Differential socialization of girls and boys in light of cross-cultural research

13
l%e settings for social deuelopment provided by a culture are an important but neglected aspect of socialization. Differential Socialization of Girls and Boys in Light of Cross-Cultural Research Carolyn Pope Edwards Beatrice BIyth Whiting Any review of current research on sex differences in children’s behavior suggests that there are three major competing explanatory systems for the origins of sex differences in social behavior: the biological model, which points to constitutional (usually hormonal) factors as the major source of sex differences; the cognitive-developmental stage model, which considers that girls’ and boys’ cognitive discovery of their gender identity leads them to choose different patterns of behavior; and the socialization model, which looks to the direct and indirect actions of socializing agents as the primary cause of sex role differentiation. The third model, predominant for a long time in the social sciences, has A report of the collaborative research of Beatrice Whiting, Carolyn Edwards, Lawrence Baldwin, Charlene Bolton, Ralph Bolton, Carol Ember, Gerald Erchak, Sara Harkness, Amy Koel, Carol Michelson, Ruth Munrce, R. L. Munroe, Sara Nerlove, Barbara Rogoff, Susan Seymour, Charles Super, and Thomas Weisner. New Directions for Child Dcuclopmmt, 8, 1980 45

Transcript of Differential socialization of girls and boys in light of cross-cultural research

l%e settings for social deuelopment provided by a culture are an important but neglected aspect of socialization.

Differential Socialization of Girls and Boys in Light of Cross-Cultural Research

Carolyn Pope Edwards Beatrice BIyth Whiting

Any review of current research on sex differences in children’s behavior suggests that there are three major competing explanatory systems for the origins of sex differences in social behavior: the biological model, which points to constitutional (usually hormonal) factors as the major source of sex differences; the cognitive-developmental stage model, which considers that girls’ and boys’ cognitive discovery of their gender identity leads them to choose different patterns of behavior; and the socialization model, which looks to the direct and indirect actions of socializing agents as the primary cause of sex role differentiation. The third model, predominant for a long time in the social sciences, has

A report of the collaborative research of Beatrice Whiting, Carolyn Edwards, Lawrence Baldwin, Charlene Bolton, Ralph Bolton, Carol Ember, Gerald Erchak, Sara Harkness, Amy Koel, Carol Michelson, Ruth Munrce, R. L. Munroe, Sara Nerlove, Barbara Rogoff, Susan Seymour, Charles Super, and Thomas Weisner.

New Directions for Child Dcuclopmmt, 8, 1980 45

46

recently lost ground, especially in developmental psychology. A recent review by Maccoby and Jacklin (1974), for instance, concludes that there is no strong evidence that parents consistently act differentially toward sons versus daughters. Although not all authorities agree with Maccoby and Jacklin (see, for example, Birns, 1976; Block, 1976; Brooks-Gunn and Matthews, 1979), the socialization explanation has tended to lose force while the other two models have received much more attention than before.

We believe that the study of socialization has been faulty because of its lack of an integrated cross-cultural perspective. This per- spective is required first, to broaden the foundation of empirical infor- mation in a crucial way and second, to offer a more complete analysis of the environments in which children learn than can be gained from working within our home culture only. In this paper we shall discuss how these two gains of the cultural perspective can greatly strengthen our understanding of the differential socialization of girls and boys. As an example, we shall offer a few of the findings of an ongoing collabora- tive cross-cultural project under our direction and funded by the Ford Foundation (Whiting, Edwards, and others, forthcoming). While our data do not address the validity of the biological and cognitive stage models, they do illustrate the advantages of a cultural perspective when conducting research on sex role socialization.

The first gain of the cultural perspective, broadening the data base, has been addressed by many earlier authors, including ourselves (Whiting and Edwards, 1973). Cross-cultural research provides a greater range of variation than does single-culture research and thus offers a way to gain leverage on antecedent variables. For example, cross-cultural research has suggested that children in U.S. households, relative to children in some other cultural settings, may show a reduced degree of sex difference in certain kinds of social behavior, notably aggression, nurturance, responsibility, and help and attention seeking. These findings were obtained in an analysis of the social behavior of 134 children aged three to eleven, who were interacting in natural set- tings in six communities- Taira (Okinawa), Juxtlahuaca (Oaxaca, Mexico), Tarong (Phillipines), Qhalapur (Uttar Pradesh, India), Nyan- songo (Kenya), and Orchard Town (in the New England area of the United States). (A detailed description of the methodology and results can be found in Whiting and Edwards, 1973; see also Edwards and Whi- ting, 1974; Whiting and Whiting, 1975). In the Six Cultures study, social interaction scores of boys and girls were compared for fifteen beha- viors of prosocial responsibility, dominance, dependency, compliance,

