law (%)

44.5

N

N

Genetic or canon law(%) N

woman:

141

45

51.0 49

who agree 65

34.7 193

25.3

71 45.6 92

the statement. Hence, the data on premarital cohabitation reinforce the inter­

pretation that orientations regarding kinship ties express conceptions of family

life along a dimension of corporateness-with Parentela Orders symbolizing the greatest salience of status in a corporate group and Canon Law and Genetic models the least.

Kinship Conceptions and Family Roles: Summary This section has focused on the connections between conceptions of kinship

and family roles in two ways: ftrst, by examining whether the family life cycle played an important part in determining kin ship orientation and, second,

by comparing the salience of marital and domestic roles with other commit­

ments. The findings with regard to family life cycle influences indicate the

foll owing :

l . Age of respondents is associated with kinship orientation. Whereas younger men in the sample are overrepresented in the Genetic and Canon Law classification, older men tend to fall more often into the Parentela Orders category. But while older women are also clustered more than expected in the Parentela Orders class, younger women more frequently are classified as Standard American.

2. Among men, age and education converge in their influence on kinship orientation, but for women, their effects are divergent. Instead, among

women, high educational level (i.e., graduate or professional school) is as­

sociated with the Standard American model. This finding reflects the increas­ ing educational levels among younger women.

3 . In contrast to the marked differences among kinship orientations found for age and educational level, analys is by ages of children-intended to

145

. te family life cycle influences on orientations--fails to yield compa1 di ca ble dis tinction s. Hence , family life cycle variat ions by themselves seem in determining kinship orientation. nsigni ficant

:

� 1

4. The on ly notev.'orthy findings w ith regard to household composition pertain to the large number of Geneti c and Canon Law fimi l ies with step and t ch ildren and the relati vely small percentage of famil ies in the Par­ adop ed entela Orders category vY'ith adu lt married-children. These data , how·ever, appear to reflect i nfluences of kinship orientation rather than household at­ tribute s as c ausative factors .

The contrast between findings by age (coupled Ylith edu cation ) and those

by fam i ly life cycle su ggests that the historical factors affecti ng enti re birth cohort s are at least as influential as life-course trans itions w ithin the family

in dete rmi ning shifts in kinship orientation . The conc lus ion that age differences are partl y a function of the historica! context in which a birth cohort grew up implies that kinship orientation will

itself give mean ing to life cycle transiti ons and family role s. Findings per­ taining to family roles inc lude:

l. Among younger women (i.e . , born in 1940 or l ater), those respondents

in the Parentela Orders category are underrepresented in the labor force , while Standard A merican women are overrepre sented .

2. Among women in smal l famil ies , respondents in the Parentela Orders

category most often favor giving priority to familial commitments over job demands , V.'hereas women i n the Genetic and Canon Lav.' grouping least often do

so.

In l arge families, the trend is in the opposite direction. But for Y�·omen

in the Parentela Orders category , size of family appears to make l ittle dif­ ference in choosing home over job , while for Genetic and Canon Law women , family size itself has a cons iderable impact. Presumably , among Parentela Orders women weighing loy alties between home and job is a matter of principle, and among Canon Law and Genetic model women situational factors play an important part i n determining priorities .

3. As opposed to women in the Parentela Orders class , \\'omen in the Ge netic and Canon LaY�· category endorse the notion that , ''It is a good idea for a man and a woman to live together for a while before they marr y . �' These findings suggest the strong significance of the status of marr iage for women with a Parentela Orders perspective in contrast to the vie�·s of w·omen classified as Genetic or Canon Lav-'. 4. The data for men on the relati onship betwe. e n kinship orientation and family role are generally inconclusive . Hence, the findings on fami l y role, l i ke those pertain ing to age and sex differences, support the view that the kin ship orientations represent different degrees of corporateness in family and ki nship ti es. \Vhile generally kinship orientations tend to be transmitted in families, cohort and educational effects

are more important than family life cycle influences in modifying kinship

146

conceptions; meanings of \\l·ork and marriage are then structured in ways

co ns i ste nt with kinship orientations in protec ti ng family boundaries.

KINSHIP ORIENT ATION AND FERTILITY In

c ont e

m porary society, every married couple faces the

decisi on

of the

e v entual size of its family. Insofar as kinship orientations represent ideologies

govern ing family relationshipst these orientations would also have implica­

tions for fertility decisions. l\.1oreovcr� since kinship orientations seem to

pers i s t in family lines over generations, one would anticipate that fenility

rat e s would vary not only among respondents themselves but also among their parents, siblings, and children in ways consistent \'lith the respondent.�' kinship orientations.

T\\'o n1odeis of collaterality und er investigation differ in the extent to \vhich

they distinguish bet\\'een closeness of ascending versus descending genera­ tions. The P a ren t el a Orders model assigns

a

greater proximity to Ego �.s des­

cendants then to his ascendants� \\'hile the Standard American n1odel �tresses

the close. ness of ance�tors. On the basis of differential v.-eighting of ascend­

ants and de�cendants, one would ex pec t that persons \vith a Parentela Orders

orientation \\'Ould have a larger-than-average number of children� \\·here as those with

a

Standard .An1erican perspective \\'Ould have relatively s1nall

families. The remaining models-Civil Lav.', Genetic, and Canon La\\'-give

equal weight to proximity of ascendants and desce.ndants; as a result, these models do not provide any cl ues per se of i mpl ica tions for fe rt ility rates. Instead, they appear to accommodate a greater flexibility in fertility rates,

and birth prevalence \\l·ould depend to a greater extent (than wou ld be the

case for Parentela Orders and Standard American) upon the specific social contexts.

This section is divided into two parts: the first part deals with the fertility

of close relatives of the respondents and the second pertains to the respon­ den - ts' projections with regard to the i r own fe rti lity .

Fertility of Close Relatives The ability

of

kinship

orientation

to circumscribe

fertility

among close rei a­

tives is described below in t\\'O tables, the first referring to fertility over

ge nerat ions and

the

second to the multiplier

e ffec t

of kinship orientation.

Table 7-15 pre sen ts data on the fertility of the respondents' parents, sib­

lings, and c hildre n (as well as the

and projections).

respondents' own

child-bearing experiences

There are several regularities in the data. Across genera­

tions, with minor ex ceptio ns : l. Close relatives of persons

with Parentela Orders or with Genetic or Ca non

Law orientations tend to have high fertility levels.

147

se relatives of persons v.·ith a Standard American orientatio n tend to 2. Clo have low· fertilit y levels. 3. Close relat ives of persons \'lith a Civil Law ori en t at ion are intermediate

in fertility.

The ex c e pti on s to these g eneral tendencies occur whe re birth histories are perhaps most incomplete-among the G en et ic and Canon Law respondents

and among the married children of Pare nte l a Orders respondents. Hence., the findings on the fertility of re s pond ent s ' parents , s iblin gs , and children appear to suppo n the proposition that pattern of collateraJity ex­ pressed by k ins hi p models is related to fertility level. Moreover, data on age at marriage (Tables 6-5 and 7-8) indi c ate that ki n shi p orientation ope rat es independently of marital age in a ffec tin g fertility.

Possible effects of d i fferen tial fertility upon the distribution of kinship

orien tations in

a

p opulat io n are suggested in Table 7-16. This table displays

the mean number of the respondents� nieces and n eph ev. s for the different ·

kinship orientations and takes into account the respondents' rel igion and father's oc c upat ion . The data on re ligion show c l e arly the long term conse­

quences of loY.' fertility

TABLE 7-lS.

amon

g Je"Y.'ish respon dents (2.3 n i ece s and ne phew s)

Kinship Orientation and Mean Number of Children Born to Respondents' Parents, Siblings,

Respondents and and Children

Standard american

Civil

Genetic or

law

canon law

3.15

2.56 334

2. 91

3.27

116

141

2.82 126

2.23

2.60

2.33

334

116

141

2.86 106

2.49

2.75

261

97

2.63 96

3.02

2.71

2.70

2.87

302

59 1

2.54

322

2.00

1.35

19

43

2. 13

3.04 23

Parentela

orders

Parentage Mean number of respondents' full

N

siblings

126

Mean nwnber of children born to

N

respondents

to

Mean nwnber of children

anticipated eventually by respondents Na Mean number of children born to

respondents� siblings Number of siblings with children

Mean number of grandchildren per

married child Number of married children

" The respoadents

n ot included

pens,"' "It's up to God.'' ··As mao)· dents also excluded.

as

IS

ga\le nonquaotifiable answers. such as "'Wlultever happossible.'' ..\'I.e haven't decided )·et, . . and so on. Divorced respon-

in t hese calculations

1 48

as contrasted with the traditionally hi gh Catholic birth rate (7. 2 nieces and

nephews). Similarly, the data indicate the high contribution that bl ue collar families have made to the po pula tion of Phoenix. As for kinship or ien t atio n' with only minor dep artures (among Catholics, Jews, and blue collar workers) , the results are similar to respondents' own fertility rates; the mean number

Gene tic

of niec es and ne phe w s is highest in the Parentela Orders,

,

Law categ ories and lo we st in the Standard American class.

and Cano n

Insofar as the Standard American orientati on is most pre va lent among the nonminority , educated middle class , the findings on numbers of nieces and

nephe w s of respondents (taken in conjunction with the re spo nd e nts ' own

fertility) seem to have significant implications for social stratification. The low prevalence of grandchildren of the Standard American res pondents par­ ·

ents presumably operates to diminish the number of middle-class candidates for desirable occupational positio ns and leaves at least some slots open for upwardly mobile individuals. More than that, the low productivity of children

also mini mi zes the possibility of having surplus childr en who cannot maintain

a gi ven socioeconomic status. It thereby enhances the probability of the family's retention of a

high soci oecon omi c status, while si multan eous ly per­

mitting u pwar d social mobility of other famil i es .

Orientation and Mean Number of Nieces and Nephews of Respondents by Respondents' Religion and Occupation of Fatbersa

TABLE 7-16. Kinship

Parentela orders

Religion and fathers' occupation Religion No preference N

4 .43

Jewish

14

N

Protestants N

Catholics

Civil law

Genetic or canon law

7.60

2.08

20

13

1.21 19

7.73

5.09

63

1 75

8.35

6.26

37

N

Standard american

85

Total

6.33 33

2.38

1.14

2.33

8

7 7.35

6.28

6.91

68

69

48 315

5.90

8.78

7. 15

29

41

192

Fathers' occupationcl

Professional, managerial, or administralive

N Clerical, sales, N Blue collar

or

craft

FBihen'

occupation

6. 1 3

6.62

5 .43

51

105

39

34

5.58

4.05

5 . 34

5 . 36

229

35

44 9.46

31 36

TotaJ with adult siblings

a

4 . 37

10.3 1

N

N

6. 29

96

7 . 83 93

7 .68 34

48

4.78

206

8 .60

211

7.33

5 .36

6.36

7.26

6.26

118

294

108

126

646

at time respondents were growing up.

149

projeded Fertility of Respondents a In smuch as the sample in this investigation is limited to persons between ages of 18 and 45 virtually all re spon dents are c apable of bearing more they already have . Although most respondents may have already children than c

ompleted their families , some of them may anticipate having additional

children. Consequently, in the interview, the respondents were asked the number of children th ey antici pated having eventually . (This question was

omitted from interviews with divo rced persons.) While the number p rojected b)' the respondents may n ot in all instances correspond to the actual number they ultimately produce, the projected number may provide a more realistic estimate of ferti lity than the number of ch ildren they now have . Table 7-17 describe s the levels of fertility eve ntual l y an ticipated by respo n­ dents with differ ent k i nship orientations and diverse social characteristics. For the various sections of the table, the results are as follows:

1. Men and Women. Except in the Standard American category, women generally anticipate having a larger number of ch i ldren than men do. Both men and women in the Parentela Orders category tend to antic ipate a large number of children and those in the Standard American class a small number, with individuals in the Civ il Law grouping between them. Among kin s hip

orienta ti ons, h owever, men and w omen in the Genetic and Canon Law cat­

egory d iffer considerabl y in expected fertility. In that category, me n provide the lowest mean number of children anticipated (as compared with the means in the other categ ories) ; but the mean for women in that category i s as high as

that for Civil La w respondents. This differenc e in an ticipated fertility

between men and wom en in the Genetic and Canon Law grouping persists

in the other sections of the table de spite independen t influences of other social

characteristics.

2. Religion. Within each kinship orientation, for both men and women, there is a general tendency for J ew i sh respo ndent to anticipate fewest children and Catholics mo st , with Protestants between them. Overall, e x cept for Prot­

estant men in the Parentela Orders category , differe nces within religion fol­

low the pattern s of fertility according to kinship orientation. Hence , although religi ous identity i tself doe s af fect fertility , this i nfluence seems less profound than that of kinship orien tat ion . 3 Education. Like religion, education exhibits a moderate influence on fertility as compared with kinship orientation. The most significan t impact ·

of education is that those persons with at least a hi gh school education an­

ti cipat e having fewer children than i ndivid uals who have not graduated from high school. As for kinship orientations, the Parentela Orders men and

Women who had had some college (but not graduate study) expect to have

150

TABLE

Orientation and Anticipated Eventual Fertility, by Various Social Characteristics

7-17. Kinship

Mean number of children

expected eventuaUy

Parentela

Standard

Civil

orders

american

law

Men

2.72

2.51

2.60

114

\\'omen N

55 3.02

2.48

40

2.86

51

147

57

Social

characteristies

ALL MARRIED RESPONDENTS

N

Genetic or canon law 2.18 34

2.87 62

RELIGION

Men Jewish

3.00

N

6 2.44

2.51

2 52

25

58

3.06 16

25

2.71

2.60

Jewish

2.28

1.82

2 . 38

7

8

Protestant

2.96

11 1 .97

2.82

27

92

34

3.86

2.56

2.93

14

32

15

2.83

2.36

Protestant N

Catholic

N

2.00

2.20

9

31

5

2.28

.

18

2.33

12

5

Women N

N

Catholic N

2.97 31

3.23 22

E DLICA TION

Men Graduate study N

6

Attended college

�

Less than N

high school

3.00

2 58

2.46

15

31

13

4.25

2.90

2.50

4

10

10

30

graduate

2.05 19

3.06

.

High school

7

2.44

2 33

N

1 . 43

11

62 .

18

2 .43

7

(continued}

fewer chi ldren than other persons with that orientation. Otherwise, the data (with minor exceptions) fall into the overall fertility pattern by kinship ori­

entation found for men and women generally.

4. Woman's employment status. The data on the employment statu s of

women respondents and wives of men respondents reinforce the often rep­

licated findings that homemakers have (and anticipate) more children than mothers who are employed outside the home . Yet , even with maternal em­ ployment taken into account, the force of kinship orientation on anticipated

15 1 ) tABLE 7-17 (continued Mean number or children expected eventually

Parentela

Genetic or

SoCial characteristics

orders

Standard american

Women Graduate study

2.60

1 .79

2.20

14

5

N Auended college s High school graduate

N Less than high school graduate N

5 2.22

2.46

N

Homemaker

N

Women respondents

Now employed N

Homemaker N

canon law

law

2 . 65

2.37

18

61

20

19

3.38

2 . 38

2.88

2.4 1

16

58

24

22

3.92

3.21

3.36

4.06

12

14

II

18

2.35

2.38

2 . 09

1.53

28

56

22

15

3.1 2

2 . 74

3 . 38

2.73

16

15

2 . 52

2.33

2.38

2.63

32

32

3.67

2.73

3.42

2.88

21

52

19

26

2.5 1

2.47

2.12

45

92

2.61

33

26

3.60

2.68 22

2.57

2.38

7

8

2.44

2.90

2.7 5

41

48

WOMAN'S EMPLOYMENT STATUS

Men respondents� wives Now employed

Civil

26

27

47

84

MARITAL STATUS

Men

Now in riJ"St marriage N �ow remanied N

\\'omen Now in first marriage N

Now remarried N

10

3 . 03

37

1 22

3.00

2.64

2.75

3.29

14

25

16

14

ferti l ity is a significant one. The only departure from the general pattern is

that homemakers in the Civil Law category expect to have an especially high

number (and conversely working mothers a low number) of children.

5. Remarriage status. Table 7-17 also includes a tabulation comparing persons in th eir frrst marriage with those in s ubsequent marriages. Apart from the Civil Law respondents, remarried persons (both men and women ) antic­ ipate having a greater number of children eventually than do those i n their first marriages. Inasmuch as fertility pertains to children from both marriage s ,

152

presumably some remarried persons conside r h aving two sets of children-­

one set in each marriage. Since this number represe nts anticipated (rathe r than actual) fertility, age does no t appear to be an important f acto r in this estimate. However, even when remarri age status is held constant statistically

kinship orientation still affects anticipated fertility. The patterns of fertilit

expectations by kinship orientation of remarried persons found for the entire sample.

;

are s i m il ar to thOSe

6. Contact with mother. Another possibly confounding variable related to kinship is the frequency of contact with one ' s own mother. Perha ps , insofar as one ' s mother often remains as a significant figure in influencing family decisions (either directly or as a reference person), it seems pl ausible that frequent contact with one's parents would reinforce fertility in the adult chil­ dren. (This statement assumes that parents are usually pron atal in outlook.) Data pertaining to the relationship between anticipated fertility and contact with mo thers appear in Tabl e 7-18. For the purpose of this anal ysis , frequent contact is i nterpreted as ''at least once a month . '' As the table clearly shows, for each kinship orientation (outside of women in the Genetic and Canon Law grouping), those men and women who see their mothers at least once a month expect to have more children than those persons who see their mothers less often. Yet, al though maternal contact does affect anticipated fertility, kinship orientation retains it s mark upon childbe aring.

and Antkipated Eventual FertiUty, by Frequency with Which Respondent Sets Own Mother

TABLE 7-18. Kinship Orientation

Number of children expected eventually Frequency with which

mother

is seen

american

Civil law

Genetic

orders

Standard

3.14

2 .S9

2.S8

2 .2 1

21

4S

19

14

2.37

2.29

2.28

2 . 13

27

48

18

IS

3.07

2.5 1

3.07

2.62

15

(Jj

29

29

2.58

2.41

2.50

61

22

2 . 72

Parentela

Men

Mother seen at least once

N

a month

Mother seen less than

N

once a month

Women Mother seen at least

once a month

N Mother seen less than once a month N

26

or

canon law

18

153 In brief , the percentages reported in table 7-17 and 7- 18 verify the impor­

tant role of kinship orientation in fertility rates even when other characteristics

which ordinarily affect childbearing are con trolled . These other characteris­ tics-sex, religion, education, maternal employment, remarriage status, and t rnal contact-while they each exert an independent constraint, are for ma e the most part subordinate to the impac t of kinship orientation upon anticipated fertility.

Kinship Orientation and

Fertility: Summary

The data on fertility generally reflect the weighting of ascendants and des­ cendants in the collaterality models over generations of families. Parentela Orders, which emphasizes closeness of descendants, tends to y iel d fairly large families; the Standard American model, which stresses ascendance,

produces small families; and the remaining models, which are eg o-ce ntered , tend to be flex i ble in fertility-sometimes (for Genetic and Canon Law models) resulting in large families in the parental generation and smal l fam­ ilies among the respondents' siblings.

The continued influence of kinship orientation on fertility over a series of generations has a multiplier effect on the proportions holding each orientation within a population. The di fferential fertility among the kinship orientations is magnified with each new generation (as the data on nieces and nephews indicate) . (In none of the social categories in Table 7-17 is the expected

fertility highest in the Standard American classification.) The consistently

low fertility rates of persons w.ith a Standard American orientation carry various implications for social stratification. The small number of children in these families permits maximum concentration in expenditure of resources upon these children and thereby serves to sustain a desirable socioeconomic position for the family members (or to enhance chances of upward social mobility) in the succeeding generation. In addition, the continual lack of a large surplus of children minimizes the probability of marrying someone of lower socioeconomic status . In a growing economy, the small number of

children produced by persons with a Standard American orientation would

necessitate recruiting new families into the middle class as managerial, ad­ ministrative, and professional positions multiply. In a contracting economy, families with a Standard American orientation would not be burdened by an

�xcess of children whose marital choices and careers would be prob le matic m hard times. Overall, the relatively low fertility rates which characterize the Standard American families yield long-run effects over generations in ways Which sustain the socioeconomic positions of these families.

The pervasive influence of kinship orientations on ch ildbearing is sug­

�e sted not only by the stability of family si ze for respondents' parents, sib­ lin gs, and children (as well as for themselves) but also by the data on number

154

of children anticipated by the respondents . The impact of variable s nonn an associated with fertility rate s--sex, religio n, education , maternal empt o ment , remarria ge , and mate rnal contact-is o rdinarily less than the i nfluence of kinshi p orientation in determining the number of children anticip ate d by the respo ndents .

i.

SUMMARY: KINSHIP ORIENTATION AND FAMILY ORGANIZATION

The typology of centripetal versus ce ntrifu gal kinship organi zation ge nerates a co rresponding typology of unatural" versus "legal" family mode ls . Tra­ ditional ly conceptio ns of the family have pre supposed a fixed set of fun ctions derived from ''nature''. Law i n co ntemporary society, however, is fa ced with the problem of dealing with continuity in social structure s without presup­ posing fiXed e nds or c orporate ''reality" for these structures. Consequen tly, family law assumes that social structures e xist for whateve r re asons individ­ uals have . As a result , laws have to accommodate a variety of ends and to con sider different value systems as bein g of e qual wo rth . The idea th at all domestic groups have equ al rights and standing in the society thereby o ppos es the no tion that con ge ries of families should be encouraged to s ustain ho­ moge neous c lusters o n the basis of "naturally " derived criteria. The disti nction between ''natural'' and ''legal'' family models implies also a diffe rence between corporate and individual p rivacy . The concept o f privacy itse lf rests upon the deg ree of autonomy from the general commun ity. To the extent that conduct is not subject to public surveill ance and accountability, it remains p rivate . In co rporate privacy , i t is the '' natural '' family v.'hose autonomy is to be protected; i n perso nal privacy , i t i s individu al au tonomy that is paramount . The dichotomy of corporate versus individual privacy denotes a relation­ ship between no rms pe rtaining to family privacy and modes of kinship ori ­ entation. The emphasi s on line of descent in the Standard American and particularly in the Parentela Orders model su ggests that people who co nfonn to these models tend also to con ceive of family org anization mai nly in te nn s of corpor ate privac y; the Ego-centeredne ss of the Ge netic and Cano n Law models, however, appear to refl�t a stress upon individual privacy in family rel ationships . The data prese nted here e xamine the e xtent to which kinship o rientations e xpre ss varyi ng degrees of corporateness in the respondents' families of procreation. Prevalent sociol og ical conception s of the fami l y general l y focus on the here-and-now an d the domain of the family is frequently delimited to the

househol d. Attention is then usually directed to attributes of the household family: (a) social pathologies, such as personal abuse , alco holi sm, deli n­ que ncy, men tal illne s s , or not "getting along," (b) e xplo rations of alternative

155

l ife sty le s , (c) soci al ization and social placement , (d) effects of socioeco­ nom ic or ethnic variables upon interpersonal ties , and so on . Although the se con cerns with the household family are often interesting and of practical i mportance , they fail to take into account the fact that relevant fami ly ties may exte nd beyond the household in time and space . Indeed , the very c rucial el ements for understanding interaction within the household may inhere in characteristics of kinship that have little to do with residence . When th e household family is regarded as the basic unit o f analy si s , then the tempo ral span i s l imited to life course transitions of the husband and wife . 1be sp an nonnally covers the period from the establishment o f a neYl house­ hold by a married couple to the time that the household is dissol ved by their death . B ut if the field of family re -l ationships is defined by the organ ization of kinship statuses , the household family is merely one segment of an entity that can stretch over several generations. For example , ind ividuals who stress line of descent as a source of identity may incorporate ancestors from the dim past-long dead-into the operational fami ly boundaries . Si milarly , yet-un­ born descendants-though necessarily anonymous-may be pulled inside the family boundaries as ' �kinship gives life meaning by prov iding people w ith a mechanism by which both their substance and their personal values . . . survive them (Craig , 1 979, p . 95)." The symbol ic extension of the. family i n time yields a coherent set of goals and norms that organizes the conduct of the family members into a corporate body and constrain s their conduct. The material in this and the previous chapter has suggested that the kinship models represent family ideologies (implying varying degrees of corporatene ss) that exert the-ir influence. over generations in significant ways . Latent consequences of separatist (or plu­ ralist) models (e . g . , Parentela Orders) differ con siderabl y from those of com ­ munal models (e . g . , Canon Law) . The corporateness implied b y the separatist models influences the homogenization of basic values in the family , stabil i­ zation of the marri age , inhibition of disill usion ment , establishment of the priority of family status over other (perhaps more personal) commitments , and the maintenance of fertility nonns over a series of generations . The se findings support the vie.w that the kinship models represent pervasive ideo log ies that influence the conduct of the con stituent household families among c losely related kin . Much of v.·hat i s happen ing in the household of one person probably has its ideological roots in the parental household; very l ikely this conduct is being repeated in the households of siblings; and i t stands a good chance o f affecting the succeed ing generation .

Relatives and Strangers

There is considerable ev idenc e th at incre a singly in modern urban settings

we are becoming s tran ger s to ou r rel­ ati ves--e v en to our spouses, parents ,

and children . Thi s a l i enati o n often

occurs not so mu ch as a result of ide.. ologi cal revisions , but bec ause the hig h degree of mo bil ity in modern soc i­ eties does not permit many of us to have an intimate rel ationship w i t h rela­ ti ves , and our d i ver s ity of personal interes t s draws us away from domestic l ife itself. There is, th us , some tendency for social scientists to consider family and kinship roles as continually dimi nishing in si gnific a nce in modern life (Bahr, 1 97 6 , p . 6 1 ) . Accompanyi ng thi s view reg ard ing the langu i s hi ng of kin s hip , there has been a prol o nged controversy over whether the fam i l y i s bec omi ng • 'struc­ turally isolated ' '-a tightly bounded , re lativ ely autonomous sub-sy stem in society ( Morgan , 1 975) . This c o ntroversy has s timu l ated considerable inves­ ti gati o n into the ties between the nuclear family and relatives. These inves­ tigations have shown that , generally , families rely on relatives in t im e of crisis ( Lopata , 1 973 ; Adam s , 1 968� Bahr, 1 97 6); w omen fonn the maj or com mu ni c ati on agents with ki n (Adams , 1 968; Bahr, 1 976); parent s tr ansfer wealth to married children over the li fe course ( S u s sm an and Burchina1 , 1962) ; people u se siblings as reference markers for evaluating their own de­ gree of succe s s (Adams, 1 968) ; and so on . The major point, however , is that, despite the persistenc e of these u se s of kin , the w eig ht of evidence sh ow s a movement toward more and more structural isolation . This trend has indicated to Schneider ( 1 968) that American kin sh ip is not based on a unifonn s et of social roles, but that i t is co ming to be defined as a system of statu ses based on biogenetic ties and emptied of specific rights and obligations upon which most people agree .

Although the trend toward structural isol ation of the nuclear fami ly has

changed modes of interaction with rel atives , the institution of kinsh i p endures as

a soc ial category in modem culture . But, having acknow ledged its en dur ­

ance , we are obl i ged to ask: Given the broad range of life circumstances , hovl

does the fact that people are re.lated to one another ge nealogicall y impinge.

upon their lives? Once having a ss igned "kinship " ' an existential status , peo­ ple mu st deal with it and integrate it in to their outlook in the world . I sh al l

1 57

now c onsider topics relevant to ways in which people handle kinship

as

a

social category .

KINSHIP AND SOCIAL EXCHANGE If kin ship statuses are divorced from specific role prescriptions in modem

society , then one can anticipate a broad range of orientations with regard to

appropriate interactions with relatives . Depending upon circumstances , at one

ex treme, people can regard their kinship ties as uniquely important and, at the other extreme , they can ignore these tie s . This variation suggests that kinship ties can be regarded as a commodity in soci al exchange-valued highly in some situations and easily expendable in others . Peter B l au ( 1 964) proposes that, "There are a number of similarities be­ tween social exchange and economic exchange . . . The principle of the even­

tually diminishing marginal utility applies to social as well as economic commodities (3 14-- 315 ) . ' ' One would anticipate that where the persi stence

of the family expresses a special interest, marginal uti l ity in the maintenance

of ties would be quite different from the situation in which the family must

compete with other institutions for supplying these satisfactions to the indi­ vidual . "Competition occurs not only among like social units that have the

same objecti ve but also among unlike units with d ifferent objectives (Blau ,

1964 , p . 33 1 ) . "

Past research on the family and kinship offers a wide range of assumptions

regarding the uniquesness of kinship as a commodity of social exchange .

lbese conceptions of kinship can be placed on a continuum to produce a series ranging from the uniqueness of kinship ties to the irrelevance of such

ties. Each point on the continuum seems to be expressed by a metaphor of

giving and taking . Meyer Fortes depicts kinship nonns in tenns of generosity ; Lewis Coser portrays institutions

as

greedy; Michael Anderson describes kin­

shi p in terms of risk , uncertainty , and profit; while Erving Goffman discusses social relationships as fraudulent manipulation . These conceptions are de­

scribed below as they relate to the Parentela Orders , Standard American , Civil

Law , Genetic , and Canon Law models . Kinship and Generosity Using a political an alogy , Meyer Fortes ( 1 969 , pp . 97- 99) concerns himself

with kinship as a unique domain of social order:

I suggest that the social and cultural elements and processes that make up

a given social system fall into determinate sectors of organization. Each sec­

tor-wh ich l call

a

domai n--c omprises

a

range of soc ial relations, customs ,

norms , statuse. s, and other analytically d iscriminable elements linked up in

1 58 ne� uses and unified by the stamp of distincti ve functional feature s that

are

common to all . . . A domain i s not merely a classificatory construct . It is

a

matri x of social organization in the sense that its members derive their spe ci­

fi c ity from it .

.

.

A status can be defined

as

a po s ition held by a person i n

a

gi ven domain , specified by the soc ial relati ons distinctive of that domai n . and

deployed in activities and attitudes con formi n g to the norms and customs and

material apparatus that are distincti ve of that domain .

