Domestic Worlds and Public Worlds 65
In 1974, in an attempt to document a universal subordination of women, Michelle Rosaldo (1974:18) proposed a paradigm relating "recurrent aspects of psychology and cultural and social organization to an opposition between the 'domestic' orientation of women and the extra-domestic or 'public' ties, that, in most societies, are primarily available to men." The domestic-public model led Rosaldo to suggest that women's status is highest in societies in which the public and domestic spheres are only weakly differentiated, as among the Mbuti pygmies. In contrast "women's status will be lowest in those societies where there is a firm differentiation between domestic and public spheres of activity and where women are isolated from one another and placed under a single man's authority, in the home. Their position is raised when they can challenge those claims to authority ..." (Rosaldo 1974:36). Accordingly, women may enhance their status by creating a public world of their own or by entering the men's world. In
addition, the most egalitarian societies will h those in which men participate in the domesti domain.
Correspondingly, Sanday (1974) suggests that women's involvement in domains of activity such as subsistence or defense may be curtailed because of their time and energy commitment to reproduction and mothering. Men, on the other hand, are free to form broader associations in the political, economic, and military spheres that transcend the mother-child unit. While the linkage of women with the domestic and men with the public domains may imply a biological determinism based on women's reproductive roles, Rosaldo (1974:24) argues that the opposition between domestic and public orientations is an intelligible but not a necessary arrangement.
One aspect of women's domestic responsibilities is that it is women who primarily raise children. Nancy Chodorow (1974) develops a theory linking adult sex role behavior to the fact
that children's early involvement is with their female parent. Chodorow argues that girls are integrated through ties with female kin into the world of domestic work. Age, rather than achievement, may define their status, while boys must "learn" to be men. Unlike girls, boys have few responsibilities in childhood and are free to establish peer groups that create "public" ties. To become an adult male a boy is often obliged to dissociate himself from the home and from female kin. According to Rosaldo (1974:26), "the fact that children virtually everywhere grow up with their mothers may well account for characteristic differences in male and female psychologies" as well as setting the stage for adult organization of activities.
As scholarship devoted to an understanding of gender issues has evolved, the influential domestic-public model has been the focus of considerable controversy, revolving around three related issues: whether male domination is universal, whether male domination is explained by the domestic-public dichotomy, and whether the concept of domestic-public has relevance in all cultures.
Lamphere (in this book) reviews the formulation of the domestic-public model, and discusses the subsequent critiques of its applicability. Rosaldo herself, rethinking her original position, said that while male dominance appears widespread, it does not "in actual behavioral terms assume a universal content or a universal shape. On the contrary, women typically have power and influence in political and economic life, display autonomy from men in their pursuits, and rarely find themselves confronted or constrained by what might seem the brute fact of male strength" (1980:394). While the domestic-public opposition has been compelling, Rosaldo suggests that the model assumes too much rather than helping to illuminate and explain (see, for example, Mathews in this book).
As Lamphere observes, it has become increasingly clear that the domestic-public opposition is the heir to nineteenth-century social theory rooted in a dichotomy contrasting home and woman, with a public world of men, and reflecting an understanding of political rights based on sex. It has also been noted that conceptualizing
social life as dichotomized into domestic and public domains does not make sense in societies in which management of production occurs within the household and in which household production itself involves the management of the "public" economy (Leacock 1978:253).
In contrast, in an industrial society, where home and workplace are clearly demarcated, the domestic-public opposition may have explanatory value. For example, Murcott (in this book) analyzes one domestic task, cooking, as part of an exploration of economic relations in the family and of the division of labor between spouses. Interviews with Welsh housewives indicate that ideas about home cooking reflect understandings of the relationship between domestic and paid labor. The informants shared the view that proper eating must occur at home and that a cooked dinner is necessary to family health and well-being. Murcott suggests that the emphasis on having a proper dinner waiting for the husband when he comes in from work underscores the symbolic importance of the return home: "the cooked dinner marks the threshold between the public domains of school or work and the private sphere behind the closed front door" (Murcott, this book).
Alice Yun Chai (in this book) considers the relevance of the domestic-public distinction to Korean immigrant women in Hawaii and examines the adaptive strategies of these women in response to their disadvantageous political and economic position in the larger society. She shows that Korean immigrant women attempt to gain status in the public sphere often in sales or clerical jobs; eventually, because of structural barriers, women create a public world of their own in family businesses engaging male and female relatives. These businesses integrate home and workplace, the public and private spheres, as husband and wife strive together to improve the family economy and educate children.
This research suggests that the public-private distinction has implications for the interplay among gender, status, and power. While traditional conceptions of power emphasized formal political behavior and authority associated with a status conferring the "right" to impose sanctions {Lamphere 1974:99), informal power
64
66    Domestic Worlds and Public Worlds
strategies such as manipulation and maneuvering are also important aspects of political activity. The analysis of Korean immigrant women indicates that a thorough understanding of women's strategies requires an examination of multiple factors: status in the public world, interpersonal influence, and the relationship between them.
Cynthia Nelson (in this book) also examines the concept of power, focusing on images of women and power in the domestic and public domains in the societies of the Middle East. Ethnographies of the Middle East have commonly differentiated two social worlds, a woman's private world and a man's public world. Women's concerns are domestic, men's political. Neison argues that the assignment of private and public reflects the imposition of western cultural categories on the Middle East; the meaning of power is influenced by these categorizations, as well as by the limitations of data obtained by male ethnographers from male informants. This is a point made forcefully by Annette Weiner (1976) in her reanalysis of Trobriand exchange. She argues that "we un-questioningly accept male statements about women as factual evidence for the way a society is structured. . . . Any study that does not include the role of women—as seen by women— as part of the way the society is structured remains only a partial study of that society. Whether women are publicly valued or privately secluded, whether they control politics, a range of economic commodities, or merely magic spells, they function within that society, not as objects, but as individuals with some measure of control" (1976:228).
Similarly, Nelson argues that by asking such questions as "How do women influence men?" "Who controis whom about what?" "How is control exercised?" it becomes apparent that women exercise a greater degree of power in social life than is often appreciated. In addition, she challenges the idea that the social worlds of men and women are reducible to private and public domains, with power limited to men in the public arena. Nelson's review of ethnographies addressing the role and position of women in Middle Eastern society suggests that women play a crucial role as structural links be-
The Domestic Sphere of Women and the Public World of Men 67
tween kinship groups in societies in which family and kinship are fundamental social institutions. Women are in a position to influence men through ritual means, to channel information to male kin, and to influence decision making about alliances; consequently, women do participate in "public" activities, and women's exclusive solidarity groups exercise considerable social control and political influence. The conceptions of power as defined by the western observer are particularly challenged by literature on women done by women that offer a perspective on the position of Middle Eastern women derived from the actors themselves.
In the course of her critique of the application of the domestic-public opposition to social organization in the Middle East, Nelson challenges long-standing assumptions regarding women's subordination and male dominance. Yun Chai's argument also demonstrates that women engage in strategies that have political implications within the ethnic community. Thus, the association between political power and a public domain that excludes women is called into question. Similarly, in studies of peasant societies Rogers (1975) and others (Friedl 1967; Reigelhaupt 1967) contend that the sector of life over which peasant women have control—the household—is in fact the key sphere of activity, socially, politically, and economically. Men occupy public and prestigious positions of authority within the village sphere, but these activities do not have the impact on daily life that household activities have. In light of these analyses demonstrating women's power and influence, we are reminded that the universality of male dominance appears untenable.
In reflecting on feminist research in anthropology, Rosaldo critiques the very tendency to look for universal truths and origins. Rather, anthropologists need to develop theoretical perspectives that analyze the relationships of men and women in a broader social context (Rosaldo 1980:414), involving inequality and hierarchy. As Henrietta Moore emphasizes, while women in many societies share some experiences and problems, women have had very different encounters with racism, colonialism, the penetration of capitalism, and international development. We need to move from assumptions of
the shared experience of "women" to a critical analysis of "concepts of difference" (Moore 1988:9).
REFERENCES
Chodorow, Nancy. 1974. Family Structure and Feminine Personality. In Michelle Z. Rosaldo and Louise Lamphere (eds.). Woman, Culture, and Society, pp. 43-67. Stanford: Stanford University Press.
Friedl, Ernestine. 1967. The Position of Women: Appearance and Reality. Anthropological Quarterly 40: 97-108.
Lamphere, Louise. 1974. Strategies, Cooperation, and Conflict Among Women in Domestic Croups. In Michelle Z. Rosaldo and Louise Lamphere (eds.). Woman, Culture, and Society, pp. 97-113. Stanford: Stanford University Press.
Leacock, Eleanor. 1978. Women's Status in Egalitarian Society. Implications for Social Evolution. Current Anthropologyi 9(2): 247-275.
Moore, Henrietta L. 1988. Feminism and Anthropology. Minneapolis: University of Minnesota Press.
Riegelhaupt, Joyce. 1967. 5aloio Women: An Analysis of Informal and Formal Political and Economic Roles of Portuguese Peasant Women. Anthropological Quarterly 40: 109-126.
Rogers, Susan Carol. 1975. Female Forms of Power and the Myth of Male Dominance: A Model of Female/Male Interaction in Peasant Society. American Ethnologist 2: 727-756.
Rosaldo, Michelle Z. 1974. Theoretical Overview. In Michelle Z. Rosaido and Louise Lamphere (eels.). Woman, Culture, and Society, pp. 17-43. Stanford: Stanford University Press.
-. 1980. The Use and Abuse of Anthropology:
Reflections on Feminism and Cross-Cultural Understanding. Signs 5(3): 389-418.
Rosaldo, Michelle Z. and Louise Lamphere (eds.). 1974. Woman, Culture, and Society. Stanford: Stanford University Press.
Sanday, Peggy R. 1974. Female Status in the Public Domain. In Michelle Z. Rosaldo and Louise Lamphere (eds.). Woman, Culture, and Society, pp. 189-207. Stanford: Stanford University Press.
Weiner, Annette B. 1976. Women of Value, Men of Renown: New Perspectives in Trobriand Exchange. Austin: University of Texas Press.
Louise Lamphere
Since 1974 there has been a burgeoning interest within anthropology in the study of women, sex roles, and gender. Anthropology has long been a discipline that contained important women (Elsie Clews Parsons, Ruth Benedict, and Margaret Mead among the
Original material prepared for this text.
most famous) and a field in which women have been studied as well (e.g.. Kaberry 1939, 1952; Landes 1938, 1947; Leith-Ross 1939; Underbill 1936; and Paulme 1963). However, with the publication of Woman, Culture, and Society (Rosaldo and Lamphere 1974) and Toward an Anthropology of Women (Reiter 1975) women scholars, many of whom were identi-
68    Domestic Worlds and Public Worlds
The Domestic Sphere of Women and the Public World of Men 69
fieci as feminists, began to critique the androcentric bias in anthropology, to explore women's status in a wide variety of societies, and to provide explanatory models to understand women's position cross-culturally.
One of the most powerful and influential models was proposed by Michelle Rosaldo in her introductory essay to Woman, Culture, and Society (1974). Her argument began by asserting that although there is a great deal of cross-cultural variability in men's and women's roles, there is a pervasive, universal asymmetry between the sexes. "But what is perhaps most striking and surprising," Rosaldo writes, "is the fact that male, as opposed to female, activities are always recognized as predominantly important, and cultural systems give authority and value to the roles and activities of men" (Rosaldo 1974:19).
One of the quotes we chose to appear at the beginning of the book, a passage from Margaret Mead's Male and Female, sums up what we saw in 1974 in all the ethnographies and studies we examined. "In every known society, the male's need for achievement can be recognized. Men may cook, or weave, or dress dolls or hunt hummingbirds, but if such activities are appropriate occupations of men. then the whole society, men and women alike, votes them as important. When the same occupations are performed by women, they are regarded as less important" (Mead 1949:125). Not only were there differential evaluations of women's activities, but, Rosaldo argues, "everywhere men have some authority over women, that [is] they have culturally legitimated right to her subordination and compliance" (1974:21)
Having argued for a pervasive sexual asymmetry across cultures, not just in terms of cultural values, but also in terms of power and authority, Rosaldo accounted for this difference between men and women in terms of a dichotomy.1 She argued that women are associated with a "domestic orientation," while men are primarily associated with extra domestic, political, and military spheres of activity. Bv "domestic" Rosaldo meant "those min-
imal institutions and modes of activity that are organized immediately around one or more mothers and their children." In contrast the "public" referred to "activities, institutions, and forms of association that link, rank, organize, or subsume particular mother-child groups. Put quite simply, men have no single commitment as enduring, time-consumir _ and emotionally compelling—as close to seeming necessary and natural—as the rela-tion of a woman to her infant child; and so men are free to form those broader associations that we call 'society,' universalistic s -tems of order, meaning, and commit me that link particular mother-child groups."
Rosaldo, along with Sherry Ortner and Nancy Chodorow who also wrote essays in Woman, Culture., and Society, insisted that the -I connection between women's role in repro- < duction (the fact that women everywhere lac- > tate and give birth to children) and their do- I mestic orientation is not a necessary one. In other words biology is not destiny. Women's -f domestic orientation was structurally and cul- -rurally constructed and "insofar as woman is ; universally defined in terms of a largely ma- s ternal and domestic role, we can account for \ her universal subordination" (Rosaldo 1974:7).
"Although" Rosaldo writes. "I would be the ^ last to call this a necessary arrangement or to J deny that it is far too simple as an account of ; any particular empirical case, I suggest that ' the opposition between domestic and public f orientations (an opposition that must, in part, ! derive from the nurturant capacities of l women) provides the necessary framework ** for an examination of male and female roles in any society" (Rosaldo 1974:24). ;..
For Rosaldo, then, women were involved I in the "messiness" of daily life; they were al-ways available for interruption by children. -Men could be more distanced and may actu- \-ally have separate quarters (such as men's \ houses) away from women's activities. Men ' could thus "achieve" authority and create rank, hierarchy, and a political world away -from women. The confinement of women to ; the domestic sphere and men's ability to ere- ; ate and dominate the political sphere thus ac- t:
counted for men's ability to hold the greater share of power and authority in all known cultures and societies.
At the time Rosaldo wrote her overview and in the introduction we both wrote, we were faced with building a framework where none existed. Despite the number of monographs on women, Margaret Mead's work and that of Simone de Beauvoir (1953) were the most provocative, and perhaps the only, theoretical works we knew." The argument for universal sexual asymmetry followed in a long tradition in anthropology where scholars have sought to look for what is broadly "human" in all cultures. In addition to language, anthropologists have discussed the universality of the incest taboo, marriage, and the family. The notion that women might be universally subordinate to men thus made sense as a first attempt at theory building in this newly revived "subfield" within anthropology.
Although Rosaldo argued for universal subordination, she was careful to make clear that women are not powerless. They exercise informal influence and power, often mitigating male authority or even rendering it trivial (Rosaldo 1974:21). In addition, there are important variations in women's roles in different cultures, and variation was discussed in most of the rest of the articles in the collection. For example, Sanday and Sacks compared women's status in a number of different societies, while Leis examined the structural reasons why women's associations are strong in one Ijaw village in Nigeria, yet absent in another. Finally, in my own article I examined the differences in women's strategies within domestic groups in a number of societies, which related to the relative integration or separation of domestic and political spheres.
Since 1974 the hypothesis of universal subordination of women and the dichotomous relationship between women in the domestic sphere and men in the public sphere have been challenged and critiqued by a number of feminist anthropologists. As appealing as this dichotomy seemed in the abstract it turned
out to be difficult to apply when actually looking at examples of women's activities in different cultures. For example, in an important article written about the same time as Rosaldo's introduction, Rayna Re iter (now Rayna Rapp) described women's and men's distinct lives in a small French village in the south of France. "They inhabited different domains, one public, one private. While men fraternized with whomever they found to talk to in public places, women were much more enmeshed in their families and their kinship networks" (Reiter 1975b:253). However, two categories of public space fell into women's domain: the church and three shops, including the local bakery. Men tended to avoid women's places, entering the bakery, for example, only when several men were together and joking, "Let's attack now" (Reiter 1975b:257).
Reiter argues that men and women use public space in different ways and at different times. "The men go early to the fields, and congregate on the square or in the cafes for a social hour after work. Sometimes they also fraternize in the evenings. These are the times when women are home cooking and invisible to public view. But when the men have abandoned the village for the fields, the women come out to do their marketing in a leisurely fashion. The village is then in female hands. In the afternoon, when the men return to work, the women form gossip groups on stoops and benches or inside houses depending on the weather" (Reiter 1975b:258). Despite the powerful imagery—women associated with the private or domestic domain and men with public space—the description also shows that the dichotomy is not neat. After all women are in public a great deal; they have taken over, in some sense, the Church and the shops and even the public square in the middle of the day.