nurturance, and aggression. In each cultural sample, t-tests were per- formed to compare the boys and girls at two age levels- three to six and seven to ten. The sample from the New England community (called Orchard Town) showed the fewest number of mean differences in the predicted direction (17/30 versus 21, 23, 23, 23, and 24 for the other samples) and was also one of the samples that showed the fewest number of statistically significant comparisons (1 versus 0, 1, 2, 4, and 5; setting significance level at p< .05, two-tailed). Generally, fewer sex differences were found for the three samples in which boys and girls were treated similarly in terms of being assigned care for their younger siblings and being asked to do household and economic work. Those three samples were from communities in New England, the Philip- pines, and Kenya. In contrast, greater numbers of statistically signifi- cant sex differences were found for the three samples in which girls and boys were treated dissimilarly: Girls, much more often than boys, were asked to do household tasks, to take care of infants, or to do work that kept them close to home and in the company of adult females. Those three samples were from communities in India, Mexico, and Okinawa. Because the general American pattern in urban and suburban areas is for neither girls nor boys to be assigned many childcare, household, or economic tasks, we can expect the extent of sex differences in social behavior of American children to be minimized relative to children in many other cultures. Thus, in the present example, comparative re- search offered a way to obtain samples in which the extent of sex differ- ences was much greater than in typical American samples. The method of study provided leverage on the antecedent variable of task assignment.

The second gain of the cultural perspective, offering a more complete analysis of the learning environment of the child, concerns coming to understand better how cultural processes affect children’s development. In discussions of social development, much lip service tends to be paid to the concept of culture, but usually only the vaguest understanding of how cuhure works lies behind commonly voiced opin- ions of the type, “Children are encouraged by American culture to be competitive and independent.” How encouraged? Who or, more likely, what encourages them? Most psychologists have little working knowl- edge of the ways in which cultural processes set the stage for individual development. A more comprehensive analysis of these processes can come from a comparative cultural perspective, which enables one to specify the functional dimensions of the learning environment of the child. As J. Whiting and others (1966, p. 83) succinctly put it, “certain aspects of the childrearing process seem to have the effect of, if not cre-

48

ating, at least strengthening values far beyond the conscious intent of the agents of socialization.” Without a concept of culture as a learning environment, it is difficult to conceptualize socialization pressures in more than a very partial and fragmented way.

Using a comparative cultural perspective to understand better the important dimensions of chiIdren’s learning environments can be illustrated by the cross-cultural data from a recent collaborative project that we directed (Whiting, Edwards, and others, forthcoming). Let us first briefly explain the project and the kind of data that our group obtained (see also Whiting, 1979). Then we will discuss one set of find- ings and explain their relevance to the question of the differential socialization of boys and girls.

The current project was based on recent observations of the social interaction and daily activities of children aged two to twelve years and, secondarily, on a reanalysis of the Six Cultures data. The more recent data on social interaction were collected by a number of different anthropologists and psychologists who worked in six com- munities in Kenya, one in India, and one in Liberia (see Table 1). In 1976, the Ford Foundation sponsored a project to analyze jointly the data on both social interaction and daily activities, and a group of twelve met in Cambridge to develop a plan of attack. Thus our collab- orative project was born, to study sex differences in the social behavior of children and to examine the effects of modernization on mother- child interaction.

To assess sex differences in social interaction, the group agreed to examine twenty-one categories of social behavior, grouped into seven larger categories of nurturance, dependence, prosocial dominance, egois- tic dominance, aggression, and sociability. The categories are based on the judged conscious intent of the actor toward the person with whom he or she is observed to be interacting and whose ongoing behavior the actor wishes to affect. According to our approach (Whiting, 1979; Whiting and Whiting, 1975), social interaction is conceived to involve the seek- ing or offering of goods and services to recipients who may accept or reject them. Different types of interaction invoIve benefit to the actor, the recipient, both of them, or some other person or persons. The goods and services exchanged in social interaction include such things as food and material goods as well as intangibles (information, instru- mental help, comfort, attention, privilege, control, physical pain, psy- chological pain, friendship, social participation, and competition). For example, prosocial dominance involves the seeking of control through acts such as assigning chores or reprimanding whose apparent goal is to

49

benefit the recipient and/or the social group. Dependent behaviors involve the seeking of services such as help, attention, and information to benefit the actor.