Kinship as a domain is governed, ac cord ing to Fortes , by a norm of gener­ osity , a rule of amity . By amity , Fortes does not mean ' 'love ' ' but i ns te ad a ' ' co nsens us in acc epti n g the value of mutual support in mai n tai ni n g 'a code of good conduct ' for the realization of each person 's 'legitimate intere sts , ' . . . even by ac t s of violence reg ard ed as legitimate (Fortes, 1 969 , p . 1 1 0) . , H"e suggests further that this pre scriptive al truism is a uni versal characteri stic of ki nship (Fortes , 1 969 , p . 234) . At least in his emph asi s on prescriptive altru i s m , Fortes takes a position cons is tent with that of Le vi-Straus s . According to Lev i - Strau s s ( 1 969) , co r­ porate structure is stabil ized when the re are o bl igations that cannot be paid off-the indebtedness is alw ay s greater than the amount returned . The Fortes position implies that the debt of being a kinsman is alway s gre ater than any ac ti vi ty that the individual can perform for his rel ative . This n orm of gen­ erosity , thu s , repre se n ts an ideal condition that su stains a corporate structure of ki n sh ip . T he c onc eption of ki nsh ip as a domain governed by a norm of gene ro si ty i mp l ies that social space is partitioned in to a complex of categories (i . e . , corporate un its) , e ac h of which i s gov e rned by a unaque set of re c i proc ity norms . This connection between social c ategory and reciprocity nonns sug­ ge st s that peo pl e who apply norms of generosity to kin have a categorical rather than a gradational perspective of relatives . Accord i ngl y , one y,rou ld anticipate that peopl e conforming to the Parentela Orders model of genea­ l og ic al structure would tend to apply the norm of generosity to broad classe s of relatives . Inasmuch as the Standard American model , l ike Parentela Or­ ders , derives its structure from categories based on line of descent, it too i m plies the norm of generosity in organizing kinship ti es. Kinship and Conflict One view of kinship as a po l i ti c al do main produc e s the Fortes conception of

tie s between relatives as resting upon the nonn of amity (or of generosity ) . Another pers pec tive , hov.'ever, denotes a conflict either v.'ithin the kinship group or between kinship obl igations and those pertaining to competi ng in­ stitutional commitments . A problem in the mai ntenance of kinship ties emerges when conflicting demand s are placed on individual s by kinship ties as opposed to obligatio ns

159

deri ved in ano ther i nst ituti o n al sett i ng (e . g . , empl o y men t) . Under such con ­ d itions , pe op le may regard their k inshi p ties (as V..' el l a s other c ommitme nts ) - s s i ve . Coser applies th e anthropomorphic metaphor of ' ·greed iness ' ' a s e xc e to those org anizations and group s that ' 'make total claims on their me mbers a nd \Vh ich attempt to e ncompa s s within their circle the whole personal ity ( Coser , 1 974 , p. 4) . " E xten d ing the metaphor further, Coser sugge sts t hat the- dem and s of � 'greedy " in stitutions upon the person are · •om n i vorou s " an d in s at i able . No matter hoy; much ti me and effort the ind i v idual devotes , the institution calls for more . Suppose that familial and c i vi l institutions are " greedy " in their demands , both of them offeri ng i n return un iqu e kinds of sat i sfac t ion to the individual . Wi l l i am Goode ( 1 960 , p p . 494-495) proposes that "rol e demands made by one insti tut ional order confl ict w i t h those made by ano ther . . . . The indi­ vidual cannot fully sats ify all d e ma nds , and [he] must mo ve t h roug h a con­ tin uo us sequence of role deci sions and bargains , by w h ich he attempts to adj u s t these demands . . . The social structure dete rmine s how much freedom

in manipulation he possesses. ' ' Coser further suggests that incompatibil ity in the i nd i v id ual 's alloc ation of time is not itself the significant factor in the conflict between ' 'greedy ' ' i nstitut i on s . Rather, it i s a matter of personal i n ve stme nt : sacrifice and de­ votion . • ' S inc e it is very d i ffic u l t to re pu diate objec t s in w h i c h one has invested so mu c h . the more one invests in an object , t h e greater the hold that object has on the pers on (Coser, 1 974 , p . 9 1 ) . n If the person then invests heavily in ' 'time , energy , and affect ' ' in both family and e m p loym e nt , the po tenti al con fl ict between the " 'norm of g e nero s it y " in fam i ly m atte rs and the norm of ac h i evemen t is heig htened . The stu dy of kinshi p patterns of Jewish families in Ne\\i York by Leic hter and Mitchell ( 1 956) showed that co n fl ict general ly arose when relative s c on ­ sidered the demands or expectations of other rel ati ves as e xcessive . Th e general comp l ai nt was that the re lative either was taki ng advantage of the person , Ylas ig norin g the closeness of the relationship , or had been untrust­ worthy . For some re l atives , this breakdown in conse nsus on re c- iproci ti es re­ su l ted in a complete rupture of ties ; for others this bre akdown was fol lowed . by a long period of "not talking to each o ther . . Despite t he confl ict , there is an unde rlyin g con sensus that people ought to maintain a series of rec i proc i ti e s with relatives; the disagreement arose over the extent to which ki n sh ip ties were too 'greedy . ' ' •

The c on ce ptio n of k i n ship ties as " greedy ' ' someho\\1· echoes the concerns of the Roman civilization at the time of the fonnulat ion of the Twelve T ab l es . The rise of the Plebi ans when the Roman Repu blic was fou nded gave voice to the contention that the Patrician class too ofte n directed its effort s to the prom oti on of the special i n tere st of its own famil ie s . H e nce forth , the de votion to fam i l y was to be m atc hed by devotion to ci v il society . Impl icit

1 60

in this change was the need to stri ke a series of barg ai ns to accommo date

confl icts in demands by family as opposed to civil society . At th is t im e ' proh ibition of intermarri ages between Plebian and Patrician famil ie s \\'as lifted , and the Civil Law model somewhat weakened the emphasis upon

ancestral line of descent impl ied by the Parentela Orders model , but it ret ained (through succession per stirpes ) a focus on family continuity .

Kinship as Market If kinship ties are regarded as exchanges in market place , interaction between relatives is based on a • •market price , ' ' an equi librium of supply and demand . Relatives are assumed to be free- acting agents-no longer strongly con­ strained by corporate injunctions to perform certain acts . Without such in­ junctions , � 'the establishment of exchange relations involves making invest­ ments that constitute commitments to the other party . Since social exchange

requires trusting others to reciprocate , the initi al problem is to prove onese lf trustworthy (Blau , 1 964 , p. 98) . ' ' The degree of trustworthiness to recipro­ cate through services of equal value then defmes one 's standing in an

ex­

change relationship . Whereas two people may trust each other to reciprocate in transactions that involve immediate exchange or services of little value , they may hesitate when the reciprocation may never occur or when the value of the initial service is very high . In such instances , kinship status counts for

little in the maintenance of relations with particular persons . Specific relatives must , like anyone else , then ' 'earn ' ' their standing by gaining personal trust .

In an analysis of reliance upon kin in 1 9th century England , Michael An­

of kinship as a source of"goods ' ' like those in any other Anderson, kinship is a source of goods in competition with

derson conceived institution . For

other sources for similar goods: material help , personal or intimate ties , cer­

tainty of delivery of services , emergency assistance , and so on . There is then nothing unique about own merits deserves

kinship that establishes it as a social category that on its to endure . For Anderson, "The highest probability that

a kinship relationship of a functional kind will come into being between kins­

man

A

and kinsman 8 will be when A has resources inadequate to solve one

or more problems alone and when B also has or expects to have problems which A is or will be able to help him solve because he

(A) has

a surplu s of

the required resources , and when neither of them have open to them alterna­

tive suppliers

of resources

which they are in need who will demand a lower

expenditure of resources by them in reciprocation (Anderson , 1 97 1 , p. 17 1 ) . ' · The conception of kinship ties as a commodity that competes on the open

market with alternative rel ationships-friendships , social agencies , neigh­ bors-for meeti ng personal satisfactions erases the idea of kinship as pe r­ petual order. Instead , the endurance of any social structures rests upon its cont i nu i ng to supply a bal ance of personal gratificati ons . Soc i al di stance is

16 1

a fu nction of personal intimacy rather than similaritie s between classe s . The can on La"' model of genealogical distances seems particularl y appropriate as an e xpression of personal intimacy in that it yields gradiations of relatives by the number of collateral or lineal l inks between them and EGO 's nuclear fam ily .

Kin ship and Fraudulent Manipulation Drav,rin g upon the metaphor of the theater, Erving Goffman ( 1 956) has ana­ lyzed the '''ays by which the people try to manage the impressions they make on others . He \\'rite s that , • •When an i ndividual appears before others he wiiJ ha..,·e many motives for try i ng to control the impression they receive of the situation . . . . The issues dealt with by stage-craft and stage-management are sometimes triv ial but they are quite general ; they seem to occur everywhere in social l ife , providing a c lear-cut dimension for formal sociological analysis (Goffman , 1 956 , p . 8) . " As in the various ideologies regarding acting , impre ssion management can be si ncere (perhaps t he performance i s intended to shield or benefit the other person) or it can be cynical or fraudulent­ primarily to manipulate the other person to respond in ways which wi ll benefit the impression manager. From Goffman 's perspective , ho\\'ever, both actors are involved in one way or another in impression management . Indeed , Goff­ man 's major thesis is that people are j udged by others according to their conformity to some set of moral standards . ' 'But , qua performers , indiv iduals are concerned not with the moral issue of real izing these standards , but with the amoral issue of engineering a conv incing impression that these standards are being realized . . . . As performers we are merchants of moral ity (Goff­ man, 1 956, p . 1 62) . ' ' In kinship as drama , rel ati ves are the audience , ' ·strangers ' ' who are witnesses to a perfonnance . If in the establishment of reciprocities , we are dealing with exchanges of moral acts , \\'e intend to receive genuine moral acts in exchange for our pretense s . We seek love and good will for our payment of empty niceties . Carri ed to its extreme , impression management eventuates in a norm of fraud-a trade of morality for immorality , a trade of the genuine for the fake . In kinship , impression management often results in the obl igation to present a friendly front, as long as the price is not exhorbitant, and to apply much tact in inte.raction . The metaphor of fraudulent manipulations symbolizes estrangement from ki n .

Range of Models

and Rec=iprocity

In su mmary , each model of kinship-mapping seems to imply a spec ific mode of reciprocity . The Parentel a Orders model appears to express the norm of

162

generosity . It emerged in societies that depended for their persistence upon the endurance of ' 'houses " ' or belief systems passed dOY.'Il through the ge n­ erations from parents to chi ldre n . The Standard American model , Y.'ith atte n­ tion more to the past than the future , seems to provide a bridge bet\veen the Parentela Orders and Civil La�· models . The Civ il LaYl model appear s to be related to institutional greedine ss. It emerged in a society in which c i v il insti tutions were in confl ict with an el ite of powerfu l famil ies . The Ge ne tic model . by virtue of its modification of the Civil LaYl computational form ula , [(i - l ) +j] instead of [i +j] , stands between the Civil and Canon Law models . The Canon Law model seems to give form to the marketplace conception of kinship reciproc ity . The Canon Law model developed at a time in late me­ dieveal Europe when the Church and the state were both centralized insti­ tutions , forcing the fami ly into a subordinate position in competition with them for the sati sfaction of needs . Each mode of reciprocity thus seems to reflect a different set of norms and value s appropriate to a specific kinship model . The remainder of this chapter deals with the topics of residenti al location , interest in kinship ties, and the discrepancy between actual and desired con­ tact with kin as all of these are related to kinship orientations and metaphors of soc iw exc hange . LOCATION OF RESIDK.�CE

In the study of kinship organization , anthropologi sts give considerable atten­ tion to the location of the residence of married couples . Following l\1urdock (1949) , the position usually taken is that the rel ati ve contribution of me n and �·omen in the division of labor determines the location of the marital resi­ dence . In those societies where the man 's contribution is greater , the couple �·ill live near his people (i . e . , virilocality); where the woman 's contributi on is greater, the rule is for couples to live near the wife 's people ( i . e . , uxorilo­ cality); and where male and female contributions are equal , residence i s either neolocal or utrolocal (i . e , near either spouses 's people) . In the preindustri al cultures u sed in these analyses , however, economic production is not isolated from residence and its environs , as it generally is in contemporary society . In modem industrial society , marital residence has (if anything ) tended to\1/·ard uxorilocal ity (Adams , 1 970) . Closeness of residence to the Yt'ife ' s relatives permits not only the persistence of inti mate tie s but also the devel­ opment of coal itions and mutual Y.'elfare networks within maternal kindreds . In fact , a primary determinant of interaction with relatives is proximity of re sidence (Klatzky , n . d . ; Adams . 1 968 ) . There are t\I,'O separate components that describe spatial aspects of marital residence . The first component concerns the extent to which persons live near relatives at all ; it pertains to the degree to which movement aY-lay from one 's

1 63 o sa ng u i ne family occurs . The amount of such movement de pends u po n a v ari e.t y of fac-tors-economic oppo rtu ni tie s , cl imate , cu ltural amen ities , ac ­ cessib ility to ed uc at ional or health facilities , and so on . The seco nd c om ­ ponent deals with the disti nction between husband 's and wife 's re l atives­ u xoril ocal ity , viril ocal it y , or utrolocal ity . Although circumstances pertain ing to degree of movement may dictate neolocality ( i . e . , isolation from both sides) , there may still be a preference on the part of husband and \\'ife to l i ve

c n

near one set of re latives .

General Residential Proximity

Table 8-1 pre sen ts data on the percentages of cl ose rel ati ves residing in ..�zona . In this table , close relative s refer to members of the respo ndent s � (and their spouse s ' ) families of orientation . Arizona (rather than Maricopa County) \\'as chosen a..� describi ng nearby residence because the major pop­ ulations c enters in the state are within a two -h ou r drive from Phoen ix; the population in the remaining areas is sparse . In Table 8- 1 , the percentage for women c learly show a wide range in resi­ dential patte rn s by kinship orientation . Women in the Parentela Orders cat­ egory have the fewest parents and s i bl ing s in Arizona , and those with a Genetic or Canon Law perspective have the most . For example , whereas only 36 percent of the mothers of Parentela Orders women reside in Arizona, 75 percent of the mothers of those with a Genetic or Canon Law orientation do so. As for in-laws , whi le the women with a Parentel a Orders conception again have the smallest percentag e s of their husbands ' families in Arizona, women in the Standard American category have the largest. For i n stance , for moth­ ers-in-law , the percentage is 37 for Parentela Orders women and 54 for those women in the Standard American category . Thus , the data indicate that Par­ entela Orders women are most likely to live apart from their own and their husbands � families of orientation , while Genetic and Canon Law women are most apt to live near their own families and Standard American women near their husbands ' families .

For m e n , the data in Table 8- 1 reveal a somewhat different picture . Dis­ tinction s amon g kin ship orientations are somewhat blurred . Except for fathers and si sters second nearest in age to the respondents, relative s of Parente la Orders men (unlike women) are similar to men with other orientations in percentages living in Arizona . The only noteYlorthy finding i s the tende nc y for men wi th a Genetic or Canon Law orientation to have a comparatively hig h percentage of close relative s i n Arizona . Table 8- 1 indicates that while both men and women in the Genetic and C anon Law categories tend to have a high percentage of close consanguineal re l at i v es in Arizo na , i t i s just the v.·omen in the Parentela Orders classific at ion whose famil i es are partic ularly prone to be out-of-staters .

1 64 TABLE 8-1.

Percentages of Close Relatll'� Residing in Arizona, by Sex of Respondent Parentela orders

Relatives

Genetic or

Standard

Civil

american

Ia._·

canon law 59 . 6

Ego 's family of orientation (Women) Father

3 1 .7

5 1 .5

.50. 0

Mother Brother closest i n age to ego S ister c lo se st in age to ego Brother second closest in age to

3 5.8

54.4

55. 0

75.0

26 .0

3 2 .0

48 .0

54.5

35.3

45 . 5

42.0

55. 1

30. 3

35 . 0

34 . 3

48.8

26. 7

44 . 6

5 0 .0

6 1 .4

Father-in-law Mother-in- law

38.2

58 . 6

50. 5

56 . 0

37 . 0

54. 1

24.4

23.4

44 . 8

48 . 5 38 .9

Wife 's sister

58 5 49 . 1

45 . 6

25 . 8

Father

29. 7

55.3

52 . 8

Mother Brother closest in age to ego Sister c lose st in age to ego

so. s 5 1 .3

54 . 5

52 . 5

38. 3

32.7

35.7

48 . 5

42 . 6

38.5

45. 5

42 . 5

37 . 5

34 . 4

3 1 .2

47 .8

21 .7

42.9

42. 1

4 1 .7

ego

Sister second closest in age to ego

ln-l:aws

Wife 's

brothel" be st known to ego best known to ego

Ego ·s family of orientation c:Men)

Brother second closest i n age to ego

Sister ego

second closest i n age to

47 .2

46 . 0

.

In-Laws

61 .2

47 1

Father-in-law

59. 6

Mother-in-law Husband 's brother be st known to

56.9

54.9

52 . 5

51 .5

5 1 .2

45. 4

46 . 9

54 . 3

54.3

48 . 4

53. 1

5 1 .6

ego

Husband's sister best known to ego

.

48 . 6

Virilocal and Uxorllocal Tendencies

In modem America with its high degree of residential mobility , people have many reasons for moving to new localities-or remaining in the same on e . But apart from extent of mobility , there is the question whether be ing near "family ' ' differentially influences mov ing and staying decisions of men as compared with women . Table 8-2 presents the ratios of relatives residing in Arizona for men and women holding each particular kinship orientation . The percentages used are taken from Table 8- 1 . For example , for mothers of persons with a Parentela Orders orientation , the ratio is 47 . 2 percent over 35 . 8 percent or 1 . 32 . A

1 65

review of the data i n Table 8- 2 reveals that for the respondents ' own famil ies

of orientation , the findings are: Parentela Orders is marked by viriloc a l tend­ enc ies (mean ratio = 1 . 1 7 ); Genetic and Canon Law models by uxorilocal tendencies (mean ratio = 0.79) ; while the St andard American and Civil Law groupings are utrolocal (mean ratio = 0 . 95 for each) .

As for res idence of in-laws , an appare n t anomaly occurs. To be sure , the findings are consistent in indicating uxorilocality for persons with a Genet ic or Canon Law perspective (mean ratio = 1 . 33) and utrolocality for Standard American and Civil Law respondents (mean ratio = 1 .0) . Yet , paradoxical l y , they also sh ow an uxorilocal tendency among Parentela Orders responden ts (ratio = 1 . 88); that is, with disproportionately more men living near in-laws. 'The anomaly , however , is more apparent than real . Since the men and women in the Parentela Orders category are not marri ed to each other , this fi n din g could readily occur if the men with a Parentel a Orders orientation are utrolo­ cal ( i . e . , living near either set of relatives) in comparison

with

the neoloc al

proclivities of Parentela Orders women , which would isolate the women from both parents and in-laws .

In general , the overview of virilocal as opposed to uxorilocal tendencies

suggests that Genetic and Canon Law perspecti ve s are uxorilocal; Standard

TABLE 8-2. For Each Kinship Orientation, tbe Ratio � Percentage of Close Relatives Living In Arizona for Men to that for Women (Percent of Men 's Relatives in Arizona I Percent of WoiMII 's Relatives bt Arizona) Relatives Ego ' s family of orientation

Father

Parentela orders .94

Standard

Civil

american

law

.98

. 94

1. 11

Genetic or

canon law

. 70

Mother

1 . 32

. 99

. 70

Brother closest in age to ego

1 . 47

1 .02

.74

. 89

1 .21

. 85

1 . 08

. 77

1 . 24

. 98

.91

.98

.81

.96

.84

. 68

1 . 17

.96

. 9S

. 79

Father-in-law

1 . 56

1 .04

. 93

.7 1

Mother-in-law

1 . 54

1 .0 1

1 . 19

Spouse 's brother best known lo ego Spouse 's sister best known to ego

. 90

2. 10

1 .0 1

2 . 32

1 . 05

1 . 16

2 .00

Mean rati o

1 . 88

1 . 03

.99

1 . 33

Sister closest in age to ego

Brother second closest in age to ego

Sister second closest in age to ego

Mean ratio Ego's in-laws

.96

1 . 40

166

American and Civil Law conceptions are utrolocal ; but that , althou gh Par­ entela Orders men tend toward utrolocal ity , Parentel a Orders wo men are highly neolocal . These different pattern s in tum define those groups of re l­ atives with whom reciprocities are estab lished . Insofar as women tend to prov ide the major lin ks in interacti on \\'ith kin the data on re sidence suggest considerable d ifference s in day- to-day co nt act by kinship orie.ntation . Fami l ies with a Genetic or Canon Law orientatio n c an be expected to have the highe st levels of interacti on with relat i v e s , \\'h ile those with a Parentela Orders perspective will I ikely have the least . Ho v.lever , as the fol lowing section indicates , variations in frequency of visiting relati v es should not be confused wi th diverse degre es of interest in kin . t

INTEREST IN KINSHIP TIES

Fami lism has general ly been treated in sociologic al research as opposite of degree of isolation of the nuclear family from l arger kinshi p attachme nts . This treatment has placed much emphasi s on residenti al propinquity as a factor in the maintenance of strong kinship ties and in the persistence of a strong personal infl uence of relatives. But if one makes a distinction bet\veen the fam ily as a collec ti vity and the family as an aggregate of rel ated persons , then famil ism changes in mean ing . Insofar as the kindred is endowed with corporate qualities , i t has a .. �re a) ity ' ' beyond its individual members . Corporate entities pro vide gu idelines for con­ duc t in principle . Th i s attribute pennits corporate units to perfonn functions of a symbolic nature over and above the mundane services of any individuals . These symbolic functions often pertain to the promotion of religiou s , ethnic , moral , or material values . It is this symbol ic function that is absent from current conceptualizations of fami lism by sociologists . The i ntroduction of symbolic functions , however, i mp lie s that there are not only differences among people in degree of familism but in types of familism as Ylel l . First , a particular society may foster many mundane per­ sonal services while inhibiting symbolic functions (or vice versa) . But , the performance of sy mb olic functions may not require the same degree of actual interaction as personal services . Instead , ritual occasions � ceremoni es , and norms indicati ng the desirability of contac t may serve to su s tain the se nse of corporate ness required for the persistence of the symbolism of the ,; 'family .. ' ' the "tribe , ' � the "clan , ' ' or the numerou s other names applied to the kindred . Th is section deals w·ith four aspects of intere st in kin s hip tie s . The fir st aspect pertains to general norms regarding ki nship obl igations and a sense of trust assoc iated w ith Fortes ' concept of axi om of amity . The second aspect involves interest in information about one 's Hroots . � � The third aspect of interest in kinship concerns the e x te nt to which contact is m aintained o ver lo ng di stances when rel atives are seen infrequently . The fourth aspect is a

1 67

negati ve one: lac k of interest in k ins hip and unwillingness to respond to questi ons on ki nshi p priorities .

Trust

and Obligation

De sp ite the m any connotations of th e

terms friend and relative , both kinds

of re lationships generally in vol ve bonds of trust and a comm itment to un­ dertake ob l igations on beha lf of another person . In m any instances, friends are re garded as sufficiently clo se to warrant the use of kin terms such as Au nti e or Uncle (Laumann , 1 966) . At othe r times, e speciall y among up­ wardly mob ile i nd i \·id uals , kin and friends are carefully se gregated (Litwak , I 960a; Babchuk , 1 965) . But in any case, to be a true friend de note s the epitomy of loyalty and conform ity to the axiom of amity (Fortes , 1 969) . In the Phoenix study i nterv ie\\' , tY.'O items pertai ned to the axiom of amit y . For the first item , respondents \\ ere asked the e xtent of their agreement v-·ith the statement , "A pers o n ' s relative s usually tum out to be his or her truest friends . " Table 8- 3 de scribes the pattern for " strongly agree · ' responses . This pattern bears a relationship to kins h i p ori e ntation . Whereas 1 7 pe rcen t of the Pare ntela Orders respondents reported that the y strong l y agree with the ··stateme n t . only 6 percent of those i n the Ge neti c and Canon Law category did so ; persons with Standard American and Civil Law perspectives \\'ere intermediate , w ith slightly over te n percent of them expre ss i ng strong agree­ ment . The table al so i ndicates that the re spon de n ts generally regard their ' spou se s views as similar to their own . ·

Orientation and Percent Agreeing Strongly with Statements about Relatives

TABLE 8-3. Kinship

orders

american

Standard

Civil law

Genetic or canon law

1 6.7

1 0. 5

12. 1

5 .7

1 26

334

1 16

1 41

IS .2

10. 1

10.0

5 .0

1 12

268

1 00

1 01

Respondent 's view (%) N

1 6.8

1 4. 1

1 1.2

4. 3

1 25

1 16

141

View attributed lo spouse (%)

333

18.3

1 6. 1

1 5 .0

1 0.8

1 09

273

1 00

1 02

Parentela Statements

A person 's relative s usually tum out lo be his or her truest friendsa

Respondent 's view (%) N View attributed N

to

spouse (%)

A person has a greater duty to help a relative than to help a nonrelalive•

N a

Response caeegories

are strongly agree .

agree , not

sure,

disaaree . and

Slrongly disagree .

1 68

The second item bearing upon the axiom of amity is e x tent of agreement wi th the statement , ' 'A person has a greater duty to help a relative than to

he l p a nonrelative . .. , This item deals more directly tha n the prev ious one wi th

obligatio ns to prom ote the general welfare of kin . Table 8- 3 shows that pattern of re s po n se s for the as sistance item is s imi lar to that for the friend sh ip question . Again , Parentela Or ders respondents di spl ay the gre atest amo un t of strong agreeme nt ( 1 7 percent) , and persons with Genetic or Canon Law

orientations the least agreement (4 percent) , while those i ndiv id ual s with

Standard American or Civil Law positi ons are in the middle . And again , the

pattern for views attributed to s pouse s follows the persons ' own positions

(but with a sl ight tendency here for regard in g the spouse as more agreeab le

than one 's self in gi v ing priority to re latives over non re l ati ve s) 4

l"he fi nd ings for the q uestions on true frie ndship and on the duty to help

relatives both reveal a variation among kinship orientations in fervency with which the axiom of a mity is held . Person s with a Parentela Orders view have the most c om mi t ment to the axiom of amity , and those with a Genetic or Cano n Law perspective have the least commitment . These find ing s are con­ sistent with the po s iti on that the Parentela Orders model impl ie s the greatest degree of corporateness of kinshi p ties , while Genetic and Canon La w models express ind i v id ua l istic ideas . Without an as sumpti on of corporateness , it would be difficult to ho ld general norms havi ng to do with the assi gn m ent of trust and obi ig ati on to an entire social category -re lativ es as opposed to nonrelative s .

Infonnation about

"Roots' '

In contemporary society , kinship biography c an become an i mportant mech­ ani s m for sust a ining one 's personal identity . Insofar as the l i ne of descent ( in Parentel a Orders and Standard American mode l s) prov ides a b a sis for identify i ng family boundaries , kinship takes on a corporate character that ex tend s far beyo nd the in di v idual . One 's corporate existence both stre tc he s back to ancestors in hi s tory and projec ts forward to future g e nerat ions . The symbo l ic e state that one inherits from the past c onsists , in large part , of biograph ical infonnation of antecedents and collaterals who are soc i al l y ( a nd pre sumably genetic a lly ) incorporated into o ne ' s idea of self. This collecti on of biogra ph i es y ield s for the i n div idu a l a ftxed point in a set of coordin ates for charti ng one ' s • • wor ld " and a se nse of belongi ng to a part icu l ar domai n . But just a s pro pert y estates vary i n value , so d o s ymbolic estates . I n indi­ vidual i stic kinship model s (e . g . , Genetic or Canon Law) , person a l ties V�'i lh relatives ov ers had ow symbolic connections . When individual s have no stake in seeki n g identific ati on th rough knowledge of biogrdphies , there is little reason to collect or to communic ate b iogra phi c al k now l edge . Such commu-

1 69

nicatio n may rattle skeletons be st forgotten, uncover symbol s of fail ure and degrad ation, or disinter relics from an irrelevant past . Sti l l in corporate kin­ ship models , it is through these symbolic family e states that kinship groups persi st with their unique identities and facilitate social differenti ation-reli­ giously , socioeconomical l y , and ethnical ly. (See Farber, 1 97 1 . ) Intere st in information about one 's • �>roots ' ' is perhaps best indicated by knowledge about grandparents . Most of the indi viduals in the Phoen ix study no longer h ave grandparents; many of them never knew their grandparents we ll even when children . For a majority of respondents, i nformation about grandparents would have to have been passed along through the parents . This is p artic ularly true for paternal grandparents , who are generally older than maternal grandparents . One piece of information \vh ich many responde nts �·ou ld not have been able to obtain directly is the birthplace of their grand­ parents ( and certainly their spouse 's grandparents) . Presumably , the greater the interest of the family in biographical information of ancestors , the greater is the probability that birthplace information about grandparents would be passed along . Table 8- 4 pre sents data on the percentages of respondents wh o know the general locations of the birthplaces of their own and their spouses ' grand­ parents . In each instance-for own or spouse s ' or for maternal versus paternal grandparents-the results are similar: The percentage is h i ghest for Parentela Orders respondents and lowest for those with a Geneti c or Canon Law ori­ entation . While differences among orientations are not large , they are con­ sistent . The findings thus suggest that greatest interest in one 's symbol ic family estate is associated with the Parentela Orders model , and least interest is associated with the Genetic and Canon Law models .

TABLE 8-4 .

Percentages of Respondents Who Know General Birthplaces of Own and Spouse's Grandparents

CJassifacation of grandparents Own grandparenls

Both maternal grandparents Both paternal grandparents

Spouse 's grandparents

Both maternal grandparents

N

B oth paternal grandparents

Locations of or

Parentela

Standard

Civil

orders

american

law

canon law

92 . 1

90.6

90.4

87 . 9

94.4

85 . 6

84 . 3

83 .7

Genetic

82. 5

75.2

8 1 .7

68.8

82.5

74. 8

79. 1

64 . 5

126

3 33

1 15

1 41

For own Jrandparenls. the respondents were asked to locale binhplace ia tbc roOowiag sections of the United

States (or 10 specify country of .birth): Northeast , Soutb . West. Arizona. Alaska ('I' Hawaii , or don 't know where born . For spouse 's grandpare nls. tbe respondents were asked, ·•no you know in what couatry any of your (husbaDd 'slwife's) grandparents were born? " If the answer was affomative. the next questions were . ••\Vbich 1f8Ddpar'enl5? Whal couatry?' ·.

1 70

Contact over Long Distances Adages are generall y parti al truth s . On the one hand , "Absence m ake s the heart grow fonder, ' ' and on the other hand , ' "Out of sight out of mind . ' ' Pre sumably . the more intense i s the qual ity of corporateness in conc ept io ns of k i nship , the greater are the chances that contact with family wil l be mai n­ tained de spite long d i stances and infrequent visiti ng . The corporate ide ntity itself wou ld provide a basis for continued contact by mail or telephone . The purest measure of kinship-i nteraction norms (by e l iminating barriers such as money , distance , and competing relationships) is the extent of mai l and telephone contact betwee n rel ati ves living in different states . In this study , a basic propos ition is that kin ship groupi ngs organized on the basi s of

special (or factional) concerns favor higher leve l s of ki nship i nteraction as

a general princ iple than do those groupings organized on the ba sis of general

(or communal ) concerns . Factional grouping encourages the maintenance of

corporate- like structures . Three indices of spec ial-concern organization are

applied-kinsh ip orientation , religion , and socioeconom ic position . The the ­ oretical discus sions and findi ngs in the prev iou s chapters of th i s monograph point to Parentela Orders , J udai sm , and profe ssional-managerial-and-admin­ istrative positions as appropriate indic ators of special-concern (or plural istic) social organization , while Genetic and Canon Law models , Catholicism , and blue-collar occupations imply universalistic (or communal) conception s . Table 8- 5 desc ribe s the percentages of re spondents who have frequent co ntact by telephone or mail with parents and in- l aws Vt'hen these relatives are seen less than once a month . Generall y , these rel ati ves l ive out of state . The data on kinship orientation , wh ile falling into the antic ipated pattern � offer only \\'eak confirmation of the proposition; the percentage difference s between Parentela Orders and Genetic-and-Canon-Law respondents are quite small for the parents and , for in-la�·s , only the Genetic and Canon La\\' category differs appreciably from the other orientations (albei t in the expected direction) . For rel igion and soc ioeconomic position , however, the results are quite striking . Je\\'ish respondents have the most frequent contact with parents and in-laYlS and Cathol ic s the least, and the differences for in-laYls exceed those for parents . For example , whereas there is a spread of only 1 6 per­

centage points

(94

versu s 78) between Jews and Catholics for mothers , the

gap inc reases t� 30 points (65 versu s 3 5) for mothers- in-law . .For soc ioeco­ nomic pos ition , the find ings also defin ite ly co nfirm the. propos ition . Respon­ dents in households v.'here the hu sband is in a professional , m anagerial , or adm i ni s trative occupation are much more l ikely to have frequent contact w ith

parents and in - l aws than are those in blue-coll ar families; for fathers there is

a 27 percent difference between them (72 versus 45 ) . Thus , particularl y for religiou s and soc ioeconomic indicators the data support the proposition that

kinship group ings organized on the basis of a spec ial concern tend to promote

171

kinship interactio n as a general norm more than do groupings organized on the basis of communal , universal istic concerns . The we akness of the fi ndings on kinship orientation suggests that a second test of the proposition be applied . Salience of kinship orientation for main­ taining long-distance ties would be even more c le arly evidenced Y.'hen face­

to-face interaction rarel y takes place-le ss than once a year. Table 8-6 sho\\'S

data regarding the extent to which Ego reports frequent contact by mail or telephone with parents and si blings despite the fact that face-to-face inter­ action occurs less than annually . For all of Ego ' s family seen less than once a year, persons in the Parentela Orders category have most contact by mai l

or telephone (and Standard American respondents somewhat les s) , while generally those classified as Genetic or Canon Lavl are least l ikely to maintain long-distance contact . For persons with a Genetic or Canon Law perspective ,

the rule often seems to be ,

...