In Margery Wolf s description of women in a Taiwanese village based on data she collected in the late 1950s, she emphasizes that because researchers have focused on the dominance of patrilineal descent in the family, they have failed to see women's presence. "We have missed not only some of the system's
70    Domestic Worlds and Public Worlds
The Domestic Sphere of Women and the Public World of Men 71
subtleties but also its near-fatal weaknesses" (Wolf 1972:37). Women have different interests than men and build uterine families-strong ties to their daughters, but primarily to their sons who give their mothers loyalty and a place in the patrilineal extended family. Outside the family in the community women formed neighborhood groups—around a store, at a platform where women washed their clothes in the canal, or under a huge old tree. In a village strung out between a river and a canal, there was no central plaza dominated by men as in the South of France.
In Peihotien Wolf did not describe a cultural geography where women were in a private sphere and men in the public one; rather there was more of a functional separation-men and women had different activities and interests. They were often located in the same places but had a different relationship to the patrilineal extended family and the male-dominated community. Women's lack of power led them to different strategies, different tactics that often undermined male control of the household and even the community. As Sylvia Yanagisako (1987:111) has pointed out the notion of domestic-public entails both a spatial metaphor (of geographically separated or even nested spaces) and a functional metaphor (of functionally different activities or social roles) in the same conceptual dichotomy. Analysts often "mix" these different metaphors in any particular analysis—sometimes using domestic-public spatially and at other times functionally.
Even in the Middle East, the association of women with a private domain (and a lack of power) and men with a public domain (and the center of politics) was too simple, as Cynthia Nelson pointed out in her article, "Public and Private Politics: Women in the Middle Eastern World" (1974; reprinted in this book). Because they are born into one patrilineal group and marry into another, women are important structural links between social groups and often act as mediators. Because there are segregated social worlds, all-female institutions are important for enforcing social norms: Women fill powerful ritual roles as sorceresses, healers, and mediums; women
are important sources of information for then-male kin; and women act as "information brc -kers," mediating social relations within both the family and the larger society.
From Rosaldo's point of view, these aspecl -of women's power are primarily "informal and very different from the public, legitimat roles of men. Nevertheless, even though Nelson affirms the separation of male and female worlds (both spatially and functionally), what is "domestic" has public ramifications (the arrangt ment of a marriage, the transmission of highly charged political information) and the shadow of the family and kin group (the "domestic") is present in even the most "public" of situations. What at first seemed like a simple straightforward dichotomy, in light of actual case material seems very "slippery" and complex.
Furthermore, in many cultures, particularly those with an indigenous band or tribal structure, a separation of "domestic" and "public" spheres makes no sense because household production was simultaneously public, economic, and political. Leacock pointed out the following after reviewing the literature on the Iroquois during the seventeenth and eighteenth century:
Iroquois matrons preserved, stored, and dispensed the corn, meat, fish, berries, squashes, and fats that were buried in special pits or kept in the long house. Brown (1970:162) notes that women's control over the dispensation of the foods they produced, and meat as well, gave them de facto power to veto declarations of war and to intervene to bring about peace. . . . Women also guarded the "tribal public treasure" kept in the long" house, the wampum quill and feather work, and furs. . . . The point to be stressed is that this was "household management" of an altogether different order from management of the nuclear or extended family in patriarchal societies. In the latter, women may cajole, manipulate, or browbeat men, but always behind the public facade; in the former case, "household management" was itself the management of the "public economv." (Lea-cock 1978:253)
Sudarkasa has made much the same point about women in West African societies such as
the Yoruba. She argues that many of the political and economic activities anthropologists discuss as public are actually embedded in households (Sudarkasa 1976, as quoted in Rapp 1979:509). Furthermore, "in West Africa, the 'public domain' was not conceptualized as 'the world of men.' Rather, the public domain was one in which both sexes were recognized as having important roles to play" (Sudarkasa 1986:99).
A more appropriate conception would be I to recognize two domains, "one occupied by i men and another by women, both of which j were internally ordered in a hierarchical fash-i ion and both of which provided 'personnel' :        for domestic and extradomestic (or public)
activities" (Sudarkasa 1986:94). f: Furthermore, a careful examination of
1 "domestic domain" indicates that the catcgo-i ries of "woman" and "mother" overlap in j Western society, but the meaning of mother-? hood may be vastly different in another soci-\ ety. Women may not be exclusively defined as \ mothers and childrearers in terms of their sta-| tus and cultural value (see Moore 1988:20-29 | for a discussion of this point). * In addition to the issue of whether the do-
, s       mestic-public dichotomy can provide an ade-i        qtiate description of men's and women's spatial |       and functional relationships in our own and i       other societies, the model has problems as an I       explanation of women's status. One of these problems is the inherent circularity of the I       model. A central point is to account for the l       nature of these domains, yet they are already I       assumed to exist widely and are treated ascat-j,       egories in terms of which women's activities f       (such as food preparing, cooking, child care, |       washing) can be classified (as opposed to male j       hunting, warfare, political councils). Com-|       arofT says that the model "can only affirm \       what has already been assumed—that is, that |       the distinction between the domestic and po-\       litico-jural is an intrinsic, if variable, fact of so-|       cial existence" (Comaroff 1987:59). When the |       model is used to explain women's positions in different societies in relation to these two orientations, the reasoning is equally circular. To put it in the words of Yanagisako and Collier, "The claim that women become ab-
sorbed in domestic activities because of their role as mothers is tautological given the definition of'domestic' as 'those minimal institutions and modes of activity that are organized immediately around one or more mothers and their children'" (Yanagisako and Collier 1987:19).
Finally, we have come to realize that the concepts of domestic and public were bound up in our own history and our own categories grounded particularly in a Victorian heritage. Rosaldo, in a thoughtful revaluation of her model, came to argue this position herself.
The turn-of-thc-cen airy social theorists whose writings are the basis of most modern social thinking tended without exception to assume that women's place was in the home. In Fact, the Victorian doctrine of separate male and female spheres was, I would suggest, quite central to their sociology. Some of these thinkers recognized that modern women suffered from their association with domestic life, but none questioned the pervasiveness (or necessity) of a split between the family and society. (Rosaldo 1980:401-402)
Rosaldo traced the historical roots of domestic-public from the nineteenth century evolutionists through twentieth century structural functionalists to her own work. Instead of two opposed spheres (different and apart), Rosaldo suggested an analysis of gender relationships, an examination of inequality and hierarchy as they are created particularly through marriage (Rosaldo 1980:412-413).
The dichotomy has been usefully employed in several ways since 1974. First, several authors have shown us how it works in Western societies (e.g., France and the United States where it arose historically and still has an important ideological function) (Reiter 1975; Collier, Rosaldo, and Yanagisako 1982). In a related way analysts have explored the meanings surrounding domestic activities of women, putting together a much more complex picture of women's relation to men in this sphere (Murcott 1983; Chai 1987; both are reprinted in this book). Second, anthropological analysis has helped us to understand the historical development
72    Domestic Worlds and Public Worlds
of domestic-public spheres in socieiies under colonialism. John Comaroffs analysis of die Tshidi chiefdom in South Africa during the early twentieth century is an excellent example of this approach'(1987:53-85). Finally, some analysts have used the cultural concepts of other societies to critique our own model of domestic-public orientations. Sylvia Yanagi-sako's essay on the clear separation of "inside-outside" domains (a spatial metaphor) and "work-family" activities (a functional dichotomy) in Japanese American culture demonstrates how the anthropological model of domestic-public mixes these metaphors, which has made analysis confusing and difficult (Yanagisako 1987).
Despite these useful attempts at examining women's lives through the lens of a domestic-public opposition, many of us would agree with Rayna Rapp's 1979 summary of the problems with this dichotomy.
We cannot write an accurate history of the West in relation to the Rest until we stop assuming that our experiences subsume everyone else's. Our public/private conflicts are not necessarily the same as those of other times and places. The specific oppression of women cannot be documented if our categories are so broad as to decontextualize what "womaness" means as we struggle to change that definition. A Tanzanian female farmer, a Mapuchc woman leader, and an American working-class housewife do not live in the same domestic domain, nor will the social upheavals necessary to give them power over their lives be the same. We must simultaneously understand the differences and the similarities, but not by reducing them to one simple pattern. (Rapp 1979:511)
Thus, many of us have tired of the domestic-public dichotomy. We feel it is constraining, a "trap," while new approaches try to get away from dichotomous thinking. These approaches do one of several things. Often they take history seriously, examining women's situation as it has evolved, often in a colonial context. Furthermore, they treat women as active agents and following Collier (1974), as people who have interests, often divergent from men, and who act on them. Third, they
often focus on gender relationships, rather than only on women. Finally, they do not treat all women as part of a single universa' category of "woman." Rather women are usu ally analyzed in terms of their social location Age, class, race, ethnicity, and kinship are all likely to divide women, so newer analyses examine women's strategies and identities a; they are differently shaped. Several examples will illustrate some of the different ap^ proaches taken in recent years.
Collier's examination of Comanche, Cheyenne, and Kiowa gender relationships (1988) illustrates the recent focus on gender and on the multiple positions that men and women hold in societies in which the domestic-public dichotomy seems inappropriate. This is because these "spheres" are integrated, and there is no firm line between domestic and public space (see Lamphere 1974 and Lea-cock above).
The Comanche are an example of a bride service society in which, like many hunter-gather societies, men and women were relatively autonomous, the concept of femininity was not elaborated, and the greatest status differences were between unmarried and married men. Marriage established men as having something to achieve (e.g., a wife), leaving women without such a cultural goal. Young men, through providing meat for their in-laws (bride service), become equal adults, and older men, through egalitarian relations and generosity, become the repositories of wisdom and knowledge. Politics focused on the issue of sexuality and on male-male relationships, which often erupted in conflict and violence. Women celebrated their health and sexuality, and hence the roles of "woman the gatherer" or even "woman the mother" did not emerge as cultural themes.
Among the Cheyenne, an equal bride-wealth society, and among the Kiowa, an unequal bridewealth society, marriage relationships were structured in a much different way in the nineteenth century, so gender relationships had a much different content, politics were more hierarchical, and ideology played a different role. Collier's interest is not in the subordination of women in these three socie-
The Domestic Sphere of Women and the Public World of Men 73
ties, because in all three there are several {dnds of inequality: between men and women, between older women and girls, between un-niarried men and married men, and between kin and affines. An interest in "spheres" and "domains" has been replaced by an emphasis on relationships and an analysis that focuses on the ways in which inequality gets reproduced through marriage transactions, claims on the labor of others, and giving and receiving of gifts. Dominance and subordination become a much more layered, contextualized phenomenon—more interesting than the simple assertion that women are universally subordinated. The processes through which women's inequality (and that of young men) is constructed are laid bare, rather than flatly asserted.
Mary Moran's study of civilized women (1990) explores the historical beginnings and present day construction of the category "civilized," which does confine educated women among the Glebo of southeastern Liberia to a "domestic sphere." The dichotomy between "civilized" and "native" (or even tribal or country) is a result of missionization and has created a status hierarchy differentially applied to men and to women. Men, once educated and with a history of paid wage work, never lose their status as "civilized," while women, even though married to a "civilized man," may lose their status if they do not dress correctly, keep house in specific ways, and refrain from farming and marketing. Native women, who market or have farms, are more economically independent but occupy positions of lower prestige. Here we see not only the importance of historical data in examining how cultural categories evolve, but also the ways in which both civilized and native women actively manage their status positions. Civilized women, through the practice of fosterage, recruit younger women to their households to cany out the more elaborate household routines in which they must engage and to train these fostered daughters to become civilized themselves.
The civilized-native dichotomy represents the juxtaposition of two systems. One is a parallel-sex system in which native men and
women are represented by their own leaders in two linked but relatively autonomous prestige hierarchies (as suggested by Sudarkasa 1986). The other is a single-sex system (based on a Western model) in which men in political positions represent both sexes, and women have little access to prestige except through their husbands. Thus, this is a much more complex system than one based on a domestic-public dichotomy. There are dichotomous categories—civilized-native, male-female— but they do not fit neatly together. Moran speaks of categories as "gender sensitive" and suggests that "The Glebo have inserted gender into the civilized/native dichotomy to the point that women's status is not only more tenuous and vulnerable than men's but also very difficult to maintain without male support." In some respects civilized women trade off dependency for prestige, but Moran provides a sympathetic picture of how both civilized and native women manage their lives.
Lila Abu-Lughod's study (1986) of Bedouin women's ritual poetry gives us further insights into the complexity of women who in 1974 we would have simply thought of as "confined to a domestic sphere." Among the Bedouin women's marriages are arranged; wives wear black veils and red belts (symbolizing their fertility); and women must behave within a code of behavior that emphasizes family honor and female modesty and shame. When confronted with loss, poor treatment, or neglect, the public discourse is one of hostility, bitterness, and anger. In the case of lost love the discourse is of militant indifference and denial of concern. In contrast, Bedouin poetry, a highly prized and formally structured art, expresses sentiments of devastating sadness, self-pity, attachment, and deep feeling (Abu-Lughod 1986:187). Although both men and women recite poetry for women it may express conflicting feelings concerning an arranged marriage, a sense of loss over a divorce, or sentiments of betrayal when a husband marries a new wife. The poems are used to elicit sympathy and get help, but they also constitute a dissident and subversive discourse. Abu-Lughod sees ritual poetry as a corrective to "an obsession with morality and
74    Domestic Worlds and Public Worlds
The Domestic Sphere of Women and the Public World of Men 75
an overzealous adherence to the ideology of honor. . . . Poetry reminds people of another way of being and encourages, as it reflects, another side of experience. . . . And maybe the vision [offered through poetry] is cherished because people see chat the costs of this system, in the limits it places on human experiences, are just too high" (Abu-Lughod 1986:259). Bedouin women in this portrait are not simply victims of patriarchy confined to a domestic sphere; thev are active individuals who use a highly valued cultural form to express their deepest sentiments, acknowledge an alternative set of values, and leave open the possibility of subverting the system in which they are embedded.
A large number of studies have been conducted in the United States that loosely focus on what used to be termed the domestic sphere and the public world of work. As in the Native American, African, and Middle Eastern cases cited previously, when one begins to examine a topic in detail, global notions like domestic-public seem too simple to deal with the complexities of women's lives. Clearly work and home are distinctly separated spheres in the United States. Women who have been employed in the paid labor force have experienced the disjunction of spending eight or more hours of the day in a place of employmentwhere they are "female workers" and the rest of their time in the home where they are daughters, wives, and/or mothers. With this comes responsibilities for cooking, cleaning, and providing nurturance, care, and intimacy for other family members. Several recent studies have examined the contradictions women face when combining work and family, the impact of paid employment on family roles, and vice versa. I will refer to only three examples of this growing literature.
Patricia Zavella's research on Chicana cannery workers examines women's networks that link the workplace and the family (Zavella 1987). Calling these "work-related networks," Zavella describes groups of friends who saw each other outside work and who were members of a kin network employed in the same cannery. Women used work-related
networks as sources of exchange for information, baby sitters, and emotional support. Networks operated in more political ways as workers organized a women's caucus and filed a complaint with the Fair Employment Practices Commission. Women's cannery work was seasonal and had relatively little impact on power relations in the family or the household division of labor. On the other hand work-related networks of friends or kin were an important "bridging mechanism" helping women to deal with the contradictions and demands that came from two different spheres.
Karen Sacks' study of hospital workers at the Duke Medical center examines the ways in which black and white women brought family notions of work, adulthood, and responsibility to work with them and used these values to organize a walk out and subsequent union drive (1988). Sacks focuses on the activities of "center women"—leaders in the union drive. Unlike the men who were often the public speakers at rallies and events, the center women organized support on an interpersonal, one-to-one basis. Rather than emphasizing the bridging aspect of women's networks, Sacks shows how the family is "brought to work" or in the old terminology how the "domestic" influences die "public."
In my own research I have traced the changes in the relationship between women, work, and family historically through the study of immigrant women in a small industrial community, Central Falls, Rhode Island (Lamphere 1987). Using the twin notions of productive and reproductive labor, I examined the rise of the textile industry in Rhode Island and the recruitment of working daughters and later of working mothers to the textile industry and to the other light industries that have replaced it since World War II. Rather than seeing production and reproduction as a rigid dichotomy (like public and domestic), I have used these categories to study relationships and to examine the kinds of strategies that immigrant women and their families forged in confronting an industrial system where wage work was a necessity and where working-class families had no control
over the means of production. Such an approach revealed a great deal of variability both between and within ethnic groups—the | Irish, English, French-Canadian, and Polish
t families who came to Central Falls between
f; 1915 and 1984 and the more recent Colom-
bian and Portuguese immigrants. Examination of strikes and walk outs in the 1920s and 1930s and my own experience as a sewer in an ? apparel plant in 1977 led me to emphasize
the strategies of resistance the women work-|' ers used on the job, as well as the impact of
f women's paid labor on the family itself. When I daughters were recruited as workers in textile
i mills, the internal division of labor within the
s household did not materially change because
s. wives and mothers continued to do much of i; the reproductive labor necessary to maintain
the household. Fathers, teenage sons, and I daughters worked for wages. In the current
J1 period, in contrast, as more wives have be-
come full-time workers, immigrant men have begun to do some reproductive labor, parties',        ularly child care. Immigrant couples often work different shifts and prefer to care for i. children themselves rather than trust baby sit-
ters from their own ethnic group. In my study ; I argue that "the productive system as consti-
r tuted in the workplaces has shaped the family I more  than  issues  of reproduction have
; shaped the workplace" (Lamphere 1987:43).
t More recently Patricia Zavella, Felipe Gon-
\ zalez, and I have found that young working
I mothers in sunbelt industries have moved
much further than Cannery women or New | England industrial immigrant women in 7 changing the nature of the household divi-i sion of labor (Lamphere, Gonzalez, and
f Zavella nd). These new committed female
j:        workers have been employed since high school and do not drop out of the labor force for long periods of time to have children. Thus, they and their husbands construct a f        family life around a two-job household. Al-! though some couples have a "traditional" di-
vision of housework (women do the cooking and the majority of the cleaning and husbands take out the garbage, do minor repairs, and Fix the car), many husbands participate in "female chores" and do substantial amounts
of child care (often caring for children while the wife is at work). Here we see the impact of what we used to call the "public sphere" on the domestic one, but in our analysis we have focused more on the varied ways that Anglos and Hispanics (including single mothers) have negotiated household and child-care arrangements, viewing husbands and wives as mediating contradictions. Subtle similarities and differences among and between working class Anglo and Hispanic women have emerged from this analysis, making it clear that the impact of work in the public work is not a monolithic but a variegated process.