It was assumed that the intention of the actor could be judged in the majority of instances by observers who were native to the culture. The observations for all samples were done in naturally occurring situ- ations, usually in the house or yard. Observers were trained to record, in running English sentences, social interaction between the sample child and/or the sample mother and the other individuals in the envi- ronment identified by age, sex, and kinship relation. The length of observation periods varied from one sample to another (the longest per- iod, one hour, was in Bubaneswar, India; the shortest in the Six Cul- tures samples, five minutes). There was also variation in the focus of observations (in most cases the sample child was the focus; in India the mother or mother surrogate was the focus; and in the Ngecha, Kenya sample, the focus was on children in most observations but on mothers in others). Finally, coding procedures varied in terms of who coded the running protocols into behavioral categories. In most cases, local observers performed this task; however, field researchers Harkness and Super, Erchak, and Seymour coded their own data, and in the Six Cul- tures study, the narrative records were forwarded from the field to the Laboratory of Human Development at Harvard for coding. Reliability of coding was monitored by all of the researchers. On the average relia- bility probably never fell below the 70 percent level (see Whiting and Edwards, 1977). Table 1 displays the number and ages of children in each sample.

The set of findings concerning nuzturunce (defined as behavior in which a person offers a service or resource to meet the needs of another) can be used to demonstrate the way in which we conceptualize culture, as an independent variable, in terms of learning environments for chil- dren. The failure of American research to elucidate the learning of nur- turant behavior can be traced to the fact that, as Maccoby and Jacklin (1974, p. 220) point out, “existing research has seldom focused on a child’s offering of nurturance to an infant or younger child.” Rather, naturalistic studies of nurturance and all other kinds of social behavior of children are based almost entirely on mother-child and same-aged peer observations. Little or no American material exists that is directly comparable to our project’s data- that is, observations including chil- dren’s interaction with infants and older and younger children as well as with their mothers. Certainly we are not the first in recent years to express the need for more study of children’s cross-age interaction (see,

50

Table 1. Characteristics of the Samples

Ywr of Linguistic LOCation Field Researcher Field Work SmPh Group

~ _ _ _ _ _

1. Kien-taa, Gerald Er&k 1970-1971 15 households; 20 Kpelle Liberia children aged 1-6

(360 minutes observatiodchild)

-~ ~~~

2. Ngecha, Beatrice Whiting 1968-1970 42 homesteads; 104 Kikuyu Kenya 1973 chiidren aged 2-10

(at least 45 minutes observatiodchild)

3. oyugis, Carol Ember 1968-1969 10 children aged Luo Kenya 8-12 (135 minutes

observatiodchiid)

4. Kisa, and Thomas Weisner 1970-1972 24 urban and rural Abaluya 5. Kariobangi, families matched

Kenya by age, education, and kinship ties; 68 children aged 2-8 (120 minutes observatiodchild)

6. Kokwet, and Sara Harkness 1972-1975 Kokwet: 64 Kipsigis 7. urot, and Charles children aged 3-10;

Kenya Super Urot: 64 children aged 3-10 (120 minutes observatiodchiid)

~ ~ ~ ~~ ~

8. Bubaneswar, Susan Seymour 1965-1967 36 households (24 Oriya India (State upper class, 12 of Orissa) lower class); 103

chiidren aged 0-10 (16 hours of observatiodhouse- hold)

for example, Edwards and Lewis, 1979; Hartup, 1976; Konner, 1975), but here the point concerns the importance of cross-age peer studies for understanding the development of sex differences in social behavior.

Our findings indicate that sex differences in nurturant behav- iors are more. prominent in some kinds of dyadic interaction than in others. We made nurturance comparisons by the following procedure. We constituted a summary category offering of material goods, physi- cal care, help, comfort, and attention. Within each cultural sample we

51

then added together all of the nurturant versus non-nurturant social acts done by girls versus boys to specific target groups of people. For example in the Kien-taa, Liberia community sample, the girls, (n = 8) made a total of 7 nurturant versus 227 non-nurturant acts to mothers whereas the boys (n = 7) made 2 nurturant versus 179 non-nurturant acts (phi nonsignificant for boys versus girls). In that same sample girls made 44 nurturant acts versus 64 non-nurturant acts to infants (under 1.5 years of age), whereas boys made 22 nurturant versus 52 non-nur- turant acts (phi = .lo, nonsignificant). Thus in both comparisons girls scored proportionately higher than boys but nonsignificantly so. Tables 2, 3,4, and 5 display the results of all of the comparisons, that is, they indicate which sex in each cultural sample displayed proportionately more nurturance to a specific target group, as well as whether the phi value of the comparison is statistically significant. Although our meth- odology has the weakness of having lumped together all the acts of girls (and boys) in a sample regardless of their ages, it has the strength of having separated out acts to various target groups-mothers, infants, children a defined number of years younger than the actor child, chil- dren a defined number of years older than the actor child, and children relatively close in age to the actor child.