Out of sight , out of mind , " even '-''hen the

TABLE 8-S. Percentages of Respondents Having Frequent Contact by Telephone or Mail with Parents and In-Laws in Cases Where these Persons Are Seen Less than Once a Month, by Kinship Orientation and Social Characteristics Kinship orientation and

Father

Mother

Father-in-Law

Mother- in-Law

Parentela orders (� )

67 . 3

82 . 3

55

62

46.5

55.6

Standard american (%)

66. 1

81.3

41 .5

1 39

1 23

50 . 3

52 . 5

social characteri stics

Respondent 's kinship orientation

N

N

1 24

43

54

1 45

Civil law (%)

64 .9

82.2

37

45

Genetic or canon law (%)

62.0

74.4

28 . 0

30. 2

50

43

so

53

80 . 8

93 .5

60.0

65 . 4

26

31

20

N N

40

52 . 3

44

Respondent 's reli gi on

Jewish (%) N

Protestant (% ) N Catholic (�·) N

67 . 3 1 53

5 8 .6 70

26

81 .3

46 . 1

53 . 6

171

1 52

1 92

32. 3

34 .7

77.9 68

62

72

Occupation of male co-head in respondent 's household Professional , managerial . or

f\

administrative (%)

72 . 3

82 .4

1 19

1 36

Cleric al , sale s , or craft (� )

65 . 1

Blue collar (%) [\;

44 .7 47

N

86

80 . 6 1 08

16.9 39

55.6

62 .0

99

1 29

46 . 3

57. 3

82

40.4

57

96

38 . 9 54

172

relative is a parent or a sibl ing , and reciprocity seems to represent for them

a per so nal rather than a corporate phenome n on in kinship .

The Refuseniks

One percent of the responde nts failed to prov ide valid answers to the que s­ tions about prioritie s among relati ves in cases of intestacy . Although these persons were not unu sual in some ways , a majority have come from broken families , none has gone to college, generally they hold blue-collar jobs, and they never attend church service s . As far as kinship is concerned , they ha ve smal ler k ind red s than do other respondents in the sample . Of the 24 relatives about whom information is sought in t he interview , the refu seniks have an average of 1 0 . 6 re l ati v es in their kindred as compared with 1 3 .4 for the total s ampl e . Most significant , ho weve r , is the fact that, with one exception , the refuseniks do not care to see even those relatives the y have . Whereas in the total sample , respondents want to see a majority of thei r relatives (58 percent) at least ' 'once a month , ' ' the seven remaining refuseniks want to see only 20 percent of thei r relatives that often . In answering questions about desired fre­ quency of contact w i th relatives , the refuseniks react strongly:

TABLE 8-6.

Percentage of Respondents Having Frequent Contact by Telephone or Mail with Parents and Siblings in Cues Where these Relatives Are Seen Less than Once a Year american

Civi l law

Genetic or canon law

66.7 9

24. 1

30 . 8

28 . 6

29

13

14

Mother (%) N

75.0

60.9

45 .5

44.4

16

23

11

Brother c lose st in age to ego (%)

28 . 1

27 . 6

4.8

lS .O

58

21

Sister closest in age to ego (%)

43 . 3

3 1 .0

23.5

20 27 . 3

30

42

17

22

27 . 8

30.4

7.7

14 . 3

18

23

13

28

44.4

26. 9

0 .0

0.0 13

Parentela

Standard

orders

Father (%) N

Relati ves

N

32

N

Brother second closest in age to ego

(% ) N Sister second closest in age to ego

( %)

18

N

• For relllli ve& seen less than once per month . die respondents

by mail

or

phoae ,

occasional contact

of ""frequent' ' is tbus

subjective.

by

mail

or

26

were

9

10 asked, · · Do )'OU

phone. or no en tact II all

by mail

or

have frequeot

pbone?

to

coatael

The defmition

1 73 " 'No, I don 't want to see any of them . " (Int. [After responding

..

299)

contact] , "That 's it if you reall y want to know . without them. ' ' (Int . 220) "[My brother) (Int . In

1 07)

sorting

lives

question on frequency of I can get along with them or

Never ' ' for all rel ati ves in the

up the block, but I only see him about once a year. "

cards to ind icate frequency v.,·ith which she wanted to see relatives,

at frrst the respondent laughed and placed the cards for al l relatives in the "Never " ca tegory . Then , referri ng to her brother,

g uess I could stand hi m once

a

s he

said , UNo, let 's see , I

month . " The interviewer noted , however,

that the respondent betra ye d much antagonism toward her brother whe n she

answered questions about hi m .

(Int . 54 1 )

Apparently , any interaction between refuseniks and their re l ati ves invol ves

expl o itation , pretense , and the norm of avarice .

Summary : Kinship Interest

This section ha s e xpl ored various aspects of interest in kinship ties. Focusing

on sy mb olic elements in kinship , I have found that person s holding a Par­

entela Orders perspective reveal the greatest i nterest in kinship ties and per­

sons with Genetic or Can on Law views show least interest . Those taking

Standard American or C i vi l Law pos itions are interme diate in their concerns .

These relation s hi p s between kinship orientation and interests hold for such

topi cs as: (a) the axiom of amity (i . e . , trust and obl igation) , (b) infonnation

about one 's ance stral roots , and (c) maintenance of contact with kin ov er long distance s (with little face -to - face i nteraction) . The data also indicate

that, when there i s little co ncern with kins hip , respond en ts refuse even to

consider e stab lishing any series of pri ori t ies a mo ng rel ativ e s .

ACTUAL AND DESIRED CONTACT

Research on kinship ties ordinarily rests upon the implied as s umption that if people really want to see their relative s often , they will arrange their lives

to do so : they do not move away from parental homesteads; they org ani ze

their calendars to fit social occ asio ns around family visits; and they maintain

a maximum of pers on a l interdependence with their re lati ves . Thi s assumption

of a social market treats contact with relatives as a ' 'good' ' in compe t iti on with other goods : economic achievement , friends , cultural e vents , and so on .

But the competitive ass um ption does not take into account the d iverse qual­

ities of family i n terest in different groups . One can hav e a pro fo u nd sen s e

1 74

of obl igation to relatives without giving them the highest pri ority in time expen d iture . f\1oveover, one may dislike certai n ki n . Rather , interest in fam­ ily and kin appears to be expressed by an ab s olu te scale . Reg ardl ess of co mpeting interests and goods , to what extent does one want to have co n­ si de r able i nteraction with re- l at i ves ? For t he individ ual who has few competi ng interests or obst acles , this i ntere st i n fam i ly ties can be re ad il y accommo­ dated , and much time can be spent with rel atives . For the individual vY'ith numerou s competing intere sts and obstac les , this interes t in fam ily mu st gen­ eral l y be held in abeyance � other interests may be more pres si n g . The interpretati on of kinship ties as a form of social exchange requi res a mode l of an interactional marketpl ac e comparable to the co nception of ideal cond i tions for perfect c ompetitio n in economic s . The mo del of a free market a s sum e s (among other th i ngs) that all buyers are eq ual ly access ib le to a l l se ll ers (and vice versa) and that none o f the buyers o r sel lers i s sufficientl y l arge and powerfu l enou gh to domi nate the general market c ondition s . Yet in the anal ysis of ki nship ties in con temporary urban popul ations , most ob­ servers base their i nterpretations upon frequen cy of contact, and they ig n o re se l ection factors in and c onseq uences of differences in the proximity of res­ i d enc e and the size of the kin uni verse . ( See Adams , 1 970 . ) Kl atzky (n . d . � p . 2 ) , however, notes that , • 'The studies which have exp lic itl y inc luded distance as a variable . . . show that kin contact i s a monotonically decreasing functio n of distance-that i s , the frequency of contact d ec li nes steadil y as d istanc e i ncrease s . ' ' Moreover� " I n tenns of daily liv ing , kin seem more sal ie nt to working-class people (Adams , 1 970 , p . 5 86) , ' ' who do have l arger fami l ies and there fore a greater chance to find relativ es vlith Yw'hom to inter act . If ou r interest i s to investi gate norms operati ng in social e xchange , t hen the degree of residenti al proximity and the si ze and c o mpo sit i on of a person 's kinshi p un i verse l imit the use of actual contact with relatives as a me ans for ascertai n i ng wh ich nonns are operative . In the intervie\\· , the respondent s �·ere asked not only how often they saYw' their rel atives b u t also to de scribe the freque ncy v.'ith which they "''ou ld l i ke to see eac h relative "if money or distance. didn �t matter. ' ' The disti nction between ac tual and desired interaction with k in penn its ex am i n a t ion of v arious kinds of discrepanc ies . Under ideal cond itions , some peop le wou ld l ike to see their relatives more often than they do , while others would pre fer to see kin less often . Both al ternatives w ith regard to de sired c o ntact v.'ith relatives are explored be loY.' . Wanting �lore Face-to-Face Interaction Ge nerall y , each individual has a re servoir of rel atives in exc.ess of those w ith whom it is nece ssary to interact on a regul ar basi s . How the n doe s an indi­ vidual decide that he v.·ould like to incre a se the amou nt of face - to- face i n-

175

te raction with a relative? The nonn of exchange with which th e indiv idual operates s eems to provide a clue to the nature of th is decision . If t he individual operates with a norm of marketplace exchange in inter­ ac tio n \a..r ith a rel ative , then any increase in interaction is moti v ated only when th e ind iv i dual perceive s that rel ative as c apable of m e eti ng s pec i fic personal needs . Metaphorically , the purc h ase of that re. l ati v e 's inte racti on occurs when there can be an immediate , concrete exchange of services . If n eed s can be met more adeq u atel y el sewhere , then the i ndiv i du al does not have to choose an addi tional relative from the kin re servo ir . Since d istance and other inconven iences i mpede the practicality of immediate exchange , the individual limits i nte racti on to those kin at hand . As a result , the amount of interact io n with re lati ves tend to co rre spond to the amount that the individual des ire s ; in general , de sired contact would not exceed actu al contact . However , if the individuals o perate in their kinship universe w ith a norm of generosity (or axiom of amity) , then the relationship between the amou n t of desired as compared w it h actual c ontac t shou ld be d iffere nt . According to this norm , specifics of g i v en and take need not be identified; rathe r , the emphasi s i s o n optimi z i n g one ·s po s itio n to ' 'give ' ' in any relationsh i p with kin. ( See Joffe , 1 949 . ) Distance and inconvenience are not supposed to dis­ tract one from entertaining the pos s ibil ity of rec iprocal relations . The norm of ge ne ros it y th u s im p l ies that an individual ou ght to want more interaction with relatives than actual circumstances d ictate . Ide al l y , this excess of de sired over .actu al interaction should occur re g ardle ss o f pe rson al qualities or im­ mediate av ai labil ity of the relative . In the Ph oenix study , i n order to ascertain the exte nt to which kin ship or ientati o n s inc orporat e norms of s ocial exchange with kin , I sou gh t to d e ­ termme whether levels of de si re d interaction (e . g . , weekly , month ly , and so on) would exceed the actua l levels reported . The findings for parents and sibl ings and for re l at i ves outside the nuc lear family are described below . Table � 7 di spl ay s information abo ut actual and desired interac ti on w ith one � s fam i I y of orientation . The findi ng s on actual in teraction indicate a general progression in freq uency of contact , y,rith Parentela Orders persons seei ng their p arents a nd sibl i ng s least often , and those w ith a Genetic or Canon Law pers pec ti ve seeing them most often . For e x ampl e , on ly 23 percent of men in the Parentela Orders category see their mothers weekly . whereas 42 pe rc en t in the Geneti c and Canon Law group ing see their mothers that often , and about 3 1 pe rcent in the Standard American and Civil Law c l as­ si fications see the i r mothers at least once a week . While there are minor de viations from this p attern , overall the relationship appears to hold . The percentages on actual co ntact in Table 8- 7 might lead one to concl ude that respo nden ts in the Pare ntela Orders cate gory are the le a st fami l i stic , and those with Genetic or Canon Law orie ntat io ns are most familistic . However, these vari at ions in contact can not be attributed to a differential lack of desire

1 76

TABLE 8-7 . Parents and Siblings Seen Weekly and, for Those Seen Less Frequently, Percentages of Parmts and Siblings Whom Respondents Would Uke to See More Often Genetic or canon law

Parentela

Standard

Civil

orders

american

law

18.4

29 .5

33.3

65 . 8

62 .9

3 1 .6

57.9

92.4

89 .5

94.4

38

l OS

38

36

4 1 .5

Parents and siblings MEN RESPONDENTS

Fathers

Seen weekly (%)

Desire to

see

more often (%)

84.2

Total (%)

N

61.1

Mothers

Seen weekly ( %)

2 2 .6

30. 5

3 1 .0

Total (% )

64.2

59. 3

57 . 1

56. 1

86.8

89. 8

88. 1

97 . 6

53

1 18

42

41

1 6.7

1 4.9

28 . 6

26. 5

70 . 8

77.2

67 .9

58.8

Desire to

see more often (%)

N

Brothers c lose st in age to ego Seen weekly (%)

Desire to

Total (%)

see

more often (%)

N

Sisters

closest

in age

Seen weekly

(%)

to ego

87. 5

92 . 1

96. 5

85 . 3

48

10 1

28

34

14. 9

1 7 .7

26 . 5

20.0

Desire to see more often (%)

66.0

66 . 7

52.9

70.0

Total (%)

80 . 9

84 .4

79 .4

90. 0

34

40

30.4

N

47

Brothers second c lose st

Seen weekly (%)

Desire to see more

Total (%)

in age

to

ego

often (%)

N

Sisters second closest i o age to ego

Seen weekly

(%)

Desire to see more often (%)

Total ( %) N

96

25 . 0

1 5 .6

1 5 .8

66.7

7 1 .9

73 . 7

56.5

9 1 .7

87. 5

89. 5

86. 9

24

64

19

23

4.2

1 8 .4

26 . 3

20.0

87.5

75 . 5

68 .4

72.0

91 .7

93 .9

94 . 7

92.0

19

25

24

49

(rontinMed)

to see fathers , mothers , si sters , and brothers . A reverse pattern is n oted for desire to see parents and siblings more often . Among perso ns who see the ir nuclear-family relatives less than weekly , those in the Parentela Orders cat­

egory express the g reates t desire to see them more o ften, and people in t he

Genetic-and-Canon Law class voice the least desire . In fac t , when the per­

ce ntage of persons who want to see their parents and si b I ings is adde d t o

177 TABLE 8-7 (continued)

Parentela orders

Parents and siblings

Standard

american

Ovil law

Genetic or canon law

WOMEN RESPON DE NTS

Fathers Seen weekly (%)

1 4 .6

34. 1

35 . 6

30. 5

Desire to see more often ( %)

75.6

53 . 3

55.6

Total ( % )

90 . 2

8 7 .4

9 1 .2

41

1 35

45

57.6 88. 1

32 . 1 62 .3

42 . 2 49. 1

45 .9

91 .3

93 4

1 73

61

90 . 3 72

1 3 .6

17 6

22. 7

7 8 .4

6 3 .6 86 . 3

N Mothers See n weekly (%) Desire to see more often (%)

94 .4

Total (%) N

53

Brotben closest in age to ego

Seen weekly (%) Desire to see more often (%)

4.0

82 0

73 . 6

8 6.0

87 . 2 1 25

.

Total (� ) N

so

Sisters closest to age to ego

Seen weekly (%) Desire to see more often ( % )

Total (%) N

Brothers sec ond closest in age to ego Seen v.·eekly (%)

47 . 5 .

.

96 . 0

51

59 50.0 40 .3

66

27.1

1 9.6

1 8 .9

26 . 0

64.7

72 . 7

66. 0

62 .9

84 .3

91 6

92 . 0

90 .0

51

13 2

50

70

.

Desire to see more often ( %)

3 .0

1 1 .7

1 4. 3

9.3

87.9

76.7

90.9

88.4

74 .4

Total (%)

74 . 3 35

43

N Sisters second closest in age Seen weekly ( %)

33 to

Desire to see more often (%)

Total (%)

N

60

88. 6

83.7

ego 1 6.7

28 . 8

33 . 3

3 1 .8

76.6

66. 7

60 . 0

54 5

93 . 3

95 . 5

93 . 3

86.3

30

44

30

66

.

those who actually see them frequently (i . e . , at least weekly) , differences among kinship orientations virtual ly dis appe ar Reg ardl es s of orientation , about 90 percent of the re sponde nt s either see or want to see their parents .and siblings freque ntl y . Hence , variations among k i ns hi p models in actual interaction with parents and siblings seem to be a functi on of diversity of co mpeti ng intere sts and opportunities rather than a result of differing degrees of absolute i ntere s t i n mai ntaini ng nuclear family ties . .

17 8

The complexity of infl uen ce s impinging on con tact with fami ly of orie n ­

tation is suggested further in Table 8- 8 on perce n t ages of local ly re si d i ng p aren ts and s ib l i ng s seen at least weekly by respondents. Overall , mo re respondents tend to see mothers and sisters weekly than they do fathe r s or brothers . In particular, ties to fathers of Parentela-Orders re s po nde n ts s e e m espec i a l ly weak , while the bon ds to mothers in blue-collar familie s ap pear notab ly strong .

TABLE

8-8. Percentages of Close Relatives Living in Maricopa County Whom tbe Respondent Sees at Least Weekly, by Kinship Orientation, Occupation of the Male Co-Head of the Household, and Religion Second Closest in age IO ego

Ki n ship orientation , male

co-head ·s occupation , and

closest in age to ego

Fat her

Mother

Brother

Sister

Brother

Sister

Parentela ordtn (%)

52 . 5

70. 3

38 . 5

50 . 0

54 . 5

37

26

32

43 8

Standard american (%)

70. 8

76. 6

42 . 4

46. 1

59 . 5

N

1 06

39

42

Civil law (%)

73 5

66

89

72 . 2

1 37

4 1 .0

53 . 3

64. 5

50 . 0

36

49

30

63 6

77 . 8

53 . 7

31

59. 1

44

63

41

64 . 0

64 . 3

42 . 9

religion

Kinship orientation

N

N Genetic or canon law (%) N

23

.

.

.

16

16

11

22

53 . 5

34 . 6

56 . 7

43

26

30

46 . 7

39 . 1

36 . 0

23

25 65 . 7

Occupation of male co-head Profe ssional . manageri al . or administrative

(%)

N

Clerical . sales, or craft N

(%)

Blue collar (%)

N

60

75

98

49

62 . 9

68 . 2

37 . 5

49. 2

40.6

70

85

56

59

32

62 . 3

83. 1

48 . 1

5 1 .7

36 . 6

54 . 8

77

54

60

41

42

93 . 3

75 . 0

77 . 8

53

Religi on

Jewi sh (%)

84 . 6 13

15

Protestant ( %)

65 . 0

73 . 6

N

1 17

1 63

Catholic (%)

65 . 2

69

N

N

12

(4l)

9

<1

2

4 43 . 4

40. 7

42 3

36 . 0

1 04

50

53

77 . 4

44. 9

64. 0

46 . 9

69

75

49

67 . 3

93

46.4

55 . 6

58 . 8

28

36

17

39. 1 23

27 . 3 II

No religi ous preference

N

8

35

Sumber too small to compute pc:rcen�e .

86

.

652

42. 9 1

179

In general , though , the data in Tabl e 8- 8 on kinship ori entati on and so­

mi c statu s reveal little about interaction with family of orien tation . c ioeco no T he pe rcentages on re l igion , however, i ndicate an ideolog ical impact . �mong rel ig io us groups , Jewish respondents exhibit the greatest amount of co ntact ..

\\ri th p arents an d brothers and sisters , and those persons with no rel i giou s

nce show the least contact . With roughly 90 percent of the Jew i sh pre fere re spo nd e. nts hav ing weekly i nteraction with local ly residing pare.nts , the find­ in g s appe ar to express the central ity of l ineal ti es in Judai s m . By way of con trast , on l y about hal f of the re spon dents \vithout a rel igious preference ha ve freq uent contact with their parents in f\1 aricopa County . Among these secul ari sts i n particu lar. kinship ties seem to depend upon chance factors (50- 50 probability) for frequent occurrence . These find ings together po int up the complexity of influe nces upon interaction with parent s .

For relati ves outside the respondents · famil ies of orientation , the data (re­

ported in Table 8- 9) regarding desire to incre ase interaction

arc

more elo­

quent. Fi rst ac tual contact with these relatives V.'ill be descri bed and aften-...· ard de sired contact . Because most people in the Phoe nix area have immigrdted from other sections of the Un ited State s s i nce 1 95 0 , only a sm al l m i nority of uncles , aunts , and cousi ns reside local l y , and onl y a t i ny proportion of these i s see n monthly . B ut since persons 'N·ith Genetic and Canon La\\' orientations

are

the

most likely to be Arizona native s , they tend to see uncles and au nts more often than do persons Ylith other orientations . As Table 8- 9 in dicates , d if­ fere nces i n i nteraction among kinship orientations are slight , hoV\I·ever, sel ­ dom more than 10 percent . As in other investigations , wome n see maternal relative s more often than paternal relati ves , but unl ike mo st other studies ,

this tendency i s not fou nd for men (Adams , 1 968 , 1 970; Schneider and

Cottrell , 1 975 , is an exception . ) Since the percentages seeing relatives smal l , no spec i al s ign ificance c an be attached to thi s finding . Thu s ,

are

as a

whol e , little signific ance can be determined from the data on ac tual inter­ action V.'ith aunts , unc les, and cousins . Despite the paucity of interpretable findings on actual i n teractio n \Vith rel ati ves outsi de the nuclear family , the information on desired interaction is productive , at least for women respondents . Table 8-9 shows that , for women , the desire to ·i ncrease interaction with relati ves is more a function of kinship orientation than ( unlike parents and sibl ings) degree of actual

contact. Except for the mother 's brother, the tendency for wanting to incre ase face-to-face interaction is stronge st in the Parentela O rders group and least among \\'omen classified as Genetic or Canon La'N· . For exampl e , whereas

six percent of Parente la Orders V.'omen actually see their be st-kno\\'n paternal

aunt (Father 's S ister) monthly , about 74 percent want to see her more often ; i n contrast, and v;h ile eight percent of Genetic or Canon Law women see.

1 80

TABLE

8-9. Percentages of Persons Desiring More

Contact with Uncles and Aunts than ActuaUy Oc�urs, b�· Kinship Orientation Parentela

Unc le� and aunts

Ci\iil

american

law

77 . 5

71 . 8

66 . 0

57 . 4

1 1 .6

'WOMEN RESPONDENTS

Father 's brother best koown to ego Desire more contact 4: % )

Actual contact at least once monthlv (%) N

Father ·s sister best known to ego

Desire more contact (%)

Actual contact a t lea st monthly (� )

�1otber·s brother best known to ego

Desire more contact (� )

Actual contact at monthly (%)

least once

Mother 's sister be st known to ego Desire more contact (� )

N

4.8

1 2.7

1 24

47

73 . 5

70.4

76.7

52 .4

5.9

5. 1

4.6

34

135

43

7 .9 63

69 . 8

72 . 3

88.9

61 .4 17.6 57

7.0 43

N

Actual contact at monthly (%)

0.0

40

once

N

Genetic or canon law

Standard

orders

54

7 .0

1 1.1

1 30

45

8 1 .4

7 7 .4

71 .1

68 . 1

11 .6 43

14.5

6.6

13.0

least once

1 24

45

Note: Data for cousiDS arc comparable to those for aunts aod uncles . There was a tendenc)· to place a cousin in tbe same category of desired contact as that cousin ·s parent.

69

fcoruirwed)

among respondents

the ir best-known paternal aunt that frequently , only 52 percent want to in ­

crease their interaction v,·ith her . Clearly , the two percent difference in actual

contact between the t\\'o kinship categorie s cannot account for the 22 percent differe nce in desired contact . (The data for other rel ati ve s yield si milar inter­ pretati on s . ) G i ven the assumpti ons made regarding desired frequenc y of in ­

teraction and norms of reciprocity , the data indicate that the Parente la Orders models represent the highest proc livity toward pre.scriptive altruism and the Genetic and Canon Law orientations the least .

For men , the data on desired interaction with ki n outside the nuclear fam i l y

d o not re veal consi stent trend s . Instead , they are ambi guou s and see m to

reflect the inc ursi on of a variety of competing elements into desi re for kin shi p

ties . Kinship nonns may not be as compe l l ing for men as they are for women .

(Perhaps me n use more nonk.inship criteria than do v.·o men i n interact ing vJith

18 1

tABLE 8-9 (continued) orders

american

Standard

Civil law

66. 7

65 .2

74 . 2

7.8

4.2

Parente] a

uncles and aun t s

Genetic or

canon law

MEN RE SPONDENTS

Father 's brother best known to ego Desire more contact (% ) Actual contact at least once monthly ( %)

N

Father's sister best known to ego

Desire more contact {%) Actual contact at l e ast once monthly (%)

39 58 . 5

0.0

N

95

81 .5

2.8

67 . 6

3.2

8.8

31

34

66 . 7

6 1 .0

61 .

9 8 .

41

1 08

33

41

72 . 5

75. 7

65 . 6

52 . 8

2 5

1 2 .6

40

1 11

12 6

1 9. 4

72. 9

73.3

66.7

75 . 0

4. 2 48

16.2 1 05

Mother 's brother best k now n to ego

Desire more contact {%)

Actual contact at least once monthly (% )

�

Mother 's si ster best known

to

Desire more contact ( % ) Actual c-ontact at least once monthly {%)

N

.

.

32

36

ego

0.0

27

3. 1 32

kin . For me n , suc h interaction may indeed involve choices made on the basi s of relati ves-as-persons . ) I n summary , the findings on des ire for increased interaction \\'ith kin reveal that:

l . Despite the greater amount of actual interaction of Genetic and Canon Law persons with parents and sibl ings, d ifference s among kinship ori enta­ tions are dependent upon competing interests and differential opportunities rather than variations in absolute interest in the maintenance of nuc le-ar family ties . 2 .For persons with parents l iving in Maricopa County , Jew s have most c ontact \vith fathers and mothers , while persons with no re.J igious preference have the least contact . Kinship orientation and socioeconomic status shov.' little overall influence on actual interaction with parents . 3 . For �·omen in panicu lar, desire for increased contact w ith kin outside the nuclear family is associ ated with kinship orientation ; those with a Par­ entela Orders perspective show the greatest desire for i ncreased interaction , and wome n in the Genetic and Canon Law grouping show the least .

1 82

Wanting Less

Face-to-Face Interaction

Ne g at iv e aspects of soc i al ex chan ge are ordinari ly discussed in tenns of i mbalanced rec ip roc itie s , i neq u al ity , exploitation , and power differe n ti a ls (Ekeh , 1 974) . I ndeed , studies of modem kinship have general l y ign ored potentially exp loit ive features i n the maintenance of kinship ties . (Le ic hter and Mitche ll , 1 956 , is an exception . ) Forc e. d ex change i s generally considered to be , at a m i n i m um , " unfai r. ' � The idea of exchange ordinarily implies an abi lity to dispose of one 's goo ds at will . Under conditions of forced e xch an g e , t h o ugh , an ind iv idu al partici­ pate s only to decel erate de pletion of hi s stock of resources . Often . b u t not necessari ly , an individual may submit to e xp loitat i on by one party w i t h an eye to reco up ing those losses thro u gh increasing the indebtedness of a th ird party . For i n s t ance , a man may endure contact with his authoritarian father to reap the benefits of internction Y.'ith his mother, or he may submit to visiting his dom i ne eri n g mother- in-lav; to m o l l i fy his spouse . But thi s forc ed exc hange places a premium upon the added value of the reY.'ards expected from the third party (i . e . , in these examples , from the mother or spou se ) . (See Ekeh , 1 974 . ) If the increment of rev;ard for the se sacrifices is not forthcoming , the relationship betv.'een the individual and the third party may s uffe r disproportionately . This de scr i p t ion of force d interaction , however, assumes that all social goods are personal (e .g . , the services of the spouse or mother) . But in some kinship groups , the costs of personal ties can be compensated for by corporate rewards , such as the perpetuation of the religious identity or of socioeconomic po s i t i on . Corporate conditions , then , opernte to inhibit the breakdown of systems of reciprocity . This inhibition could occur in two ways . First, si nce the compe nsation s for corporate interaction have no thing to do directly with the offending individual , the offences (or costs) of such i nteracti on are mini­ mi zed in their consequences; rather, it i s the offending individual 's role in the corporate structure that cou nts and not his obnoxious c haracteris tic s . Second � there is little additional obligation pl aced upon a third party to compen sate for the forced rel ati o n sh ip with an offen din g party . Consequently , the probab i l i t y of disillusionment with the third party for fail i ng to supply an increment of satisfactions would be reduced. The earlier chapters have i n dicated that Ge­ netic and Canon Law individuals h ave a tende ncy toward disillusionment Ylith the spouse; at least a part of this tendency may res ide in the inability of the spouses to compenste (under individualistic c on ditions ) for the dep letion of o ne ' s social and psychological resources in unw anted kin sh i p interaction . In­ stead , a corporate conception of famil y and kin s hi p (as i n Parentela Orders) may shield the spouse from such demands . Authority and power seem to provide conditions for exp l o itatio n in soci al interaction . Generally , in k i nsh ip , such inequality occurs in re l ationship be­ tween generations . The authority of parents over children and the subservi-

1 83

ence of the children wou l d characterize a corporate kinship structure . I n s o­ cie ti es with a disintegrating patrilineal traditio n . such as ours , one would expect thi s residual corporate authority to reside in the father and h is rel ativ e s . In corpo rate kin s hip , the father and p aternal relatives m i gh t be seen as over­

steppi n g the normal bounds of influence-as meddlin g in one ' s affairs . How ­ ever, i n kinship s ystems without organized corporate aspects , the crucial bond

tying families into networks is the husband-wife relationship (Parsons , 1 949;

S c hnei d er, 1 968) . In those systems , intru sion into the affairs of the couple would then be seen often by the co u p le as in-law i nterfere nce . Table 8- 1 0 shows comparisons among rel ati ves whom the re sponden ts

want to see less often than they actually see them . In the section of the tabl e

on gene rational distinctions , parents are compare d with the siblings with whom the respondents have the most intimate rel ationships ( Adams , 1 968) .

Since ties Ylith parents and s i blings are usually close , percentages are s m all ;

but the trends in the data are fairly cle ar. For both Parentela Orders and Stand­ ard Ame ri can orientations , more respondents would like to redu ce their inter­ action with parents than with siblings-regardless of sex of res pond e nt of or parent. Sex d iffere nce s occur, however , in the other orientations . Findings TABLE 8- 10. Comparisons of Relati,·es Whom the Respondents Want to See Less Often than They Actually See Them, by Generation and Maternal Versus Paternal Tie orders

american

Standard

Civil law

M en Women

Men Wo men

Men Women

Parentela

Relatives

Genetic or c anon

Men

By generation

Father (%)

Brother closest in age to ego (%)

6.7

4.0

5 .0

Fa

2.6

4.4

0.0

5 .6

7. 1

3 .9

1 1 .8

9.5

7 .2

6. 1

3.9

6.2

8.8

0 .0

.5 . 0

J'JO

Mo

.r..to

Mo

Mo

Si

Mo

1 7 .0

Father 's brother ( % f

Mother 's brother (%)3 Greater percentage Father 's sister (%)a

Mother •s sister (%)3

Greater percentage(�

• Best

known to

10. 3

15.2 2.3

1.5

14.7

12.9

6.5

2.5

4 .7

9 .0

.5 . 7

6.3

FaBr

FaB r

FaBr

8.3

0 .0

8 .6

9.8

FaSi

1 1 .8

FaSi

1 3 .0

FaSi

FaBr 1 1. 1

7.3

FaSi

Equal 9. 1 3.7

FaSi

Fa

Fa 1 3. 8

J'JO

6.8

Br

Br 2.4

4. 3

Maternal versus paternal tie

Fa

8.5

ego (%) Greater percentage 1 '

Fa 5.9

8. 1

Mother (%)

Sister closest in age to

c:

6.2

4.9

�'omen

Fa 5.7

Greater percentage

b

15.8

law

5.9

20. 4

0.0 FaBr 9. 3

11.1

10.5

MoBr

MoSi

FaSi

FaBr 1 9 .0 10. 1

14.6 12.5

13.3

respondent.