In summary the dichotomy between the public world of men and domestic world of women was, in 1974, an important and useful starting point for thinking about women's roles in a cross-cultural perspective. As anthropologists have written more detailed and fine-grained studies of women's lives in a wide variety of other cultures and in our own society, we have gone beyond the use of dichotomies to produce analyses of the complex and layered structure of women's lives. We now treat women more historically, viewing them as social actors and examining the variability among women's situations within one culture and in their relationship to men.
NOTES
1. Rosaldo says that "the opposition does not determine cultural stereotypes or asymmetries in the evaluations of the sexism, but rather underlies them, to support a very general . . . identification ofwomen with domestic life and of men with public life" (Rosaldo 1974:21™ 22). Thus, I would argue, Rosaldo did not attempt to explain women's subordination through the dichotomy, but saw it as an underlying structural framework in any society that supported subordination and that would have to be reorganized to change women's position.
2. It is interesting that we did not know of Elsie Clews Parsons' extensive feminist writing during 1910 to 1916, much of which is reminiscent of the kind of position we took in Woman, Culture, and Society. In another article I have
76    Domestic Worlds and Public Worlds
Food, Mealtimes and Gender in Some 5outh Wales Households 77
noted the similarities between Shelly's prose and that of Parsons (see Lamphere 1989 and Parsons 1913, 1914, 1915).
REFERENCES
Abu-Lughod, Lila. 19813. Veiled Sentiments: Honor and Poetry in a Bedouin Society. Berkeley and Los Angeles: University of California Press.
Brown, Judith. 1970. Economic organization and the position ofwomen among the Iroquois. Eth-noliislon 17(3/4): 131-167.
Chai, Alice Yun. 1987. Freed from the elders but locked into labor: Korean immigrant women in Hawaii. Women's Studies 13:223-234.
Collier, Jane. 1974. Women in politics. In Michelle Z. Rosaldo and Louise Lamphere (eds.). Woman, Culture, and Society. Stanford: Stanford University Press.
-. 1988. Marriage and Inequality hi Classless
Societies. Stanford: Stanford University Press.
Collier, jane, Michelle Rosaldo, and Sylvia Yanagisako. 1982. Is there a family? New anthropological views. In Barrie Thorne and Marilyn Yalom (eds.). Rethinking the Family: Some Feminist Qjiestions. New York and London: Longman.
Comaroff, John L. 1987. Sui genderis: Feminism, kinship theory, and structural "domains." In Jane Fishburne Collier and Sylvia Junko Yanagisako (eds.). Gender and Kinship: Essays Toward a Unified Analysis. Stanford: Stanford University Press.
de Beauvoir, Simone. 1953. The Second Sex. New York: Alfred A. Knopf. Originally published in French in 1949.
Kaberry, Phyllis M. 1939. Aboriginal Women, Sacred and Profane. London: G. Routledge.
-■—. 1952. Women of the Grnssjields. London:
H. M. Stationery Office.
Lamphere, Louise. 1974. Strategies, cooperation, and conflict among women in domestic groups. In Michelle Z. Rosaldo and Louise Lamphere (eds.). Woman, Culture, and Society. Stanford: Stanford University Press.
-■—. 1987. From Working Daughters to Working
Mothers: Immigrant Women in a New England Industrial Community. Ithaca, NY: Cornell University Press.
-. 1989. Feminist anthropology: The legacy
Zavella. (eds.). Working Mothers and Sunbelt Industrialization: New Patterns of Work and Family, Submitted to Cornell University Press.
Landes, Ruth. 1938. The Ojibaw Woman, Part I: Youth. New York: Columbia University. Contributions to Anthropology-, Vol. 31.
-. 1947. The City of Women: Negro Women Cull
Leaders of Bahia, Brazil. New York: Macmillan.
Leacock, Eleanor. 1978. Women's status in egalitarian society: Implications for social evolution. Current Anthropology 19(21:247-275.
Lcith-Ross, Sylvia. 1939. African Women: Study of the I bo of Nigeria. London: Faber and Faber.
Mead, Margaret. 1949. Male and Female. New York: William Morrow and Co.
Moran, Mary H. 1990. Civilized Women: Gender and Prestige in Southeastern Liberia. Ithaca. NY: Cornell University Press.
Moore, Henrietta L. 1988. Feminism and Anthropology. Minneapolis: University of Minnesota Press.
Murcott. Anne. 1983. "It's a pleasure to cook for him": Food, mealtimes and gender in some South Wales households. In Eva Gamarnikow, D. H. J. Morgan, June Pin-vis, and Daphne Taylorson (eds.). The Public nud the Private. London: Heinemann Educational Books.
Nelson, Cynthia. 1974. Public and private politics: Women in the Middle Easl. American Ethnologist 1:551-563.
Parsons, Elsie Clews. 1913. The Old Fashioned Woman. New York: G. P. Putnam's Sons.
-. 1914. Fear and Conventionality. New York:
G. P. Putnam's Sons.
-. 1915. Social Freedom: A Study of the Conflicts
of Elsie Clews Parsons. American Ethnologist 16(3):518-533. Lamphere, Louise, Felipe Gonzales, and Patricia
Between Social. Classifications and Personality. New York: G. P. Putnam's Sons. Ong, Aihwa. 1987. Spirits of Resistance and Capitalist Discipline. Albany, NY: State University of New York Press.
Paulme, Denise (ed.). 1963. Women of Tropical Africa. Berkeley: University of California Press.
Rapp, Rayna. 1979. .Anthropology. Signs 4(3): 497-513.
Reiter, Rayna (ed.). 1975a. Toward an Anthropology of Women. New York: Monthly Review Press.
-. 1975b. Men and women in the South of
France: Public and private domains. In Rayna Reiter (ed.). Toward an Anthropology of Women. New York: Monthly Review Press.
Rosaldo, Michelle. 1974. Woman, culture and society: A theoretical overview. In Michelle Z. Rosaldo and Louise Lamphere (eds.). Woman, Culture, and Society. Stanford: Stanford University Press.
---. 1980. The uses and abuses of anthropology- Signs 5(3): 389-417.
Rosaldo, Michelle Z. and Louise Lamphere (eds.). 1974. Woman, Culture, and Society. Stanford: Stanford University Press.
Sacks, Karen. 1988. Caring by the Hour: Women, Work, and Organizing at the Duke Medical Center. Urbana and Chicago: University of Illinois Press.
Sudarkasa, Niara. 1976. Female employment and family organization in West Africa. In Dorothy McGuigan (ed.). New Research on Women and Sex Roles. Ann Arbor: Center for Continuing Education of Women.
-. 1986. The status of women in indigenous
.African Societies. Feminist Studies 12: 91-104.
Underhill, Ruth. 1936. Autobiography of a Papago Woman. Supplement to American Anthropologist 38(3), Part II. Millwood, NY: .American Anthropological Association.
Wolf, Margery. 1972. Women and the Family in Rural Taiwan. Stanford: Stanford Universitv Press.
Yanagisako, Sylvia Junko. 1987. Mixed metaphors: Native and anthropological models of gender and kinship domains. In jane Fishburne Collier and Svlvia Junko Yanagisako (eds.). Gender and. Kinship: Essays Toward a Unified Analysis. Stanford: Stanford University Press.
Yanagisako, Sylvia junko and Jane Fishburne Collier. 1987. Toward a unified analysis of gender and kinship. In Jane Fishburne Collier and Sylvia junko Yanagisako (eds.). Gender and Kinship: Essays Toward a Unified Analysis. Stanford: Stanford University Press.
Zavella, Patricia. 1987. Women's Work and Chicano Families: Cannery Workers of the Santa Clara Valley. Ithaca, NY: Cornell University Press.
Anne Murcott
INTRODUCTION
I think it lets him know that I am thinking about
.... a
him—as il he knows that I am expecting him.
But it's not as if 'oh I haven't got anvthing
ready' . . . Fair play, he's out all day . . . he
doesn't ask ibr that much . . . you know it's not
as if he's been very demanding or—he doesn't
come home and say 'oh, we've got chops again',
it's really a pleasure to cook for him, because
Reprinted with permission from Eva Gamarnikow, David H.J. Morgan, June Purvis, and Daphne Taylorson (eds.), The Public anil the Private (Lundnn: Heinemann Educational Books, Ltd., 1983), pp. 78-90. Copyright 9 1983 British Sociological Association.
whatever you ... oh I'll give him something and I think well, he'll like this, he'll like that. And he'll always take his plate out. . . and he'll wash the dishes without me even asking, if I'm busy with the children. Mind, perhaps his method is not mine.
Every now and then an informant puts precisely into words the results of the researcher's analytic efforts—providing in the process a quotation suitable for the title! The extract reproduced above, explaining the importance of having the meal ready when her husband arrives home, comes from one of a series of interviews on which this paper is
78    Domestic Worlds and Public Worlds
Food, Mealtimes and Gender in 5ome South Wales Households 79
based.1 The discussion starts by remembering that 'everyone knows' that women do the cooking: all the women interviewed—and the few husbands/boyfriends or mothers who came in and out—took it for granted that cooking was women's work. Informants may not enjoy cooking, or claim not to be good at it; they may not like the arrangement that it is women's work, or hanker after modifying it. But all recognise that this is conventional, some volunteer a measure of approval, most appeared automatically to accept it, a few resigned themselves and got on with it.
Studies of the organisation of domestic labour and marital role relationships confirm that cooking continues to be a task clone more by women than men; this is also the case cross-culturally (Stephens, 1963; Murdock and Provost, 1973). Emphasis in the literature has shifted from Young and Willmott's (1975) symmetrical view of sharing and marital democracy. Now rather more thoroughgoing empirical study suggests their assessment is little more than unwarranted optimism (Oakley, 1974a and b; Edgell, 1980; Leonard, 1980; Tolson, 1977}. This work improves on earlier studies of the domestic division of la-bour by going beyond behaviourist enquiry about 'who does which tasks' to consider the meanings attached to them by marital partners. The distribution of work turns out not to correlate neatly with assessments of importance or allocation of responsibility. (Oakley, 1974b; Edgell, 1980).
Part of this effort (in particular, Oakley, 1974a and b) has in addition attempted to analyse domestic work as a 'job like any other, considering housewives' work satisfaction, routines, supervision and so on. While this line of enquiry has undoubtedly made visible much of women's lives conventionally rendered invisible, it has perhaps not gone far enough. The study of housework as an occupation needs to attend in addition to features such as quality control, timekeeping, client as well as worker satisfaction, and perhaps further consideration of who, if anyone, is a housewife's boss. As will be seen, each of these is implicated in the discussion that follows.
These occupational aspects of housework provide, moreover, additional means of examining the relationship of the domestic division of labour to the economic structure as a whole. Recent commentary has also proposed that the view of the family as stripped of all but the residual economic function of consumption is ill-conceived and over-simplified. Domestic labourers refresh and sustain the existing labour force and play a key part in reproducing that of the future—as well as providing a reserve of labour themselves. The precise manner in which the political economy is to be accounted continues to be debated (West, 1980; Fox, 1980; Wajcman, 1981). For the moment, however, the general drift of that discussion can be borne in mind by recalling the everyday terminology of eating; food is consumed, meals have to be produced. The language favoured in cookbooks echoes that of industry and the factory (Murcott, 1983a). Homecooking may nicely embody the terms in which the family and household's place in the division of labour has to be seen. It may also provide a convenient arena for the further exploration of the economic and labour relations in the family and the relation of the marital partners to the means of production of domestic labour (Middleton, 1974).
Examination of the household provision of meals in these terms is, however, some way in the future. This paper does no more than offer some empirical foundation on which such study might build. It brings together informants' ideas about the importance of cooking, their notions of propriety of household eating and indicates their relation to gender. It starts with views of the significance of good cooking for home life, and goes on to deal with the place of cooking in the domestic division of labour. The familiar presumption that women are the cooks is extended to show that their responsibility in this sphere is tempered with reference to their husband's, not their own, choice. The paper concludes with brief comment on possible ways these data may illuminate some of the questions already raised.
HOME COOKING
Aside irom love, good food is the cornerstone of a happy household . . . (Opening lines of a 1957 cookbook called The Well Fed Bridegroom).
Right through the series of interviews three topics kept cropping up; the idea of a proper meal, reference to what informants call a 'cooked dinner' and the notion that somehow home is where proper eating is ensured. Moreover, mention of one like as not involved mention of another, sometimes ail three. The composite picture that emerges from the whole series suggests that these are not merely related to one another in some way, but virtually equated.
It first needs to be said that informants seemed quite comfortable with a conception of a proper meal—indeed the very phrase was used spontaneously—and were able to talk about what it meant to them. Effectively a proper meal is a cooked dinner. This is one which women feel is necessary to their family's health, welfare and, indeed, happiness. It is a meal to come home to, a meal which should figure two, three or four times in the week, and especially on Sundays. A cooked dinner is easily identified—meat, potatoes, vegetables and gravy. It turns out that informants displayed considerable unanimity as to what defines such a dinner, contrasting it to, say, a 'snack' or 'fried'. In so doing they made apparent remarkably clear rules not only for its composition but also its preparation and taking. I have dealt with their detail and discussed their implications in full elsewhere (Murcott, 1982). But in essence these rules can be understood as forming part of the equation between proper eating and home cooking. And, as will be noted in the next section, they also provide for the symbolic expression of the relationship between husband and wife and for each partner's obligation to their home.
The meal for a return home is, in any case, given particular emphasis—a matter which cropped up in various contexts during the interviews. Thus, for some the very importance
of cooking itself is to be expressed in terms of homecoming. Or it can provide the rationale for turning to and making a meal, one to be well cooked and substantial—not just 'beans on toast. . . thrown in front of you'.
The actual expression 'home cooking'—as distinct from 'cooking for homecoming'—received less insistent reference. Informants were straightforward, regarding it as self-evident that people preferred the food that they had at home, liked what they were used to and enjoyed what they were brought up on. Perhaps untypically nostalgic, one sums up the point:
When my husband comes home . . . there's nothing more he likes I think than coming in the door and smelling a nice meal cooking. I think it's awful when someone doesn't make the effort... I think well if I was a man I'd think I'd get really fed up if my wife never bothered . . .
What was prepared at home could be trusted—one or two regarded the hygiene of restaurant kitchens with suspicion, most simply knew their chips were better than those from the local Chinese take-away or chippy. Convenience foods had their place, but were firmly outlawed when it came to a cooked dinner. In the ideal, commercially prepared items were ranged alongside snacks, and light, quick meals: lunches and suppers in contrast to proper dinners. Informants talked about home cooking, but used this or some such phrase infrequently; the following is an exception:
I'd like to be able to make home-made soups and things, it's just Finding the time and getting organised, but at the moment I'm just not organised ... I think it would probably be more good for us than buying... 1 suppose it's only—I'd like to be—the image of the ideal housewife is somebody who cooks her own food and keeps the household clean and tidy.
The sentiments surrounding her valuation of home-made food are not, however, an exception. Time and again informants linked not only a view of a proper meal for home-
80    Domestic Worlds and Public Worlds
Food, Meaitimes and Gender in Some South Wales Households 81
coming, but a view of the proper parts husband and wife are to play on this occasion. So cooking is important when you are married.
you must think of your husband . . . it's a long day for him at work, usually, . . . even if they have got a canteen at work, their cooking is not the same as coming home to your wife's cooking ... I think every working man should have a cooked meal when he comes in from work . . .
Cooking is important—though not perhaps for everybody 'like men who don't cook'—for women whose 'place [it is] to sec the family are well fed'.
In this section, I have indicated that informants virtually treat notions of proper meals, home-based eating and a cooked dinner, as equivalents. The stress laid on the homecoming not only underlines the symbolic significance attached to both the meal and the return home. It simultaneously serves as a reminder of the world beyond the home being left behind for that day. Put another way, the cooked dinner marks the threshold between the public domains of school or work and the private sphere behind the closed front door. In the process of describing these notions of the importance of cooking in the home, it becomes apparent that the familiar division of labour is assumed.