The findings (Tables 2 to 5) suggest that, overall, girls behave more nurturantly than boys. There are twelve comparisons in the four tables in which girls score significantly higher than boys but only three of the reverse. What is further evident is that the sex difference in behavior favoring girls is most consistently found in the behavior of children to infandtoddlers (see Table 2). Table 4, presenting behavior to children older than the actor, has the fewest number of significant sex differences. To summarize, sex differences in nurturance are most con- sistent in behavior directed to very young children (infanvtoddlers), least prevalent and consistent in behavior directed to mothers and older children, and intermediate in behavior directed to younger children or to children close in age to the actor. In no type of dyadic interaction are boys consistently more nurturant than girls; rather, sex differences favoring girls on this dimension appear in all dyadic contexts but to a greater or lesser extent depending on the specific type of interaction.

What might be the cause of girls’ greater display of nurturance? It does not appear to be the result of extensive differences in treatment by mothers of daughters versus sons. Like Maccoby and Jacklin (1974), we find few behaviors that mothers consistently direct more to one sex than to another (although repimanding or correcting is one such behavior; see Whiting, 1979). Nevertheless, while mothers may behave

52

Table 2. Sex Differences in Nurturant’ Behavior I

Tvpc of Social Intmacrion To Motfur& To Infannl/Toddlcrs

Kien-taa urot Kokwet Kisa Kariobangi Ngecha

Bubaneswar-Lower Class Bubaneswar-Upper Class Nyansongo Tarong Juxtlahuaca Taira Khalapur Orchard Town

oyugis

B G G G B G

G = G B B G G

G B

-

-

G

G - - G’ G

Note: This table presents the sex of child showing the higher percentage of nurturance in beahvior directed to mothers and behavior directed to infmt/toddlem. ‘Nurturance is the summed category including the offering of food, material goods, physical care, help, comfort, and attention. bFor each cultural sample, nurturant versus non-nurturant acts of all girls (G) to their mothers were summed; the same was done for boys (B). The resulting scores for girls versus boys were tested for significance using the phi coefficient. A dash (-) indicates that there were fewer than twenty social acts by girls to the target or less than twenty by boys to the target, for a particular sample. *p< .05, * *p< .01, ***p< .001. “Infant/toddlers were less than eighteen months old for the first set of samples and lcss than two years old for the Six Cultures samples. Sex differences were tested following the same procedure outlined in footnote b.

relatively similarly to boys and girls, it is clear that in many cultures girls spend much more time in the dyadic contexts that elicit nurtur- ance- that is, in the company of infants and young children (Whiting and Whiting, 1975). Our data do not suggest that this is necessarily due to mothers assigning more child care chores to girls but rather that, when mothers do request such tasks, girls are more likely to comply. Thus, we suggest the following two-part scenario for the socialization of sex differences in nurturance: (1) During the early years (perhaps ages three to six), girls are not more nurturant than boys but they are more willing to play with and take responsibility for younger children, especially infants, toddlers, and younger sisters. (A large number of studies conducted on U.S. samples have found females of various ages, including preschoolers, to display more interest than males in babies. See, for example, Berman, Monda, and Myerscough, 1977; Berman and others, 1978; Edwards and Lewis, 1979; Feldman and Nash,