Parenlal Percentage greaJer: 1 3 : sibling percentage greater: 3 _

Father ' s side greater percenaage :

1 3 ; mothec·s

side

1 .4

�atel" perceotage : 2; equal percen tage 1 .

:

FaSi

1 84

for women in the Civil Law , Genetic , or Canon Law categories are similar to those for Standard American and Parentela Orders cl as ses . But men in these other categories are more likely to want to decrease interaction with sibl ings than with parents . These findings are clearly in line with expectations reg ard­ ing the impact of residual authority of parents in families with a large com­ ponent of corporateness in organization . The contrastive desire to d iminish interaction w·ith paternal versus mate rn al kin is indicated in the se.c ond section of Table 8- 10. As for parents , the findings in this section also suggest the residual effect of patrilineality o n Parentela Orders and Standard American kinship . For both of these orienta­ tions , there i s a greater tendency to wish for dimin ished interaction \\'ith paternal uncles and aunts than \vith maternal relative s . Departures from th i s pattern of responses are found again parti cu larl y among men in the Civ il Law , Genetic , and Canon La\\· categories . Hence , for aunts and uncles � as for parents , the d ata support the in terpretation that (particularly where co r­ porateness prov ides a basis for kinshi p organ ization) the des ire to cut down on interaction i s associated with persi stence of efforts at sustai ning authority surviving from a patrilineal tradition . Tab le 8- 1 1 presents data on the de sire to decrease interaction \Vith i n ­ law·s , reveal ing that , regardless of kinship orientation , roughly 20 pe rcent of the respondents would like to diminish the amount of face-to-face contact they nov.· have v.'ith their fathers- in-law . On the other hand, mothers- in-la\.,.. of Genentic or Canon Law men and women are disti nctly unpopular . Thirty­ two percent of the men and 27 percent of the women with these orientations wou ld like to see their mothers-in-lav.· less often ; in contrast , only 1 4 percent of the men and 20 percent of the women in the Parentela Orders category prefer to i nteract y,rith their mothers-in- law less frequently . Clearly , contact with mothers-in- l aw by persons with a Genetic or Canon Law perspective i s

TABLE 8-1 1 . The In-law Men Father-in-law ('i ) N

Mother-i n-law (%) N

Women Fatber-in-law (%)

N Mother-in-law (%)

N

Desire to Decrease Interaction

with

In-Laws

Parentela orders

Standard american

Civi l

2 1 .3

22.5

14.7

1 11

34 19. 3

47

law

1 3 .8

20 . 5

58

1 22

20. 6

1 8 .9

1 4. 3

18. 1

1 7 .0

34

1 9. 6 4S

1 22

1 26

40

42

46

Genetic

or

canon Jaw 20.5

35 31 . 7 41

22. 0 so 26 . 9 67

1 85 a

soc ially costly proposition . Presumably , Genetic and Canon La\\' k in shi p i deology cannot re ad ily accommodate ties Ylith the mother- in-lav.' . The findings on �·anting le ss face-to-face i nteraction with kin reinforce the prem i se that one ought no t equate amou nt of contact wi th relati ves as a sign of v ol un tari s ti c famil i sm . \\'hereas some individuals do not see their re l atives often e no ug h , others find frequ en t contact ( particularly w ith parents � i n - l a Y�· s , or p aternal kin) d i stastefu l . H owe v er, unpleasantness i n ki nship tie s see ms to depen d in p art upon the ki nd of authority and power patterns associated w ith k in ship orientation . Persons \\'ith a Parentel a Ord e rs or Stand ard A mer­ ican orientation are more l ikel y to encounter overi nteraction with parents and paternal kin , and tho se in the Genetic or Canon Law category more often have mother- i n-law prob lem s . Summary : Actual and Desi red Contact The co mpantive i sol ation of the nuc lear fami l y from re latives in modem

indu strial society has led to the conc l u s io n among man y obse rvers that k in s h i p is becoming a rei ic distance and the competi t ion of other interests imped ing vi sitation and assi stan c e . From th i s perspective � fu nct ionally kinship loses i t s value as an i n st itution . This viev.' , however , ignores the symbol ic value of kinship and the normative p
This section has reported various conn ec tion s bet\\/·een kinship orientations and i nterac tion with rel atives : 1. Alth ou gh there are large v ari ati on s among k inship orientations in the frequenc y \ll i th which people see their parents and s ib li ngs , these di ffe rences deri ve p ri m ari l y from differenti al opportun i tie s and competing interests . When de s ire to see parents , brothers , and si sters is taken into account , about 90 percent of the respo ndents either see or v.·ant to see their n uc lear families frequently . 2 . B ecau se of their migrant s tatus , few respondents see relatives outs ide their famil ies of orientation frequently . Yet , although most women would l ike to increase their interaction with kin considerably , those \ll i th a Geneti c or Canon Lav.' orientat ion are fairly relu ctant to do so . Men apparently apply nonkinship criteria to a greater extent than women do in evaluating the amount of contact they would like with relatives . 3 . Ki nship ori entation influences the extent to v.·h ich people \\/'ou l d like to decrease con tact with particu l ar relatives . Presumably because of authorit y or unwanted intrus io n , individuals in the Parentela Orders or Standard American categori e s may want to decrease interac t ion with their parents or paternal k i n , while those with Genetic or Canon La\\1· orientations may regard contac t w ith their mothers-in -law a s forced.

186

SUMMARY: TIES WITH KIN

In this chapter I have expl ored some relationships between kinsh ip orientation and interaction with relatives. Specifical ly , I have shown how the se orien­ tations bear upon such topics as marital residence , interest in kinship matters ' and the disparity between desired and actual contact with kin . The findings here suggest that the kind of kinsh ip orientation held is related to norms of social exchange . S uppose that V.'e interpret the norm of generos ity as a long-tenn exc hange i n which repayment for assistance may be long in comi ng and the norm of market exchange as a shon-term exchange in Y.'hich immediate reciprocation is expected . Then , in particu lar, Parentel a Orders orientati on is associated \Vith stable kind reds , w hich can v,rithstand much physical separation or petty annoyances . According to analyses reported e ar­ lier, persons who are characterized by high educ ational and income le vels , Jevlish upbringing , older age s , marri ed , and with many progeny are over­ represented in the Parentela Orders category . One would then expect a nonn of generosity to operate to a greater e xtent �·here people exhibit the se attri ­ butes . Moreover , if Genetic or Canon Law orientations are indicative of short­ term reciprocities , the stabi lity of the kindred in groups with those orienta­ tions i s highly dependent upon maintaining residential propinquity . The de­ pendence of Vlorking-class kinship ties upon propi nquity is suggested by the fact that, in this sample , poor persons and/or those who are not high school graduates are overrepresented in the Genetic and Canon Law categories . In addition, the young , the unmarried , and those with few children are poorly motivated to operate with a nonn of generosity . Thus , models of kinship ori entation-Parentela Orders , Standard American , Civil Law , Genetic , or Canon Law-while referring to patterns of priorite s among kin seem also to express different norms for social exchange and consequently different bases of kinship organization in diverse population segments . (Cf. Sahlin s , 1 965 . )

In previous chapters , I have reported the findings on the distribution of kinship models in an urban population . Th is chapter v.·ill be consigned to the integra­ tion of the materials in the rest of the book and to an exploration of their implications for various theoretical and practical concerns . The frrst part of the chapter provides a summary of re­ search and its theoretical underpinnings . The second part relates the findi ngs to previous socio logical and an thropological writings about kinship and social structure . The final section i s a brief epilogue on the interpretation of ' 'kin­ ship ' ' and its impl ications for social policy .

Perspectives on Kinship: Concluding Remarks

SUl\fMARY STATEMENT

This section draY;s together various theoretical and empirical materials that ha\'e bee n described in a more extensive manner throughout the book . Its topics range from a general portrayal of kinship ties to an integration of survey re sults . There are two subsections: the basic ki nship grid as a starting point for analysis of models of collateral ity and the re lationship between the kinship models under i nvestigation and soci al structures in which they appear to thrive . THE KINSIDP GRID

The term kinship connotes a degree of relatedness through a biological (or quasi-biological) connection . One mode of representing this relatedness is to construct a grid of loci of kin ties deriving from birth and marriage . Because the genealogic al locations of kin in any society can be defined from the perspecti ve on one member , Ego , the elements in the matrix are identified by their rel ationship to Ego . Table 9- l indicates a basic grid of kinship loci from Ego 's perspective . Thi s grid , which seems to represent a scheme prev­ ale nt in Western civilization , refers only to consanguines; a comparable grid c an be constructed to sho'V'I' the loci of affines . Metaphorical l y , a basic kinship grid can be regarded as a physical map of kin loci generated from connections through birth . Upon this map , one can superimpose other infonnation de-

1 88

sc ri bi ng boundarie s of c l u ste rs or priorities a mo n g ki n-just as po l i ti c al boun­

daries a nd other soc ial characteri stics are s uperim posed upon m aps sh o\\'i ng o n ly th e phy s ic a l featu res of a territory . Hence , a basic kinsh i p gri d can be consid e red as re pre sen ti ng the physical basi s fo r the cre at io n of social ar­ ra ngenlent s based upon birth and marriage . The construction of the kinship grid i n Table 9- l re sts upon the assu mpt io ns that :

l . Rel ated ne ss of any kins hip loc us to Ego can be described in tenns o f (a) chai ns of gene ra t i on a l l inks of parentage (i . e . , l inks from children to p are n t s) and (b) chains of descent l i nka ge s ( i . e . , links from p aren t s t o c h i l ­ dren) . 2 . In T ab le 9- 1 , re l atedne ss to Ego in terms of p are nta ge (j) i s ind ic ate d by the. number of ge nerations up from Eg o to an ancestor (or , for col l ateral kin , from Ego to the close st com mon ancestor) : for example , for Ego 's grandparent , j = 2 . 3 . In T able 9- 1 , relatedne ss to Ego in tenn s o f descen t ( i) i s indicated b y the number of gene.rati o n s dov.'n ward from Ego (or Ego 's ancestor) to a de sce nd ant . For e x amp l e , for Ego 's grandch ild , i = 2 . 4 . The characte-ristics of the social structure dete rm i ne the topologi cal features of the grid: how parentage and descent are differen ti al l y w·c i ghtcd in o rg a ni zi n g ki n sh i p relations; how groups of kinship loci are part i t i oned ;

which k i n sh ip loci are ign ored in terminology and conduct; and 'Nhich ones

are m erged in larger classes , and so on . In modern s oc iety , the complex it ies of soc i a l stratification , occ u p ati o nal differentiation , re l ig io u s and ethn i c di ­ versity , frequent c on t act with forei gners , and variations in do me s t i c l ife sty les all i nteract to s ustai n a broad spectrum o f modes of ki n s hi p organization . These d i v e rse organizational modes eac h rest upon ideo l og i e s regarding the re l at io n s h ip between fam ily an d soc iety . George Peter Murdock ( 1 949 , p . l 03) has ide ntified co l l ate ral i t y as a basic d imensio n in the orde ri ng of re l at i on sh i ps among ki n in v i rtu a ll y al l societies . The term , col l ate rality , refers to aspects of ki n s hip ties as soc i ate d with ge­ n e al og ic al distances among re l ati ve s . In the Cnited States , in particular , col­ l atera l ity seems to be a significant dimen sio n in determining how rel at i ve s s hould be treated . (See Klatzky , n . d ; Litv.'ak , 1 960a, 1 960b; R o sen berg and A n spac h , 1 97 3 ; Adams , 1 968; Schneide-r and Cottrell , 1 975 . ) B ut if mod e s o f kinship organization vary in diffe re nt segments of A m eri c an soc iety , then it seems p lau s ib l e that people vary al so in the w a ys the y interpret genealogical d i st an c e s . The Phoeni x s t ud y has explored the \vays by w h i ch diverse models of co l l atera l i ty expre ss d ifferent modes o f kins hip o rg an i zati o n . The models include Parentela Orders � Standard .�erican , Civ il Law , Genetic , and Canon Law . The co l l atera l ity models , w hich pro v id e the variou s val ues for the cell s i n Table 9- 1 rep rese nt unique c ombinat ions o f i an d j . For collateral re-lat i ves ·

�

•

N umber of

deMZnding

In

3

2

0

reten. In

Standard

(descent aencration) (i )

Table, SAM

ancestors

generations from ego and ego 's

Civil: 1

2

SAM : I S P OM : 3

1

1

SA M : 1 1

POM: 6

nephews

Civil: 4

Genetic: 3

Canon: 3

Orders model.

SAM: 1 6 POM: 7

Grand,eic�s and grandnl'phews

Ci vil: 3

Genetic: 2

Canon : 2

2

Ni���s and

Civi l :

POM: S

SAM : h

POM: 4

SAM: 1

Siblings

I

Genetic: 1

Canon:

Ci vi l :

Canon : I

Genetic:

Par�nts

ascending generations

American model: POM refers to Parentela

Civil: 3

Canon: 3 Genelic: 3

Great grandchildren

Civil:

Canon: 2 SAM: 1 0 Genetic: 2 POM: 2

Grandi:IU/dren

Canon: I SAM: 5 Genetic: I POM: I

Cltildrt-n

Ego

0

Number of

Cnnun: 2

2

12

POM: 10

SA M :

Genetic: 4 Ci vil: S

Canon: 3

POM: 1 1

SAM : 1 7

childr�n

1-'ir.'il cuu.••in.'i'

Ci vi l : 4

Genetic: 3

Cunnn:

POM : 9

SAM: 7

First cowins

Civil: 3

Genetic: 2

Canon: 2

POM; 8

SAM: 2

(�ntage generation) (j )

Genetic: S Civil: 6

( �amon : 3

POM: 1 5

SAM: 1 8

cousins

SAM: 1 3 POM: 14

s�cond

Canon: 3

Genetic: 4 Civil: S

cousins

Par�nts' jlrlt

Civil: 4

Canon: 3 SAM: 8 Genetic: 3 POM : 1 3

Greal uncks

Great aunts,

Civil: 3

3 P OM : 1 2 SAM:

SAM: 14

Civil: 7

s�cond 4

----

SAM: 1 9 POM : I ll

cousin.t

Genetic: 6

Canon:

POM: 1 8 - - - · --

Parmts'

Civil: 6

4

Genetic: S

Canon:

SAM: 9 POM: 1 7

Grandpllrenls' fir:al CtiU.fin.'i

Civil: S

Uenetic: 4

Canon: 4

Grandpllrtnls ' aunts and unclt's

Canon: 4 SAM: 4 Genetic: 4 POM: 1 6 Civil: 4

Great great Rrandpar,.nts

Great grandparenll'

Genetic: J

4

3 Canon: 3

deMCendunts

A unls and uncles

Genetic: 2 Civil: 2

Grt:Utdpartnls

2

from ego and ego 's

of Kinship Lod Dellned by Parentage and Descent As Viewed from Ego's Perspective, with Lod Designated by American Terms for Relatives and Collaterality as Determined by Canon Law, Genetic, Civil Law , Standard American, and Parentela Orders Meamresa

Table 9- 1. Grid

00 \0

�

1 90

in particul a r , each of t he mode l s combines the values for parentage gen era­

t ions (j ) and descent generati o ns (i ) in a different \\'ay . For example , con­ sidering si bl i ngs (i = 1 ; j = I ) and grandchildren ( i = 2; j = 0) , we fi nd t h at Pare nte la Orders as sign s a c lose r relationship to g randch i l dre n (2 versus 5) , Civil Law reg ard s them as equidistant (both s ibl in g s and grandch i l dre n , 2) , w hi le S tandard A meri c an , Genetic , and Canon Law models p l ace sibl in gs c l oser to Ego (respectively , 5 versus 10, 1 versu s 2 , and 1 versu s 2) . (The c omputat ion a l formulae for all five measures of collaterality are summarized in the I>escriptive-Formulae co lu mn of Table 9-2 . ) A compo ne ntial analysi s of the kinsh ip models (reported i n Ch apter 4) di sc lose s that these models l ie on a co nti n u m in the extent to which they stress d egree of coll ate-ral re mo va l from Ego 's li ne of descen t in al loc a ti n g pri o ri tie s among kin . The analysis i nvolves , among other procedures ,

a

trans­ pos i t ion of the ba sic kinship grid . In the trans posed grid, t he vertical and horizontal ax e s are called G ( generational distance) and R (collateral re­ moval ) . S peci fic a l ly , G i s the number of gene ratio ns above or below Ego in which any given rel ati ve is located , and R is the l ateral distance bet\veen rel at i ve s in Ego 's l ine of descent ( incl uding direct an c e s tors and descen dan t s ) and Ego 's other kin in that same generation as those rel atives . As T ab le 9-2 reveals , all models di sp l ay the b asic two dimensions , G and R . But , in additi on , apart from Canon Law , each model reveal s at l east one other co m pon ent . Moreover, the components of the models increasingly stre ss R co l lateral removal (or , c onversel y , the centrality of Ego "s l ine of desc ent ) as one moves from Canon La\\' to the S tandard Ame ri c an and Par­ entela O rde rs co nceptions :

l . The C ano n Law model , in g iv i ng equ al weight to G and to R , p arti tions kin b y pl ac i n g them along gradi en ts ex tendi ng outward from Ego and the n uc l ear family . It thereb y essentially ignores l i ne of de scent in d esc ribing degree of re l atio n sh ip . 2 . The Ci vil Lav.' model , in a ss i gn i n g a d oub le weight to R as c o m p ared \�·ith G , modifies t he Canon La�· pattern by emphasizing the lineal ties of Ego to fam i l y of orientation (th roug h parents) and to family of proc reat i on (through chi ldren) . All other relatives are then distributed on the basi s of thei r distance from this abbreviated l ine al core , w hi ch consi sts of Ego , his pare nts , and his children . (The. Genetic model i s tran s i tiona l between the Canon Lav.· and Civil Law m od e l s and it embodie s characteristics of both . ) 3 . The Standard American mode l has three distinct co mpon ents . Th e first c om pone nt i s the basic Canon Lav.· degree (G + R); t he second compo nent i �

the R i n c re men t o f the Civil Law deg ree ; and the third is a unique facto r that is defi ne d in T able 9-2 as n (DG + R) . The unique component apportions an

equal weight to all of those kin who are in a g i ven descen t gen e rati on (i )

from Ego and h is ancestral l ine . For e xampl e , in T ab l e 9- 1 , al l kin in the

i +

Parentela

=

(n

Standard american

R R

R

+ + + -t

+

+

-t

law

R

R

R ]

(R - f) ]

conttibution

Civil

when i is JI'Caler than j , 1 for al l nther kin.

A

�

n; wben i is smaller. then A - I

(when

i

=

j,

then G

•

0� and AG must also

=

=

rn (AG

n (OO

+

+

R) ]

R) l

descent

Classes of parentelae-lines of

generations

core

Lineality emphasis in ucending

Lineal

civil law

lntcnnediate between canon luw and

Undifferentiated radiation

en1phuis

C.ornponential

=

n =

and 00

must

=

=

for whom priorities are being also = 0); A Ascendancy coefficient; when direct-line kin or half siblings am considered: f

0.

the largesl i in a set of relatives

), G 0)� I"

-

+

+

component

Unique

formulation

an ancestur (e . g . t Parent = I ; G randparent - 2)9 i - number of generational links between Egn and a ancestor 's lirlsa.'!ndant (e .g. � Child - I ; Grandchild - 2; Sibling = I � Niece = 2); G � j - i (plus nr minus sign of G ipored ); R

LG

(G

R

R ]

+

+

removal

Coll ateral

= j or I� whichever is 5Dlaller; n = lugest j in a set of relatives for whom prinrities are being computed; m compul8d; 11 = Descendancy coefficient; when j is greater than i , D • 0; when j is smaller, then D = 1 (when I

and that

l )j

j

LG

LG

[ G

d ifferenl"e

Generational

number of generational links between Ego and

+

1 )i +

(m

+

I) + j

descendant or helween Ego's ancestor

Definitions of symbols: j

orders

i +

Civil law

j

(i -

Genetic

j, whichever larger

is

i

or

fonn ul a

Caoon Jaw

Ki nship models

Descriptive

Basic grid dimensions

Componential

Table 9-2. Descriptive and Componential Jt,ormulations of Kinship Models

...... \0 ......

192 row i = 1 would have the same score on the unique component; those in ro\"' i = 2 would all have the same score; and so on. 4. Like the Standard American model, Parentela Orders also has three components: the Canon Law factor (G+R), the Civil Law contribution (R), and a unique factor. The unique component in Parentela Orders, \vhich is defined in Table 9-2 as m(AG+R), provides an equal weight to all kin descended from any particular ancestor of Ego (as well as from Ego himself). In Table 9- 1 , all relatives listed in a given column j would have the san1e score on this componen t. The exact v.·eight for each class---i- .e. , an ancestor and his line of descendant�epends upon the generational distance G be­ tv.·een Ego and the ancestor. The classification of ancestors and their sets of descendants corresponds precisely with the verbal description of the Parentela Orders conception in Chapter 1 . This series from Canon Law to Standard American and Parentela Orders models is described in terms of metaphors of social space-from concentric circles (from Ego) to categorical structure (from line of descent)-in Chapter 3. The concentric circles model rests upon the generation of grddicnts ra­ diating from a single point; the categorical model is based upon a classifi­ cation derived from a line. of demarcation. The propositions examined in the Phoenix study are: (a) the spatial metaphors symbolized by the components of the kinship models are isomorphic with communal versus segmental dis­ tinctions in rei igious, ethnic, and socioeconomic settings in the social struc­ ture, and (b) these components are associated with particular kinship norms and value-s relevant to the communal-segmental dichotomy.

KINSHIP MODELS AND SOCIAL SETTINGS

If kinship structures are perceived as modes of organization which connect

personal life to the general social structure, then Yle can interpret each pattern

of collaterality as expressing a particular set of assumptions regarding the nature of this connection. The character of these assumptions is suggested by the kinds of social structure in which these patterns seem to have crys­ talized historically: 1 . The Parentela Orders orientation appears to have e-merged in pluralistic settings (e.g., the Jewish experience, classical Greece). The ideology upon

which this orientation rests is that kinship is organized centripetally to pro­ mote the perpetuation of special political and/or religious interests embodied in the mischpokheh, the house, or the oikos. 2. The Standard American pattern of collaterality seems to have developed as Protestantism and family capitalism blossomed and the merchant class dominated economic and political life in \\'estern society. The Standard American orientation appears to be associated v,rith high achievement moti­ vation and therefore with the perpetuation of special socioeconomic interests.

1 93 3 . The Civil Lav.' map represents a compromi se in the TVw·elve Tables between t he pl uralism of the Patricians and uni v ers alism of the Plebes at the founding of the Roman Republic . Its pri nciples thus seem to express a middle ground in pluralism versus universalism as id eologi es . 4. The Canon Lav.· me asure evolved in its present form duri ng the 12th century to express the aims of the Church toVw·ard universalism in spr e ading the faith over vast domains-with centralized authority and a common body of Ia�·. The main vehicles for the perpetuation of society v.,·ere organized to promote the common interests of the entire social organism \Vith centrifugal family and kins hi p ties providing the informal supports for the cor e s of au­ thority and poVw·er. 5. The Gene tic model, which has emerged in the 20th century \Vith the deve lopm ent of population genetic s . imp lies that the preci se form of family organization is irrelevant to determ i ning the mode of the larger social s tru c­ ture . R ather , again by implication� the ideal is t he reverse-that •• ration al sci ent ifi c •' p ri ncipl es should guide governments and ot her m ajor institutions in determ i ning the chantcter of fami ly and kinship ties. The above de scription of social contexts suggests that the models of col­ laterality express differing modes of kinship orientation along a cont i nuum from centri petal to centrifugal norms a nd values. Theoretically, at least, cen­ tripetal kinship organiza tion emerges (or is sustained) in a situation in which a special interest group is vying Vv'ith other groups for economic , political, or rel igiou s ''goods.'' Centrifugal kins hi p , however, is fostered when spe cial interests are subordinated to b roader concerns extending beyond the particular group. As a general rule, centripetally organized groups appear to be mobi­ lized into tightly organ i zed corporate structures that co ncentrate their effort s on promoting their special interes t s. In contrast, the centrifugal systems are organized in ways which place individual members under numerous cross­ pressures and which serve to erase group boundaries (Farber, 1975). Jewish and Catholic nonns offer a sharp distinction between centripet al and centrifugal modes of organi zation . The contrast bet w een t he cen tripetal tendencies in Jewish kinship and the centrifugal forces in the Catholic kinship can be clarifed by a c omp arison of assumptions about social structure. AG­ cording to ��ndrew Greeley ·s ( 1977) anal ysis , central to co ntempor ary Cath­ olic social theory is the existence of all - embracing Church that is directed at un ifying the diverse segments of society . A lthough conflict and compet ition are everpresent, commo n intere st s deriv ing from the spirit of the Church outwe i gh d ive rg e nt concerns in the maintenance of social order. The Church is regarded as t he formal organization foste ring uni t y and it sustains its ef­ fectiveness through the informal organization existing among localized groups (such as families, fri ends hip netVf·orks, neighborhoods, and so on). The particularism and emphasis upon persona] re- lationships (rather that mo­ bilization of family and kinship ties into corporate en tit ies) create a cohesion,

1 94

based on cooperative norms, in these "dense and intimate interpersonal net­ \'l'orks. '�But at the same time, Greeley (1977, p.262) informs us, HCatholic theory categorically rejects the notion that one can or should sacrifice the present for the future.,, Sacrifice ignores the mutuality of interests and� be­ sides, a ·�.benign future paradise cannot be counted on to appear (Greeley�

1977, p.262). ''Instead, like the humanist, �.'The Catholic theorist is abso­ lutely commited to the \Vorth and dignity of the individual person (Greeley, 1977, p. 265). '' The '-dense networks of overlapping commitments, relation_ ships� loyalties, involvements'' (Greeley, 1977, p.260) create a myriad of cross-pressures that undercut special interests of subgroups. \Vhat remains is a

\veb of relationships expanding to envelop the entire collectivity. Fron1 the

perspective of kinship, this \\'eb extends out\\'ard, \\'ith gradients intenveaving

\'-l'ith other \vebs. This centrifugal movement in reaching out to create \\'ebs

of relationships is symbolized in Church lore by the Canon Law mode.l of collaterality.

4�s opposed to Catholic vieVw'S on social order� the Je\vish vision assumes

a merger of religious and familial obligations and minimizes the distinction

between fonnal and informal organization. The kinship community itself is identified as a religious community; birth and endogamy even more than faith mark its boundaries. Parents and children, husband and \Vife, grandparent and grandchild, cousins, uncles and aunts, nieces and nephews are mobilized in an organization w·ith numerous corporate aspects-homogenization of re­ ligious norms and values, sacrifice and benefice� household rituals, interest in ancestral "·mots "-a historical mission for family and kinship. The per­

sistence of Judaism is identified as inhering in family continuity and in Has­ idic sects even rabbinic dynasties

established. It is primarily among Re­ fonn Jev.;s that a sharp distinction emerges between formally organized are

religious institutions and informal family relationships, i.e., a gap between ties \\'ith Judaism and domestic existence; and it is primarily Refonn JeY.'S "'·ho often hold a Canon Law or Ge.netic conception of collateral - ity. Other­ \'l'ise, the separatist te.nde.ncies in Judaism are expressed by the Parentela

Orders model of kinship, \\'hich appears in the Mishnah and survives in modified form in Isra eli la�'.

Both Jev.'ish and Catholics are generally regarded as "familistic "-inter­ ested in family and kin. However, they appear to represent different ap­ proaches: Je\'-I'S with their Parentela Orders perspective. seem attuned to long­ distance kinship interaction, V.'hile kinship contact among Catholics appears to be tied more to locale (as it is with other persons with a Genetic or Canon Law orientation). These different approaches apparently represent oppos ing '6templates'' in organizing the social existence of Catholic, Jew·, Protestant,

and humanist. In this book, I have. referred to the '6corporate structure'' of kinship in

connection with the Parentela Orders model and to a lesser extent the Standard

1 95 American model. Obviously co rpor ate structures in American k i n s hi p are comp are d with the unilin e al structures in African societies (Rad­ shado wy as

liffe- B ro wn and Forde. 1950). In th ose societies, membership is c le arly demarcated by descent t hrou gh males or females; unit s are nam e d � offices a� assigned; and rights and obl i gati ons are explicitly designated. In American kinship, all of th e se criteria are weakened. Yet, there are some signs of peop le treating their kinship structures as more tha n a collection of individ­ uals: empha s i s on line of descent exten ding to ancestors no longer alive an d to desce ndant s yet to come;. a ten d en c y to homogenize values "'ithin th e group; r e c og ni tion of general nonns and values connected to kinship ties­ prior i ty in a ssi sting relatives more than nonrelatives� m aintenanc e of contact over long di s t a nces , obligations to see kin, and so on. Cnfortunately� there is no English term that falls bet\veen aggregate (or netv.'ork) of individuals and formal corporate grou p . Unless I reson to a \l.'ord lik e semi-corporate structure, I am left only \\'ith the option of referring to '·corporate aspects·' of kin ship . Apparently, the absence of appropriate tenns to express the corporate as­ pects of American kin ship sugg e s ts to some o bserv e rs that such structural elements do not exist . But such an omission seems legitimate since there is considerable variation in American society in the extent to which kins hip connotes corpo raten ess . As ethnographic reports show, this kind of over s i gh t is not unique; b asic social categories lack termin olo gical indicators in many societies. For examp le , the patrilineal Chinese have no he or she pronouns. Similarly, th e lexicon of classical Athens and of the Pen an of Borneo have no word for marriage (Needh am , 1974). Under such circumstanc es one would expect the meanings of other words to be stretched to inco rpor ate phenomena for which vocabulary is missing and to rely upon c o nt extual cues to si gnal intended usag es . For i nstanc e , the modern German wor d for mar­ riage (Ehe )is an extension of the Middle - H igh German term for lav.' or statute (ewe), and the English marriage and French mariage are derived from mnr­ itus, husband in Latin (Needham, 1974, p.43). The seman tic stretch, \\l'hich gives rise to multiple me anings , serves to transform categorical concepts into "odd-job" wo rd s (Needham, 1974). In American society, people have to rely uponfamily, cousins, roots, or a special tenn (e.g. tribe, mishpokheh, c

Ven.,·andtschaft.

clan) to express corporate aspects of kinship, but others may p ap ly these same terms merely to identify aggregates. Consequently, as in any cultural analysis, one cannot expect that every sig nifi c ant element be unambiguously named (Schneider, 1968). Nevertheless, the finding s of the Phoenix study reveal

a

relationship bet\\l·een

kinship orientation and corporateness in the structure of family and kinship ties. In marriage, this corporateness is seen in tendencie s toward (a) regarding marriage as a sanctified institution; (b) homogenization of religious values of husban d and wife; (c) low probability of a sense of disillusionment for self

196

and spouse; and (d) a high degree of marital stabil ity. In fami ly life, cor­ porateness is express by (a) priority given to domestic roles over occupat i on (b) emphasis upon status pl acement functions, and (c) stabil ity of fen il i nonns over generations. In kinship ties, corporateness is reflected in: (a) a greater tendency toward virilocal than uxorilocal residence (presum abl y an emphasis upon instrumental ties); (b) the presence of interest in kinship as a general principle; (c) the maintenance of frequent contact with kin over long distances as apparently expressing an ''axiom of amity"; (d) a desire to increase contact with non-nuclear family relatives , in spite of competing interests, and (e) a hint of authority problems with paternal kin-perhap s tied to traditional patrilineal tendencies . The clustering of these attributes among persons in the Parentela Orders and to a les ser extent the Standard American categories suggests that the se models of collaterality signify kinship struct ures with strong corporate components. These findings can be interpreted in tenns of the dimensions of the kinship models derived in the componential analysis in Chapter 4. At one extreme, the lineality component, expressed by the Parentel a Orders and Standard American unique components, is overrepresented in segmental (or factional­ type) settings, religious as well as socioeconomic, and it is related to the presence of' 'corporate'' elements in kinship ideology. At the other extreme, the undifferentiated-distance component, expressed by the Canon Law and to lesser extent by the Genetics model, is associated with communal setting s, and it is tied to indiv idualistic kinship ideologies, which min imize classes of kin. The Civil Law model, expressing the l ineal core component , lies between the extremes . Consequently, the findings in the Phoenix study can be con­ strued as disclosing an isomorphism between these components as spatial metaphors and types of social settings.