COOKING IN THE DOMESTIC DIVISION OF LABOUR
As noted in an earlier section, all those interviewed took it for granted that it is the women who cook. What they had to say refers both to conventions in general, and themselves and their circumstances in particular." There are two important features of their general presumption that women are the cooks; one indicates the terms in which it is modifiable, the other locates it firmly as a matter of marital justice and obligation. The upshot of each of these is to underline the manner in which the domestic preparation of meals is securely anchored to complementary concepts of conduct proper to wife and husband.
To say that women cook is not to say that it is only women who ever do so. It is, however, to say that it is always women who daily, routinely, and as a matter of course are to do the cooking. Men neither in the conventional stereotype nor in informants' experience ever cook on a regular basis in the way women do.1 Husband/boyfriends/fathers are 'very good really'; they help informants/their mothers with carrying the heavy shopping, preparing the vegetables, switching the oven on when told, doing the dishes afterwards (cf. Leonard, 1980). Such help may be offered on a regular enough basis, notably it is available when the women are pregnant, dealing with a very young infant, unwell or unusually tired. But none of this is regarded as men doing the cooking.
More significantly, it is not the case that men do not cook—in the strict sense of taking charge of the transformation of foodstuffs to some version of a meal. They may make breakfast on a Sunday, cook only 'bacon-y' things, can do chips or 'his' curries: all examples, incidentally, of foods that do not figure in the proper cooked dinner (Murcott, 1983c).
For some, however, competence in the kitchen (and at the shops) is suspect: he'll 'turn the potatoes on at such and such a time . . . but leave him he's hopeless' and another just 'bungs even-'thing in'. For others, it is men who make better domestic cooks than women, are more methodical, less moody. Another couple jokingly disagree: she 'not taken in' by Robert Carrier on TV, he claiming that 'the best chefs are men'. The point is that either way, of course, informants do regard gender as relevant to the question of who is to cook.
It is not even the case that all men cannot cook the proper, homecoming meal. One or two, when out of work for a while, but his wife still earning (this only applied to those having a first baby) might start the meal or even have it ready for her return. But once he is employed again he does not continue to take this degree of responsibility, reverting either to 'helping' or waiting for her to do it. Now and again, wives have learned to cook not at
school or from their mothers, but from their husbands. But it was still assumed that it was for the woman to learn. This was even so in one instance where the informant made a 'confession . . . my husband does the cooking'. But now that she was pregnant and had quit paid work she would take over; 'it would be a bit lazy not to'. Like others for whom the cooking may have been shared while both were employed, cooking once again became the home-based wife's task (cf. Bott, 1957, p. 225; Oakley, 1980, p. 132).
The issue is, however, more subtle than an account of who does what, or who takes over doing what. Men and women's place involve mutual obligation. 'I think a woman from the time she can remember is brought up to cook . . . Whereas most men are brought up to be the breadwinner.' The question of who does the cooking is explicitly a matter of justice and marital responsibility. A woman talks of the guilt she feels if she does not, despite the greater tiredness of late pregnancy, get up to make her husband's breakfast and something for lunch—'he's working all day'. Another insists that her husband come shopping with her so he knows the price of things—he's 'hopeless' on his own—but she has a clear idea of the limits of each person's responsibility: each should cook only if the wife has to earn rather than chooses to do so.
Here, then, I have sought to show that informants subscribed in one way or another to the convention that it is women who cook. In the process it transpired that it is certain sorts of cooking, i.e. routine, homecoming cooking, which are perennially women's work. The meal that typically represents 'proper' cooking is, of course, the cooked dinner. Its composition and prescribed cooking techniques involve prolonged work and attention; its timing, for homecoming, prescribes when that work shall be done. To do so demands the cook be working at it, doing wifely work, in time that corresponds to time spent by her husband earning for the family (Murcott, 1982). This is mirrored in Eric Batstone's (1983) account of the way a car worker's lunch box prepared by his wife the evening before is symbolic of the domestic relationship which
constitutes the rationale for his presence in the workplace; he endures the tedium of the line in order to provide for his wife and family. It transpired also that men do cook in certain circumstances, but such modification seems to reveal more clearly the basis for accounting cooking as part of a wife's responsibility (to the family) at home corresponding to the husband's obligation (to the family) at work, i.e. their mutual responsibilities to each other as marriage partners.
WHO COOKS FOR WHOM?
At this point I introduce additional data which bear on cooking's relation to the question of marital responsibility. Repeatedly informants indicated that people do not cook for themselves; evidently it is not worth the time and effort.1 But the data suggest implications beyond such matters of economy. Two interrelated features are involved: one is the distinction already alluded to in the previous section, between cooking in the strict sense of the word and cooking as preparation of a particular sort of meal. The other enlarges on the following nicety. To observe that people do not cook for themselves can mean two things. First it can imply that a solitary person does not prepare something for themselves to eat while on their own. But it can also imply that someone does not do the cooking on their own behalf, but in the service of some other(s). Examination of the transcripts to date suggests that not only could informants mean either or both of these, but also that each becomes elided in a way that underlines the nuances and connotations of the term cooking.
The question of a lone person not cooking themselves a meal unsurprisingly cropped up most frequently with reference to women themselves, but men, or the elderly were also thought not to bother.
Informants are clear, however, that not cooking when alone does not necessarily mean going without. Women 'pick' at something that happens to be in the house, have a bar of chocolate or packet of crisps later in the
82    Domestic Worlds and Public Worlds
Food, Mealtimes and Gender in Some South Wales Households 83
evening or a 'snack'. Men will fry something, an egg or make chips. No one said that a man would go without altogether (though they may not know), whereas for themselves—and women and girls in general—skipping a meal was thought common enough. Men—and occasionally women—on their own also go back to their mother's or over to their sister's for a meal. One informant was (the day of my interview with her) due to go to her mother's for the evening meal, but fearful of being alone in the house at night, she was also due to stay there for the next few days while her husband was away on business.
The suggestion is, then, that if a person is by themselves, but is to have a proper meal, as distinct from 'fried' or a 'snack' then they join a (close) relation's household. The point that it is women who cook such meals receives further emphasis. Indeed, when women cook this particular meal, it is expressly for others. In addition to the temporary lone adults just noted who return to mothers or sisters, women in turn may cook for the older generation, as well as routinely cooking for children or for men home at 'unusual' times if unemployed or temporarily of a different shift.
This conventional requirement that women cook for others is not always straightforward in practice. At certain stages in an infant's life the logistics of producing meals for husband and child(ren) there as well meant the woman felt difficulties in adequately meeting the obligations involved. And not all informants enjoyed cooking; most just accepted that it needed doing, though there were also those who took positive, creative pleasure in it (cf. Oakley, 1974b). Part of this is expressed in the very satisfaction of providing for others something they should be getting, and in turn will enjoy.
More generally cooking can become tiresome simply because it has to be done day-in, day-out. The pleasure in having a meal prepared for you becomes all the more pointed if routinely you arc cooking for others.' In the absence of any data for men, it can only be a guess that going out lor a meal is thus specially enjoyable for women. But for those who on occasion did eat out this clearly figures in
their pleasure. Even if it rarely happened, just the idea of having it put in front of you meant a treat: 'it's nice being spoiled'.
The question 'who cooks for whom?' can now begin to be answered. Apparently it \< women who cook for others—effectively, hus bands and children. If husbands and childrer are absent, women alone will not 'cook', in deed many may not even eat. It is the others' presence which provides the rationale for women's turning to and making a proper meal—that is what the family should have and to provide it is her obligation. Men—and children—have meals made for them as a matter of routine: but for women it is a treat. That solitary men do not 'cook' for themselves either, and may go to a relative's for meals (cf. Rosser and Harris, 1965; Barker, 1972), or that a woman on her own may also do so does not detract from the main proposal that it is women who cook for others. For it is not only that informants or their husbands will go tern-" porarily back to their mother's, not their father's, home-cooking. It is also that both men and women revert to the status of a child for whom a woman, a mother, cooks. The mother may actually be the adult's parent, but they—and I with them—may stretch the point and see that she may be mother to the adult's nieces or nephews or indeed, as in the case of cooking for the elderly, she may be mother to the adult's grandchildren.
The appreciation that it is women who cook for others elaborates the more familiar convention, discussed above, that in the domestic division of labour cooking is women's work. First of all it indicates that this work is sendee work. Cooking looks increasingly like a task quite particularly done for others. Second, when cooking for others women are performing a service to those who are specifically related (sic) rather than for a more generalised clientele known only by virtue of their becoming customers. The marital—and parental—relationship defines who is server, who served.
That said, there remains the question of deciding what the server shall serve. As already discussed in an earlier section, the conventional expectation shared, it seems, by
both woman and man. is that meals shall be of a certain sort—a cooked dinner for a certain occasion, most commonly the return home from work, or the celebration of Sunday, a work-free day. The 'rules' involved are not entirely hard and fast, or precisely detailed. Cooked dinners are neither daily nor invariable affairs (Murcott, 1982). And the cooked dinner itself can properly comprise a number of alternative meats (and cuts) and range of different vegetables. What then, determines the choice of meat and vegetables served on any particular day? Some of the factors involved, as will be seen in the next section, once again echo ideas of responsibility and mutual obligation.
DECIDING WHAT TO HAVE
A number of factors feature in deciding what to have for a particular day's meal.'1 First, a question of cost was taken for granted. This does not necessarily mean keeping expenditure to a minimum—eating in the customary manner despite hard times was highly and expressly valued by some. Second, the conventional provision of proper dinners itself contributed to the determination of choice. These two factors present themselves as marking the limits within which the finer decisions about what the precise components or the day's dinner are to he. Here reference to their husband's—and, to a lesser extent, children's—preferences was prominent in informants' discussion of such detailed choices.
It was indicated earlier that in an important sense women's cooking is service work. This sort of work has two notable and interrelated aspects affecting decisions and choice: is it the server or served who decides what the recipient is to want? Exploring the mandate for professionals' work, Everett Hughes (1971, p. 424) highlights a key question: 'professionals do not merely serve: they define the very wants they serve'. Servants, and service workers such as waitresses (Whyte, 194S: Spradley and Mann, 1975) compliantly provide for the wants identified by the served. On
the face of it, then, the professional has total and the waitress nil autonomy. Examples reflecting this sort of range occurred among informants varying from one woman apparently always deciding, through to another always making what he wants for tea. But in the same way that the maximum autonomy of the professional is continually, to a certain degree, a matter of negotiation and renegotiation with clients, and that, similarly, the apparent absence of autonomy is modified by a variety of more or less effective devices waitresses use to exert some control over customers, so a simple report of how meal decisions are reached can, 1 propose, either conceal negotiations already complete, or reveal their workings.
Thus informants interested in trying new recipes still ended up sticking to what they usually made because their husbands were not keen. Others reported that 'he's very good' or 'never complains' while some always asked what he wanted. A non-committal reply however did not necessarily settle the matter, for some discovered that being presented with a meal she had then decided on could provoke adverse and discouraging remarks. But it was clear that even those who claimed not to give their husbands a choice were still concerned to ensure that he agreed to her suggestion. It is almost as if they already knew what he would like, needed to check out a specific possibility every now and then but otherwise continued to prepare meals within known limits. Deciding what to have already implicitly took account of his preferences so that the day-to-day decision seemed to be hers.
The material presented in this section provides only a glimpse of this area of domestic decision-making. Other aspects need consideration in future work. For instance, what degree of importance do people attach to the matter (cf. Edgell, 1980, pp. 58-9)? Attention also needs to be paid to wider views of the legitimacy of choice in what one eats. In what sense do restauram customers choose and mentally subnormal patients not? Does a child that spits out what it is fed succeed in claiming a choice or not? And in apparently acquiescing to their husband's choice, are
84    Domestic Worlds and Public Worlds
Food, Mealtimes and Gender in Some South Wales Households 85
wives circumscribing their own? But it looks as if deciding what to have is of a piece with a shared view of marital responsibility whereby he works and so deserves, somehow, the right to choose what she is to cook for him.
GENDER AND THE PRODUCTION OF MEALS
I know a cousin of mine eats nothing but chips, in fact his mother-in-law had to cook him chips for his Christmas dinner and she went berserk . . .
This 'atrocity story' recapitulates various elements of the preceding discussion. Such unreasonableness is, no doubt, unusual but its artless reporting emphasises a number of points already made. Not only do chips break the rules of what should properly Figure in a Christmas meal, superior even to the Sunday variant of a cooked dinner, but it remains, however irksome, up to the woman to prepare what a man wants. The burden of this paper, then, may be summarised as revealing allegiance to the propriety of occasion such thai a certain sort of meal is to mark home (male) leisure versus (male) work-time, and that such meals are cooked by women for others, notably husbands, in deference, not to the woman's own, but to men's taste.
This examination of cooking, mealtimes and gender within the household has implications for the continuing analysis of domestic work as work. While it does not shed light on why such work is women's, only reasserting that conventionally this is so, it clearly casts the work of meal provision as service work.
The everyday way of describing dishing up a meal as serving food is embedded in a set of practices that prescribe the associated social relationships as of server and served. As already observed this involves two interrelated matters: control over the work, and decisions as to what are the "wants' the worker shall serve, what the work shall be. Each is considered in turn.
Oakley (1974a and b) reports that one of the features of housewifery that women value is the feeling of autonomy. Care is needed,
though, not to treat such attitudes as tantamount to their analysis. Just because housewives express their experiences in terms off v enjoying being their own boss does not mean that their conditions of work can be analysed in terms of a high degree of autonomy. The material presented in this paper suggests that doing the cooking is not directed by the woman herself, but is subject to various sorts * of control.
First of these is the prescription for certain kinds of food for certain occasions. The idea of the cooked dinner for a homecoming is just such an example of cultural propriety. Related to this is a second control, namely that the food is to be ready for a specific time. Mealtimes construed in this way may exert just the same sort of pressure on the cook as any other production deadline in industry. \ Third, control is also exerted via the shared understanding that it is the preferences of the consumer which are to dictate the exact van- . ant of the dinner to be served. What he fancies for tea constrains the cook to provide it. These kinds of control in the domestic provision of meals find their counterpart in the industrial concerns of quality control, timekeeping and market satisfaction. A woman \ cooking at home may not have a chargehand ; 'breathing down her neck' which is understandably a source of relief to her. But this does not mean to say that she enjoys autonomy—simply perhaps that other controls make this sort of oversight redundant. Evidence either way is extremely sparse, but Ellis (1983) suggests that failing to cook according to her husband's wishes can contribute to a wife's battering.
Linked to the issue of control of domestic cooking is the question of decision-making. Edgell (1980) has drawn attention to the de- j gree ofimportance couples attach to different aspects of family living about which decisions have to be made. He distinguishes assessments of importance from, first, whether the | decision is mainly the wife's or husband's re- j sponsibility and second, from the frequency with which the decision has to be made. So, for instance, moving is the husband's decision, perceived to be very important and in-
frequent, a contrast to the matter of spending 0n food. What Edgell does not make clear, however, is quite what either his informants or he mean by 'importance'. As an analytic device, the idea does not distinguish between family matters which partners may identify as both important and somehow major or permanent such as moving, and those identified as mundane, or fleeting but important nonetheless, such as daily eating. Like refuse collection or sewage work which is regarded as vital but low status, the importance attached to meals may not be remarked in the general nm of things, though noticed particularly if absent. But that does not necessarily mean that both husband and wife regard it as unimportant. And, harking back to the question of autonomy in decision-making, reports such as Edgell's that food spending, cooking or whatever is regarded as the wife's responsibility, cannot, of itself, be seen as evidence of her power and freedom from control in those areas. For as Jan Pahl (1982, p. 24) has so cogently observed, 'being able to offload certain decisions and certain money-handling chores on to the other spouse can itself be a sign of power'. The delegate may be responsible for execution of tasks, but they are answerable to the person in whom the power to delegate is originally vested.
The preliminary analysis offered in this paper has theoretical and political implications concerning power and authority in marriage and the relation between domestic and paid work. The exploration of ideas about cooking and mealtimes starts to provide additional approach to detailing the means of domestic production. And the sort of work women are to do to ensure the homecoming meal provides a critical instance of the juncture between the control of a worker and the (his) control of his wife. The meal provides one illustration not only of a point where the public world of employment and the private world of the home meet one another; it also shows how features of the public take precedence within the private. For the stress informants lay on this mealtime offers an interesting way of understanding how the industrial rhythms which circumscribe workers are
linked to the rhythms which limit women's domestic work (cf. Rotenberg, 1981). And women's continual accommodation to men's taste can also be seen as a literal expression of wives' deference to husbands' authority (Bell and Newby, 1976; Edgell, 1980, p. 70). This acquiescence to his choice provides the cultural gloss to the underlying economic relationship whereby industry produces amongst other things both the wage, and the raw materials it buys, for the domestic to produce what is needed to keep the industrial worker going. Part of the conjugal contract that each in their own way provide for the other, it does indeed become 'a pleasure to cook for him'.