53

Table 3. Sex Differences in Nurturant Behavior I1

7 jpe of Sociul Interactions To Younger Childc To Younger Childe

Sampleb o f fhe Same Scx ofthe Opposl Sex -~

Kien-taa Urot Kokwet Kisa Ngecha

B*** G*** G* G B* *

~~~ ~

Acts by Children in Older Age Group to Children in Yowrger Age Groupd of Same Sex of Oppositc lscx

Nyansongo Tarong Juxtlahuaca Taka Orchard Town

G G* G G* G

G G* G B -

No&: This table presents the sex of child showing the higher percentage of nurturance in behavior directed to children younger than the actor child. BComparisons and significance tests done as in Table 2, using phi coefficients *p< .05, * p < .01,

"Samples have been omitted where frequency of interaction for these dyad types were too low for comparisons to be made. It is interesting to note in this and all following tables the greater preva- lence of same-sex versus cross-sex interaction (column 1 versus 2). CHere, data include all behavior directed by actor children to targets three or more years younger than they are (excluding infants). dHere, data include all behavior directed by children classified in the older group (age 7-10) to children in the younger group (age 3-6). Thus the procedure varies from that for samples above, necessitated by the way in which the Six Cultures data were originally coded.

** *p< .001.

1978; Feldman, Nash, and Cutrona, 1977; Frodi and Lamb, 1978; Lamb, 1978.) (2) By middle childhood (age seven to eight and above), sex differences in nurturance toward infants and younger sisters begin to appear as a result of girls having learned to be responsive to the needs of others in those dyadic contexts.

The relevance of this scenario for the comparative study of human development is that cultures differ greatly in the amount of time that children, especially girls, spend in the company of very young children. As targets of interaction, infants and toddlers are much more available to children of some cultures than to children of others. For example, in the Six Cultures communities of Nyansongo, Kenya, and Juxtlahuaca, Mexico, an average of 25 percent of all children's social acts were directed to infants, whereas in Orchard Town, that amont was only 3.5 percent (Whiting and Whiting, 1975). Thus, different cul- tural groups can magnify, minimize, or perhaps even eliminate, any

54

Table 4. Sex Differences in Nurturant Behavior I11

Type of Social Inlnocliona To O h Childb To O h Childb OfthcSamSOC ofthc opposik? sa

~~

Kien-taa Urot Kokwet Kisa Ngccha oyugis

G G G G

G - B - Acts by Children in Younger Age Goup to Children in Oldn Age Groupe

ofthesamcsa of the Opposite Scx

Nyansongo Tarong Juxtlahuaca Taira Orchard Town

G B B

G - B B

- G B

Note: This table presents the sex of child showing the higher percentage of nurturance in behavior directed to children older than the actor child. 'Comparisons and tests of significance follow procedures in Table 2, *p< .05, **@< .01,

bHere data include all behavior directed by actor children to targets three or more years older than they wen (but not older than age 12). CHere data include all behavior directed by younger age group (3-6) to older age group (7-10). (See Table 5, footnote d.)

-*p< .001.

sex differences in nurturance to the extent that they place girls and boys in the settings that promote the development of nurturant behavior.

To conclude, we believe that different kinds of social behavior are learned in interaction with different sedage groups of people, and that cultures shape the behavior of their children by selecting the company they keep and the activities that engage their time. Parents do, of course, consciously attempt to transmit certain behaviors and values to their chii- dren; but without the support of that complex of dyadic contexts and activities that we call the cultural learning environment, those sociali- zation pressures would probably have little force. This is true in the area of sex role socialization as much as in any other, and consequently, ifwe limit our study of children to only a few social contexts and activities (such as mother-child interaction and same-age peer interaction at school), we may miss much of the action where sex role learning occurs.

The study of the relative distribution of different social contexts and activities in the daily lives of children is a critical but largely ignored subject in most current socialization research. The recent shift

55

Table 5. Sex Differences in Nurturant Behavior IV

Type of Social Interaction’

Same Sex Opposite Sex Same Sex Opposite Sa To Slightly Younger Childb To Slightly Older ChiMb

- G

B

Kien- taa - - Kokwet B G B G Kisa G - - - - - Kariobangi G Ngecha G+ + G G G oyugis - - G -

~ ~~~ ~~