�

KINSHIP

MODELS AND OTHER

PERSPECTIVES

In the fmal analysis, the contributions by the Phoenix study to knowled ge

about contemporary social structure and changes now occurring can be as­ sessed only by an answer to the question: Does the analysis of American kinship ties in terms of models of collaterality yield insights that are absent from other conceptualizations? In this section I wi ll discuss implication s of the fmdings of the Phoenix study for comparing the kinship-model perspe c­ tive with previous conceptualizations regarding American kinship. There are four subsections. each of them an independent essay deal ing with a different topic . The subsections include: (a) Standard American versus middle-cl ass kinship , (b) permanent availability and k inship orientation, (c) kinship ori­

entation and communal structures, and f inally (d) components in measures of collaterality.

197

STANDARD

A�IERICAN VERSUS MIDDLE-CLASS KINSHIP

one of the contributions of the Phoenix study is the serendipitous finding of the existence of the Standard American model of collaterality. This model, held by over 40 percent of the respondents, is prevalent among families \\'ith Protestant, "middle-class'' attributes, and it appears least often among per­ sons w ith the following characteristics: (a) blue-collar occupation of the male household co-head; (b) family income in 1977 under $10,000; (c) respondent not a high school graduate; (d) respondent's father born in Europe; (e) mem­ ber of a racial minority group; and (f) respondent not in a traditional Protestant denomination (i.e., is instead neofundamentalist, Catholic, Monnon, or Jev..·­ ish). Typically non-middle-class and non-Protestant characteristics are asso­ ciated with non-standard-English linguistic codes in speech. Because of this analogy betvleen kinship model and linguistic code, I have identified the serendipitous measure of collaterality as the Standard American paradigm. In order to determine whether the Standard American model is a useful tool for understanding the workings of contemporary kinship, I shall compare findings pertaining to this model with those for other conceptualizations, notably those by Talcott Parsons, David M. Schneider, and Eugene Litvlak. Prior to the comparison, I shall briefly outline their conc-eptualizations. The Parsons Conception

Talcott Parsons (1954) characterizes the standard form of American kinship as an ''open, symmetrically multilineal, conjugal system.,' The openness refers to the lack of restrictions on assortive mating (apart from limitations imposed by the incest taboo). Multilineality refers to the lack of distinction by se'l among intennediate ancestors who intervene between a distant ances­ tor and any given descendant; as a consequence, there are numerous lines of descent�ight from the great-grandparents alone. Unlike its European fore­ bears, the American system is symmetrical, giving no preference in sol­ idarity to any one line. Conjugality refers to the fact that parents and children consti tute an identifiable household unit that is regarded as a distinct collec­ tivity. Thus, according to Parsons, the system is marked by ''a maximum of dispersion of lines of descent and the prevention of the structuring of kinship groups on any other principle than the ·onion� principle, which implies pro­ portionately increasing 'distantness' with each 'circle' of linked conjugal famili es (Parsons, 1954, p.182)." Parsons proposes that deviations from the standard kinship system occur when solidarity with one line of descent is emphasized at the expense of another. Specifically, he mentions (a) tendencies in farm families to approach the pattern of the European stem family as described by LePiay, (b) strong p atrilineal proclivities in the upper class, and (c) the powerful matrilineal ties in lower-class settings . In the urban middle class, however, neither line of

198 descent is stressed, and instead one's "frrst kinship loyality is unequivo cal I to his spouse and the.n to t heir children (Parsons, 1954, p.l86). � ,

Parso y stresses that, in the middle class, upon m arry i ng , Ego "is by comparis with other k in s hip systems dr astically segregated from his family of ori n e ta tion , both from his parents and from his siblings " (Parsons , 1954, p.186 ). Parsons also attributes to th e symmetrical multilineality the nonn of sexual equality in family relationships. For Parsons, conformity to the standard kinship system (or deviance from it) derives from the American occupational system. There is a .. fit', belween the '4open, symmetrical multilineal conj ugal fami l y system'' and the require­ ment of the occupational system for a h i gh degree of social and residential mobility, unham pered by co mpeting loyal ti es. The societal need for a highly mobi le. \a.'Ork fo rce interferes with the maintenance o f close kinship ties and un i l ineal solidarity. Litv.'ak and others have criticized the Parson ian concep­ tion because of its emphasis on the iso l atio n of the nuclear family as a kinship

: �

structure.

The Litwak

Conception

In the years following World War ll, fam ily soc iol o gi sts mustered sufficient data to launch an attack on Parso ns ' description of the American kinship system (Sussman, 1 965; Sussman an d Burchinal, 1 96 2a and 1962b). The thrust of the criti cism is that Parsons is too extreme in his depiction of the nuclear fam i l y unit as segregated from the family of orientation and other forebears. Research findings indicated to these sociologists that Parsons had underestimated the extent to which families sh are functions and resources despite their residential i solation from each other. In particular, Litwak's concep t of the modifie d expend ed family calls attention to the fact that k in ship ties may be valued despite residential distance and lack of day-to-day contact. The classical extended famil y demands "•geographical propinquity, occu­ pa tional involvement, ... nepotism , [and] . . . an hier arch ica l authority structu re ," while the modified extended family "consists of a series of nu­ cl ear families bound together on an equalitarian basis, with a strong em ph as i s on these extended famil y bonds as an e nd value (Utwak, 1960a, p. 10).

''

Lit\\'ak 's paradig m for the classical exended family in the United Stat es i s the Italian family, and his model for the modified extended family is the Jev.'ish family. He indicates that "the Italian family structure. rela tively speakin g , is sti ll defined in terms of geograph ica l and occup ational proximity·

In contrast, the Je wish extended f am i ly , for historical reaso ns , cl o se ly resem­

ble.s the family structure considered to be prototypical of conte.mporary urban

bureaucratic life. ' "

But in contradiction to Utwak 's analysis, Winch and his associates (1 9 77)

1 99 report that, as compared with Catholics and Protestants, Jews sho\\' higher u onal and geographic concentrations. They ar e "more entrepreneurial, occ pati familistic than either Catholics or Protestants,'' and tess migratory, and more even when they do migrate, ''they are much more likely to live in the spo use's home community (Winch, 1977, p. 32)." The contradictory statements by Litwak and Winch suggest that it is difficult to differentiate between a non­ mobile modified extended family and a classical extended family. The fea­ tures that Litwak attributes to Jewish kinship are generally confirmed by the Phoenix study, but they express the Parentela Orders orientation rather than the Standard American view. Litwak proposes that the modified extended fa mily is the most effective form of kinship organization in modem society in that it pennits geographical and social mobility which simultaneously makes affectional and instru­ mental resources of close kin available to individual families . Ye.t, Litv.'ak 's own da ta (Table 5, 1960a) contradict the claim that the modified extended family is a more effective instrument tha n the isolated nuclear family in mee.ting the needs for most people. In his study, a maj o rity of responde nts in any occupational stratum con fonn to the ' nuc lear family orientation,'' and fewer than a fourth are classified as ''e.xtended family orientation." (Se.e also Ke rckho ff , 1965, Table 5-2.) Moreover, the major differences betVw'een upw ardl y mobile and stationary occupational classes are not in percentage.s of persons with an extended family orien tato n but in the perce ntage s Vw'ho \\'ant nothing to do "'ith · •family' (�·ith 34 percent of the stationary lo\\l·er class cla ssified an '"non-family o ri ented , as compare d with 23 percent of the upwardly mobile). These data from Litwak's investigation suggest that the mod ified extended family is more of a goal for social policy than a description of prevalent ki n ship norms. Although Litwak �s \\'ork may h ave focused at­ tention on a residue of kin ties in contemporary soc ie ty , it does not appear �

·

·

to negate Parsons' notions about American kinship.

The Schneider Conception

The position taken by Davi d M. Schneide r is that standard American kinship e xists onlv at the cultural level and that variations in American soci etv result . from the intrusion of influences from other d oma ins upon the kinship domain. Actual relationship s with kin are thereby governed by no rms \Vhich are a compromise of the demands of the various domains. Hence, for Schneider, American kinship survives only as a set of general symbols pertaining to blood ties and general moral codes isolated from specific social contexts. But Schneider"s analysis of American kinship as a normati ve phe nomenon -as compromised norms--does emphasize the significance of socioeco­ nomic cl a ss in molding ties among relatives. Like Parsons, S chneide r a ssigns

200

primary weight to the occupational structure in the development of kinship norms: "'One of the most important sets of forces affecting familial be havior is the relationship of individuals and families to the economic system (Schnei­ der and Smith, 1973, p.l97)" Yet, he weakens this assertion by indicatine that, ''The cultural orientation of the lo'.'l·er class can be found among peopl; who are quite well off, just as there are many genteel poor of undoubted middle-class orientation (Schneider and Smith, 19 7 3, p. 6 7). ' ' One must as­ sume, therefore, that for Schneider and Smith orie.ntation toward kinship is at least partly inde.pendent of social class. Schneider and Smith ( 1 973) characterize middle-class kinship by distin­ guishing it from norms associated with the lower class. For them, middle­ class kinship has these characteristics: I . Nuclear family self-sufficiency; 2. Nuclear family solidarity undifferentiated by sex (as opposed to separate­ ness of men and \'/'Omen and solidarity of mother and children in the lower class); 3 . The kindred as a chain of linked nuclear families (as opposed to personto-person relationships in the lower class); . 4. The rule of reason (with accompanying self-discipline and sel f-control) and expertise (i.e.,rationality) rather than status as basis for authority (as opposed to authority vested in males in lower class).

Comparison with Standard American Model The analysis based on the Standard American collaterality model yields in­ sights into the functioning of kinship in middle-class American that are not apparent in the Parsons, Litwak, or Schneider fonnulations. This section

compares these conceptions with the Standard American model, first in their similarities and then in their differences. The Standard American model provides a means for integrating the depic­ tions of middle-class kinship by Parsons, Litwak, and Schneider: (a) sexual de-differenti�tion (Table 8-2); (b) significance of the marital tie (Table 4-7); (c) norms of equalitarianism in family relationships (Table 6-11); (d) fa cili­ tation of upward social mobility (Table 5-9); (e) fairly frequent contact viith parents and siblings living at a distance (Table 8-6); and (t) interpretation of

relatives as representatives of collectivities rather than simply persons (Table 8-3, 8-9 , and 8-10). Like Litwak's concept of the modified extended family' the series of which the Standard American model is a part differentiates between distinctly kinship ties (regardless of residence) and the ''segregated nuclear family'' as a household unit. Nevertheless, along with integrating these attributes, the Standard American model generates additional insights

201

about contemporary American ki ns hip . lJnlike other mo dels of collaterality, the Standard American measure is associated with (a) e mp hasi s upon anc es­ tral line in co nc ep tio n of coll at eraltiy (Table 4-4)� (b) c onn ect io ns among nativity, traditio nal Protestantism, and kinship orienta tio n (Table 4-5); (c) the tran smi ssion of a norm of low fertility and co nsequently the production of small kindreds (Table 6-8, 6-9, 7-15, and 7-16); and (d) re s ponsivene ss to the changing stat us of women in American society (T abl es 7-10 and 7-1 2). Significantly, the fin d in gs indicate that the Standard A meri c an model is no t the only kinship orientation whose adherents show a prop ensity toYlard up­ Y.lard social mobility (Table 5-9). Perhaps of most i mp ortance, the representation of Standard American kinship as a model of collaterality avoids l ogical difficulties in the Parsons, Litwak, and Schneider app roac hes. All three analysts-through their empha­ sis on f unc t ionality in a soc ioeconomic setting-face the problem frrst of deriving the attri bu tes of ''middle-class 9' kin s hip from the functional require­ ments of the American occ u pational structure and then of hav i ng to account for its existence in non-middle-class sege men t s of society (as well as its absence in many mi ddle-c lass families). The pers istenc e of kinship structures that deviate from social class ··ne eds '' (or deman ds ) prompts one to consider an alternate approach. The analysi s based on kinship models, however, rests upon an alternative set of ass umpt io n s. It begins v-·ith the no tion that transcendant plur al ist (or factionalist) ideologies s ti mula t e groups to dev elop views about kin s hip that set them off ov er gen erations from other groups. I have suggested that the emergence of family capitalism (al o ng with early Protestant ideas about worldly success and worthiness of ancestors among the Elect) a cce nted the function of status-placement as a d ominant family fun ctio n. I have further proposed that this accentuation is symbolized in t he Standard American model of coll ateralit y. Certain ly, families in relatively high socioeconomic posi tio ns are m o tivated to perpetuate their stat us-to continue to set themselves off from families with inferior resources and mo des of existence. Such families tend to exhibit a Standard American con cep tio n of ki nsh ip over generations. But, in addition, similar ideologies ma y emanate in groups and movements in diverse parts of the social structure. For e x ampl e , the data indicate that the Standard American m od el is e spec ial ly preval e nt among younger women, who apparently have been sti mulated by the contemporary feminist movement and consequently exhibit high con centrat io ns in the work force and in grad­ uate level education (Table 7-9, 7-10, and 7-12). H ence , by c onc ei v in g of kinship ori entatio ns as cultural models , which are applicable to v.'ide range of social conditions and economic systems, we can avoid the logic al circu­ larity of explaining the features of middle-class kins hip by positing '·needs'' fo r such structures.

202 PERI\f.-\NENT A\' AILABILITY AND K�SHIP ORIENTATION

The fac t that mo de of kinship organization has an i mpact on norms penainin g to m arri age is well documented in anthropological l itera t ur e (e.g., Fox

1967). But the nuances in marriage rules i mplied by the A merican

kinshi;

system have scarcely be en e xplored . In 1964, in my book, Fwnily: Orga. nization and Interaction I suggested that, insof ar as the multil ine al system

is associated V.'ith assortive mati n g , not onl y marital selection but also tem­ poral aspects of marriage would be affected. Fo ll owing Pars ons , lead that

kinship in the United States is organized as a sym me tri cal multiline a) system,

I reas one d along the following lines: I. In a multil ine al system, no recog n i zable descent group or line of ances­ tors has any e x c lusiv e claim on an individual �s assests or abilities. (The

n uclear family's claims are l imited only to its life span; special claims of the family of orientation are di ssipated as the individual m arries . Any such claims that remain de.rive from personal relationships among famil y members rather than from familial status.) 2. Without such a clai m , the kin dred has no legitimate basis fo r any special rights or obligations pertaining to an individual and hence no spec ial stake in that individual's destiny , particularly his marital destiny . 3 . \\'"ithout a special stake, the re are no g rou nds for imposin g restrictions on an individuars choice of spouse or mode of ex istence (i.e., famil y life style) following marriage. 4. Since the openness extends not m e rel y to choice of marriage partner but als o to home life afterwards, individuals are free to e n d marriages and beg in new ones without the intrusion of parents or siblings. 5. As marital status as sumes the identity of a potentially transitory rel a­ t ions hip , individuals become av..·are that they are pe rmanently available for a new marriage (or a new mate) re gardl e ss of their current marital status. 6. \\rhereas for some individuals, th is pennanent avail a bi li t y may become a positive value, for many pe ople this awareness decreases the le.vel of tol ­ erance for enduring difficulties in a marri age . Moreover, it may sensitize people to focus on probl ems in their m arri a ge in order to determine ,-..·hether the critical point of toleration has been exceeded (or wi ll soon be reached)· 7. Pennanent availability as a social real ity has various implica ti ons for courtship and family norms:

a. Sinc e marr iage is po tentially a temporary relationship, the appropriate courtship process is n ot one of c are fully sifting and narrowing a field

of eligi bles until one person is selec ted for lifetime mating. Instead the proper procedur e would be to enter into a series of marital - l ike rela­ tionships. When one of these trial relationships is sufficiently satis tying to call fo r stabilizing measures, marri a ges may then be formalized.

203

b. Inasmuch as the. system assumes remarriage as a normal occurrence, there is no ·special virtue in maintaining virginity prior to marriage. Rather, appropriate socialization for permanent availability would call for early trai n in g in developing sexual competence. c. Since marriages are not assumed to be permanent, there is no s pe ci al virtue in delaying marriage or a quasi-marital tie . Fonnal marriage symbolizes a personal desire to stabilize the relationship rather than a commiunent to do so. d. Children serve as pledges to maintain the marital relationship. Never­ theless, there is no certainty that they will be effective in performing that task. e. With the presupposition that marriages may be temporary, one is re­ lucta nt to break sentimental ties '-'·ith one's parents and siblings. In the event of marital breakup, bonds to parents and siblings may resume their importance as sources of personal and financial assistance and emotional support. This vie'-'· implies that a chasm exists whereby con­ tinued support is regarded as appropriate but pare-ntal controls are not. C ontrols are attributed to family ties of a corporate nature, whereas assistance is given on a personal basis (i.e. for love). f. Even after marriage, because of the uncertainty of marital destinies, people (intentionally or inadvertently) re m ai n in a state of readiness. with an emphasis on youthfulness, keeping attractive, and maintaining sociability with the opposite sex. g. Because of the openness in choice of sequential marital partners, any given line of descent is highly vulunerable to the introduction of diverse norms and values; consequently permanent availability is associated with changes in fa mily organization over time. The opposite of per­ manent availability is orderly replacement of family nonns and values in any given I ine of descent from one generation to the next-as it might be symbolized by t he Parentela Orders orientation. Of the several kinship orientations, the Genetic and particularly the Canon Law models seem closest in approximating the conception of social space that fits the attributes of permanent availability. The Canon Lav.· model is

Ego-centered, with gradients extended outward. The members of the nuclear family occupy the innermost space and with each step away from the nuclear family in coll at era l or generational links, one degree of genealogical distan ce is added. Pictorially, the Canon Law model resembles the onion analogy a ppl ied by Parsons in his characterization of American kinship (Parsons,

1954).

In norms as v.·ell, the Genetic and Canon Law models approach the con­ ditions of permanent availability. The findings in the Phoenix study reveal the following for respondents in the Genetic and Canon Lav.· category:

204

1. Openness in choice of marital partner: Among Protestants, the persons with a G ene tic or Canon Lav• orientation ha v e the larges t percentage of religious interm arriages of any of the kinship classifications (Tables 6-3 and 7-2). 2. Prevalence of divorce: The Genetic and C anon La w category has the largest percentage of divo rced parents and siblin gs (Tables 6-1 and 6 -2) and the p re ponderance of respondents t hem selv es who have ever been divo rced (Tables 7-6 and 7-7). 3 . Living together as part of courtship: Canon Law and Genetic w·omen� in particular, yield the l arge st percentage of persons v-·ho agree that '•it is a good idea for a man and woman to live together for a w hi le before they marryn (Tab le 7-14). 4. Age at marriage: Genetic and Canon Law wo men tend to marry at an early age (as in Tabl e 6-5 for respondents' mothers and Table 7-8 for women respondents and wives of men respon dents) . 5. Having children as a pledge: Regardless of social b ackgrou nd , women in the Genetic and Canon Law category (unlike o the rs) �·ant more children than men in that category do. Since it is gene rall y the women who seek to stabilize cohabi tin g rel ation s hips , this finding is co ns isten t with the view that c hild ren serve as a pled ge to maintain t he marital tie. 6. Kinship ties: Persons in the Geneti c and Canon Law categ ory have the larges t percentage of close relat ive s residing in Arizona (Tab le 8-1) and see them freq uentl y (Table 8-7), thereby keeping kin ties in readiness if the marital relationship proves inadequate. 7. Sensitivity to tolerance limits: The respondents in the Genet i c and Canon Law cl assifi cation are als o most prone to per cei ve a hi gh deg ree of ge neral disillusionment (or "anomia") in their marriages (Table 7-4), and they tend to view their spouses as more profane than they themselves are (Table 7-5). The fmdings thus point to a convergence between cond itions for permanent availability and the Genetic and Canon Law models of collaterality. Despite this g ene ral convergence, the countervailing i nfluen ce of re ligious controls cannot be ignored. It is true that amo ng the parents of responde nts , regardless of religion, divorce tends to be more pre vale nt in the Canon Law and Genetic catego ry than for the total sample (Table 6-1). Nevertheless, for the respon­ dents themselves, the relationship between prevalence of d ivo rce and Ge­ netic -Canon Law cl assification holds only for non-Catholics. Amon g Cath­ o l ics , the percentage of persons currently in their first marriage (i.e., nei ther remarri ed nor now d ivorced) is equal to that for all Cat ho lic s in the sample ( Tab le 7-7) . But cle arly , fewer Jews and Protestants with a Genet ic or Canon Law ori ent ati on are still in their frrst marri ages than is true of the sample totals in those rel igious groups ( Table 7-7). Hence, Catholicism-with its canon i ca l limitations on divorce-appears to counteract effectively the pro-

205

cl iv ities toward marital breakup and remarri age stemming from Genetic and canon Law kinship orientations . KINSHIP ORIENTATION �� COMMUNAL STRUCTURES

The disti nction between associational and communal structures has been fru it­ fully applied in the anal ysis of religious and ethnic correlates of family life . Over a decade ago, Gerhard Lenski (1963, p . 356) suggested that American Judaism faces an uncertain future because ' 'the weakening of the associa­ tional side of Judaism [by which he meant the hold of the synagogue or te mple] may make the subculture of that group more vulnerable to environ­ mental influences . , . Since a communal subculture seems to rest , at least in part, on family and kinship ties , it would seem that the persistence of Amer­ ican Judaism as an identifiable entity requires a strong commitment to tra­ ditional family and kinship obligations . The Len ski ' 'communal " position pl aces much emphasis on those factors that have been found in sociological research to affect family ties: extent of mobility , residential scatter , and personal contact (or to use Robert Winch 's ( 1 974) tenn , functionality) . These are the elements that sustain commitment to the informal networks of relati onship required for the persistence of a subculture. However, the ' 4>associational ' • position taken by Herberg ( 1 960) and by Glazer and Moynihan ( 1 974) opposes that of Lenski . Whereas Lenski see s ethnic survival as a consequence of a balance between associational and communal elements , the associational proponents suggest that the persistence of Jewish and other ethnic groups depends more upon religious and/or secul ar collective action than upon the peculiar functionality o f the family and other communal institutions . This position leads to the expectation that degree of com mitment to J udaism and its familial concomitants rests primarily upon participation in fonnal , specialized structures--extent of synagogue attend­ ance and ritual observance and membership in Jewish organizations (e .g . • B 'nai B 'rith, Hadassah , Organization for Rehabilitation through Training) . Andrew Greeley ( 1 976) made a comparable distinction vlith regard to Cath­ olics--ecclesi astical Catholicism versus communal Cathol ici sm. Although Greeley 's dichotomy is stated in political tenns-the sources of leadership and authority in Catholic praxi�ssentially it too re sts upon asse ssing the role of associations in the persistence of Catholicism as an identifiable col­ lectivity , despite its considerable ethnic diversity in America . Ecc lesiastical Catholicism refers to the influence of Church institutions on the I i ves of people , that i s , the impact of the organized church and the various institutions under its control (Greeley , 1976 , p . 2) By contrast, communal Catholicis m refers to " 'other pow er centers withi n the collectivity that have an independ ent influence of th eir own , not necessarily in opposition to the religious institution but distinct and se parate fro m it (Greeley , 1 976, p . I 04) . ' � These must grasp

206

" 'the consi derable heterogeneity of family structures , role expectatio n s , and values concerning intimate behaviour that can be found among the di v erse communities that constitute our society . ' ' Further light on this dichotomy is shed by a study of Jews living i n K ans as

City in 1 976 . That analysis discloses two patterns of participation in the Jewish community . One segment of the community participates ass oci ati on­ all y-by fonnal and ritual means and reliance on formal institutio nal arrange­

ments for its involvement . This group , which shows the greatest i nclination toward membership in Jewish organization s , synagogue (or temple) atte-nd­ ance , and ritual observation , tends to conform to kinship model s that di sre­ gard line of descent in partitioning of relatives ( i . e . , Genetic and Canon Law models) . There i s , however, another segment that parti cipates comm unally . Th i s secon d group consi sts of people who tend to be more active in Jev;i sh

matters than their parents , live in Jewi sh areas of the city , di sapprove of

religious intermarri age , and are concerned v.'ith the persistence of Jev.'ish

identity . The communally oriented Jews conform more often to the Parentel a Orders model (Farber , 1 979) . The connection found between kinsh ip model and communal (as opposed to associational) ties to the Jewish community , it seems, can be generalized

to Catholics and Protestants as well . Both Lenski (1963) and Gree ley ( 1 976 )

suggest that Catholics convey many communal ties in organizing their daily

live s . Indeed , Greeley ( 1976) proposes that , despite the official theoretical position , as skeptici sm about Church teachings grows , Catholics are becom­ ing increasingly communally oriented in their approach to social ethics . There are , therefore , two questions to be asked about communal-versus-associa­ tional ties: (a) Do American Catholics continue to show greater communal

inc linations than. Protestants do (and les s than Jews)? (b) What role. does

kinship orientation play in associational-versus-communal organization?

Drawing from the Kansas City findings cited above , let us assume that the

Parentela Orders model expresses a communal mode of organization . The regression analysis in the Phoenix study indicates that in a general population the Civil LaYl model is associated with associational ties . (See Appendix A) . Then for any partic ular religion , the extent to which the proportion of respon­

dents in the Parentela Orders category outweighs the proportion in the Civ il

Law category describe s the degree to which that religious group can be

depicted as communal in its kinship organization . A computation based on

Table 5- l shows that the ratio of Parentela-Orders to Civi i -Lav.' respondents for the various rel igious groups is as follows : Jewish Catholic Neofundamental ist Protestant sects

1. 75

1 . 37 1 . 33

207

Reformation Era. Protestant denominations

0 . 77

Pieti stic Protestant denominations

0 .83

Monnon (Latter Day Saints)

0. 3 7

The dependence among Protestant denominations upon formal institutional ties for their coherence as social entities is apparent in the list of ratios above . Except for the sectarian s , who set themselves apart communally , Protestants , as in Lenski 's earlier analysis, remain associational in orientation . Moreo\'er, whi le Catholic s appear to be highly communal in outlook , Jews are even more. so . Table 9-3 summarizes the data presented in previous chapters pertaining to kinship characteristics relevant to the communal-associational disti nction . Presumably , a communal orientation rests upon kinship relations that (a) inhibit divorce ; (b) foster religious unity in the family; and (c) encourage frequent contact with close kin . The table disclo ses that , for both the respon­ dents and their parents, kinship characteristics expressing communal ism are general ly most prevalent among Jew s , next among Cathol ics , and least prev ­ alent among Protestants . Moreover, within religious groups , persons i n the Parentela Orders category display more evidence of communalism than do those with a Civi l Law orientation . The data therefore offer support for the ' 'stickiness ' ' of kinship sources of communal organization among re ligiou s groups in American society . The tenacity of kinship ideologies that foster autonomous group identities seems to complement socioeconomic character­ istics in structuring religious col lectivities .

COMPONENTS � MEASURES OF COLLATERALITY

In the past , analysis of kinship tenninology has provided anthropo logists V.'ith numerous insights into the mechanics of social structure . Nevertheless , when they have sought to apply conventional linguistic procedures to American kinship , they have fac-e d two major problems; 1 . Past research and speculation have indicated that changes in kinsh ip nomenc lature occur long after other modifications have taken place in nonns governing kinship ties (Naroll , 1 970) . In those soci eties in v.'h ich changes in social structure accrue slowly , the study of kinship terms offers many cl ues to the factors that give the society its form. In modem , rapidly-changing society like the United States , however, tenns may l ag far behind current family an d kinship usages . Consequently , an analysis of American kin ship terminology per se may reveal more abou t the fam i ly and kinship structure of ' •our foun d in g fathe rs � ' than it might abou t contemporary arrangements . For e xample , in thei r analyse s , several an thropologists have concluded that

2 08

Table 9-3.