ACKNOWLEDGEMENTS
I am very grateful to all those necessarily anonymous people who made the research possible and who generously gave their time to answer my questions. I should like to record my appreciation of conversations with Tony Coxon, Sara Delamont, Robert Dingwall, Rhian Ellis, Bill Hudson and Phil Strong at various stages during the preparation of this paper and of the computing help and advice Martin Read provided. Only I and not they are to blame for its deficiencies. And I must thank Lindsey Nicholas, Joan Ryan, Sheila Pickard, Myrtle Robins and Margaret Simpson very much, despite flu all round, and an unusually scrawly manuscript, for their help in typing both drafts.
NOTES
1. In order to begin remedying sociology's neglect (Murcott, 1983b) of food beliefs and of the social organisation of eating, I conducted a single-handed exploratory study (supported by a grant from the SSRC) holding unstructured tape-recorded interviews with a group of 37 expectant mothers attending a health centre in a South Wales valley for antenatal care (22 pregnant for the first time), 20 of whom were interviewed again after the baby's birth. No claim is made for their representativeness in any hard and fast sense,
86    Domestic Worlds and Public Worlds
Korean Immigrant Women in Hawaii 87
though they represent a cross-section of socioeconomic groups. For present purposes the data are treated as providing a composite picture. The prime concern here is to indicate the range and variety of evidence gathered. An instance that occurs once only thus becomes as interesting as one occurring 30 times. This is reflected in the discussion by the deliberate use of phrases such as 'some informants' rather than '6 out of 37'. In any case reference to numbers of instances is no more exact, and risks implying a spurious representativeness.
These qualifications are most important. But for the sake of a tolerably readable account I do not hedge every other sentence with reminder oF these limitations. Yet thev do actively have to be taken as read.
2. Informants referred not only to themselves but also to mothers, sisters, sisters-in-law and women friends doing cooking.
3. No informant who had children old enough to cook currently shared the household with them.
4. Market researchers know how to trade on such reports. During the period of interviewing a TV commercial was running which sought to persuade busy housewives not to neglect themselves but have a frozen ready-cooked meal at [unchtime.
5. Interestingly, no one talked of hospital meals put in front of them as a treat. {None had a home delivery.) Rather it was the quality of the food provided which informants concentrated on. Institution cooking could not be home cooking.
6. It might have been expected that nutritional criteria would figure in these decisions. Analysis so far suggests that cultural prescriptions for proper eating at home override what is known about healthy eating. (Murcott, 1983d).
REFERENCES
Barker, D. L. 1972. 'Keeping close and spoiling,' Sociological Review, 20(4), 569-590.
Batstone, E. 1983. 'The hierarchy of maintenance and the maintenance of hierarchy: Notes on food and industry,' in A. Murcott (ed.), The Sociology of Food and Eating, Gower.
Bell, C. and Newby. H. 1976. 'Husbands and wives: The dynamics of deferential dialectic,' in D. L.
Barker and S. Allen (eds.), Dependence and 1 ploralion in Work and Marriage, Longmans. Bolt, E. 1957. Famih and Social Network, Tavistock, Edgell, S. 1980. Middle Class Couples: A Study of Segregation, Domination and Inequality in Marriage Allen and Unwin. Ellis, R. 1983. 'The way to a man's heart. . . ,' in A Murcott (ed.), The Sociology of Food and Ealing Gower.
Fox, B. (ed.). 1980. Hidden in the Household, To
ronto: Women's Press. Hughes, E. C. 1971. 'The humble and the proud,
in The Sociological Eye: Selected Papers, Aldine-
Atherton.
Leonard, D. 1980. Sex and Generation. Tavistock.
Middleton, C. 1974. 'Sexual inequality and stratification theory,' in E. Parkin (ed.), The Social A nalysis of Class Structure, Tavistock.
Murcott, A. 1982a. 'On the social significance of the "cooked dinner in South Wales," Social Sei- •. ence Information, 21(4/5), 677-695.
-. 1983a. 'Women's place: Cookbook's image
of technique and technology in the British Kitchen,' Women's Studies International Forum, 6(2) (forthcoming).
-. (ed.) 1983b. The Sociology of Food and Eating, Gower.
-. 1983c. 'Cooking and the cooked,' in A.
Murcott (ed.), The Sociology of Food and Ealing, Gower.
-. 1983d. 'Menus, meals and platefuis,' International Journal of Sociology and Social Policy (forthcoming). Murdock, G. P. and Provost, C. 1973. 'Factors in the division of labour by sex: A cross-cultural analysis,' Ethnology, XII(2), 203-225. Oakley, A. 1974a. Housewife. Penguin.
-. 1974b. The Sociology of Housework, Martin
Robertson.
1980. Women Confined: Towards a 'Sociology
of Childbirth,' Martin Robertson.
Pahl. J. 1982. 'The allocation of money and the structuring of inequality within marriage,' Board of Studies in Social Policy and Administration, University of Kent, mimeo.
Rosser, C. and Harris, C. 1965. The Family and Social Change, Routledge and Kegan Paul.
Rotenberg, R. 1981. 'The impact of industrialisation on meal patterns in Vienna, Austria,' Ecology of Food and Nutrition, 11(1), 25-35.
Spradley,J. O. and Mann, B.J. 1975. The Cocktail Waitress: Women's Work in a Man's World, John Wiley.
Stephens, W. N. 1963. The Family in Cross-cultural Perspective, Holt, Rinehart and Winston.
Tolson, A. 1977. The Limits of Masculinity, Tavistock.
Wajcman.J. 1981. 'Work and the family: Who gets "the best of both worlds?'" in Cambridge Women's Studies Group. Women in Society, Vi-
West, J. 1980. Apolitical economy of the family in
capitalism: Women, reproduction and wage labour,' in T. Nichols (ed.), Capital and Labour: A Marxist Primer, Fontana.
Whyte, W. F. 1948. Human Relations in the Restaurant Industry, McGraw-Hill.
Young, M. and Willrnott, P. 1975. The Symmetrical Family, Penguin.
Alice Yun Chai
An immigrant's class in her home country plays a large part in the kinds of adaptive strategies she develops in the new. Men and women do not necessarily have the same experiences, either of class or of immigration. This paper is a study of middle class, professional and educated Korean women immigrants married to professional and student husbands in Hawaii. It focuses on the differential impacts that immigration has on women and men in changing their division of labor, their relative statuses, social identities and their class positions.
BACKGROUND TO IMMIGRATION
Urban Middle Class Ideals in Korea
Marriage in Korea is ideally a complementary relationship in which husband and wife divide
From Women's Studies 13(3):223-234, 1987. Reprinted with permission of Gordon and Breach, Science Publishers, Inc. This article is based on a longer essay entitled "Adaptive Strategies of Recent Korean Women in Hawaii." in Janet Sharistanian (ed.), Beyond the Public/Dichotomy; Con temporary Perspectives on Women's Public Lives (New York: Greenwood Press, 1987).
the labor into male and female tasks, as well as having separate friendship networks and leisure pursuits. Younger, educated couples are more likely to participate in joint leisure activities such as family outings, films and concerts, or to visit relatives and friends. Wives who live in households separate from their husbands' parents have more contact with their own parents than do those who live with their in-laws.
The wife's domain is home and motherhood, the husband's economic provision. Both are indispensible to the welfare of the family, and husbands and wives respect each other's autonomy and competence within their separate domains. It is not acceptable for married women of relatively high socioeconomic status to enter the labor force, though in poorer urban and rural families it is acceptable because the incomes of both spouses are necessary to the family's economic support.
The Korean "domestic" and "public" domains divide social labor somewhat differently than they do in the United States. Korean women's household responsibilities are much broader than their American counterparts. Many urban middle class wives in
88   Domestic Worlds and Public Worlds
Korean Immigrant Women in Hawaii 89
Korea have at least one non-nuclear family member living with them, usually a housemaid or a single female relative. Heavy tasks and repairs that in the United States are the responsibilities of a husband are not part of the cultural domain of Korean middle class husbands. Workers are hired for these tasks, but wives do the hiring and supervision.
As domestic specialists, wives make decisions about family finances from real estate to consumption. Most husbands give all their pay to their wives, who in turn give their husband an allowance to cover his daily expenses. In addition, most middle class urban wives supplement their husband's incomes by participating in mutual financing associations called Kye, as well as in other business ventures.
Koreans in Hawaii
The Korean population in Hawaii and in the United States remained relatively small until the liberalized Immigration Act of 1965 which permitted not only spouses and children, but also parents, married children and siblings of American citizens to enter the country (H. Kim 1977: 91). Between 1970 and 1976, the number of Koreans in Hawaii increased almost 300%, making them the second fas test-growing Asian group in the state (Hawaii Commission on Manpower and Full Employment 1978). Hawaii's 9,8G8 Korean immigrants between 1970 and 1978 made up 15.6% of its immigrant population.
WOMEN'S LIVES
For this study I interviewed 27 women at length, informally, in Korean and in their homes. All are Christians (as are most Korean immigrants), and belong to one of Honolulu's two largest Korean Protestant churches. Most of the women had been urban full-time homemakers, high school or college educated, Protestants or Buddhists prior to emigration to Hawaii no more than seven years prior to being interviewed (the average was
four years). All the women had married in Korea when they were between the ages of 20 and 30—those with less schooling tended to marry younger than college-educated—and
were between the ages of 21 and 48, living in |
Honolulu with husbands and school-aged [ children when I talked with them.                  - \
Most women said they came to Hawaii to | join their own or their husband's relatives. [ Many hoped for economic improvement, and I for educational and occupational opportunities for their children (Koo and Yu I9S1: 11). f About half said they would like to move to the U.S. mainland for their children's higher ed- t ucation, or to return to Korea after their chil- : dren finished their schooling. |
Almost all (23 women) had already become |
citizens or were applying to do so. Their rea- *
sons were pragmatic: to vote for Korean elec- \ toral candidates, to get financial aid for their
children, to invite relatives to visit or emi- *
grate. Only one woman said she would not I
obtain American citizenship because she did I not want to forsake her allegiance to Korea.
The remaining three had not come to any de- |
cision about citizenship. i
MIDDLE CLASS WOMEN IN WORKING CLASS JOBS
Unlike the immigrants of the early 1900s, who had been largely uneducated, poor and of rural origin (Chai'l981: 328-344), recent Korean immigrants are well-educated. One-third of the household heads surveyed in 1975 had at least one year of post-secondary education, and more than three-fourths of that group had completed college (Hawaii, Commission on Manpower and Full Employment 1977: 27). Of the women who migrated to Hawaii between 1968 and 1975, 68% had at least a high school education (Gardner and Wright 1979: Table 6). Over two-thirds of the immigrants who reported an occupation were in the professional, technical or managerial categories.
However, few of these workers have the opportunity to practice in the fields for which they trained. This is primarily because of dis-
crimination in employment and the refusal of local professional accreditation agencies to Issue licenses on the basis of their previous education and training, and only secondarily because of language barriers. Recent immigrants often work at low-status, low-paying unskilled or semi-skilled jobs (Hawaii, Commission on Manpower and Full Employment 1977: 38). More than half of all recent Korean immigrants to Hawaii worked in sendee jobs. The median income of immigrant men ($7,400) was less than half that for .American born Korean men (SI6,600). For immigrant and American-born Korean women, the differences were even greater (32,750 and $8,857), while the contrasts between men's and women's earnings are equally striking (Gardner and Wright 1979: Table 16).
This situation is even sharper for Korean women with college degrees. Here, an estimated 60% of liberal arts graduates work as operatives, sales or clerical workers (Hyung-chan Kim 1977: 107). In Los Angeles, where Korean immigrant women also have higher labor force participation rates than native born women with children, most found work in garment and bead factories. An alternative pattern, to be explored below, was for self-employment in small retail shops (Bok-Lim C. Kim 197S: 186).
Many Korean immigrant women who identified as full-time homemakers in Korea moved into the labor force after they came to Hawaii (Hawaii, Commission on Manpower and Full Employment 1977: 33; H. Kim 1974: 23—42). Two-thirds of the women I interviewed did not work outside the home when they first came to Hawaii. Half of these had been teachers, in technical jobs or small businesses before and after marriage, but had stopped working after the arrival of children. The other half had never been employed. These women's husbands' occupations were evenly distributed among "clerical, technical, professional and business fields. Their median monthly familv income in 1979 was $1,000.
Two-thirds of the women I interviewed said they had to change from being home-makers to being working women for eco-
nomic survival. All 27 women remembered their first jobs in Hawaii. Nine worked as kitchen helpers in Korean or Japanese-owned restaurants; six women worked in dress-making factories; and three women worked as hotel maid5 in white or Japanese-owned hotels; three werejewelrv makers or farm workers; and three were sales or clerical workers for Korean or Japanese employers. Their second jobs were not significantly different, in that there was more horizontal than vertical mobility, from one to another entry-Ieveljob.
Available jobs arise from the interplay among English proficiency, skills, age, sex and marital status. English was the most common language used at work, followed by Korean and Japanese together, and then by English and Japanese together. Women faced a variety of obstacles in getting jobs and trying to work continuously. They tend to be confined to jobs with the lowest status and wages, to short-term, fluctuating employment, and they are often the last hired and the first to lose their jobs in the more volatile tourist, garment and jewelry industries.
Immigrant women's need for employment is also related to discrimination faced by their men. The women's husbands, most of whom had been professional, managerial or white collar workers in Korea, found themselves working as janitors, gardeners, painters, and dishwashers. Some men were able to move into more technical and semi-skilled work, or became store-owners.
By the time these women described their third jobs, there was a noticeable shift toward being unpaid workers in family-owned businesses—typically grocery stores, restaurants and gift shops. To start such a business by herself, with her husband or another relative, was preferable to moving from service or factory work to clerical or sales work. Women preferred the freedom, decision-making power and flexibility of working hours (though it sometimes meant longer hours and harder work) that a small business provided. Those women who aspired to such businesses usually had worked as kitchen helpers or seamstresses for several years, frequently working 10-16 hours a day, seven days a week
90    Domestic Worlds and Public Worlds
Korean Immigrant Women in Hawaii 91
often at more than one job in order to save enough money to start their business.
DOMESTIC STRATEGIES
Economic demands of immigrant life changed women's attitudes in some ways. Because of their experiences in Hawaii, three-fourths of the women came to believe that it was psychologically and socially beneficial (beyond the economic benefits) for married women to work outside the home. However, most women interviewed said that it was very difficult for mothers of small children to work in Hawaii because of the absence of extended family support and suitable babysitters.
Even mothers with older children feel badly about not being able to spend more time with their children. A mother of four children ranging in age from 7-20, who worked at a sewing factory during the day and as a kitchen helper at a club at night, and whose husband worked as a janitor, noted:
Even though the life here is much harder than Korea and my husband wants to go back to Korea or move to the mainland for the easier life and the better job opportunities, I have to bear it because of the children's education. I think Korean women in Korea who are full-time housewife/mother are much happier and better off than Korean women here. I have developed a bad headache and dizzy spells since I came due to fatigue from overwork.
In Hawaii, the division of labor between husbands and wives is more like the American ideal than the Korean. Men do outdoor and technical chores, women the indoor ones. Wives, assisted by daughters and/or female relatives do most of the cooking, cleaning, dishwashing, laundry and child care. Husbands help with grocery shopping if the wife does not drive, or if the family has only one car. They also assisted with yard work and home repair. In only a few families, husbands and sons helped occasionally with cleaning, laundry and dishwashing.
Older women and those who had been
married in Korea before immigrating to the -
United   States   complained   about   their '•
husbands' refusal to help with household 1
tasks, since their husbands still expected to be i waited on as they had been in Korea.
Life is very hard in America. We have to work as ^
hard as men at work place and have to conic 1
home and do more work. Many Korean h I
bands in America still want their wives to se I
them as they did in Korea, and marital conflict *
occurs because of this. In Korea, since we usu- f
ally had some domestic help we did not work as !
hard physically as we do now and did not have I
as much mental tension as we have here be- I
cause of double work loads and the language : barrier.
Finances too are arranged differently. In ;
Hawaii, equal numbers of husbands and wives '',
were responsible for family budgets and took ;
care of paying bills or going to the bank. Be- \
cause wives had less time, and tended to have [
fewer language and driving skills, many hus- ;
bands were forced to share some of these fam- \
ily responsibilities. However, husbands were ;
doing a traditional wife's work reluctantly and \
temporarily until they gained the needed \
skills. ii
In the midst of changing areas of responsi- \
bility, most women felt that their husbands re- \
tained final authority to approve or reject \
their wives' decisions about their behavior, ^
Several wives said that they could not take [
night shift kitchen work because of their \
husband's objections. One middle-aged wife [
with a teenage daughter explained: i
■ i
I suffer most because of the language barrier. ... |
Next hardest thing is too much work, both at i
work and at home because my husband still I
wants to be served and he does not even give f me freedom to go to English classes at night. But I go to English classes against his will anyway. I can only endure for so long, but if I cannot take it any more, I will have to leave him.