Acts Between ChiUren Both in Oldn Age Groupc

Acts Between Children Both in Younger Age Groupc

Same ser Opposite Sex same sex Opposi& six

Nyansongo G G B B Tarong G+ G B+ c Juxtlahuaca G’ - G G Taira B B G G Khalapur B - G Orchard Town G G G B

- Note: This table presents the sex of child showing the higher percentage of nurturance in behavior directed to children close in age to the actor child. aComparisons and tests of significance follow procedures in Table 2, *p< .OS, +*p< .01, ***fj< .001. bHere slightly youngcr and slightly older are defined as two years or less age difference.

(See Table 3, footnote d.) age group includes children aged 7-10, younger age group includes children aged 3-6.

in developmental psychology away from socialization explanations of sex differences may, therefore, reflect more the inadequacy of present research strategies than a basic weakness of the socialization model itself for predicting and explaining sex differences in children’s social behavior.

References

Berman, P. W., Goodman, V., and Sloan, V. L., and Fernander, L. “Preference for Infants Among Black and White Children: Sex and Age Differenceti.” Child DeveI@-

Berman, P. W., Monda, L. C., and Myerscough, R. P. “Sex Differences in Young Children’s Responses to an Infant: An Observation Within a Daycare Setting.” Child Deuelopment, 1977, 48, 711-713.

Birns, B. “The Emergence and Socialization of Sex DifTerences in the Earliest Years.”

Block, J. H. “Issues, Problems, and Pitfalls in Assessing Sex Differences: A Critical Review of The Psycfwlogy ofsex Da~erences.”Merrill-Palmer Qmly, 1976,22, 203-300.

ment, 1978, 49, 917-919.

Mmill-Palmer Q~arterly, 1976, 22, 229-254.

Brooks-Gunn, J., and Matthews, W. S. Heand She: How Children Devef~p Tluir Sur-Role Zdmfity. Englewood Cliffs, N.J. : Prentice-Hall, 1979.

Edwards, C. P., and Lewis, M. "Young Children's Concepts of Social Relations: Social Functions and Social Objects." In M. Lewis and L. A. Rosenblum (Eds.), The Child and Its Family. Genesis of Behavior, vol. 2. New York Plenum, 1979.

Edwards, C. P., and Whiting, B. B. "Women and Dependency." Politics and Sock@,

Feldman, S . S., Nash, S. C., and Cutrona, C. T h e Influence ofAge and Sex on Res- ponsiveness of Infants." Developmental Psychology, 1977, 13, 675-676.

Frodi, A. M., and Lamb, M. E. 'Sex Uitterences in Kesponsiveness to Infants: A Developmental Study of Psycholphysiological and Behavioral Responses." Child Development, 1978, 49, 1182-1 188.

Hartup, W. W. "Cross-Age Versus Same-Age Peer Interaction: Ethological and Cross- Cultural Perspectives." In V. L. Allen (Ed.), Chikiren OF Tmhers. New York Aca- demic Press, 1976.

Konner, M. "Relations Among Infants and Juveniles in Comparative Perspective." In M. Lewis and L. A. Rosenblum (Eds.), F r i d h i p and Peer Relations. Origins of Behavior, vol. 4. New York: Wiley, 1975.

Lamb, M. E. T h e Development of Sibling Relationships in Infancy: A Short-Term Longitudinal Study." Child Deuelopmmt, 1978, 49, 1189-1 196.

Maccoby, E. E., and Jacklin, C. N. The Pychology of& D#mences. Stanford, Calif.: Stanford University Press, 1974.

Whiting, B. B. "Maternal Behavior in Cross-Cultural Perspective." Paper presented at Annual Meeting of the Society for Cross-Cultural Research, Charlottesville, Vir- ginia, February 20, 1979.

Whiting, B. B., and Edwards, C. P. "A Cross-Cultural Analysis of Sex Differences in the Behavior of Children Aged Three Through Eleven." Journal of Social Psychology, 1973, 91, 171-188.

Whiting, B. B., and Edwards, C. P. "The Effect of Age, Sex, and Modernization on the Behavior of Mothers and Children." Report to the Ford Foundation, January 1977.

Whiting, B. B., Edwards, C. P., Ember, C. R., Erchak, G., Harkness, S., Seymour, S., Super, C., and Weisner, T. The Company T h Keep. Manuscript in progress, Harvard University.

Whiting, B. B., and Whiting, J. W. M. Children ofsix Cultures: A Psycho-Cultural Analy~is. Cambridge, Mass.: Harvard University Press, 1975.

Whiting, J. W. M., Chasdi, E. H., Antonovsky, H. F., Ayres, B. C. "The Learning of Values." In E. Z. Vogt and E. M. Albert (Eds.), Peoplc of Rimrock. Cambridge, Mass.: Harvard University Press, 1966.

1974, 4, 343-355.

Carolyn Pope Edwards is Assistant Professor of Education at the University of Massachusetts at Amherst and Director of its Early Childhood Education Laboratory School. Her professional interests center on social development, both cross-cultural and in preschool children in America.

56

57

Beatrice B. Whiting is Professor of Anthropology and Education at the Harvard Graduate School of Education. She is widely known for her work on the development of children’s social behavior and its relationrhip to culturalfaatures such as the economic rote 4 women.