Communai-Assodational Dlsdncdon , by Orientation or Respondents

Family and kinship characteristics and

Source

kinship orientation

table

6- 1

Summary or Family and Kinship Characteristics

·

Religion and

Kinship

Catholic

Protestant

Prevalence of divorced parents of respondents a

Parentela orders (%)

7-6

Jewish

Relevant for

Civil law (%)

6.7

1 5 .8

22 .4

1 1 . 8b

26 . 7

20. 8

74 . 1

69 . 3

64.9

9.7

22. 7

Percent of respondents now in their f1rst marri age (i .e. , neither rem arried or curre ntly d ivorced)

( %)

6-3

Parents who intermarried across religious groups<. Parentela orders (%) Civil law (%)

7-3

Religioos intennarriages which remain' 'mixed ' . a

Parentela orders (%) Civil law (� )

64

31 .2

37 . 5

63 . 9

52. 2

1 00.0

73 . 4

83 . 6

85.7

72 . 9

54 . 6 52 . 9

Respondents ' siblings are all in the same religiona

Total sample (%) Parentela orders (%)

86. 7 7 1 .4

Ci vii law (%)

Respondents ' chi ldren are all raised

8- 8

25 . 9

in the same religion ( %)

97 . 5

58.8

88 . 6

82.9

77.4

73.6

44 . 9

40. 7

64.0

42. 3

46 . 9

36.0

67 . 3

43 .4

Close relatives I ivi ng in Maricopa County who are see n at least weekly

M other ( % )

93 . 3

84 . 6

Father ( %)

Brother c losest in age to ego (%) Sister closest i n age lo e go (%)

75 . 0 77 . 8

Brother second closest in age to ego (%)

Sister second clo sest in age to ego

(%)

•

By religiou in wbicb respondent was raised .

b

Because of small number of

c

By religion

in which

cases,

Geaetic and Caaon law

category

65 . 0

6.5.2

combined with Civil

Law

category .

mother was raised .

the traditional American-English terminology still reflects the s tructure of the "typical " conjugal famil y in the United States , that is ,

a

family in which

husband and w ife and their children share a household ( Brown , 1 974 ; Casson , 1 973 ; Schneider, 1 968) . Yet, for decades U . S . Census data , papers in lav.' journals , research on kinship and family I ife styles , and court decisions have

209

de scribed a shift in both legal and community norms toward an increasing conceptual separation between (a) family and kinship as social entities and (b) famil y and kinship as residential and ecolog ical arrangements . Ob­ viously , the conventional anal y sis of traditional nomenclature cannot discern

this ch ange in American culture . 2. A second deficiency in the past research on American k i nship termi­ no logy has been its failure to take into account the heteroge neity of family and kin ship structures in American society . Unlike mo st societie s studied by anthropologists , American society consists of a multitude of peoples with different ethnic and religious backgrounds . The countless waves of immigra­ tion that occurred throughout American history have introduced diverse and often contradictory ideas about kinship into the U nited States . Each ethnic group has imported those conceptions of kinship prevalent in its native land . But despite this heterogeneity , the ideal of Americ anization has , in the past , made it incumbent upon immigrants and their children to use Americ an­ English as a vehic le of communication in al l domains , including kinship . However, the use of the English language does not imply that the newc omers and their descendants have automatically assimilated norms and values which molded Engl i sh kinship nomenc lature . Rather , it seems plausible that stand­ ard English terminology has been stretched to accommodate the diversity of kinship usages current in the v arious ethnic and religious groups . To the extent that the United States remains a pl uralistic society , one would expect this heterogeneity to persist in spite of the common use of English to describe kinship statu ses . Empirical investigation b y anthropologists has indeed revealed that people apply disparate semantic models in their categorizations of kin statuses (e .g . , Sanday , 1 968 ; Wexler and Romney , 1 972; Brown , 1 97 6) . An analysis of the models of collaterality in the Phoenix study has demonstrated that each model expresses a unique component in kinship classification . These components

are described in the last column of Table 9-2 ( i . e . , componential emphas is) . The Canon Law component stresses gradients of kin from Ego; the Genetic component is intennediate between the Canon Law and the C iv i l Lav.· models; the Civil Law component is based upon a minimal lineal core consisting of Ego , his parents , and his children ; Standard American component emphas izes the role of lineality in categorizing kin; and the Parentela Orders component places relatives in classes based upon their l ines of descent . This brie.f review of the componenti al analysis indicates that the kinship models differ in their emphasis upon distancing , lineality , and generational differentiation . But it al so reveals that e ach model of collaterality produces additional information about dimensions applied in the structuring of Amer­ ican kinship ties . The first po int to be made in the di scussion of the components concern s the Civil La w model . As Chapter 2 indicated , the Civil Law model eme rged

210

as a form of accommodation between the Plebians and Patricians in formu­ lating the Twelve Tables when the Roman Re-public was founded. In the Phoenix study , the Civil Lav.' model is overre-presented where accommodation betwe-en social classes might be most prevalent-in the lower-middle cl ass . This class is characterized by the following combination of attributes ; (a) families of clerical, sales , and craft workers (Tables 5-7 and 5- 8); (b) high school graduates who have not attended college (Table 7- 1 0); (c) non -mi ­ nority families with a relatively high income (Tables 5-4 and 5-6); (d) pers on s in Protestant denominations , particularly Mormons and Neofundamentali sts (Table 5 - 1 ); and (e) individuals with a history of frequent church attendance (Table A- 1 ). As for kinship structure , the data disclose that the Civil Law model represents conventional vie\\·s we normally associate with relative ly affluent , stable lovler-middle-class family life: (a) intergenerational stability in social status (Table 5-9) ; (b) low prevalence of marital breakup (Table 7 6); (c) utrolocality in marital residence (Table 8-2); (d) a high percentage of religious intennarri ages that remain � ·mixed . , afterwards (Table 7-3) ; (e) a tendency to disapprove of premarital cohabitation (Table 7 -14); (f) a trddition of maternal employment�o maintain income. level (Table 6-7) , and (g) the maintenance of strong ties with close relatives (particularly siblings) living in Arizona (Table 8-8). These findings suggest that there is a distinct pattern of kinship prevalent in lower-middle-class American society , which differs from both the upwardly mobile professional and managerial class and the lov.'er blue-collar class . This mode of kinship organization , structured by the principle of an abbreviated lineal core (i . e . , Ego, parents, and children) , seem to have escaped the attention of sociologists and anthropologists study­ ing modem kinship. A second point e.voked by the data in the Phoenix study, especially by the uncovering of the Standard American model, pertains to the application of the concept of generation in American kinship. Wexler and Romney ( 1 972) have identified generational distance of relatives from Ego as a major di­ mension in the semantics of American kinship tenninology. But unlike Wal­ lace and Atkins ( 1 960) , they have dismissed the concept of ascending versus descending generations as a central cognitive principle in partitioning kin . Instead , the diagram provided by Wexler an d Romney gives priority to gen­ erational distance from Ego over the ascending -descending distinction . Sup­ porting Wallace and Atkins , the componential analysis of collaterality models shows that the Standard American and Parentela Orders models (each of which has its own nonns and social correlates) assign differential weights to Ego 's kin in ascending generations as opposed to those in descending gen­ erations . Since the Standard American model is the most prevalent one in the Phoenix study , the omission of this cognitive dimension from a listing of primary principles in the semantics of American kinship terms appears to be a serious deficiency .

21 1 A BRIEF EPILOGUE: MODELS OF COLLATERALITY AND SECULAR I�ORTALITY

In the past dec ades , the American family has undergo ne m any change s at all stages of the l ife course . Yet , of the nonnati ve sh ifts that have occurred, most have taken place among young adults-those persons u nd er 40 . The se change s have incl uded fluctuation s i n (a) age at frrs t marr i age , (b) the pop­ ularity of marri age , divorce , and remarri age , (c) the preval e nc e of one-parent families and concomitantly young parents living apart from th ei r children , (d) the entrance of married women in childbearing y e ars into the labo r force , (e) the decline in ferti lity , and (f) the widespread phe- no menon of hou seholds consi sting of unmarri ed co uples (Glick , 1979) . Many pe-opl e h ave interpreted this proliferatio n of alternative life styles as a b roadeni n g of normative options o ver the life s p an . Although these opt i on s have obvious consequences that follow people into their old age , nevertheles s , they derive from c hoices made i n youth . Despite the l o ng - ru n effects of these early decisions , the ideas that people hold about kin ship and its si gn ifi c ance seem to shift as they mature . Data from a 1 976 Kansas City study , which includes the elderly , supplement the

fmdings of the Phoen ix invest igat io n in discl o s ing a tren d in patterns of kinship orientation . Table 9-4 shows the age distribution of kinship models in a survey of the Kansas City Jewish co mm un ity . The table describes a dramatic drop in Geneti c and C anon La\\' responses after 40 , V.'ith a corre­ sponding increase in Parentela Orders and Standard American c onc eption s . Whereas the percentage for persons with a Paren tela Orders orientation is similar to that of perso ns classified as Gen etic or Canon Law (about 25 percent) in the age cate gory of 40 or under , there is a large gap between them in the 65 or over age grou�35 percent for Parente la Orders and only 1 2

Table 9-4. Kinship Orientation of Respondents in 1976 Kansas City Study of Jell·ish Community , by Age

Age of respondent

40 years

Kinship Orientation

or under

Parentela orders (% )

27 . 3

Standard american (%)

36. 4 . 92

.81

Ratio to total

Ratio to total Civil law

Ratio to total Genetic model and canon Rati o to tolal

l'

12. 1

law

1 . 10 24. 2

1 . 52 33

40- 64

65 years of over

35 . 3

34 . 6

1 .05

1 . 03

38.8

43 . 1

.98

1 .09 9.8

1 1 .2

. 89

1 . 02

1 1 .8

15.3

33.5 39.6 1 1 .0 1 5.9

. 74

. 96

51

98

Data recomputed to accommodale Standard American model . See Farber

Total

( 1 979)

1 82

for descri ption of lhe study :

212

percent for Genetic and Canon Law . Thus , the data for Kan sas City e xtend th e trends apparent in the Phoenix study on the relationship between age and model of coli aterality . But what is the significance of the age trend for social policy? The findi ng s in Chapter 7 on children 's ages and residence (Table 7- 1 1 ) s uggest that stages in the life course have little impact on kinship orientation . But the data for both men and women do show that as individuals ripen into later maturity . they inc reas i ngl y regard kinship as a secular means for perpetuating them­ selves-symbolically as well as biologically . (See Table 7-9) . It is pl au si ble that the trend over the life course toward the Parentela Orders conception of collaterality expres ses this continually growing concern with secular immor­ tality . While the data permit alternative interpretations , they do not rule out sec u l ar i mmortality as an explanatory concept. (Cf Craig , 1979; Farber, 1 97 1 on symbolic family estates ; and Needham , 1 974 , on definition of ukinship . ") The age sh i ft in kin s hi p orientation toward an increasing concern with secular i mmortal ity evokes que s tions about the adeq uac y of bas i n g fami l y po l icies upon a s i n g l e kinship perspecti ve . As the predom inant conception , the Standard American paradigm probably should be applied to establi sh c ri te ria for defming familial rights and obligations . Characteristics of Stand­ ard Americ an kinship are in accord with current trends in public pol ic y . Consistent \vith the Standard American model, proposals have been made by legal scholars and students of public po l icy that: (a) in intestacy Ia�· , the entire estate should devolve upon the surviving spouse , with its res i due even­ tually trickli ng downward to the children (Table 4-8 ); (b) in ferti lity , small families should be encourged by law and regulatory agencies (Table 7- 1 6); (c) i n educational an d occupational matters , active partic ipation of married women should be facilitated by an extension of affirmative-action and child­ care programs (Tables 7- 1 0 and 7- 1 2); (d) fmally , measures should be taken to make early marriage difficult (Table 7-8) . Thus , insofar as there is a mo v em en t toward a comprehen sive family policy in the United States , it appears to follow guidelines consistent with the Standard American model . But there is another problem . One of the principles buttressing recent legal vie�·s is that the law should not i m pose a single family life style upon the entire pop u l ation Instead , it should admit a wide range of religious and ethic al systems in its application . Despite this libertarian principle , however, the data of the Phoenix study suggest that the Standard American model most readi ly accommodates a Protestant, upper-middle-class population segment and serves religious , ethnic , and rac i al minorities least . (See Table A- 1 . ) Less than half of the Phoenix sample conforms to th is kinship perspective . Therefore , one can construe use of the Standard American model as a bas is for social policy as favoring the family norms and values of one religious group and one social class over others . �

.

2 13

The problem to be solved in public policy is how to accommodat e a di ­ versity of kinship ideologies associated with age , religion and elhni c ity w hen

(a) the dominant kin shi p ideology

itself is overrepresented in ·a singl e religiou s

and socioeconomic clas s and (b) even the least restrictive version of kin sh ip i deo logy itself is expressed by a particular kinship model (i . e . , Canon Law coll aterality) held by only a small minority in the society . A dedication to pl ural ism logically involves granting a measure of legal autonomy to re li­ gious , racial , and ethnic communities-a state of affairs that the Enlighten­

ment dispelled. Perhaps , all that can be hoped for is a continued dialectic between pluralist and universalist doctrines in the creation of eclectic publ ic

policie s , a dialectic that would involve cognizance of the full range of concern with secular i mmo rtality in kinship ideologies .

Appendix A

Tabular pre se ntation s of percentages are useful in that they communicate information in a straightfor� ward manner with minimal loss of important detail s . Multiple Unlike more complex forms o f analysis , they pennit Regression much flexibility in sel ecting subsamples for display ; they accommodate nonlinear relationships re ad ily � Analysis and they are easily comprehended . Their greate st deficiency is that they pl ace limits upon the extent to which the interacti on s among a large series of variables can be taken into account in an analysis . Consequently , use of tabul ar analysis ba.� on percentages may produ ce spurious i nterpretations of a set of data . In order to avoid this danger , it appears advisable to supplement the tabul ar analyses of percentages with a multi variate procedure . Thi s appendix c onsi sts of three sections . First, the procedure follov.'ed i n the multivariate analysis i s described . Then , the findings are discussed , and finally conclusions are offered . THE MULTIVARIATE PROCEDURE

A stepwise multi ple regression analysis was undertaken to monitor the inter­ pretation s yielded by the tabul ar analyse s in the variou s chapters of the book . The regression procedure weights the contribution of each independent vari­ able on the. basis of a linear rel ations h ip to the kinship orientations , and it accommodates a l arge number of independent variables in computation . The regression analysis thereby allows for simultaneous treatment of many vari­ ables �·hose direct infl uence on one another may not have been an ticipate d Table A- 1 reports the results of four regression analyses--one for each kinship orientation ( with Genetic and Canon Law treated as a s i ngle cate­ gory) . The dependent variable in each case is a dummy v ari able in which a given kinship orientation is assigned a value of one and al l others are gi ven a zero value . A single set of independent variables is used for all four anal­ yses . Inasmuch as the main purpose of the regression analysis is to avoid spurious interpretations regardi ng the rel ation sh i p bet�·een the kin ship ori ­ entations and the soc ial structure . questions pertaining to opin io ns or subjec­ tive states pertai ning to family and kinship are omitted . The items selected as independent variables thus con sist of the more objective types of infor­ mation presented in the tables of the monograph Except \\the re dummy vari.

.

215

abies have been constructed , the range of categories for these items is de­ scribed in those tables . The variables in the regress ion analyses incl ude the fol lowing: A.

Social and demographic background variables 1 . Occupational status of male co- head of household 2 . Educational level of respondent

3. 4. 5.

6. 7.

8. 9.

B.

Family income in 1 977 Minority-group status Religion : du mmy variables for General Protestants , Refonnation Era Protestant denominations , Pietistic Protestant denominations , Neofundamental ist Protestants , Catholics , Mormons , and Je\\'S Size of community in which respondent was raised Was respondent raised in Arizona? Age of respondent Male respondent = 1; female respondent = 0

Family and Kin of Orientation 1 0 . How well the respondent knew each grandparent while groVw·ing up : 1 Oa . Father 's Father 1 Ob . Father 's Mother lOc . Mother 's Father l Od . Mother 's Mother 1 1 . Nativity of Father 's Father (U . S . 1 ; Other = 0) . 1 2 . Were parents divorced? 1 3 . Number of respondent ' s sibl ings 14. Is mother living in Arizona? 1 5 . Actual contact vlith mother 1 6 . Actual contact Vlith father 1 7 . Desired contact with mother -

1 8 . Desired contact v,rith father C.

Self and Family of Procreation 1 9 . Is this the respondent 's first marri age? 20 . Is respondent now divorced? 2 1 . Number of children born to respondent 22 . Frequency of church or synagogue attendance 23 . Importance of religion to respondent

An exami nation of Table A - 1 indicates that the Beta coefficients , as well as explained variance (R-square) , are small . The use of dummy variables as the dependent variables tends to attenuate both the regres sion coefficients and the explained variance . Moreover, the incorporation of 32 independent vari ­ ables (inclu ding dummy variables) into the equation itself has a depressing effect on the individual Beta coefficients , though not on the explained vari­ ance . However, the purpose of this regression analysis is not to produce a

216

statistical model but to i n s ti tu te a check for spurio usness of in terp ret at i ons based upon the i ndi vu du al tables . Consequently , my concern is on l y whether those items with the larges t Betas (arbitrarily tho se exceeding . 05) correspond to the factors that I have concluded are i mpo rtant for kinship orientations . The re gre ssion anal y s is utilizes the 772 cases in the Phoe n ix study . As po in ted out in Appendix 8 , the criteria for inclusion in the sample are (a) permane nt residence in the c i ty of Phoenix , (b) 1 8-45 age range , and (c) nov.' or prev iously married . There are t wo subsamples : one of them , consisting of 723 responde nts , was obtained by use of prob abil i t y sampling procedures , and the other 49 cases are Jewish persons in a convenience sample In order to retai n all 772 case s in the regre ssi on analy ses , when data for spec i fic items are m iss i ng for any individual res po n dent , mean val ues for the total s ample are app l i e d . .

FINDINGS Table A- I presents standardi zed regres s i on coefficients (Betas) in stepw ise mu ltiple re gre s sion equ ati on s where the k inship orientations are the dependent variables . Only those Betas of .05 or over are displayed and those items for which there are no Betas of that magnitude are omitted . The in depe ndent variabl es are classified under three rubrics: (a) social and demographic back­ ground variables , (b) family and kin of orientation , and (c) self and fam i ly of procre at ion . Social and Demographic Background

The Beta weights in the multivariate analyses reveal the fol low ing results with regard to the responde nts ' sociaJ and demographic background: 1 . By t hems e lve s , se- veral i tem s hav e l ittl e influence o n kin sh ip ori entati on . Those variables with no Beta weig hts of .05 or over include: (a) family income , ( b) Catholic religio n , (c) size of community in which the respon den t was raised , and (d) whether or not the respon dent was (& sed in Arizona . Their i n teract io n with other v ariab les in the analyses d i s s i pate their effect on ki n shi p orientation . In p arti cu l ar , much of the variance in k i ns hip orientations e xpl ai n ed by family inc ome is reduced by correlat ions with edu c ati on (r = . 36) , oc c upati o n (r = . 33) , Jewish re l i gion (r = 2 6 ) , minority group s tat u s (r = - . 24) , and currently divorced (r = - . 24) . 2 . The data in Table A- 1 indicate that d iverse kinds of variables are as­ soc i ated \vith the vario us orientations . But this d i vers ity is consistent w ith interpretations in the text . Tab le A-2 , which compares the res ul ts of the multiple re g res s ion an al y ses with the tabular presentations , shows a strong conirrrnation of the tabular findi ng s based o n percentages . The major revis ion in interpretation pertain s to family i ncome and m i nority-group status . .

217

Table A-1. Standardized

Regregion Coefficients (Beta)

of

.05 or Over in

Multiple Regre.ion Equations, with Kinship Orientations as Dependent Variablesc.

Independent variables

Social and demographic Occ upational status

Parentela

Standard

Ci vil

orders

american

law

background variab1es

Educational leve1

Minority-group status

- . 06 1

. 060 . 073

Religion Refonnalion Ent Protestant

Pietistic Protesant Protestant

.046

. 049

- . 04 5

Monnon (LOS)

. 056

Jewish

. 078

Age Male gender

canon law

- 060 - . 1 34 .

. 1 26

.049 .054

General Protestant

Neofundamentalist

Genetic or

. 062

. 058 - . 074

- . 070

.079

- . 085

- .06 1

Family and ki n of orientation

How well respondent knew

grandparent while

Father 's father

growing up:

Father 's mother Parents divorced

Number of siblings

- .05 1 .

Mother living in Arizona

05 2

.099

- . 1 08

Actual contact

Desired contact with father

.079 - . 067

098

. 080 - . 046 -

.057

Desired contact with mother

.

.080

. 050 - .047

with mother Actual contact with father

-

. 1 24 093

.

- . 048 - .

05 5

. 095

Self and family of procreation Is this the respondent 's frrst

marriage?

. 053 - . 067

Is respondent now dh'orced?

Number of chi ldren born lo

.076 - . 1 16

respondent

Frequency of church or

synagogue attendance

. 1 12

Importance of religion , to

- .068

respondent

Multiple R

.23

. 25

.2 1

. 30

R- square

. 05

. 06

.05

. 09

•

Only those independent variables with at least one Beta of . 05 or over are listed i n lhis table . All smallel' are omitted . The complete variable list used in the stepwise regression analy�is appe ars in tile leJtl of Appendix A. EJlcept for dummy variables, the categcries of tbe indcpendeut variable s are tbo� found in the various tables throughout the moaosraph. Kinship orientations are dummy variables deri ved from the classificalion procedure described ia Chapter 4.

Beta weights

2 18

TABLE A-2.

Comparison Tabular

of Results Presentations

of

Multiple Regression Analyses

S pecifi c

Overrepresentati on ( + ) or

analysis: Kinship

underrepresentation ( - )

Relevant

of \'ariable: Sign of

table

regression

orientation

Variable

with

Compari son with

beta

in text

(+)

5- 1

Confirms

(+) ( -)

5- 1

Confinns

table

SOCIAL AND DEMOGRAPHIC BACKGROUND VARIABLES

Parentela orders

Jewish Neofundamentalist Protestant Monnon Age

Male gender

Racial minority-group

{-)

status

Standard american

Occupational

status

Educational level

Protestantism: Refonnation-Era General

Pietistic Age Civil

law

Pietistic Protestant

or

c anon Jaw

Confinns

7-9

Confmn s

7-9

Confmns Clarifie s

S-5

ambiguity

(+) (+)

4-5

Confmns

4-5

Confums

( +)

4-5

Conftnns

(+)

4-5

( +)

4-5

(-)

7-9

( -)

5-5

Conrann s Conftnns

Con rums

Racial minority-group

status Educational level

Genetic

(+) (+)

5- 1

R ac ial mi nority-group status

Male gender

(-)

7- 10

(+)

5- 1

(+)

5-5

Con firms Confirms Co n firms Confirms C onfi rms

(-)

7-9

( ... ) (-)

6-8

(-)

8-7

Confirms

(+)

5- 1 2

(-)

6-8

Confmns Confirms

(-)

F.� MIL Y -� ND KIS OF ORIE� TA TION

Parentela

orders

Number of siblings Knew father ·s mother

Desired contact w ith father

Standard

american

father of siblings

Knew father 's N umber

Mother living in Ari zona Contact with father

Civ il law

Mother l iv i ng in

Arizona

Contact with mother

5-12

•

Confums Clarifies

8- 1

Contirms

( +)

8-7

Confirm s

(+) ( -)

8- 1

Confinn s

8-7

Confirms

(co1flirwed)

219 TABLE A-2 (continued) Specific


regression analysis :

( + ) or underrepresentation ( - )

Relevant

Comparison

Kinshi p orientation

of variable : Sign of bela

table in text

with

I\ umber of siblings Parents divorced Contact with mother

(�)

6-8

Confinns

6- 1 8-7

Confirms Confirms

Contact with father

(-) (-)

8-7

(+)

5- 1 2

Di sconfums Confirms Clarifies

(+) ( -)

8 -7

( +)

7-6

Clarifies

( -)

7-6

Confmns

( +)

7- 1

Clarifies

Genetic or

cannon law

Variable

Knew father ·s father Knew father·s mother Desired contact with: Father

�1other

( +) ( -)

5- 1 2

table

C lari fies Confi rm s

8-7

SE L F A N D FAMILY OF PROC REATIO�

Parentela orders Standard

american Civil law

Is thi:s respondent 's first marriage ? Is respondent now divorced ? Frequency of church attendance Importance of religion to

respondent Geneti c or

canon law

(-)

Not reJX>rted

Is respondent now divorced ? Number of chi ldren

(+) ( -)

7-6 7- 1 5

Confmns Confmns

• The following variables did not yield any Beta coefficienh of .OS or over: (a) Family iocome iD 19n; (b) Size or community in which respondent was raised; (c) Wu respondent raised in Arizona'?; (d) Respondent's religion is Catholic ; (e) How well respondent knew Mother 's Father while growing up; (f) How well respondent kaew Mother's Mother while growing up; and (g) Nathily of Father's Father.

Whereas the analyses based on Tables , 5-5 and 5-6 yields the conclusion that income outweighs minority-group status in affecting kinship orientations , the regres sio n analysis demands the reverse judgment . Family and Kin of Orientation

As shov.'n in Table A- 1 , the regression coefficients for items referring to parents , siblings , and grandparents indicate that:

2 20 I . When interactions among variables are taken into account, nati v ity of patern al g nmd father and hav i ng kno wn maternal grandpare nts are inconse ­ qu en t i al for the dev e lopmen t of kin shi p orien tati ons . 2 . The co mp ari son s in Tabl e A-2 shoYl that the reg ress i o n a nal ysi s tends to c on firm the tabular interpretations of the role o f the fam ily of orientati on in the deve lopment of kinship i deo logy . In addition to su pporting the per­ vasive findings that n u m ber of siblings is associated with ki n sh ip ori e nt ati on , the regre ss i on findi ngs also reveal negative weights for p atern al kin in Par­ en te l a Orders ( in contrast to Standard American kinship ) and a strong female skew in kin ti es in the Ge ne ti c and Canon Law model s .

�If and Family of Procreation

The Beta weights in Table A- 1 c l ari fy the fol lowing tendencies about the respondents and their fami l i es of procre ati on :

I . Pare nte l a Orders and Standard American model s , with one e xception , do not have an y reg re ss ion coe ffic ien ts ex c eedi n g . 05 among items pertaining to fami lies of p roc reati on . 2 . The Civil Law model is characterized not on ly by high marital stability but also by frequent church attendance (thou gh no t rel igios i ty ) . 3 . The data on Genetic-Canon Law e mph as i ze the prec ari ou snes s of marri age in these ori entatio ns . 4. Table A-2 indicates that the family-of-procreation in terpretati ons based on the tables in the text are g enerally confrrmed . CONCLUSIONS Overall , the data provided in Tables A- I and A-2 strongly buttress the inter­ pretations i n the text . Moreov er , the fi nding s of the regres s i on analyses tend to clarify some of the ambigu iti e s of the tables that report perce ntages . For e xample , in contrast to the tables , the regre s sion equations show that (a) mi nority-group status is more salient than income in infl uenc.ing kinship ori ­ entation ; (b) Civil Law ori en tatio n is ex pre ssi ve of associational tendenc ie s in rei i g ion ; and (c) there m ay be some resentment of the frequent contact among Genetic-Canon Lav..· res ponden t s \vith their mothers .

Perh aps t he majo r contribution of the regression an al y se s is re vealed in the distribution of B etas of . 05 or o ver among the different c ategories of inde ­ pendent variables . In Table A-3 , the Beta coeffients display ed in Table A- I are c l a ssi fi ed as either Social and Dem ograph i c ( S D) or Family Variables (F) (which i nc l u de s the items pertain i n g t o fam il y of o ri en t at ion and to family of proc reat ion) . The bottom row of Table A-3 p re sents ratios of the nu mber of ' 'high ' ' Betas in the Social and Demographic category to those in the

22 1

Table A-3.

Coefllcients .OS or Over for Variables Classified Demographic Background or as Family Variables, by

Number or Beta as

Social and

Kinship Orientation

Standard american

Civil

Genetic or

orders

law

canon law

6

7

4

3

3

s

5

11

2 .00

1 .40

. 80

. 27

Parentela Classification of variables

Soci al and demographic (SD)

Family variablesy (_F)

Ratio 8

(SD)/(F)

Includes \'Miables classified as FamiJy

aDd Kia of Orientation

and

as

Self and

Family of Proc�ation.

Gamma = . 5 1

Family V ariabl e s category [(SD)/(F)] . As one proceeds along the col um ns from Parentela Orders to Gene tic Canon Law , the size of t he ratios decreases regu larl y . My interpretation is that this decrease expres ses a shift in the comparative influence of pluralistic (in contrast to universalistic) ideologies . Parentela Orders and Standard American models reveal effects on ki nship -

org ani zat ion derived fro m class-membership in the larger institutional struc­ ture of soc iety (e . g . , religion , socioeconomic grouping) , whereas the more Ego-centered kinship orientations are shaped by a communal outlook and they stress the character of familial interactions . Th u s , the overall pattern of Betas provides additional support for the theoretical po s i ti on taken here .

Appendix B Method of Data Collection Morris Axelrod and Edward A . Greenberg

This appendix describes the method by which the data were collected for the Phoenix kinship study . In order to limit the population to those person s for

whom the questions were most me an­ ingful in terms of the immediacy of their experience , the target group consisted of residents of Phoenix , Arizona , who had been marri ed and who �·ere now between the ages of 1 8 and 45 . These restrictions were instituted ( l ) to preclude asking elderly respondents abou t their parental household during adolescence and (2) to avoid querying persons about marri age and family norms when they had never been married . The 772-case sample , yielding the data upon which the anal ysis is based , consi sts of two components-a probability sample of Phoenix and a supple­ mentary sample of Jewish hou seholds . The frrst part of this appendix de­ scribes the technique by which the probability sample was sel ected . The sample frame consi sted of all households within the city limits of Phoenix that ( I) are served by a telephone and ( 2) contain at l east one permane nt re sident , between 1 8 and 45 years of age , inclusive , Ylho is either no w or has once been married . Respondents were selected in two steps : First , a random-digit-dialing (RDD) te lephoning procedure was applied to screen for eligible households . Next , with in each hou sehold that contained at least one person eligible for a hou sehold interv iew , a respondent was randoml y selected from among all eligible persons . Since the probability sample would yield too few cases for ethnic and religious comparative analysis of Jewish house­ _ holds , it was decided to augment the sample with cases dra'-'·n from an available list of Jevlish households . Telephone screening procedures simi lar to those used for the probability sample were followed to select cases for the supplement. The second part of the appendix depicts the procedure for se­ lecting respondents in the Je wish supplementary sample. The final section discusses the field procedures applied in the telephone screening and hou sehold interviews . It touche s on the variety of interviewing techniques , requ ired to el icit the data for analysis , and presents an enumer­ ation of topics included in the questionnaires .

223 PHOENIX PROBABILITY SAMPLE

To eliminate ineligible persons , we screened potential respondents by tele­ ph one before inve sting heavily in hou sehold interv iew s . Acc o rding to the U . S . Cen sus , 84 percent of Phoenix hou seho lds are served by at least one te l ephone (United States B ure au of the Census , 1 970); so the biasing effects of non-subscribership are assumed to be minimal . (Cf. Tuchfarber and Klec ka , 1 976. ) A ddi ti o nal l y , the use of random-digit-dialing (ROD) alloY�·s us to reach unlisted and unpublished as well as listed tele p hone numbers .

The telephone informant at each randoml y selected number was asked several Telephone Contact Quest ion s to determine whether the number served a re sidence in Phoenix . If the telephone number was associated with a Phoe­ nix residence , a Telephone Screening Interv iew was u sed to ascertai n the

c om po s it io n of the ho use ho ld and the eligibility of the house ho l d for incl usi on in the final sample . In addition , several attitudinal items of curren t interest were inc luded in the telephone interview . These served to gain the cooper­ ation of the informant , establish rapport , and prov ide some topical data about the community . The fol l owing sections report the sample selection proc edure and the anal y s i s of re sponse rates .

Sample Selection Procedure The te l eph one sampling procedure was based on a design s ugge sted by Waks ­ berg ( 1 976) and elaborated on by Groves ( 1 977) . This technique cons ists of dividing the entire list of po s sib le telephone numbers in a c ommu n i ty into clu sters , choosing a sam ple of clusters , and then determining which of these

clusters are likely to contain h ou seho lds with specified characteristic s . If the first number called in a clu ster is eligible, chances are that the cl uster contains a greater-than-average proportion of eligible households , and that c l u ster is retained in the sampl ing frame . The probabi l i ty that a clu ster is kept in the sample is directly propo rtional to the number of e l i g ib le households in that cluster (Waksberg , 1 976). The procedure thus raise s the percentage of pro­ ductive calls above that yielded by a simple random sample . The c l us ter tec hnique was applied in the fo ll o wing manner: for Phoenix , the total hypothesized telephone coverage was divided into cluste.rs of I00 n umbers each . Particular clusters were then selected for in c l us i on in the sample as a result of a call to a single randomly selected telephone number in each cluster . If the number reached was as sig ned to a residence in Phoeni x , the entire clu ster was considered to be in the sample and n additional numbers within the cluster �·ere called . (In th i s study , n = 5.) If any of these were non-sample , they were replaced by other numbers from the same cluster until

2 24 n e l ig i b le

h ou se hol ds are identified . The s pec i fic steps in the selection of the sample are graphic all y presented in Figure B- 1 an d are detailed below . Step 1 . Determining the range

of telephone numbers . The c ity of Phoenix 63 different prefiXe s (the frrst three digits of the

is served by telephones with telephone numbers) . These i nclude 46 prefiXe s specifically design ated as Phoenix prefixes an d 1 7 prefixe s that joi n tly serve Phoenix and neighboring communities . If e very number for each of these prefi.Xes were in use , there would be 1 0 ,000 numbers for e ach prefix (e .g . , 555-0000 through 555-9999) . Thus , there are 63 x 1 0 ,000 = 630 ,000 pos sible Phoenix telephone numbers .