■■ .(■
More wives said that their marital relation- j
ship improved in Hawaii. (Only two women j
said that the relationship had been better in f
Korea.) This was because their husbands I
came home directly from work, rather than going to stag parties with male friends or colleagues as they did in Korea, and because there were no elderly relatives living with them who might inhibit their spontaneous reactions to each other. Sources of conflict centered around fatigue and irritation from hard physical labor, and around husbands' refusal to help with housework while demanding wives' services. One-third of the women said they had less frequent sexual intercourse in Hawaii because of long working hours, different shifts and fatigue. One woman lost all sexual interest because she was so pessimistic and depressed, but gave in to her husband's sexual demands because it was the only ego booster left to him.
Almost half the women said that their health deteriorated since immigration, and that they experienced insomnia, stomach disorders, headaches, chest pain, dizziness, loss of appetite and weight, eye ailments or frigidity. Eight reported no real changes in health, and five had improved health.
Korean wives feel especially burdened because they are not accustomed to double responsibilities of work and home. Though they only take on wage work from economic necessity, the fact of earning a regular wage has affected both their perceptions of and their actual relations with their spouses. They come to doubt their husbands' right to dominate them, and they insist on their sharing in the housework. They also refuse to meet their husbands' demands for the preparation of Korean food, which is quite time-consuming when added to cooking American food for the children. Women's wage earning may lead to a more flexible division of labor, decisionmaking and parental responsibility, as well as to less sex segregation in social life and public places.
In Hawaii, couples tend to engage in joint family, social and religious activities to a greater extent than they did in Korea. The greater isolation of the nuclear family and the wives wage-earning status account for this change and explain why female immigrants tend to make greater demands on their husbands for joint activities. Another factor is
changed kinship networks, more specifically, women's separation from their mothers and other close female relatives.
On the one hand, since spouses need each other more, they have learned to share and cooperate in ways they would not have learned in Korea. Men whose work shifts were different from those of their wives often cared for children while their wives worked. Husbands were especially likely to help with household chores when their wives worked and when no older child or adult female was available at home. On the other hand, the absence of kin also makes life more difficult because women have no one to depend on for day-to-day child rearing help.
Relations between parents and children are also stress points. Half the women said they disciplined their children, while half said that both parents did so, husbands dealing with older sons and wives with daughters. Many women were unhappy about lack of control over their children as they got older.
They themselves control their own lives. Do you think they would listen to their old-fashioned immigrant parents?. . . . The only way to control our children in America is by withdrawing material things and money.
Seventeen of 27 women said they were very much concerned about their children being influenced by negative aspects of American youth culture, such as premarital sex (especially for daughters), smoking, drinking and drug use.
Apart from their problems with language and communication, the children of immigrants themselves felt that Hawaii's schools were academically relatively easy compared to Korean schools. All students felt greater freedom and independence after moving to Hawaii.
Women put great stress on their children's education. Children provide future emotional support and financial security for their mothers. While mothers experience racial and sexual discrimination at work, and are often demeaned by it, they enjoy great prestige in the family in their valued role as moth-
92    Domestic Worlds and Public Worlds
Korean Immigrant Women in Hawaii 93
ers. Giving birth to sons is still regarded as the most important function of married urban Korean women. Moreover, women whose lives have been devoted to child-bearing feel psychological security and vicarious status achievement through their sons' academic and occupational success (Chai 1978: 46).
However, economic burdens prevent women from giving children the kind of care for which they have immigrated. At the same time, their roles as homemakers and mothers are reinforced by sexual discrimination to severely restrict their employment opportunities. This is a frustrating double bind for many women.
ADAPTIVE STRATEGIES: NETWORKS AND ORGANIZATIONAL PARTICIPATION
Koreans form a highly organized ethnic community in Hawaii, with many overlapping associations that generate and distribute money, labor, jobs, clients and information {Bonacich, Light and Wong 1976: 443). As a group, middle and upper middle class Korean immigrants participate in a variety of organized activities: professional and social associations, Korean ethnic churches, church-sponsored English classes, and Korean language newspapers.
New immigrants often live temporarily with resident relatives until they can find housing. Patterns in Hawaii departed from Korean patterns of living with the husband's family. More new immigrants lived with the wife's kin than the husband's during resettlement. Likewise, the number of women who said they were sending money to their own parents in Korea was greater than those who said they were sending money to their husband's parents. This may be because they earn their own money and do not need to ask their husbands for it.
Eighteen of the women I interviewed attended Korean Protestant churches regularly: one attended the Catholic church and another was a Buddhist. Church and English
classes were women's main organized socializing, though some did participate in auxiliary activities of their husbands' alumni, professional or recreational associations. Most women had been members of high school or college alumni associations in Korea. These met at least monthly for lunch and a kye meeting. Those with whom immigrant women had most frequent contact were church friends, relatives, co-workers and English-class mates. Indeed, these classes were important sources of companionship and emotional support for these women.
When the women were asked what advice they would give to relatives or friends planning to come to America, one-third said they would advise against it if the person were more well-to-do than average in Korea. The remainder would encourage immigration, provided the person were well-informed and genuinely prepared to face hardships, or if they were young and had skills that were useful in America. Significantly, no one suggested immigrating for the sake of their children's education and their future.
Over half said that they would not have come had they known of the hardships they were to experience. Over half said that their only hope for the future was their children's success. Nevertheless, they were prepared to live independently from their children in old age unless the children wanted their parents to live with them. One-third wanted to return to Korea in their old age to live with their relatives and friends because they saw around them the loss of status, the isolation and the loneliness of the elderly in America.
SUMMARY AND CONCLUSION
Research for this paper began with the premise that particular domains of behavior and institutional arrangements in specific societies must be observed before cross-cultural generalizations about women can be made. I examined the multiple dimensions of middle class, married Korean immigrant women's roles in both the private and public domains
of behavior and in specific institutional contexts in order to show women's economic, domestic, social and psychocultural adaptive strategies from the perspectives of the women themselves.
As a family strategy for survival in the new society, and because of a lack of viable alternatives, Korean middle class immigrant women obtain employment in unskilled or semi-skilled service jobs. The decision that these women should work is explained in terms of family survival, since Korean immigrant men are unable to transfer their professional and vocational training and experience to comparable positions in Hawaii.
Korean immigrant women attempt to enter the American labor market in female sales or clerical jobs. Since this channel is open to few immigrant women because of English language and structural barriers, the women create a public world of their own by developing and engaging in family businesses with the cooperation of male and female relatives. They seek small business opportunities where wives and husbands can become their own bosses and work together with their children nearby. Thus they integrate the public and private spheres and enjoy more flexible working hours than they would otherwise in the poorly paid menial jobs that keep them away from home and that require a separation between the workplace and home.
Success in this family strategy establishes a separate ethnic public sphere with its own ladders of achievement and brings immigrant women a certain economic status and some power in the majority culture. This in turn enables their children (who have acquired language skills and American credentials) to achieve social and political status. This is not so different from the situation in Korea, where these women's status would increase as they reared occupationally successful sons, and where mothers lived vicariously in the public sphere through their children's achievement.
In Hawaii, because of the double burden of home and work and the absence of an extended kin network, immigrant women feel that they cannot be effective mothers. Their
inadequate English, limited knowledge of American cultures, and their work at degrad-ingjobs have lessened their maternal authority. These may be the major reasons for their psychosomatic symptoms—which actually reaffirm the centrality of motherhood in their lives. Because traditional cultural values emphasize the importance of motherhood—and this is reinforced by low status, low paying jobs—women consider themselves mothers and wives first. They regard their menial jobs as only temporary means of earning supplementary income for their children's education. The goal is still to raise successful children. Despite the fact that they contribute significantly to the economic well-being of the family, their meager earnings and the structural limitations placed on them by American society force their continued dependence on their husbands. Furthermore, because of the structural barriers in the larger society, both husbands and wives are forced to rely on each other and to look to the Korean community for security and support. This study affirms Heidi I. Hartmann's observation that "The conflicts inherent in class and patriarchal society tear people apart, but the dependencies inherent in them can hold people together" (Hartmann 1981: 394).
REFERENCES
Bonacich. Edna, Ivan Light and Charles Choy Wong. 1976. Small Business Among Koreans in Los Angeles. In Emma Gee, ed. Counterpoint: Perspectives on Asian America. Los Angeles: Asian American Studies Center, University of California, pp. 437-449.
Chai, Alice. 1978. Korea: Urban Profile. Impulse (Honolulu: East-West Center). Winter: 44-46.
-. 1981. Korean Women in Hawaii: 1903-
1945. In Hilah F. Thomas and Rosemary Skinner Keller, eds. Women in New Worlds. Nashville: Abingdon, pp. 328-344.
Gardner, Robert W. and Paul A. Wright. 1979. Asian Female Immigrants on Oahu, Hawaii. Paper prepared For the Women in the Cities Working Group, 5-23 March, Honolulu, East-West Population Institute.
94    Domestic Worlds and Public Worlds
Public and Private Politics: Women in the Middle Eastern World 95
Hartmann, Heidi I. 1981. The Family as the Locus of Gender, Class, and Political Struggle: The Example of Housework. Signs, 6, 3:366-394.
Hawaii, Commission on Manpower and Full Employment, Immigration Services Center. 1977. Immigration in Hawaii: 1965-1975. Prepared by Irene D. Makiya.
-■— 1978. Immigrants in Hawaii Fourth Annual
Report.
Kim, Bok-Lim C. 1978. Asian Americans: Changing Patterns and Changing Needs. Association of Korean Christian Scholars Series, No. 4. Urbana: AKCS Publication Services.
Kim, Hyung-chan. 1974. Some Aspects of Social Demography of Korean Americans. International Migration Review, Spring: 23-42.
-. 1977. Ethnic Enterprises Among Korear
Immigrants in America. In Hyung-chan Kim ed. 77« Korean Diaspora. Santa Barbara: ABC Clio, Inc., pp. 85-107.
Koo, Hagen and Eui-Young Yu. 1981. Korean Im migration to the United States: Its Demo graphic Pattern and Social Implications for Both Societies. Papers of the East-West Population Institute No. 74. Honolulu: East-West Center Population Institute.
Cynthia Nelson
THE ETHNOGRAPHIC IMAGE
One of the most commonly held assumptions found in the ethnographic literature discussing the political significance of women in the society at large is that the decisions that women take do not have repercussions on a very wide range of institutions. The general argument is most clearly stated by Mary Douglas:
The social division of labour involves women less deeply than their menfolk in the central institutions—political, legal, administrative, etc.—of their society. They are indeed subject to control. But the range of controls they experience is simpler, less varied. Mediated through fewer human contacts, their social responsibilities are more confined to the domestic range . . . their social relations certainly carry less
Reproduced by permission of the American Anthropological Association from Ameiican Ethnologist 1:3, August, 1974. Not for further reproduction.
weight)' pressure than those which are also institutional in range. This is a social condition they share with serfs and slaves. Their place in the public structure of roles is clearly defined in relation to one or two points of reference, say in relation to husbands and fathers. As for the rest   _ [ of their social life, it takes place at the relatively { unstructured interpersonal level, with other women. . . . Of course I would be wrong to say i that die network of relations a woman lias with others of her sex is unstructured. A delicate patterning certainly prevails. But its significance
for society at large is less than the significance of ......i
men's relations with one another in the public rote system (1970:84).
Nowhere is this assumption more uncritically taken for granted than in the ethno- j graphic descriptions of pastoral and seden- h tarj' societies in the Middle East in which the assertion is made that there are dual and separate worlds of men and women in which the former world is public and the latter world is private. Typical of such assertions is the following:
The women's world is not merelv more narrowly circumscribed like in most civilizations; it is also provided with a complicated system of devices for cushioning off: i.e., safeguards which provide limited access to each other's worlds. The women's world has two major manifestations: the home (tent) and the private communication patterns between women of several homes. For men there is limited access to the former and practically no access to the latter and this is paralleled by lack of interest by the men about the women's world. . . . The men's world has two major manifestations: the sphere of earning a living and the public sphere of communications including public affairs. Access of women to the former is limited and formally none in the latter— old grandmothers being an exception. But women are keenly interested in male affairs!!! No doubt that the two worlds have their regular meeting point in the home, for this is where a good deal of clearing goes on continuously (van Nieuwcnhuijze 1965:71).
Inherent in this statement as well as most other discussions on sex roles found in the ethnographies of the Middle East, and particularly those centering on nomadic societies, is not only the commonplace notion that the human universe is segregated into two social worlds marked out by the nature of the two sexes, but also that these two social worlds are by definition characterized as being private (the women's) and public (the men's) (Asad 1970; Barth 1961; Cole 1971; Cunnison 1966; Marx 1967; Pehrson 1966; Peters 1966). The former world is invariably described as domestic, narrow, and restricted, whereas the latter is described as political, broad, and expansive. Authority is also segregated in terms of this dichotomy. The home is regarded as the woman's for all internal purposes. Her authority in domestic affairs is an established fact. For external purposes, the home is the man's, the assumption being that whatever articulates the household to the public sphere is by definition political and thereby a male concern. And the inference drawn from this assumption is that women are far more interested in men's affairs than vice versa.
Also inherent in the ethnographic accounts of these two social worlds is the notion
that what is the concern of women is the domestic and not the political. This raises the whole question of the meaning of "power" and "the political" and why this should be linked to such notions as "private (domestic)" and "public (political)." By assigning private and public to the different social worlds of men and women described for certain Middle Eastern societies, I would argue that western social scientists have imposed their own cultural categories onto the experiential world of the Middle East and that the whole discussion of "power" in these societies is influenced by these categorizations.
Most anthropologists working in the Middle East tend to view power in the classic functionalist tradition. Following RadclifFe-Brown, they define the political system as the maintenance or establishment of social order within a territorial framework by the organized exercise of coercive authority through the use or possibility of use of physical force.
Barth, for example, explores the kinds of relationships that are established between persons (only males as it turns out) among the Swat Pathans and the way in which these may be systematically manipulated to build up positions of authority and the variety of political groups. The main sources of authority/power available to persons are ownership of land, the provision of hospitality, and a reputation for honor. Most statuses and rights are usually defined by contractual agreements between persons. In these circumstances, each man's aim may be seen as the adoption of the strategy that will best serve his interest. "Physical force or the threat of it is in Swat a characteristic sanction in a great many relations" (Barth 1959:53). Barth sees Swat Pathans as being driven by self-aggrandizing passion and maximizing rationale and represents the political activity of the dominant land-owning class as the foundation of social order. As Asad has so cogently pointed out, Barth's model is an anarchic, conflict-ridden, violent society which reflects Barth's Hobbesian model of human nature (Asad 1972:74-94).
Criticizing the functionalist position, Asad argues for a distinction between power and authority. For Asad, power refers to the rela-
96    Domestic Worlds and Public Worlds
Public and Private Politics: Women in the Middle Eastern Worid 97
□'on between agent and an object as a means, that is, to the opposition of exploiter and exploited, whereas authority refers to the subordination of human consciousness to a legitimate rule (and contingently to those who determine the rule) (Asad 1972:86). Asad sees the problem of political domination in terms of a dialectical relationship and raises an important alternative in terms of the way ethnographers tend to look at political systems. But he, too, tends to view power and authority as the exclusive concern of men. "The overall authority of the household head is based on the fact that he has greater power and moral responsibility than any other member of the household" (Asad 1970:100-101).
Given the fact that most ethnographers of the Middle East have been European or American males who, by virtue of their for-eignness and maleness, have had limited if no access to the social world of women, we seem to be confronted with the normative image of the society as reported to male ethnographers by male informants." Lienhardt expresses this dilemma cogently:
And though the segregation of women from men not closely related to them is one of the things that must at once meet the eye of any visitor to the towns and villages of the Trticial Coast, this segregation makes it difficult for a visitor to gain any precise knowledge of the women's position. Apart from its being difficult for a man to talk to women there, it is not even proper for him to ask very much about them, particularly to ask in any detail about specific cases . . . one can easily be misled, particularly in assessing the extent of male dominance (1972:220).
From the ethnographic literature, we know precious little about how women in these societies view their situation, whether they feel they have "power" and how they wield it. If we had better knowledge of the "lived-in-world" of nomadic women, we might come up with different images of the society and the definitions of power. Also, we might ask what could the contributions to our knowledge and un-
derstanding of the relationship of women and power be, if we were to re-think the notions of "power" and recognize its special feature as a particular kind of social relation rather than as an embodied quality institutionalized i types of social structures. Some social sciei -tists have argued along the following lines:
The initial problem of defining social power is to recognize its special features as a particul; kind of social relation, as reciprocity of influence. Rťciprocilv of influence—ú\e defining crit rion of the social itself—is never entirely d strayed in power relations, except physical violence when one treats another as object. \\ cannot sever power relations from their roots social interaction. One actor controls the otb with respect to particular situations and spher of conduct—or scopes—while the other actor regularly dominant in other areas of situated conduct (Wrong n.d.).
Olesen has suggested in an unpublishe paper that the concept of "the negotiated order" is a useful idea for understanding rec procity of influence in interactive situatioi -(Olesen 1973). Persons in an interactive situ; tion, she argues, negotiate the rules that define and circumscribe that relationship.