Of co urse , not all of these numbers are i n service . In addition , many of the W?rki n g numbers are as signe d to telephones outside the Phoenix city limits or to non-residenti al teleph o n es . Step 2 . Division into clusters . The next step in the telephone survey was

to detennine which of the 630 ,000 pos si ble numbers are l ike ly to be both in

Phoenix and residential . Each "bank " of 1 0 ,000 numbers was di vided into 100 groups , or clusters , of 1 00 nu mbers each . The nu mbers 555-0000 through 55 5-0099 constitute one such cl uster , as do the numbers 55 5-0 1 00 through 555-0 1 99 , and so on . T here are 6300 c l u sters of 1 00 numbers each in our popul at i on of tel epho ne numbers . Step 3 . Selection

of

cluster sample . We w ro te

computer program to generate a random number from each cluster. That number was d esign ated the ' 'primary · ' nu m ber for its c luster . At random , 2750 of the 6300 primary numbers were selected for contact. If, on the basis of the Telephone Contact Question s , a primary number w a s found to be both in Phoenix and residential . the cluster was i nc lu ded in our sample . If the pri m ary number was ascertained to be not i n service , out of Phoe ni x , or non-residential , the entire cluster Vlas droppe d from the s ampl e . a

Step 4 . Contacting primary numbers . The primary numbers were u se d

onl y to detenn i ne whether a cluster remained in or was excluded from the sample (via the Telephone Contact Questions) . Telephone Screening Inter­ views were not taken at the primary numbers . The 2750 selected primary numbers were called d uring a seven-week peri od from October to December , 1 977 . Of these , 975 primary numbers (35 . 5 pe rcent ) were in Phoenix and residential ; 1 775 numbers (64 . 5 percent) were non-residenti al , not in service , or outside Phoenix city limits . Step 5 .

Selection of numbers for Telephone Screening Interviews .

We es ­

timated that about half of Phoenix households would contain a respo nd e nt

225

Step

Retained

1 . Prefixes servi ng PHX Numbers per prefix

Numbers serving PHX

2.

Numbers per cluster

Clusters in PHX

3 . Clusters screened

4.

In-Sample clusters

(ln = PHX , residential)

S . Clusters selected for

Telephone Scree ning

Eliminated

63

X 1 0,000

630,000 + 1 00 6,300

� - ---l

2 .750

3 ,550

(43 . 7�)

(56. 3%)

� - -- �

97S

(35 . 5 %)

1 , 775

Non-Sample clusters

(64. 5%)

� - - -1

406 (4 1 . 6%)

Unused clusters

569

( 58 . 4%)

Unused cl usters

Interviews Telephone numbers selected for contact

in each cluster

6.

x5

Telephone sample n

2,030

Com pleted interviews

1 ,826

7 . Households containing an eligible respondent

� - - -,

(90. 0%)

204

( 1 0 . 0%)

�-- - �

916

(50 . 2� )

907

(48 . 8%)

�- --1

8.

Address ascertained

868

(94.8%) 9. Selection of a respondent in each household

I 0. Completed Household Interviews

Figure 8-1.

Non-lnterv ie\lf· s

4S

Households co ntaining

no el igible res pondent

Addres s not ascertai ned

(5 . 2%)

868

�---

716

( 8 2. S% )

1 52

Non-Interviews

( 1 7 . 5%)

Disposition of cases in selection of Phoenix probability sample

226

eligible for a Household Interviev.· ( i . e . , 45 or under and ever-married) . Our proje ct ed sample size was 800. Allov.'ing for nonre sponse in both te l ep hone sc reen i ng and household interviewing , we felt it necessary to scree n about 2000 telephone numbers . We decided to take Telephone Screening Interviews at five ' 'secondary ' � numbers selected from each cluster retained in the sample . We t here fore needed to select about 400 clusters from the 975 clusters (from Step 4) to yield 2000 numbers for screening . In fact, 406 clusters were s elected , for a sample of 5 x 406 = 2030 telephone numbers serving residences in Phoenix . Step 6. Telephone Screen ing Interviews . A total of 1 826 Telep hone

Screening Interviews were taken from N ov ember, 1 977 through May , 1 978 , w �en a 90.0 percent response rate was achieved. Duri ng this period , another 1 844 numbers were ascertained to be non- working , non-residential , or outside the Phoenix city limits . Each of these 1 844 nu mbe rs was replaced v.·ith a randomly chosen number from the same cluster. Step 7. Determining eligibility for home inten·iew . In order to determine

whether a family member was eligible for inclusion in the Household Inter­ view sample , we asked the telephone informant during the Telephone Screen­ ing Interview to report the household com pos iti on by age , sex , marital status , and relationship to the informant. If the ho u sehold contained a person be­ twee n 1 8 and 45 who was ever marri e d , it was considered eligible for a household interview . Of the 1 826 hou seholds at which a Telepone Screening Interv iew vlas taken , 9 1 6 contained at least one eligible respondent. The remai ning 9 1 0 households contained only persons who were over 45 years of age or who had never been marri ed. Step 8 .

Ascertaining address of eligible

Step 9 .

Selection of respondent .

households . If

a

ho u sehold con­

tained a person eligible for a Household Interview , the telephone informant was a ske d to provide a name and an address to which a letter outlining the project and requesting an interview could be sent . Addresses were. obtained for 868 of the 9 1 6 eligible households (or 94 . 8 percent) . A few informants declined to divulge their names, but they were willing to provide an address to which a letter could be sent . The next step was to select

respondent randomly from among e l igibl e persons in each of the 868 households for which we had an address . The procedure used was developed by Kish ( 1 949) . According to this technique , for every household , eligible adults of each sex are numbered in o rde r of decreasing age . A series of 1 2 selection tab les is used to designate a respondent in e ach h ou sehol d , so that every eligible adult a

227

i n each household has an equal chance of being chosen . At the beginning , one table i s chosen at random , and the tables are then used in sequence with each succeeding household . Step 10. Interviel1-'ing the respondents . After the selection of respondents , the cases were assigned to household interviewers . The interviewers were not pennitted to substitute other household members for designated respondents; instead , they were instructed to make as many callbacks or appointments as feasible within the field period in order to complete. the interv iew with the respondent-designate . The last page of the Telephone Screening Interview served as the cover sheet for the household questionnaire . It contained a listing of household membe rs (including the designation of the sel ected respondent) , the surn ame of the family (i f given) , the telephone nu mber , and address . When each completed interview was received , the cover sheet was removed from the household questionnaire in order to ensure confidenti ality of the interviews . Of the 868 households for which we had an address , household interviews were completed at 7 1 6 . Response Analysis

Response rates to the Telephone Screen ing Interviews and household inter­ views are summari zed in Table B - 1 . The response rate for the probability sample is a product of three stages in the interv iew i ng procedure . The first factor affecting the response rate was the sample loss due to failure to obtain a telephone screening interview . It is obviously not possibl e to determine whether those persons who declined at this point were eligible for a household i nterview . The Telephone Sc reen in g Interview (T) response rate is: T

C om pleted Interviews

=

=

Completed Interviews + Non-Intervie.w s

1 826

2030

- =

90 0% ·

The second factor affecting the response rate was the ability to obtain an

address of househ olds found to hold an eligible respondent. The addres s (A) response rate among households containing an eligible respondent is: Households with addresses

868

94 8% 916 El i gible households The final factor governing the response rate was the res pond e nt ' s decision to take part in the Household Interview . This decision can be evaluated on the basis of two different denominators: (a) households for which addresses are known (Ht) and (b) households with eligible respondents (H2) . A

= - =

·

228

is:

The response rate for households for which addres ses were obtaine d ( H a ) H. _

Completed Household Interviews

=

Households with addresses

7 1 6 = 82 S % 868 ·

The eligible Household Interview response rate (H2) equals the num ber of household interv ie w s completed divided by the number of household s in the sample that contain eligible persons , that i s : Completed Ho usehol Interviews 2 H = . Households wtth elig1ble persons

�

=

716

=

916

78 . 2%

An overall , adj usted Household Interv iew response rate incorporates the response rate for H ousehold Interviews and the response rate for the Tele-

Table B-1. Suney Response Analysis

Probability sample

Stage

Stage 1 :

Response category Sample

Jewish supplemenl

II :

Stage 1:

Stage II:

Telephone

Household

screeni ng

interviews

Telephone

scree ning

Household interviews

2030

916

425

58

(91 6)

(7 1 6)

(58)

(49)b

n

Interviews

Eligible for household interviewsa

Ineligible for household

interviews

Total

Non-interviews Non-sample Response rate.. •

( 1 1 8)

(9 1 0) 1 826 204 J 844d

200

one persoo

between the ages oi l S

49 9

149

. 7 82

.900

Households conrailli.Jll • least

1 76 1 00:

716

.63 8

. 845

aod. 45 BDd �\'er married .

taken io l he Jewish household supplemeatary sam ple were disoova-ed to be non-- J ewish. were used to replace non-interviews aad were combined with tbe 7 16 inten·iews from tbe probllbility sample to increase the number of cues to 723 ror analysis The seven added cases �·ere excluded from the computation of response rates. " Se-.·en inraviews

These seveo

.

phase of the study i��eluded 44 households Vw·hich were found to be eligible ror a addition . tbece were 86 for wbicb eligibility was not ascertained (EN A). It is assu med that the proportioo of non-sample households to sample househo lds is tbc same for ENA cases as for the �maiDder of lhe sample . Thus , 3' percent of the EN As (30 households) were cormdered 10 be DOn- sample and were added to die lmown non-sample c.ses (30 + 1 1 9 1 49) . Similarly, 6S percent of lhe ENAs (56 boliSeholds) were treated as non-inlerviews and were .sc:ted to the actual non-irder'Yiews (44 + 56 100) . Sillce moe t of the EN A.s are probably in facl DOD-sample , the non-illlerv icw estimate is a conscr.·ative one .

c

Non·iaten·iews for

this

household interview. lo

=

=

d In the random-dqpl dialinB sampling procedure,

Pboeni Jt aad'or oon-rcsidential) is replaced by r

Response rate

a

a

telepbooe

number claaifJed

as

non-sample

new randomly-selected number from

Completed iaten•iews . Completed ioterviews + Non-i ater.'te\\' S

.

the same

(i .e. ,

outside

cluster.

229

phone Screening Interviews . The adj usted Household Interview response rate ,

which take s into account telephone screening non- interviews , withholding addresses, and house hold non-interviews is the product of the Telephone Screening Interview respon se rate (T) and the eligible Household In terv i ew response rate (H2) , or: T X H2

=

90 .0

X

78 . 2

=

70 . 3 %

JEWISH SUPPLEMENTARY SAMPLE

Previous analysis has shown a relationship between the Pare ntela Orders model and Judai sm (Farber, 1 977) . For this reason , it w as considered des ir­ abl e to select a supplementary samp l e of Jewi sh respondent s to be inter­ viewed . This supplemen t was necessary to augment the smal l proportion of interv ie\\ S '-''ith Jews in Phoen i x yielded by the t e lephone probability sampl e (approximately 2 percent) . The probability sample produced fev.·er than 20 Jewish households , too few for analy sis . We elected to supplement the prob­ ability sample by s el ec ting Jewish households from available lists . The response analy sis for the Je wi sh Supplementary S am ple is described in Tab l e B - 1 . We began v;ith a li s t of 425 names of potentially eligible households ( i . e . , containing at least one Jewish , ever-marri ed perso n bet ween the ages of 1 8 and 45) . Telephone numbers were available for some of the names on the list � but not al l . For those whose telephone numbers \\'ere not prov i ded , other sources (in particular, the Phoenix Telephone Di rectory) were used to a sc ertai n the n um ber when possible . For the S upp le me ntary Sample , telephone numbers '-'·ere ob tained for 3 39 of the original list of 425 households . These numbers were te l epho ned and the Te lephone Screening Interview w as administered . Of the 3 3 9 numbers , Te l ephone Screening I nterv iews were completed for 1 76 . One hundred of these num bers yielded non-intervie\\'S (refusals , incomplete interviews , ad­ dre s s not ascertained , etc . ) , and 1 49 numbers were found to serve non sampl e households (outside Phoenix , non-residential , or no n J ewish ) . Of the com­ pleted Telephone Screen ing I nterviews , 58 yield ed households that contained at least one eligible respondent who met the age , marital status , and religiou s criteria . Household interviews were taken at 49 of the 58 eligible h ou se ho l d s (or 84 . 5 percent) . Although the Jewi sh Supplemen tary S ample was selected on the basis of availabil ity , it is believ ed to be reasonably representative of the kn own Je\\·ish community . An additional adj ustment was made to the set of completed Household Intervie w s after the data collection for the supplementary sampl e w as com­ pleted . Seven interviews taken at households selected from this sample frame did not contain an eligible Jewish respondent . These intervie ws were taken with responden ts w ho we re in most respects s imilar to the Phoe n ix prob ab ili ty '

-

-

230

sample with respect to religion , marital history , family size, and other ch ar­ acteristics . These interviews were retained as replacements for non- interv i e\\·s in the probability household sample . In summary , the final set of comp leted household interviews consists of: Probability S ample Jewish Supplementary Sample Non-Jewi sh from supplementary sample Total Intervie\v

7 16 49 7

772

FIELD PROCEDURES

The interviewing for the Phoenix kinship study was done in two major stages : (a) Telephone Screening Interviews v-·ere conducted with a sample of Phoenix households selected by random digit di aling ; (b) face-to-face intervie�·s were conducted with selected eligi ble respondents in their homes . H ou se ho l d in ­ terviewing took place from January throu gh July , 1 978 . On the average , the interviewing staff consisted of about 20 persons at any one time . Because the telephone i nterviewing was done on our premis·e s , we had close , continuous contact bet\lleen project staff and interviewers . Th is contac.t extended interviewer instruction i nformally on a regular basis after the in itial training periods , and it pennined the interviewers to gain a heightened un­ derstanding of the total project . As a result , they were well equipped for the subsequent household interviewing . The fonn s used in the household interviews were intended to elicit infor­ mation in a variety of areas , and several techniques were applied . The house­ hold questionnaire was rather complex in that it used charts , self-admini stered , forms , " 'map games , .. and many sets of questions that were contingent upon marital statu� , number of children , number of siblings , and so on . For the most part , however, the interview consisted of a straightforward face-to-face oral administration of a questionnaire , with the interv iewer recording the responses on the form. The orally administered sectton s pertained to materiaJ on the following : a. Respondent 's and spouse ' s children-vital statistics, religion , reside nce � and marital and parental status . b . The respondent 's brothers and sisters-vital statistics .. rel igion , and mari­ tal and parental history . c . Teen age values-a series of questions aimed at ascertaining the role ex­ pectation s in the respondent �s family of orientation . d . The respondent 's parents-vital statistics , marital history , relig ious back­ round , educational and occupational backroun d .

23 1

e . Views about family and society-41uestions on topics relevant to family relationships ( including views on abortion , unmarried couples living to­ gether , obligations to relatives , religion and the family , and comparative loyally to home versus work) and the Srole Anomia scale . The respon­ dent 's speculations about the spouse ' s views were also solicited . f. Respondent 's and spouse 's background-demographic information and marital histories . A second series of questions was included in a self-adminis tered written

fonn on closeness of kin . Responses to these questions were used to clas sify respondents according to kinship orientation . This form also referred to ac­ quaintanceship with relati ves prior to adulthood . A third segment of the interview involved a card-sorting technique . In this segment , the respondent pl aced cards , on which were written specific rela­ tives , on maps to indicate where kin w ere born and where they no�· live . In another series of card sorts , the respondent showed how much contact there i s with these relatives and how much contact is desired . The Household Interviews averaged about an hour i n length . De spite the apparent complexity , we were able to develop a questionnaire form that was virtually self-guiding for the interviewer and proved to be effective in eliciting the necessary information with l ittle error or confusion .

References Ackerman, Charles 1 963

Affil iations: Structural Dete rm in ants of Differenti al Di vorce Rates . A m e ri ­

can Journa l of Sociology 69 : 1 3- 20 .

�.\dams, Bert N . 1 968 1 970

Kinship in A n Urban Setting . C hicago : Markham .

Isolation, Fu nctio n , and Beyond : American Kinship in the l 96o ·s . Journal

of Marriage and the Family 32: 57S- 597 .

Alsop, Stewart 1 968

The New American Family-1 1 : I t \\'asn �t Born Yes te rd ay . Sarurday ning Post 24 1

,

issue 1 5 , J u ly 27 .

Eve­

Altheim, Franz 1 93 8

A History of Romnn Religion . London: Methuen and Company .

Anderson, 1\lichael 1 97 1

Family Structure in �Vineteenth Century Lancashire . New York: Cambridge Uni versity Press .

Aristotle 1 942

The Student' s

Oxford Ari.�totle . Edited by

W . D . Ross . New York: Oxford

University Pres s . Vol . V , Ethics .

Arizona 1 975

A rizona Revi.\·ed StaJutes . St. Paul:

�rest Publishi ng Company . (Pocket Sup­

plement , 1 979) .

Atkins, John 1 974

R.

On the Fund amental Consangui neal Numbers and The ir Structural Basi s .

American Ethnologist I : 1 - 3 1 .

Augustine 1 966

The

City of God Againsr the Pagans . Cambridge , Massachusetts : Harvard

University Press . Vol . 4 .

Babchuk, Nicholas

1 965

Primary Friends and Kin : A Study of the Associations of Middle-Class

Coupl e s . Social ForceJ 43: 483-- 493 .

Bahr, Howard 1976

M.

The Kinship Role . In Ivan F. Nye , ed . , Role Structure and Analysis of the Fan1ily . Beverly Hills, California: Sage Publ ications . Pp . 6 1 - 79 .

2 34 Beale ,

Howard

1 964

K.

Fami ly Culture a nd Genealogy . In Edward N . Saveth , ed . , A merica n His­ tory and the Social Sciences . Nevl York: Free Press .

Blau , Peter 1 964

Exclulnge and Po�'er in Social L{(e . New York: \\'iley .

Bogardus, Emory 1 959

S.

Social Distance . Los Angeles: Pub l i s hed by author_.

Brown, Cecil H. 1 97 4 Ps ychologica1 , Semantic , and Structural Aspects of American Eng I ish Kin­ s h ip Terms . A merican Ethnologist I : 4 1 5- 436 . 1 976 An Examinati on of the Ordinary Use of American English Kin Terms and K in Te.nn Bo u nd Fonn s : · �semantics " as Necessary Meaning . A nthropo­ /igica/ Linguistics 1 8 : 1 29- 1 56.

Buchler, Ira R. and Henry A. Selby 1 968

Kinship and Social Organ ization . New York: Macmillan .

Burgess , Ernest W. 1 973

On Community . Family, arul Delinquency . Chicago: Uni versity of Chicago Press .

Burgess, Ernest W. , Haney J . Locke, and Mary Margaret Thomes 1 963 The Fam ily : From Institution to Companionship . New York: American Book Co . Burgess , Ernest W. and Leonard S. CoUreU 1 939

Predicting Suceess or Failure in Marriage New York: Prentice Hal l .

Burgess , Ernest W . 1 953

Engagement and Marriage . New York: Lippincott .

Burr , Wesley 1 973

and Paul WaDin

R.

Theory Co nstr uction and the Sociology of the Family . New York : \\'iley .

CampbeD, Bruce L . and Eugene E. CampbeU 1 976

The Mormon Famil y . In Charles H. Mindel and Robert W. Habenstei n �

eds . , Ethnic Families in America . New York : Elsevier. Pp . 379- 4 1 2 .

Carr,

Lestie G .

1 97 1

The Srole Items and Acq uiescence . A merican Sociological RevieK' 36 : 2 8 i293 .

Casson , Ronald W. 1 973

An Equivalence Rul e Analysis of Ameri c an Kin ship Terminolog y . Anthro·

polog ica l Linguistics 15: 1 89- 202 .

Chodorow, Stanley 1 972

Ch ristian Political Theory and Church Poli tics in the �'fid-TMle/fth Century.

Berkeley and Los Angeles : Universily of Cal i fornia Press .

Church Educational S)'stem 1 976

Achieving a Celestial J\1arriage . Salt Lake C ity: Department of Se min aries

and Institutes of Rel igion , Church of Jesus Christ of Latter-day Saints .

23 5

Coser, Lewis 1 974

Greedy Institu tions: Patterns of

Press .

l..lndi�·ided

Commirment . New York : Free

Craig, Daniel 1 979

Immortal ity through Kinship : The Vertical Transmission of S ubs tance and Symbol ic Estate . American Anth ropologist 8 1 : 94-- 96 .

Cruz-Coke, Ricardo 197 7

�A. Genetic Refonn o f the C iv i l Law Computation of the Degrees of Rela­

ti onsh i p in Fami l ies. Social Biology 24: 93- 99 .

Danby , Herbert (translator) 1 933

The Mi.-.hnah . f'ew York: Oxford U n iv ers i ty Press .

Dizzard , Jan 1 968

Soc ial Change in rhe Fan1ily . Chic ago : Comm unity and Fami l y Study Ce n ­ ter , Cniversity of C h ic ago .

Durkheim, Emile 1 91 5

Elementary Forms of Religious

Durkheim, E mile an d Marcel Mauss 1 96 1

Soci al Structure and the Structure of Thought . In Talcott Parsons , Edw ard Shils. Kaspar Naegele , and Jessee R . Pitts , eds . , Theories o.l Society . New York : Free Press .

Ehrenberg, Victor 1 946

Life. New York : Macmillan .

A spects of the

Pp.

A ncient

1 065- 1 068 .

World. Nev.· York: Will iam S alloch .

Eisenstadt , Shmuel N . 1 977

Soci o logical Theory and an Analysi s of the Dynamics of Civil izations and of Revolutions . Daedalus 1 06 : 59- 7 8 . (Discoveries and Interpretat ions : S tudies in Contemporary Scholarship , Volume II . )

Ekeh, Peter 1 975

Social

Excha11ge Theory and rhe T"'o Sociological Traditions . Cambridge �

�1assach usetts: Harvard University Press .

Epstein , Isadore ( ed.) 1 935

The Babylonian Talmud . London : Soncino Press . Volume 2 .

Farber, Bernard 1 964 1 97 1 1 972

1 973

Family : Organiwtion and In teraction . San Franci sco: Chand]er. Kinship and

ClaJ�i:

A Midtvestern Study. Kew York: Basic Books .

Guardians of Virtue: Salem Families in 1800. Nev.,· York: B asic Books .

Family and Kins hip in ��odern SO£'iety. Glenview. Dlinois : Scott, Fores-

man .

1 97 5

B il ateral Ki nship: Centripetal and Centrifugal Types of Organization . Jour­

nal o.f Marriage and the

1977

Family 37:

Social Context, Kinship !\�lappi ng , and Family Norms . Journal of Marri age

and the Family 39: 227- 240 .

1 979

87 1 - 888 .

K i n ship Mapping Among Jews In A M idv.·estem City . Social Forces 57: 1 1 07- 1 1 23 .

236 Farber, Bernard , Charles H . 1\olindel , and Bernard Lazenvitz

1 976

The Jewish �A.merican Family . In Charles H. Mindel and Robert w· . Ha­ benstein . eds . , Ethnic Families ;, Amen'ca . New York : El sevie-r. Pp. 347378 .

Feng, Han-yi 1 937

The Chinese Kinship. Ha rva rd Journal of Asiatic Studies 2: 1 4 1 - 2 74 .

Firth , Raymond 1 956

TK'O Studies

of Kinship

in London . London : Athlone Press .

Fortes, Meyer

1 969

Kinship and the Social Order. Chicago: .4Jdine .

Fox, Robin �

1 967

Kinship and .�arriage . Baltimore: Pengui n .

Freedman , Ronald , Paul K. \\belpton, and Arthur A. Campbell

1 959

Family Pllln ning, Sterility . and Population

Hi l l .

Grol�'th . New York: MacG raw­

Glazer, Nathan and Dan iel P . 1\foynihan 1 974

Why Ethnicity . Commentary 58: 33- 39 .

Glick, Paul C .

1 979

The Future of the American Family . Current Population

Studies . Series P-2 3 , No . 7 8 . U . S . Bureau of the Census .

Reports : Special

Glock, Charles V. and Rodney Stark 1 965

Religion and Society in Tension . Chicago: Rand McNally .

Gluckman, Joel

1 976

R.

Intestate Success ion in New Jersey: Does It Confonn tations? Columbia Journal

Goffman, Erving

1 956

Presentation

of the

of La'"'·

To

Popular Expec­

and Social Problems 1 2: 253- 294 .

Self in Everyday Life . Edinburgh: Social Sciences Cen­

ter , University of Ed i nbu rgh

.

Goode , \Vill iam J. 1 960

A Theory of Role Strai n. American Sociological Re1"iew 25 : 483-496 .

1 962

Marital Satisfaction and Instability : A Cross-culrural Class Analysis of Di­ vorce Rates . International Social Science Jou,;,al 1 4 : 307- 526 .

1 963

World Revolution and Family Patterns . New York.: Free Press .

1976

Production and

Good�·, Jack

Reproduction .

Cambridge: Cambridge University Press .

Gordon, Michael 1 977

Primary-Group Differentiation in Urban Ireland . Social Forces 55 : 743-752.

Greeley, Andrew

1 976 1 977

M.

The Communa l Carholic . A Persona/ .Manifesto . New York: Seabury Press . The A merican Catholic . A Social Portrait . New York : Basic Books .

23 7 Groves, R . M. 1977

An Empirical Compari son of T\\'0 Telephone Sample Design s. Unpublished

report of the Survey Research Center of the Univers ity of M ich igan , 1 977 .

Harrison, Alick R. W. 1 968

The Law Press .

of A thens, Fam ily

and Property . New York: Oxford Uni vers ity

Haskins, George Lee 1960

Law and A uth o rity in Early J\fassachusetts . New York: �tacmillan.

Health Resources Administration 1 974

Teenagers : Marriages , Divorces , Parenthood , and Mortal ity . Vital a1ui

Healrh Statistics . Series 2 1 men t Pri nting Office.

, Number 23 .

Washington , D . C . : U . S . Govern­

Heitland, W . E . 1923

The

Roman Republic . Cambridge , Eng la nd:

Cambridge Cniversity Press .

Herberg , \\?ill 1960

Protestant, Catholic, }eYt/ . Ne w York: Anchor.

Hertz, J . H . (eel.) 1 960

The Pentateuch and Haftorahs . London: Soncino Press . Second Edition .

HiD , Daniel 1 977

Labor Force Participation Decisions of \Vives. In Greg Du ncan and James

N. Morgan , eds . , Five Thousand American Families - Patterns of Ec on om ic

Progress , Volume V. Ann Arbor: Institute for Social Research , University

of Michigan.

Horst, Paul (ed . ) 1 94 1

The Prediction

searc h Council .

of

Personal A djustment . New York : S oc i al S cience Re­

Ho�·t, Homer 1 933

One Hundred Years

C hi cago Press .

of Land

Values

in

Chicago . Chi cago : C ni v ersi ty of

Huebner, Rudolf 1 968

A History of Germanic Private La�1--· . New York : Augustus �1 . Kelley Pub­ lishers . (Published originaUy in 1 9 1 8 . )

JotTe, Natalie F. 1 949

The Dynamics of Benefice htmo ng Easl European Jews . Social Force.s 27 : &

238- 247 .

Johnson, E.L.

1 969

An Introduction to the So viet Legal System . London: pan y .

l\1ethuen and Com­

Jolowicz, H.F. 1 967 Historical Introduction t o the Study of Roman La»' . Eng land: Cambridge University Press .

Cambri dge ,

238 Kerckhoff, Alan C .

Nuclear and Extended Family Relationships: A Normati ve and Behavi ond Analysis . In Ethel Shanas and Gordon F. Strieb , eds . , Social Structure and

1 965

the Family: Generational Relati(}ns . Englewood Cliffs , New Jersey : Pren­

tice-Hall . Pp . 93- 1 1 2 .

Kish, L .

A Procedure for O bj ecti v e Response Sel ecti on within the Household .

1 949

na/

of the A merican Statistical Ass(}ciatio!J 44: 38� 387 .

lor� r­

Klatzky , Sheila R.

n .d.

Pattern.\·

of Contact

'rvith Relatives . \\'ashington , D .C . : American Sociolog­

ical Association .

Kohler, Wolfgang �

1 947

Gej·talt Psychology . New York: Li veright .

Laumann, Edward

ri ll . 1 97 3

Bonds

of Pluralism :

�ev.· 'r"ork: W iley .

Leichter, 1 956

0.

Prestige und As:�ociarion in an Urban Community . New York: Bobbs - �1er­

1 966

Hope J .

The Form and Substan£·e of lirban Social JVern•orks .

and William E. MitcheU

Kinship and Ca.se K'ork. �ew York : P ussell S age Foundation .

Ge rha rd E. 1 963 The Religiort.\'

..

Lenski ,

Factor . New York : Doubleday .

Lenski , Gerhard E . and 1 960

John C . Leggett

Caste , Class , and Deference in the Re searc h Interview . A merican Journal

of SocitJ!ogy 65 : 463- 46 7 .

u,·i-Strau� , Claude 1 963

Structural A nth ropology . New York : Basic Books .

1 969

The Elementary Structures of Kinship . B oston : Beacon Pres s .

1 966

The Savage Mind . Chi c ag o : Uni'f·ersily of Chic a go Press .

Lewin, Kurt 1 936

Pri11ciples of Topological Psychology . New York : McGraw- Hill .

Lewis , Oscar 1 967

Further Observations on the Folk-Urban Continuum and Urbanization , \\'ith

Special Reference to Mexico Ci ty . Cited in Ell iot Liebow , Tally' s Corner .

pp . 204- 205 .

Liebow, EUiot 1 967

Tally' s Corner. Bo ston: Little . Brown .

Litwak, Eugene 1 960a Geographi c M obi l ity and Extended Family Cohe s ion American Sociolog­ .

ical Re viehl 2 5 : 3 85- 394 .

1 960b Occupational Mobility and Extended Family Co he s i o n Amen·can Sociolog­ .

ical Rel-·iew 25: 9- 2 1 .

239

1 965

Extended Ki n Relations in an Indu strial Democratic S oc iety . In Et hel Shanas

and Gordo n F. Streib , eds . , Social Structure and the Family: Generationa/

Re la tions . Englewood Cl iffs , Ne\\' Jersey : Prentice-Hall . Pp . 290- 32 5 .

Locke, Harvey J . 1 95 1

Predicting

Adjusunent

in Marriage: A Cotnparison of

a

Dit-·orced and a

Happily Married Group . New York: Holt.

Lopata, Helena Znaniec:ki 1 973

Widtnvhood in an America1J City . Cambri dge , Massachusetts : Schenkm an (General Learning Press) .

�tabnonides, l\1oses 1 963

1 967

The Guide of rhe Perplexed. Chicago: University of Ch ic ago Press .

J\fi.shneh Torah . New York: H ebre v.· Publish in g Company .

�taranda, Pierre 1 974

French Kinship, Structure and History . The Hague: Mou ton .

1\tcFarland , David D . and Dan iel J. Brol'·n 1 973 Social Di sta nce as a Met ric : A Systematic Introduction to Smallest Space

Analysis . In Edv.·ard 0. Lauman n , ed . , Bonds of Pluralisnz : The Fonn and

Substance of Urban Social 1Venvorlcr . �e \\' York : ( Appendix A).

\\ri ley . Pp . 2 1 3-25 3

Mlller, Delbert C. 1 977

Handbook of Research Design and Social Mea ... urement. Ne"'· York : David :f\,fcKay .

Model, Otto 1 964

Testamentsrecht. Berlin: C . H . Beck �she Verlagsbuchhandlung.

ModeU, John, Frank F. Furstenberg and Douglas Strong 1 97 8

The Timi ng of Marri age in the Transition to �"-dulthood: Conti n u i ty and

Change , 1 8� 1975 . In Jo h n Demos and Sarane S pence Boocock , eds Turning Poin u . Chicago: American Jourool of Sociology, S l 2a- S I 50 .

. •

Supplement 84:

Mommsen , Theodor 1 905

The History of RonJe. f\ew York: Charles Scribner 's Sons .

Morgan, D .H.J . 1 975

Socia l Theory and the FanJily. London: Routle dge & Kegan Pau l .

1 87 1

Systems of Con.\·anguin ity aru/ Affinity of the Hunum Family . Washi ngton �

1\torgan, Lewis Henry

D . C . : Smithsonian I nstitution .

Murdock , George Peter

New York:

1 949

Social Structure .

1 970

Cross -sex Patterns of Ki n Behavior. Ethnology 9: 3 5 9- 368 .

Naroll , Raoul 1 970

�1acmillan.

\\'hat Have \\'e Learned from Cross-Cu ltural S urveys'] American A nth ro ­

pologi!�t 1 2 : 1 227- 1 288 .

240

Needham , Rodney 1 974

Remarks and Inventions: Skeptical Essays about Kinship . Londo n : Tavi stock

Publications .

Paige, Jeffery l\1. 1 974

Kinship and Pol ity in Stateless Societies. American Journal of 80: 30 1 - 320 .

Sociology

Parsons, Talcott 1 949

The Social Structure of the Family . I n Ruth Nanda Anshen, ed . , The Family:

Its Function and Destiny. New York: 1 954

The Ki nship

ed . ,

Harper. Pp. 24 1 - 274 . System of the Contemporary United State s . In Talcott

1 77- 1 96 .

and lJarriage . Ne\\' York: Oxford

L"niversity

RadclitTe-Brown , A.R. and Dar)'ll Forde 1 950

African System s of Kinship

Parson s,

Pp.

Es.rays in Sociological Theory . New York : Free Pre ss .

Press .

Radin, Max 1915

The Jews Among the Greeks and Romans . Philadelph ia: Jewish Publication Society of America .

Rainwater, Lee

Poor

Get Children . Chicago: Quadrangle Books .

1 960

A nd the

1 960

The Extended Kinship System in The American Middle Class . Ph . D dis­

Reiss, Paul J.

se rtati on ,

Rbeinstein, Max 1 955

The Law

tion) .

Harvard Uni versity .

of Decedents• Estates . Indianapolis: Bobbs Merrill

Romney , A. KirnbaU and Ro)' 1 964

(Second

edi­

G . D' Andrade

Cogni tive Aspects of English Kin Terms . American Anthropologist 66: 1 4Cr 1 70 .

Rosenberg , George S . and Donald F. Anspach 1 973

Jtlorking Class

Rosenfeld, Jeffrey P. 1 974

Kinship.

Lexington , Massachusetts: Lexington Books .

.. Inheritance: ..A Sex-Related System of Exchange . In Rose Laub Coser, ed . ,

The Family, Its Structures and Functions . Ne�· York.: St . Martin 's

Pp.

Press .

400- 4 1 1 .

SabUns, 1\farshaJI 1 965

On the Sociology of Primitive Exchange . In Michael Banton , ed . , The Relevance of Models for Social Anthropology. Nev.' York: Frederick

A.

Praeger . Pp . 1 39- 236 .

Sandars, Thomas Collett 1 874

The Institutes of Justinian . London: Long lllaJlS , Green (Fifth

Sanday , Peggy R. 1 968

The · ·Psychological Real i ty " of America� Engl i sh Kinship

Edition) .

Terms: An

formation-Processing Approach . American Anthropologist 70 : 508- 52 3 .

In­

24 1 Saveth, Edward N .

1963

The American Patrician Class : A Field of Research . A me rica n Quarterly 1 5 (Summer Supplement) : 235- 252 .

Schlesinger, Benjamin 1 974

The Jew i s h Famil y and Religion . Journal of Comparative Fam ily Studies 5 : 27- 36 .

Schneider, David 1 968

M.

A merican

Kinship:

Cultural

A

Account .

Cliffs ,

Englewood

New

Jersey: Prentice-Hall .

Schneider , David

1 975

M.

and Calvert B.

CottreU

The American Kin Universe : A Genealogical Study . Ch i cago : Department of Anthropology , Uni vers ity of Chi c ago .

Schneider, David 1\1 . and Raymond T. Smith 1 973

Class Differences and Sex Roles in American Kinship and Family Structure . Englewood Cliffs New Jersey: Prentice-Hall . ,

Schulz, Fritz 1 95 1

Classical Roman LaK' . London: Oxford at the Claredon Press .

Selden, John 1 636

De Sui:cessionibus in Bona Defuncti! ad Legis Ebraeorum . London : Richard

B i shop . 1 640

.�'aturali

Ju re

De

Gentium ,

et

Juxta

Ebraeorum .

Disciplinam

London: Richard Bishop.

Shennan , C .Bezalel •

1 964

Demographic and Soc ial Aspects. In Oscar I . Janov.lsky , ed

. ,

The A merican

Jew : A Reappraisal. Philadelphia: Jewish Publication Soci e ty

Shils, Edward 1 97 5

Center

and Periphery,

Chicago Press .

.

Pp . 27- 5 2 .

Essays i n Macrosociology. Chicago : Universi ty of

Simmel, Georg 191 1

How Is Society Possible? American JourtUJl of Sociology 1 6 : 372- 3 9 1 .

Simon , Rita J., William Rau, and 1 980

Mary

Louis FeUows

Property Distributi on at Death . Social

Skolnick, Arlene S . and J erome H. Skolnick 1 97 1

Family

1 97 6

The

Smelser, Neil J . Hall

in

Transition . Boston: Little , Brov.rn .

Sociology oj· Economic

S mith, Charles Edward 1 940

Attitudes about Forces 58: 1 263- 1 27 1 .

Public Versus Statutory Choice of Heirs: A Study of Public

Papal

Enforcemen t

of

Life . Englewood Cliffs , New Jersey : Prentice­

Som e

...Medie1-·al

A-1arriage

La ws .

Baton

Rouge: Lou isiana State Uni versity Press . ( R e i ssued in 1 972 by Kennikat Press, !\ew York . )

242

Sorokin , Pitirim A . 1 959

Social

and Cultural Mobility . New York: Free Press .

Srole , Leo

Soci al Integration and Certain Coroll aries : An Exp l orat ory Study . A merican

1 956

Sociological Re�';e..,tt· 2 1 : 709- 7 1 6 .

Staples, IWI

Robert

The BUick Family , Essays and

Stone, Lal\Tence

Studies . Belmont , C alifornia:

Wadsworth

The Ri se of the Nuclear Fam i ly in Early �1odem England : The Patriarc h al

1 97 5

State .

In

Charles

E.

Rosenberg ,

ed . ,

The

Fam ily

Philadelphia: Un i vers i ty of Pennsylvania Press. Pp. 1 3-- 5 7 .

in

Hi... tory .

S ussman , Marvin B.

Re l ations h ip s of Adult Children with their Parents in the U nited States . In

1 965

Ethel Shanas and Gordon F. Streib, eds . . Social Structure and the Family : Generational Relations . Engle\llood Cli ffs , New Jersey: Prentice-Hall . Pp .

62- 92 .

S ussman , l\1arvin B .

and Lee G .

Burc:hinal

1 962a Kin Family Network : Unheralded Structure in C urrent Conceptional izations

of Family Functioning . J.,larriage and Family Living 24: 23 1 - 240. 1 962 b Parental Aid

to

Married Children: Implications for Fami ly Function ing .

Journal of J\larriage and the Family 24: 3 20-- 332 .

S ussman , Mar,yin B . , Judith N . Cates, and 1 970

David T.

S mith

Th e Family and Inheritance . New York : Russell Sage Foundation .

Swanson, Guy E. 1 96 7 1 969

R eligio n and Regime. Ann Arbor: Uni vers it y of �1ichigan Press .

Rules oj· Descent: Studies in the

Sociology

of Parentage . Anthropological

Papers, No . 3 9 . Ann Arbor: !vluseum of Anthropology , University of .[\lfich­ tgan .

Tedeschi , 1 966

Guido

Studies in Israeli Private La""· · Jerusalem: Kiryat Sepher . Lewis

Terman, 1 938

Psychological Factors in Marital Happiness . N�w York: McGraw-Hill .

Tucbfarber, A. J . and 1 976

KIMka, W .R.

Random Digit Dialing : Lowering the Cost of Victimization Surveys . Wash­ ington , D . C . : The Po li ce Foundation .

LTUmann, 1 955

Walter

The Gro""'·th of Papal Government in the Middle Ages .

Ithaca, New

York : Corne ll University Press .

1 975

Lat-.l and Politics in the Middle Ages . Ithaca, New York: Cornel l University

Press .

243 United States Bureau of the Census. 1 970

Census

of Housing :

Vol . 1 , Hous ing Characteristics for States , Cities, and

Counties , Part 4, Arizona . Washington , D .C . : U . S . Government Pri nting Offi ce . \Vaksberg, J. 1 976

Sampling Methods for Random Digit Dialing. Unpubl ished report . Rock­

ville , Maryland: Westat , Inc .

Wallace, Anthony F C and John R. Atkins .

1 960

.

1be Meaning of Kinship Terms .

Waller, Willard and Reuben Hill 1 95 1

Anthropologist 62: 58- 80.

The Family. New York: Dryden .

\\'atson , Alan 1 97 1

American

The Lal1l

of Succession

in the Later Roman Republic . London: Oxford at

the Cl arendon Pres s .

\\'eber, Max 1 96 1

General EcononJ ic History. New York : Coll ier Books .

\\'eigert , Andrew J . and Ross Hastings 1 977

Identi ty Loss , Family , and Social Cbange . 82 : I 1 7 1 - I 1 8 5 .

A merican

Journal

of SocioloRY

W"'estofl', Charles F . and Nonnan Ryder I fT77

The Co n tracepti ve Revolution . Princeton , Ne w Jersey: Princeton University Press .

\\'exler, Kenneth N. 1 972

and A . Kimball Romney

Individual Vari ations in Cognitive Structures . In A . Kimball Romney ,

Roger N . Shepard , and Sara Beth Nerlove , eds . , Multidimensional Scaling (Vol . 2 , Applicarion.f) . �ew York: Seminar Press . Pp . 73-92 .

Winch, Robert F. 1 974

S om e Observations on Extended Familism in the United States . In Robert

F. Winch and Graham B . Spanier, eds . , Selected Studie.� in Marriage and

1 977

the Family .

New York: Holt , Rinehart, and Winston .

Familial Organizarion,

Wypyski, Eugene M. 1 976

A

Pp .

1 47- 1 60 .

Quest for Determinants . New York :

Free Press .

The Law ofInheritance in All Fifty States . Dobbs Ferry , New York : Oce ana .

Index

A Ackerman, Charles. I 06 Adams, Bert N., 42, 83, 87, 1 56, 1 62, 1 74. 1 79, 1 83, 1 88 Affinity, 2, 4 1 Agnatic group, 2Pr-29

Alienated collectives, I 5

Alsop, Stewa� 55 American Indian respondents, 80.. 8 J

A mity, axiom of. 26, 1 58- 1 59, 1 67, 1 68- 1 69..

1 75, 1 96 Anderson, M ichael, 1 5 7, 1 60 Anomia ( Srole index), 96-98, 1 0 1 - 1 02 Anspach, Donald F., 42, 1 88 Aristotle� 25-26 Artificial family ties, 1 2 1 Assimilation into American society, 89-94, 99

7, 45, 46, Augustine ( City Q( God), 2�30 Atkins, John R. , 3, 5,

59, 67, 2 1 0

B

Babchuk, Nicholas, 1 67 Bahr, Howard M., 1 56

Beale, Howard K., 55 Bernard of Pavia, 3 1 Black respondents, n, 80, 82

Blau, Peter, 1 57, 1 60

•stood" a nd genealogical d i stance , 1 0, 1 5 Bogardus, Emory, 36 Brown, CcciJ H. , 208, 209 Brown, Daniel J., 36

Buchler, Ira R . , 59

Burchinal, Lee G. , I 56, 1 98

Burgess. E rnest V.'., 36, 37, 1 04, 1 29 Burr, Wesley R . , 83

Canon Law model centrifugal tendencies, 1 05- 1 06

exchange norms, I (]()a 1 6 1 permanent availa bility, 202-205 social setting, 1 7- 1 8, 20, 29-33, 34, 43, 70. 1 93. See also Collaterality models Carr, Leslie G., 1 0 1 Casson , Ronald W . , 208 Cat� Judith N., 59 Cath olic Church, 1 2

canon law, 30..32, 33, 42, 70, 95

Catholics, 7S-76, 8 1 , 95� 1 93- 1 94 contact with relatives, 68 � 1 70 divorce, I 07

familiarity with gra ndpa rents. 92-93. 94. 99

levels, 77-78

fertility, 1 1 (,.1 1 9. 1 4&- 1 53 i ncome

occupations, 84 religious e ndoga my, 72. 1 1 0, 1 27- 1 28 siblings' religion, I I ().. I l l

Ccntnlization of culture.. 1 5- 1 7

Centrifugal kinship tendencies. I 04- 1 06. I 54,

1 93- 1 94

age at marriage. 1 1 2, I 34- 1 36 common interests. 106, 1 1 9

marriage and

legal family model, 1 2 1

Ca mpbell, Bruce L., 73-74 Ca mpbell. Eugene E. , 73-74

family nonns. I OS

age at marriage, 1 1 2, 1 34- 1 35

marriage and

maternal role, I 1 5, 1 4 1 - 1 45

natural family model, 1 2 1 privacy, 1 25 special interests. 1 05, I 1 9

Child ren and collaterality. 1 44- 1 4 5 adoption, 1 39

c

fa mily norms, I 06

maternal role, l i S, 1 4 1 - 1 45 Centripetal kinship te ndencies� 1 6. 1 03- 1 05. 1 93

ages of children. I 39- 1 40, 204 mari taJ status of grown children, 1 40 maternal employment. 1 42 residence of children, 1 39

246 Chodorow. Stanley. 30

Ch urch Educational System. 74

Ci vil Law m ode l , 209-2 1 0 accom modation o f dass con flicts, I 59- I 60 social setting. 1 7. 27-29. 32-33. 34, 43, 70, 1 93. See a/:io Collaterality models

Collateral removal, 59-M Collaterality (definition), Preface, 1 -3, 1 00 Collaterality in Ia•;, 2- 1 0, 20, 42, .57-5&. 2 1 2 Collaterality models age and se" of res pond ents , 1 36- 1 38, 1 42- 1 44 age at marria� 1 1 2- 1 1 3, 1 34- 1 3 5 componential analysis, 60-67, 1 90-- 1 92 contact with relatives, 1 74- 1 85 description. 7- 1 2, 4849 disillusionment, �98 distribution in Phoeni� 50-5 1 , 56-51 familiarity with grandparents, 90-94 incom e, 77-82 interest in kin, 1 67- 1 69 marital residence� 1 63- 1 66 measures, 9 . 1 2. 46-50, 1 9 1 minorities, 77-82, 96 occup ations , 8� 96-98 religion, 72-76 religious endogamy, 126- 1 28 social mobility, 87-89, 98 social structure, 1 7-1 8, 40-43, I OS- I 06, 1 70, 1 92- 1 96 socioec ono mic status, 84-86, �97, 1 92 spatial metaphors, 42-43. See abo Canon Law, Civil Law, Gcneti� Parentela

Orders, and Standard American models Communal stateless societies, 1 3

Communal welfare (com mon interests, common welfare), 1 3- 1 4, 1 8 communalism, I 03- 1 04 gradient conception of social space, 39-40 kinship, l 06, 1 70 religion. 7 1 Componential analysis, 58-66, 1 90- 1 92, 1 96, 207-2 1 0 Consa nguineous marriage, 2 1 , 29-30. 1 09 Consa nguinity, 2-3, 4 1 Coparents-in-law, 1 09 Core a nd periphery in social structure, I � I S Corporateness in family and kinship. 1 66. 1 93- 1 96 collaterality models, 1 30, 1 32, 1 35, 1 45, 1 96 privacy, 1 23- 1 25� I 54 social exchange, 22, 24, I 57- 1 62, 1 82 Coscr, Lewis� 1 57, 1 59 Cottrell Calvert B. , 4 1 -42, 68, 7 1-72, 1 00, I 56. 1 79. 1 88 Cottrell. Leonard S. , 1 29 Cousinship, 3 1 , 54 Craig, Da niel. 1 5 5, 2 1 2 Cruz-C o ke . Ricardo. 10, I I

D

Danby, Herbert, 24 o·Andrade, Roy G., 59

De Gradibus Cog111lf ionis, 29

DisiDusio nmcnt with spo use, 1 3� 1 32

Divorce, 1 0� 1 06, 204

age at marriage, 1 34

collaterality, 1 07- 1 08, 1 20, 1 32- 1 3 5 Dizard, Jan, 1 29 Durkheim, Emile D., 3 7-38, 40, 7 1

E Ec ono mic depressi o n, 1 3 7

Ehrenberg. Victor, 24-25

Eisenstadt. Sh muel N 1 2- 1 4. 1 7 E ke� Peter, 1 82 Englis h ki nsh ip, 54 , 1 60 English law, 8- 1 0 Epstein, Isadore, 24 Ethnicity, 16, 1 8, 68, 75, 79. 82. 94, 1 04, 209, 212 .•

F

Fa ctionalis � 1 3- 1 4, 1 6, I OJ- I 06, I 07, 1 70. See

also Pluralism, S peci al interests

Family capitalism, S4

Family history and oral tradition, S> S6

Family life cycle, 1 37- 1 40, 1 44- 1 45 Family policy , I I , 57-58, 2 1 2-2 1 3 Family, symbolic aspects, SS, 1 54, 1 66, 1 68- 1 69

Farber, Bernard. pa.uim Fellows, Mary Louise, 58 Feng. Han-yi, 3

Fertility and collateraJity, 1 05- 1 06, 1 1 6- 1 1 9. 1 46- 1 53 close relatives of respondents.. 1 1 7- 1 1 9, 1 46- 1 48 project io ns by respondents, 1 49- 1 53 Firt� Raymond, 55 Forde, Daryl� 1 95 Fortes, Meyer, 1 57- 1 58, 1 67 Fox. Robin, 202 French kinship, 54 Furstenberg; Fra nk F. , 1 1 2

G

Generational distinctions in com ponents of

Genealogical distance, 4 1 -42, 68-70

models, 58� I

Genetic model centrifugal tendencies.. I 05- 1 06 �ociaJ setting. 1 8, 20-2 1 . 1 93. See also Collat­ eralit y m ode ls German kinship (J'Jogschafr), 1 09 Gestalt psychology, 34-35

247 Glazer, Nathan, 205 Glick, Paul, 2 1 1

conflicting kin ties, 1 59 divorce, 1 06

Glock, Charles Y., 1 25 Gl uckma � Joel R., 58

familiarity \\ith grandparents, 92-94

Goffina� Erving, 1 57, 1 6 1

fertility, 1 1 6, 147- 1 48

Goldber& Arthur, 1 2 3 Goode. Willia m J 83, 1 04, I S9

income levels. 78-79. 82

Gordon, M ichael , 83

models of collaterality, 43. n-73, 76, 95 occupations. 84

.•

Goody, Jack, 2 1

Gratia� 30

Greece, classical peri od, 1 2, 22. 2�27, 42. 70 At hens, 20, 2 1 , 24, 3 2, 42

intest acy Ia w, 26-27

significance of contracts, 2S

Greeley, And rew M . , 1 9.3- 1 94, 20S-206 G roves.. R. M. , 223

H

Harrison. Alick R . Yl., 25 . 26-27 Haskins .. George Lee , 1 2 1 - 1 22

H ast ings, Roa, 55

Health Resources Administration, 1 34 H eirs i n wills. 58

H eitland, V.'. E., 20., 27 H eJvetians, 4, 5

Heroerg. Will, 205 Hereditary associations, 54 He rtz, J. H 22 Hillt Da niel . I I S Hill, Re u ben , 1 29 Homans, George, 1 S H orst , Paul, 1 29 H o� Homer.. 36 .•

H uebner, Rud olf. S, 32, I 09

1

Idealization in courtship. 1 29 Imperial regimes, 1 3- 1 4 I mpression management, 1 6 1

I ntestacy laws,

Arizona, 6 Ceylon , 8

Georgia. 9 Israel. 6 J'i a poleonic Code. 8

familism, 68, 1 94

kin contact, 1 79

law, 6, 22-24, 33\ 82, 94

religious endogam y, 72, 1 09- 1 1 1 . 1 27- 1 28

siblings' religion, I J 1

j/etl life, I 09

J offe, Natalie F., 1 74 John de Deo, 32

John of Orleans, 3 1

Johnson, E. L., 20 Jolowicz, H. F., 28. 29

Justinian Code, 29 K

Kerckholf9 Alan C., 1 99 Kin contact, 204

distance9 83, 1 62, 1 70-1 7 1 . I 74, 1 79, 1 85 fertility. I 5 2 forced contact, 1 82- 1 85 increased contact desired� 1 74- 1 8 1 iso l ated nuclear family, 1 56 rcfuseni ks. 1 72- 1 73

sex d iffe rences.. 1 79- 1 8S

Kin universe, 4 1 , 68

Kinship continuity.. 23

American kinship, l , 69-7 1 , 99- 1 0 I . l 9S- 1 96

dimensions, 3, 4 1

and house hold family, I S4- 1 5S, I 56

kins hip grid, 1 87- 1 89

middle class � 1 97-20 1

nomenclature, 2, 4 1.. 54, 58, 1 95, 207-2 1 0. See

also Centrifugal kinship te ndenci�. Cen­ tripetal kinship tend enci es , and Collateral­ ity models

Kinship ties

confticting commitments, 1 58- 1 60 family roots , 1 68- 1 (I)

impressi on management, 1 6 1

R oma � 28-29

interest in, 1 66

West Germany, 6, 3 1

market no rm s. 1 60- 1 6 1 , 1 75

U nion of So uth Africa, 8

I nterviewing and questionnaires, 229-23 1 Isidore of Seville, 3 1

Israel, a ncient, 1 2, 22, 32-33, 42. 70 modem Jaw, 6

J Jews, 2 1 , 1 09. 1 94 concept of Covenant, 22

marital residence, 1 62- 1 66

norm of generosity, 1 57- 1 .58, 1 7 S

propinquity , 1 63

trust and obligation, 1 67- 1 68 types of social exchange, 1 57, 1 74, 1 86. See

al.w Intestacy law

Kish , Leslie, 227

Klatzky, Sheila R., 42, 66. 83, 1 62, 1 74. 1 88 Klecka, W. R., 223 K ohler, Wolfgang, 34-35

248 Mommsen. Tbeodor , 21

L

Laumann, Edward 0., 1 6, 1 67

Law

colonia� 1 2 1 - 1 22

contemporary, 1 23- I 24

Wenttadr VS. Baird, family, 1 2 1 - 1 25, 1 S4

1 24

Gruwo/d vs. Connecticut, 1 23- 1 25 PlanMd Parenthood of .�issoW'i vs. Danforth, 1 24- 1 25 Roe

vs.

R-'ade. 1 24

l .azerwitz., Bernard, 94

Leggett. John C. . I 0 I Leichter H ope J . , 1 59, 1 82 Lensk� Gerhard E., I 0 I , 205, 206 LePJay, Frederic. 1 97 Levi-Strauss, Claude, 23, 35. 3 8- 39, 40, 1 58 Lewin, K urt, 35 Lewis. Oscar. 56 L.iebow. Ell iolt, 56 ..

200

42, 83,

� Napoleonic Code, 8,

29

Narol� Raoul, 207 N atural farn ily mode� 1 2 1 - 1 2S, 1 35-1 36, 1 54 T\eedham, Rodney, 1 95 , 2 1 2 1\ew Engla nd, family roots, 54, 1 2 1 - 1 22 -

Legal-family model, 1 2 1 - 1 25. 1 3 � 1 36, 1 54

Li twak, Eugene.

Morgan, D. H. J., 1 6, 1 56 Morgan, Lewis Henry, 2-3, 4, 5, 4 1 � onnn oos , 73-74, 76, 95 Moynihan, Da niel P., 205 M urdock, Gerse Peter, 3. 4 1 , S9, 1 88

1 67,

1 88.

1 97, 1 98- 1 99.

Nicomachean ethi� 25-26

Nonrnarita1 c oh abi tati o n, 1 43-145

0 Oikos, 22, 24-25 Oriental respo nde n ts 80, 82 ,

p

L oc ke, Harvey J . , 1 04, 1 29

Paige. Jeffel')· M . , 1 3- 1 4, 1 7

Locke, John. 1 22

Pare ntage., degree of, 10. Se� also Genetic model

L o pata. H e le na Znaniecki, 1 56

Parentela Orders model

centripetal tende ncies , I OS norm of generosity, 1 58

1\1

social sett iq, 1 7- 1 8, 2 1 -27, 32-3 3. 34. 43,

M a imonides, M oses, 22, 23

49-50, 70. 1 92 Set also Collaterality models

M a randa. Pierre. 30, 3 1 , 54 M a rital roles. 1 4 1 - 1 45

Pa rs ons, Talcott, 37, 38, 1 83, 1 97- 1 98 ' 200.

�larriagc age at marriage, 1 1 2- 1 1 4, 1 34- 1 35, 204 e nd ogamy, 1 0� 1 07, 1 � 1 1 1 , 1 20, 1 26- 1 27 Judaic and Athenian svstems 2 1 legal i n terpretations, I 23- 1 2 5 ' levirate and sororate, I 07 mate selection and adj ustment, 1 29 remarriage and fertility. 1 5 1 . See also Consangui neous marriage M ate rnal role. 1 05., 1 06. l i S. 1 42. 145 employment 1 1 5, 1 42. 145 fertility projecti ons� 1 50 MalBS, !darcel. 37, 38, 40 McFarland, Da\oid D., 36 Mexican American respondents, 75-76, 80-82 Middleman minorities. 1 6 Miller, Delbert C . , 96 Mindel. Charles H ., 94 Minorities, 1 6, 52, 53, 66, 77-83, 95-96 97, ..

2 1 2, 2 1 4, 220

.�ishnah, 23, 24, 26, 95, 1 94

.Mishpokheh. 22, 1 09, 1 92, 195

Mitchell \\'illiam E., 1 59, 1 82

..

.

202, 203

Paterfamilias, 28 29 ..

Patrim onial regimes, 1 4, 32

Permanent availabi litv 202-205 Pluralism, 1 7, 20 1 ,

Xil'. See also Factionalism

Privacy., corporate and individual, 1 23, 1 25

Protestantism, 1 3, 5 1-54, 70

Protestants, 43, S I , 75, 80, 95

chW"Ch attendance, 1 25- 1 26. 2 1 0

divorce. 1 07

familiarity ¥tith grandparen� 92-93, 94

fertility, 1 1 6, 149

i ncome, 77-78

Neofundamentalists,

occupations, 84

72.

7S, 76, 82, 9S

religious endogamy, 109- 1 10, 1 27- 1 28

siblings' religion, I I 0-1 1 1

Public interests, 1 6- 1 7

Q

Questionnaire� 229-23 1

..

M odel, Otto. 6

M odelL John. 1 1 2

M oieties, J9

R

Radcliffe-Brown., A R., 1 95 Radin, Ma � 23, 27, 28

249 Social exchange. 1 57, 1 75 . Se� also Reciprocity

Ra ndom digit dialing, 223 Ra u. \Villiam, 58 Reiss, Paul J

.•

1 36

Reciprocity, 22, 25, 59� 1 57- 1 58, 160-1 6 1 � 1 7 5 Reformation, Protestant , 1 3

Regressio n analysis, 8 1-82, 2 1 4-22 1

Relatives as persons, ff}, 90, 92, 99- 1 00

genealogical technique, I 00- 1 0 I

Religion, 7 1 -76, 95, 1 97. 204

Social space categorical represe ntation, 35-36, 38-40, 1 92 concentric circle models,. 36, 37. 39, 1 92

core and periphery, 37-38, 39 gradient representation, 37-39, 40 metaphors, 34-40 special interests, 38-39 Socioecono mic status age and sex differences , 1 3 8, 144

age at marriage, 1 1 3

church attendance, 1 25- 1 26

age at maniage, 1 1 3- 1 1 4

fertility, 1 1 6- 1 1 7, 1 47- 1 48, 1 49

kin contac4 1 70

divorce� I 07 fertility, 1 1 6, 148, 1 49- 1 50

divorce, 107, 1 3 1 - 1 33 familiarity v.itb grandparents, 9 1 -92

sociaJ mobility, 87-89. See also Collaterality models

intermarriage� 1 26- 1 28 irreligiosity of spo ._, 1 30- 1 32

kin contact, 1 70, 1 79- 1 80

Sorokin. Pitirim A.,

occupations, 84-86

siblings, I I 0- 1 1 1 . See also Catholics, Jews, M ormons� and Protestants

Religi ous groups, associational versus communal tendencies, 75, 76, 205-2()7

Residence (ncolocality� virilocality, utrolocality�

Srole, Leo, 96

35-361

42

Standard A merica n model, 70. 98, 99. 1 96- 1 97. 200.20 1 centripetal tenden cies 105 ,

comparison with Parentela Orders model, 49-SO

middle class fertility, 147, 1 48- 1 49

uxorilocality), 162- 1 66 Response rates, 229

nativity of parents and grandparents, 5 1 .

Rheinstein, Max, 8- 1 0

Roman Republic, 7, 1:1, 32-33, 42, 70, 1 58, 21 0.

See also Twelve Tables Romney, A. Kimbal� 59, 209, 2 1 0

9 1 -94

norm of generosity. IS8 social characteristics, �52, 66 social placement. 56 source, 1 20 See also CoDaterality models Staples. Robert, n, 82 .

Rosenberg. George S., 42. 1 88 Rosenfeld, Jeffery P., S4-S5

Stark, Rodney, 1 25

Stone, La wrencc, 54

Ryder, Sorman, 1 1 9

Strong. Douglas, 1 1 2

Sun·eys on collaterality.

s

Arizona State University, 43, 72

Sablins, Marshall, 1 86 Sample selection procedure, 223-227, �2 3 1

SandaB, Thomas Collett, 19

Kansas City, 43 . 72, .206, 2 1 1 Phoenix, 2, .56-51. 222-23 1

Suss man, M arvin B. , S8, 156. 1 98

Sanday, Peggy R., 209

Swanson, Guy E . , 1 2- 1 4, 17, 7 1

Sa \ldh, Edward :'1 . , S5

Schlesinger, Benjamin, f8

Schneider, David M., 1 0, 40, 4 1 -42, 68, 69, 7 1-72,

77, 82, 96, 100- 1 0 1 , 1 2 1 , 1 5 6, 1 79, 1 83 , 1 88, 1 9S, 1 97, 199, 208 "Second Treatise of Civil Govern ment 1 22 Secular immortality, 21 1 -2 1 2 Selby, Henry A., 59 Selden, John. 5-6, 20 Shared chromosomes a nd kinship distance, I 0- 1 1 . Sn tllso Genetic model Sherman, C. Bezalel, 79 Shils, Edward, 1 4- I S, 1 7, 37-38 Si mon, Rita James. S8 Smelser, Neil J., 83 Smilh, Charles Edward, 20, 3 1 -32 Smilh, David T. , S9 Smith, Raymond T., t/J, Tl, 82, 96, 1 00. 200 Social distance, 3S-36, 40 .•

T

Tedeschi, Guido, 6

Terman, Lewis, 1 29

Thomes, Mary Margaret, 1 04

�Tree of Consanguinity� 3 1 -32 Tuchfarber, A. J., 223 Twelve Tables, 20, 27, 28-29, 42,. 70, 1 59, 2 1 0 ..

u

U Uman Walter, 7, )).3 1 U nited States Census, 82, 223 U niversalism, 18.,. 1 93. See also Communal welfare

w

Waksberg.. J., 223 Wallace, Anthony F. C., 59 2 1 0 ,.

250 Waller, WillarcL 1 29 Wallin, Pa ul, 1 29 \\tar, 1 14, 1 37 Watson, Ala� 28. 29 Weber, Max. 104

Weigart, Andrew J. , 55 Westoff, Charles F.� 1 1 9 \\'exler, Kennet h N . , 59., 209, 2 1 0 \\'inch, Robert F. , 68-69, 198- 1 99., 205 Wypyski, Eugene M 6 .•