Assuming that men and women are ii volved in "negotiating their social order"-i.e., the rules and roles of social interaction-we must not lose sight of the fact that social action is always "situated action" and circumscribed by culturally given constructs of soci reality, the social stock of knowledge at hand, as Schutz would say (1962:120-134). What becomes relevant for the purposes of our discussion is to recognize that despite the existence of segregated social worlds and the implication that there exists a differential distribution of social knowledge—the man's and the woman's—this knowledge is structured in terms of relevances, and women's relevance structures intersect with those of men at many points. Applying this to the ethnographic situation, we must ask how women can and do influence men to achieve their own objectives. The notion of power implied in the concept of the negotiated order is the
potential for levying sanctions, the potential for influencing further actions of others (as well as one's own). Sanctions are not just threats of physical force but capacities for influencing the behavior (action) of others. They are ways of creating possible lines of action for others as well as for oneself.
Looking at power from this perspective, we are forced to raise a different set of ethnographic questions, questions that have been neglected perhaps due to not recognizing the ongoing dialectical process of social life in w'hich both men and women are involved in a reciprocity of influence vis-a-vis each other. What are the normative constructs that facilitate, limit, and govern "negotiation"? What are the sanctions open to women? In what ways can and do women set up alternatives for men by their own action? How do women influence men? Who controls whom about what? How is control exercised? How do women control men? Other women? How conscious are women of their capacity to influence? In other words, what are women doing in this reciprocity?
In the remainder of this essay I would like to challenge the notion that the social worlds of men and women, despite the element of segregation, are reducible to spheres of private and public with power limited to males in a so-called public arena. By using data from ethnographic studies by both men and women concerning women in the Middle East, I shall suggest that women can and do exercise a greater degree of power in spheres of social life than has heretofore been appreciated.
THE ETHNOGRAPHIC EVIDENCE
The detailed and scholarly work of Use Lichtenstadter (1935) on Women hi the Aham Al-'Arab offers an interesting analysis of the role that the Arabic women played in the warfare of her tribe and thus presents us with a view of the life and position of women in pre-Islamic Arabia.1 Although the material deals  with  nomadic  society  during the
Jahiliya, it is still informative about the manner in which women were depicted in everyday life and how they exercised political influence in the man's world.
According to the Aiyam narratives, the women's influence was felt beyond the tent. Through marriage she played an important role in Arab policy by being the link and mediator through which powerful alliances between tribes were accomplished. As matron she acted as counsellor to her son who very often submitted to the advice of his mother and, whenever it was possible, she tried to gain influence over her son in order to bring an enmity to an end. Lichstenstadter points out:
That Fatima bint al-Khurshub tried, though unsuccessfully, to mediate between her son and Qais b. Zuhari shows that she could be sure that her opinion would at least be heard. In this case, however, the son did not accept his mother's advice; the events proved that she was right (1935:65).
In warfare the woman very often was the cause of quarrels and great feuds. She was also employed as a spy and, if captured in war, was the source of great ransoms. As the women were not far from the spot where the battle took place, they were able to watch the bustle of the fight and incite their men by acclamations. When in the greatest distress and danger, the Arabs had recourse to a device which was meant to excite their desire of fighting to the highest degree: they exposed their women, particularly noble women, to danger by forcing them to fall from their camels and litters in order to show the warriors that they must fight or die (Lichtenstadter 1935:43).
In summarizing her analysis, Lichtenstadter suggests that pre-Islamic nomadic society was a society that treated women with esteem and one in which they were allowed to take part in public life. "From the Aivam tales pre-Islamic Arab women played a part in the life of their tribe and exercised an influence which they lost only later in the development
98    Domestic Worlds and Public Worlds
Public and Private Politics: Women in the Middle Eastern World 99
of Islamic society" (1935:81). "But as during the time ofjahiliya women were not separated from men but lived in close intercourse with them they could readily get to know their plans and projects" (1935:83). "In addition the conditions of life were such that in times of distress clever advice was eagerly accepted and followed regardless whence it-came, even if offered by a woman. In this sense we are justified in speaking of the 'influence' of a woman without exaggerating her importance in the public life of an Arab tribe" (1935:85).
Emrys Peters also describes the manner in which women can and do exercise influence over men in pastoral societies, for example, as mediators between natal and affinal groups in marriage alliances, as controllers of the products or the property, and as wielders of authority in the domestic sphere.
The pivotal, points in any field of power in this, a superficially dominant patrilineal, patrilocal and patriarchal society where the male ethos is vulgar in its brash prominence, are the women. What holds men together, what knots the cords of alliances are not men themselves, but the women who depart from their natal household to take up residence elsewhere with a man. and who, in this critical position communicate one group to another (Peters 1966:15).
Among the Bedouin of Cyrenaica men may boast of their dominance over women (and certainly this might be expected with male informants channeling information to male ethnographers), but they are constrained in their actions by the control women possess over the preparation of food, the provision of hospitality, the comforts of shelter, and the reputation for honor. Men may control the economic resources—land, water, animals—the durable properties, but women control the products. Utilization of the products, the dividends of their investments, are granted to women, and through these they acquire legal rights in men. Bereft of controlling rights in property, women are nevertheless critical in its manipulation. They possess the legal right to protection and support against husband as daughter or sister of the
man who holds the bridewealth. Women mediate between the two, make demands on the men as a right, and are given public support (Peters 1966; Mohsen 1967).
Marx makes the same point about Bedouin of Negev—that women in multiplex role situations increase their potential for negotiation:
A marriage link acts as a very effective communicative device between groups because the woman who conveys the communications is so intimately bound up with both her husband and sons and with her father and brothers. She has the interest of both groups at heart and would suffer most from an estrangement between the two groups. At the same time she is on the inside of both groups and thus able to assert Iter influence over the sections through the men to whom she is closely connected, as well as through women (1967:157).
Cunnison, while arguing that women occupy no formal position of power or authority among the Baggara Arabs, does suggest that:
Women have a profound influence on politics in two respects. Firstly, they are arbiters of men's conduct, and they can make or break a man's political career. They do this by singing songs of praise or alternatively of mockery. The brave man and the cowardly man have their fame spread. The man who is undistinguished in either direction goes unsung. The songs sweep the country, and the reputations are made and broken by them. Secondly, a policy decision that the men of a camp or a surra (kin group) make is influenced by the kind of reactions that the women of the group are likely to have(1966:117).
In these respects, at least for the Baggara, women have a significance in the public role system of the society. Cunnison also points out that the Baggara ideas about the value of manliness involve the closely related aims of wealth, women, and power.
Cattle attract women and allow a man to marry more than one wife. The possession of cattle plays an important part in the relations of men and women. It implies that a man is endowed
with those qualities that Baggara men and women alike regard as most admirable. Herd building means easier access to women who play a positive part in spreading a man's virtue or challenging man's honour. Although there is often argument about amount of bridewealth to be paid, debate is not between two families; instead the men of both families agree and unite in argument to try to beat down die price. The bride's modier backed by die women of the family is demanding. Final word is that of the bride's mother. She can try to stop a marriage that she or her daughter don't want by refusing to lower the price. "Let us chase him off with our demands" (1966:116).
The Pehrsons in their study of the Marri Baluch describe the strategies used by women to achieve influence over men: (1) playing men off against each other; (2) seeking alliance and support from other women; and (3) minimizing contact with the husbands (1966:59).
In another monograph on La Femme Chaouia de L'Aures (1929), Gaudry draws our attention to the power that women can exert over men in their capacity as saints, sorceresses, magicians, and healers. The Auresian woman has, like the man, the cult of mzara, places sanctified by the passage of a saint, and the woman, just as the man, affiliates herself to religious organization. Gaudry describes a female Marabout, Turkeyya:
There existed a Marabout of great virtue named Turkeyya who was most pious and exerted a great influence on her many clients and adepts. Tiiere was also a male Marabout, Sidi Moussa, seeing that his authority had diminished with a number of his followers felt peeved. He decided to put an end to this competition with his dangerous rival and he devised a simple plan to get rid of her. He appealed to one of his devouts and entrusted him to kidnap Turkeyya and marry her. As a Marabout can only enter into a family of Marabout she lost, from this mis-marriage, all the authority she had (1929:235).
Sorcery is another means by which women can be said to exert their influence over males in Chaouia society, and as a sorceress woman
has more power over the man than as a saint because of her ability to divine the future, enhance love, deter evil, and heal illness. According to Gaudry:
It can be said that the superstitious fear of the women which filled the Berber mind allowed the women to impose on inferior religion of which they arc the priestesses which is a response to a collective need (1929:246).
All old women are more or less sorcerers, learning their craft from their mothers. They teach women to prepare lotions which can "tame any man." Men fear them, and some forbid their wives to receive them. A sorceress has power over the male through the women. Says a male Chaouia proverb:
The child of male sex comes to the world with 60 jnoun in his body; the child of the female sex is born pure; but every year, the boy gets purified of ajinn, whereas the girl acquires one; and this is the reason that old women, 60 years and with 60 jnoun arc sorcerers more malignant than the devil himself. Blind she sews more material, lame she jumps over rocks and deaf she knows all the news {Gaudry 1929:267).
Gaudry Ts work raises a fundamental issue that has 710/ been the focus of much recent ethnographic field research among Middle Eastern societies; that is, the degree to which men perceive women exercising power over them through the idiom of the supernatural. Crapanzano's recent work on the Hamadsha in Morocco, however, is suggestive of the powerful significance of the "camel-footed she-demon, A'isha Qandisha" on men in the curing rituals (1972:327-348).
Should one of A'isha's followers disobey her, he is immediately struck and suffers grave misfortune or bodily harm. The Hamadsha, her special devotees, are said to be favored by A'isha and are very proud of the intimacy of their relations with her (1972:333).
It must be noted that both Gaudry's and Crapanzano's work was conducted among Islamicized Berber cultures of North .Africa
100   Domestic Worlds and Public Worlds
Public and Private Politics: Women in the Middle Eastern World 101
among whom saint cult worship is predominant (Gellner 1969; Geertz 1968). Nevertheless, both ethnographers underscore the fear and veneration expressed by men toward these female supernatural figures and suggest lines of inquiry for further investigation. The paucity of ethnographic description surrounding the relationship of women to the religious system, in general, and the supernatural, in particular, suggests more a lack of interest on the part of ethnographers than it does a lack of concern on the part of the actors in Middle Eastern Islamic societies. This does not seem to be the case when we look at studies of women in societies of sub-Saharan Africa (Lebeuf 1971; Hofer 1972). Perhaps if we were to turn to the more recent ethnographic studies focusing specifically on women in pastoral and sedentary societies of the Middle East, we might discover evidence that suggests women do exercise control in society in a variety of ways.'1
Farrag, in her study of social control among the Mzabite women, demonstrates convincingly how moral, social, and religious control is exercised over women by women through a specific all-female religious institution called the Azzabat. Both men and women have a very important stake in the conduct of their women in that there is a firm belief that god's anger befalls the whole community as a result of any sexual misconduct on the part of women/' Because of the frequent and prolonged absence of the men in the community for purposes of trade, the women have assumed an increasing importance in the mechanisms of social control over women.
Although social changes since independence have also affected die power of the Azzabat, they still exercise a far stricter control over the women than the Ozzaba (male religious group) do over the men (Farrag 1971:31 S).
Yet, at the same time, increasing demands are now being made by the men for the "modernization" of the women, thus creating a situation in which certain inconsistencies and ambivalences are created. The thrust of Farrag's argument is to show how breaches of
certain norms are still effectively and formally sanctioned through all-female religious institutions coupled by informal sanctions of the power of mothers-in-law, public opinion, and gossip, regardless of social status. The implications of Farrag's article to the main thesis ol this paper are obvious. Instead of an image ol segregated social worlds of men and women, in which women are relegated to the private domestic sphere, we find all-female institutions responsible for the sanctioning of breaches of social norms—certainly a most public concern.
Approaching her study of women from the perspective of social stratification, Vanessa Maher, in a recent study among townswomen in the Middle Atlas of Morocco, makes an exhaustive analysis of the social mechanisms, both political and ideological, by which women are confined to the traditional status-based mode of social relationship "where women are not working for wages because participation in the public sphere of the market is considered immoral" (1972:15). Given this segregation of men's and women's roles, the dependence of women on women becomes all the more necessary, especially in late pregnancy and early childbirth. These feminine links of cooperation that form are independent of those formed by the male's kindred of cooperation and operate to redress the balance of power between men and women. Maher argues that the market principles and prerogatives of kinship and status struggle for hegemony with the result that there is a structural conflict between the social necessities of marriage and the superior rewards of kinship (status-based relations), especially for women. The chief locus of conflict is marriage, and Maher argues that lacking political control over their own lives and lacking religious worth as second-rate Moslems, women are forced to turn to intrigue and witchcraft—weapons used in the power struggle between men and their wives.
Nancy Tapper (1968) explores this theme in more depth in her study of a women's sub-society among the pastoral Shahsavans in Iran. In this women's sub-society she describes how women establish among them-
selves a range of relationships in which women may gain achieved status in the community as rriidwives, ceremonial cooks, and religious leaders.
In each lira (clan or family group). . . were one or two women held to be knowledgeable on religious matters. Commonly, they are women who, with a male relative, have made the pilgrimage to the shrine of Imam Reza at Mashhad and are thereafter referred to by the title atMashadi. In fact the position of Mashadis among women is comparable to that of a Hajji among men. The Mashadis are among the few women who pray regularly; their position is a highly conservative one and they firmly support traditional Shahsavan customs and moral attitudes, sometimes by reference to imaginary Koranic injunctions. The opinions of such a woman in matters of family law and custom are sought by both men and women and her advice is given equal weight with that of a man (Tapper 1968:17).
Al-Torki, in her pioneering study of townwomen injeddah, Saudi Arabia (1973a), postulates that in societies where the segregation of women prevents their participation in public affairs, elaborate networks of friendship and gift-exchange will be found. These networks are likely to enable the participants to gather vital information, which gives them considerable informal control over decisions that are nominally the exclusive prerogative of males. Injeddah the women exercise considerable control over marriage alliances.
This control has far-reaching consequences in a society where kinship dominates as a structural principle in local and national politics. Obviously, their influence must derive from a different source of power than control of resources. I suggest that the women's eminent control of information in matters relating to the arrangement of marriage constitutes this source. The very nature of the women's exchange networks gives them an almost exclusive access to information on which the decisions of their male relatives depend. By manipulating their knowledge to accommodate their own interests in potential marriages, the women actually manage to direct or impede the men's efforts to establish marriage alliances (1973b:5).
Lienhardt (1972) underscores this proposition in his description of marriage and the position of women in Trucial Coast society. Since shaikhs, great and small, are political personages, their marriages have a much clearer political dimension than the marriages of other people. Here some women achieve remarkable influence and power. Lienhardt suggests that the marriage of shaikhs with Bedouin women may be one of the reasons why the women of shaikhly families in general seem to lead a less secluded life than most others. When they are women of strong character, the senior women of ruling families can play an important part in affairs. One remarkable woman of the Trucial Coast, Shaikha Hussah bint al-Murr, the mother of the present ruler of Dubai and wife of his predecessor, came so far out into open public affairs as to hold her own majlis ('public meeting'), not for women but for men, sitting receiving visitors, as people said, like a shaikh, and when her husband ruled it is said that more men visited her majlis than visited his. This remarkable lady was an outstanding figure in both politics and business. On the one hand, she played a leading part in a political struggle that led to civil war and the subsequent expulsion of the reformist party. On the other, she restored her husband's family fortunes by property development, trade, and, one gathers, that profitable but risky enterprise of Dubai, smuggling with Persia and India (1972:229-230).
This may be an extreme example of the potential importance of leading women in public affairs; however, it suggests that we must re-evaluate the metaphors of private and public in terms of domestic and political.' What Lienhardt's and the other ethnographic material suggest is that women do approach public affairs but they do so from private positions. In public, women are separated from men, and men mix widely in the public circle of the market. The women, on the other hand, mix in a large number of smaller groups, more exclusive than the society of men and consisting largely of closer and more distant kin affines, and other women who are friends of the women of the family, e.g.,
102    Domestic WorSds and Public Worlds
I
Public and Private Politics: Women in the Middle Eastern World    103
azzabat, female sub-societies, and kindred of cooperation as discussed above. In the societies we have been discussing, families are one of the basic groupings in its economic and political, as well as its moral aspects. Here, in some senses, the range of women is greater than that of men, and it is the very segregation of women and the impropriety of discussing them in male company that makes this so. Women, in general, are a necessary part of the network of communications that provides information for their menfolk, and at the head of the social hierarchy are some women who form a focus for the smaller groupings of women and a bridge between their concerns and the public concerns of men.
Once we begin to examine the role and position of women in Middle Eastern society from the standpoint of the woman and to describe the woman's view of the social worlds in which she lives and interacts, it becomes clear that our ethnographic imagery about domestic spheres being private and female and public spheres being political and male is misleading. This is not to argue that only women can understand or do ethnography about women, but only to suggest that by taking the standpoint of the woman, by examining her taken-for-granted assumptions about her social worlds, we discover another image of power and influence operative in society. I take the position that to place oneself imaginatively into the inner self of another (including a set of interacting "others" of which the ethnographer is one) is not only necessary, but it is the very foundation of social life (Ber-ger and Luckman 1967; Schutz 1962; Mills 1967). Also, by re-evaluating the notion of power from the standpoint of reciprocity of influence, we can specify ethnographically those particular situations in which women can and do exercise influence over men. Based on the evidence presented, what are these situations?
One dominant theme repeated throughout the ethnographies is the crucial role women play as structural links between kinship groups in societies where family and kinship are the fundamental institutions of ev-
eryday life. Simultaneously the woman as daughter, sister, wife, and mother acts as an "information-broker," mediating social relations within the family and larger society. The implications for power (reciprocity of influence) are obvious in that by these networks of relationships, the woman is in a position to channel or withhold information to the male members of the kindred. And in this position the woman influences decision-making about alliances, actually sets up marriage relations, and informs male members of the household what is going on in other homes. But of course the "home" in question is not that of a tiny nuclear family, but of a wider family group. And this family group is one upon which many of the affairs of the society—social, economic, political—turn.
The ethnographies do support the idea of segregated social worlds but rather than seeing this as a severe limitation on women, the evidence suggests that the segregation of women can alternatively be seen as an exclusion of men from a range of contacts which women have among themselves. This emphasizes a second major theme emerging from our data, particularly from sources on women written by women, and that is that women form their own exclusive solidarity groups and that these groups exercise considerable social control (Farrag 1971; Maher 1972). Also, by seeking alliance and support from other women in the community, certain women achieve high social status in the community and consequently exercise political influence (Aswad 1967; Tapper 1968).
From the ethnographic data summarized above it is evident that women do participate in public activities, activities which have their reverberations and intentions in large-scale societal networks.
A third realm in which women emerge as having influence over men is through the religious or supernatural. That is, in those situations where women are publicly acknowledged as having power, it is associated with the supernatural and the fear that men have of women's sexuality or, better expressed, as the felt threat to male esteem of women's sex-
ual misconduct. Through witchcraft, sorcery, divination, and curing, women are instrumental in influencing the lives of men.
A final point to be mentioned is the degree to which the public image of a man is influenced by the particular behavior of his women—through ridiculing, through gossip, through honor and shame (Schneider 1971; Cunnison 1966).
|        IMPLICATIONS FOR ETHNOGRAPHY
J The main thrust of this essay has been to chal-
lenge the prevalent ethnographic image of j the position of women in selected pastoral
\        and sedentary societies of the Middle East. Specifically, it has addressed itself to the ques-tion: In what sense can we speak of women exercising power in those societies which are 1 avowedly patrilineal, patrilocal, and patriar-
chal? This raised the whole issue of the con-l ception of power as viewed by Western cth-i nographers in their writings about Middle Eastern society—a view that defines the j sphere of masculine activity as the public—in I the Greek polis sense, and the sphere of femi-f nine activity as the private and domestic. \ What becomes defined as the public and pri-| vate spheres, however, are less the categoriza-{ tions of the world by the actors living in these \ societies than they are the metaphors of the I observers who are recording the actions of ? men and women in these societies. t The fundamental implication for ethnog-
^        raphy emerging from this essay rests on the i        following question: What are the kinds of data I        generated in a field situation in which accessi-[        bility to the social worlds of men and women $        is predicated primarily on the sex. role of the I        observer? As sex roles circumscribe the way (        the actors interact with each other in society, I        it is of fundamental importance to realize that sex roles also circumscribe the way in which actor and ethnographer interact with each other. The sex-linked aspect of social interaction places the actor and ethnographer in situations of communication in which the so-called "raw data" of ethnography are the very
products of this communication. In other words, it is the interactive situation itself which produces (or generates) the data from which ethnographies are written. An awareness of this phenomenon suggests to me a quite different stratég)' of fieldwork than has hitherto been expressed in the ethnography of the Middle East, particularly on the important question of the position of women. What I would suggest is that we, as ethnographers, become more imaginative in the creation of our models of the activities, norms, and interpersonal linkages that make up a society's political processes. My descriptions of women as information brokers, as reputation builders and maintained, as "power" of their own, all concern a better filling in of all the links. We must get away from the simplistic, mechanical models (either those implicit ones of the ethnographers—such as Barth's Hobbesian view of politics—or the official ones of the informants). We must become conscious that our data arc not "gathered" but manufactured and grounded in the interactive foundations of the research process itself. Not being conscious of this phenomenon has led to an ethnographic image of women and power in the Middle East that is both incomplete and misleading.' We must ask ourselves how do we come to understand the world of the opposite sex? And as ethnographers, what are the criteria by which we have selected data to record the "true image" of society?
For most ethnographers of the Middle East, there exists an implicit assumption that "reality," the data, exist "out there" waiting to be described and that it matters little what the social position of the informant is since "as a bearer of the culture" he (or she) is as informed as anyone else. Unfortunately, this view overlooks one of the most critical problems in fieldwork—the importance of the social distribution of knowledge (Berger and Luckman 1967; Mannheim 1936; Schutz 1962). It also denies the relevance of the interactive situation as the source of our knowledge about the society and the implications of these shortcomings have been the focus of this paper.
104    Domestic Worlds and Public Worlds
Public and Private Politics: Women in the Middle Eastern World    1 OS
On the one hand, I have pointed out that when we examine the literature of the Middle East on women done by women, our ethnographic image of women and power is considerably different from that of the male ethnographers. On the other hand, when I examined the literature of the male ethnographers, I discovered hints that the women in these societies are not as powerless as the ethnographer themselves had concluded. In other words, there exists internal evidence that our images are incomplete, that the dynamics of power and authority are much more subtle than we have been led to believe, and that our theoretical perspectives about the position ofwomen in Middle Eastern society must be the common-sense world of the actors themselves.
NOTES
1. This is a much revised version of an earlier paper delivered at a symposium on nomadic-sedentary interaction in Middle Eastern societies held at the American University in Cairo. March 17-21, 1972 (Nelson 1973). The emergence of the present paper, however, is in large part due to the many stimulating and provocative discussions while on sabbatical at the Institute of International Studies, 1973-1974, with colleagues at the University of California, May N. Diaz, Lucile Newman, Elvi Whittaker, and, particularly, Virginia L. Olcsen, whose incisive and critical comments forced me to sharpen my own ideas considerably. To Hildred Gecrtz, a note of appreciation for suggesting a closer link between the central argument of the paper and its title.
2. The argument might be made that women share the man's construction of the social world, but this seems to me more of an implicit taken-for-granted assumption than an ethnographic fact. Until we know that this is the case, based on ethnographic evidence emerging from studies of women's views about their social world, we are making generalizations from information channeled from male informants to ethnographers.
3. Although Lichtenstadter's analysis is drawn from written materials rather than from per-
sonal observations, her insights are suggestive of themes emerging from more recent ethnographic studies (Lienhardt 1972). Dr. Abdou (personal communication) of the University of Riyadh drew my attention to the fact that the Prophet Mohammed worked for a female merchant whom lie later married—Khadidja. He commented that "generally speaking the Bedouin women of Saudi Arabia participate more as men in society than do their urban counterparts." Abdou attributes this to the greater Turkish influence on the seclusion of women in the urban centers. He also noted that in the Saudi .Arabian town of Taclrnur there are special women's markets where only women are sellers. This phenomenon is also recorded among the Berbers of the High Atlas and Rif Mountains of Morocco (Benet 1970:182, 193; Hart 1970:38).
4. Evidence that my generalizations about women and power in Arab society also held true for Arabicized Berbers in North Africa is found in several recent published and unpublished manuscripts (Alport 1970; Bcnct 1970; Hart 1970; Murphy 1970; Mason 1973; Joseph 1973). Subsequent to the submission of this essay, a useful exploration of this topic among pastoral nomads in the Sudan came to my attention. I wish to thank Elinor Kelly of the University of Manchester for generously sharing with me her unpublished thesis from the University of London.
5. Supporting this image ofwomen from yet another part of Moslem Africa is the following excerpt from an Agence France-Presse news item quoted in the June 23, 1973 issue of the San Francisco Chronicle:
Kano, Nigeria—Single women arc being ordered to get married immediately or leave Northern Nigeria because religious authorities here say that the current drought in West Africa is caused by prostitution and immorality. . . . Many unmarried women are reported to have fled their homes following the get-married-or-leave order by the Emirs in the Moslem area. Landlords have been ordered not to let out rooms to single women since, according to one Emir, "because of prostitution there has been no rain."
6. One historian of the Middle East has suggested that among nomads there is no equivalent Arabic term for the concept "public"
arena. The tent and the camp are not synonyms for public and private (Dols, personal communication). 7. The recent exchange between Abou Zahra (1970) and Antoun (1968) on the meaning of the modesty code in the Middle East is an excellent example of conflicting ethnographic images.
REFERENCES CITED
Abou-Zahra, Nadia. !979. On Modesty of Women in Arab Muslim Villages: A Reply. American Anthropologist 72(5): 1079-1088.
Alport, E. A. 1970. The Mzab. In Peoples and Cultures of the Middle East, Vol. II. L. Sweet, Ed. New York: The Natural History Press, pp. 225-241.
Al-Torki, Soraya. 1973a. Religion and Social Organization of Elite Families in Urban Saudi Arabia. Unpublished Ph.D. dissertation. University of California, Berkeley.
--. 1973b. Men-Women Relationship in .Arab
Societies: A Study of the Economic and Political Conditions of the Status of Women. Unpublished research proposal.
Antoun Richard. 196S. On the Modesty of Women in Arab Muslim Villages: A Study in the Accommodation of Traditions. American Anthropologist 70( 1): 671-697.
Asad, Taial! 1970. The Kababish Arabs: Power, Authority and Consent in a Nomadic Tribe. London: C. Hurst.
-. 1972. Market Model, Class Structure and
Consent. Man 7(l):74-94.
Aswad, Barbara. 1967. Key and Peripheral Roles of Noble Women in a Middle East Plains Village. Anthropological Quarterly -10(3): 139-153.
Barth, Fredrik. 1959. Political Leadership among Swat Pathans. London School of Economics Monographs on Social Anthropology, 19. London .
-. 1961. Nomads of South Persia: Basseri
Tribe of the Khameseh Confederacy. Boston: Little, Brown.
Benet, Francisco. 1970. Explosive Markets: The Berber Highlands. In Peoples and Cultures of the Middle East, Vol. I. Louise Sweet, Ed. New York: The Natural History Press, pp. 173-203.
Berger, Peter, and T. Luckman. 1967. The Social Construction of Reality. New York: Doubleday (Center Book).
Bujra, Abdullah. 1966. The Relationship between the Sexes amongst the Bedouin in a Town. Unpublished paper delivered at the Mediterranean Social Science Conference, Athens.
Cole, Donald. 1971. Social and Economic Structure of the Al Murrah: A Saudi Arabian Bedouin Tribe. Unpublished Ph.D. dissertation. University of California. Berkeley.
Crapanzano, Vincent. 1972. The Hamadsha. In Scholars, Saints and Sufis: Muslims Religious Institutions Since 1500. Nikki R. Keddie, Ed. Berkeley: University of California Press, pp. 327-3-18.
Gunnison. Ian. 1966. The Baggara Arabs: Power and Lineage in a Sudanese Nomad Tribe. Oxford: Clarendon Press.
Douglas, Mary. 1970. Natural Symbols: Explorations in Cosmology. New York: Random House (Pantheon).
Evans-Pritchard, E. E. 1971. The Senusi of Cyrenaica. Oxford: Clarendon Press.
Farrag, Amina. 1971. Social Control amongst the Mzabite Women of Beni-lsguen. Middle Eastern Studies, Winter (3):3 1 7-327.
Gaudry, Mathea. 1929. La Femme Chaouia de L'Aurcs. Paris: Geuthner.
Geertz, Clifford. 1968. Islam Observed. New Haven, CT: Yale University Press.
Gellner. Ernest. 1969. Saints of the Atlas. Chicago: University of Chicago Press.
Hart, David M. 1970. Clan, Lineage, Local Community and the Feud in a Rilian Tribe, hi Peoples and Cultures of Middle East, Vol. II. Louise Sweet, Ed. New York: The Natural History Press, pp. 3-75.
Hofer, Carol. 1972. Mende and Sherbro Women in High Office. Canadian Journal of African Studies 6(2):151-164.
Joseph, Roger. 1973. Choix ou Force: Une Etude sur la Manipulation Sociale. Unpublished manuscript.
Lebeuf, Annie. 1971. The Role of Women in the Political Organization of African Societies. In Women of Tropical Africa. Denise Pauline, Ed. Berkeley: University of California Press, pp. 93-119.
Lichtenstadter, Use. 1935. Women in Aiyam al-Arab. London: The Royal Asiatic Society.
Lienhardt, Peter A. 1972. Some Social Aspects of the Trucial States, hi The Arabian Peninsula: Society and Politics. D. Hopwood, Ed. London: Allen and Unwin, pp. 219-229.
Maher, Vanessa. 1972. Social Stratification and the Role of Women in the Middle Atlas of Morocco.
106    Domestic Worlds and Public Worlds
Unpublished Ph.D. dissertation. Cambridge University.
Mannheim, Karl. 1936. Ideology and Utopia. New York: Harcourt, Brace (Harvest Books.)
Marx, Emmanuel. 1967. Bedouin of the Negev. New York: Praeger.
Mason, John. n.d. Sex and Symbol in a Libyan Oasis. Unpublished manuscript.
Mohsen, Safia. 1967. Legal Status of Women among the Awlad Ali. Anthropological Quarterly 40(3): 153-166.
Mills, C. W. 1967. Situated Actions and the Vocabulary of Motives. In Power Politics and People. Oxford: Oxford University Press, pp. 439-452,
Murphy, Robert F. 1970. Social Distance and the Veil. In Peoples and Cultures of the Middle East, Vol. I. Louise Sweet, Ed. New York: The Natural History Press, pp. 290-314.
Nelson, Cynthia. 1973. Women and Power in Nomadic Societies of the Middle East. In The Desert and the Town Nomads in the Greater Society. Cynthia Nelson, Ed. Berkeley: Institute of International Studies, University of California, pp. 43-59.
Nieuwenhuijze, C. A. O. van. 1965. Social Stratification in the Middle East. The Hague: E. J, Brill.
Olesen, Virginia L. 1973. Notes on the Negotiation of Rules and Roles in Fieklwork Studies. Unpublished manuscript. San Francisco: De-
partment of Social and Behavioral Science; University of California.
Pehrson, Robert. 1966. The Social Organization o the Marri Baluch. Viking Fund Publications ii Anthropology, 53. New York.
Peters, Emrys. 1966. Consequences of the Segre gation of the Sexes among the Arabs. Unpub lished paper delivered at the Mediterraneai Social Science Council Conference, Athens.
Rosenfeld, Henry. 1968. The Contradictions be- . tween Property, Kinship, and Power as Reflected in the Marriage System of an Arab Village.   In   Contributions   to Mediterranean Sociology. ]. Peristiany, Ed. The Hague: Mou- : ton, pp. 247-260.
San Francisco Chronicle. June 23, 1973.
Schneider, Jane. 1971. Honor, Shame and Access to Resources in Mediterranean Societies. Ethnology !0(l):I-24.
Schutz, Alfred. 1962. Collected Papers, Vol. II: : The Problem of Social Reality. The Hague: Martinus Nijhof.
Sweet, Louise. 1970. Peoples and Cultures of the Middle East, Vols. 1 Sc 1. New York: The Natural History Press.
Tapper, Nancy. 1968. The Role of Women in Selected Pastoral Islamic Society. Unpublished M. A. thesis. S.O.A.S. London.
Wrong, Dennis H. n.d. Some Problems in Defining = Social Power. Unpublished manuscript.
We all live in a world of symbols that assign meaning and value to the categories of male and female. Despite several decades of consciousness raising in the United States, advertising on television and in the print media perpetuates sexual stereotypes. Although "house beautiful" ads are less prominent as women are increasingly shown in workplace contexts, body beautiful messages continue to be transmitted. In children's cartoons women are still the helpless victims who the fearless male hero must rescue. Toys are targeted either for little boys or little girls and are packaged appropriately in colors and materials culturally defined as either masculine or feminine.
To what extent are these stereotypes of men and women and the symbols with which they are associated universal? If they are universal, to what extent are they rooted in observed differences about the biological nature of men and women that are made culturally significant? These questions have interested scholars as they
have attempted to account for both similarity and difference among the people of the world.
Making the assumption that the subordination of women exists in all societies—a "true universal"—Ortner (1974:67) sought to explain the pervasiveness of this idea not in the assignation of women to a domestic sphere of activity, but in the symbolic constructions by which women's roles are evaluated. Ortner argues that women, because of their reproductive roles, are universally viewed as being closer to nature, while men are linked with culture. She defines culture as "the notion of human consciousness, or . . . the products of human consciousness (i.e., systems of thought and technology), by means of which humanity attempts to assert control over nature" (1974:72). That which is cultural and subject to human manipulation is assigned more worth than that which is natural; hence, women and women's roles are denigrated or devalued, whether explicitly or implicitly.
107