Under Western Eyes: Feminist Scholarship and Colonial Discourses
- First Online:
- Cite this article as:
- Mohanty, C. Fem Rev (1988) 30: 61. doi:10.1057/fr.1988.42
It ought to be of some political significance at least that the term ‘colonization’ has come to denote a variety of phenomena in recent feminist and left writings in general. From its analytic value as a category of exploitative economic exchange in both traditional and contemporary Marxisms (cf. particularly such contemporary scholars as Baran, Amin and Gunder-Frank) to its use by feminist women of colour in the US, to describe the appropriation of their experiences and struggles by hegemonic white women's movements,1 the term ‘colonization’ has been used to characterize everything from the most evident economic and political hierarchies to the production of a particular cultural discourse about what is called the ‘Third World.’2 However sophisticated or problematical its use as an explanatory construct, colonization almost invariably implies a relation of structural domination, and a discursive or political suppression of the heterogeneity of the subject(s) in question. What I wish to analyse here specifically is the production of the ‘Third World Woman’ as a singular monolithic subject in some recent (western) feminist texts. The definition of colonization I invoke is a predominantly discursive one, focusing on a certain mode of appropriation and codification of ‘scholarship’ and ‘knowledge’ about women in the third world by particular analytic categories employed in writings on the subject which take as their primary point of reference feminist interests as they have been articulated in the US and western Europe. My concern about such writings derives from my own implication and investment in contemporary debates in feminist theory, and the urgent political necessity of forming strategic coalitions across class, race and national boundaries. Clearly, western feminist discourse and political practice is neither singular nor homogeneous in its goals, interests or analyses. However, it is possible to trace a coherence of effects resulting from the implicit assumption of ‘the west’ (in all its complexities and contradictions) as the primary referent in theory and praxis. Thus, rather than claim simplistically that ‘western feminism’ is a monolith, I would like to draw attention to the remarkably similar effects of various analytical categories and even strategies which codify their relationship to the Other in implicitly hierarchical terms. It is in this sense that I use the term ‘western feminist’. Similar arguments pertaining to questions of methods of analysis can be made in terms of middle-class, urban African and Asian scholars producing scholarship on or about their rural or working-class sisters which assumes their own middle-class culture as the norm, and codifies peasant and working-class histories and cultures as Other. Thus, while this article focuses specifically on western feminist discourse on women in the third world, the critiques I offer also pertain to identical analytical principles employed by third-world scholars writing about their own cultures. Moreover, the analytical principles discussed below serve to distort western feminist political practices, and limit the possibility of coalitions among (usually white) western feminists and working-class and feminist women of colour around the world. These limitations are evident in the construction of the (implicitly consensual) priority of issues around which apparently all women are expected to organize. The necessary and integral connection between feminist scholarship and feminist political practice and organizing determines the significance and status of western feminist writings on women in the third world, for feminist scholarship, like most other kinds of scholarship, does not comprise merely ‘objective’ knowledge about a certain subject. It is also a directly political and discursive practice insofar as it is purposeful and ideological. It is best seen as a mode of intervention into particular hegemonic discourses (for example, traditional anthropology, sociology, literary criticism, etc.), and as a political praxis which counters and resists the totalizing imperative of age-old ‘legitimate’ and ‘scientific’ bodies of knowledge. Thus, feminist scholarly practices exist within relations of power - relations which they counter, redefine, or even implicitly support. There can, of course, be no apolitical scholarship. The relationship between Woman - a cultural and ideological composite Other constructed through diverse representational discourse (scientific, literary, juridical, linguistic, cinematic, etc.) - and women -real, material subjects of their collective histories - is one of the central questions the practice of feminist scholarship seeks to address. This connection between vomen as historical subjects and the re-presentation of Woman produced by hegemonic discourses is not a relation of direct identity, or a relation of correspondence or simple implication.3 It is an arbitrary relation set up in particular cultural and historical contexts. I would like to suggest that the feminist writings I analyse here discursively colonize the material and historical heterogeneities of the lives of women in the third world, thereby producing/representing a composite, singular ‘third-world woman’ - an image which appears arbitrarily constructed but nevertheless carries with it the authorizing signature of western humanist discourse.4 I argue that assumptions of privilege and ethnocentric universality on the one hand, and inadequate self-consciousness about the effect of western scholarship on the ‘third world’ in the context of a world system dominated by the west on the other, characterize a sizable extent of western feminist work on women in the third world. An analysis of ‘sexual difference’ in the form of a cross-culturally singular, monolithic notion of patriarchy or male dominance leads to the construction of a similarly reductive and homogeneous notion of what I shall call the ‘third-world difference’ - that stable, ahistorical something that apparently oppresses most if not all the women in these countries. It is in the production of this ‘third-world difference’ that western feminisms appropriate and colonize the constitutive complexities which characterize the lives of women in these countries. It is in this process of discursive homogenization and systematization of the oppression of women in the third world that power is exercised in much of recent western feminist writing, and this power needs to be defined and named. In the context of the west's hegemonic position today, of what Anouar Abdel-Malek calls a struggle for ‘control over the orientation, regulation and decision of the process of world development on the basis of the advanced sector's monopoly of scientific knowledge and ideal creativity’ (1981: especially 145), western feminist scholarship on the third world must be seen and examined precisely in terms of its inscription in these particular relations of power and struggle. There is, it should be evident, no universal patriarchal framework which this scholarship attempts to counter and resist - unless one posits an international male conspiracy or a monolithic, transhistorical power structure. There is, however, a particular world balance of power within which any analysis of culture, ideology, and socio-economic conditions has to be necessarily situated. Abdel-Malek is useful here, again, in reminding us about the inherence of politics in the discourses of ‘culture’: Contemporary imperialism is, in a real sense, a hegemonic imperialism, exercising to a maximum degree a rationalized violence taken to a higher level than ever before - through fire and sword, but also through the attempt to control hearts and minds. For its content is defined by the combined action of the military-industrial complex and the hegemonic cultural centers of the West, all of them founded on the advanced levels of development attained by monopoly and finance capital, and supported by the benefits of both the scientific and technological revolution and the second industrial revolution itself. (1981: 145-6) Western feminist scholarship cannot avoid the challenge of situating itself and examining its role in such a global economic and political framework. To do any less would be to ignore the complex interconnections between first- and third-world economies and the profound effect of this on the lives of women in all countries. I do not question the descriptive and informative value of most western feminist writings on women in the third world. I also do not question the existence of excellent work which does not fall into the analytic traps I am concerned with. In fact I deal with an example of such work later on. In the context of an overwhelming silence about the experiences of women in these countries, as well as the need to forge international links between women's political struggles, such work is both pathbreaking and absolutely essential. However, it is both to the explanatory potential of particular analytic strategies employed by such writing, and to their political effect in the context of the hegemony of western scholarship, that I want to draw attention here. While feminist writing in the US is still marginalized (except perhaps from the point of view of women of colour addressing privileged white women), western feminist writing on women in the third world must be considered in the context of the global hegemony of western scholarship - i.e., the production, publication, distribution and consumption of information and ideas. Marginal or not, this writing has political effects and implications beyond the immediate feminist or disciplinary audience. One such significant effect of the dominant ‘representations’ of western feminism is its conflation with imperialism in the eyes of particular third-world women.5 Hence the urgent need to examine the political implications of our analytic strategies and principles. My critique is directed at three basic analytical presuppositions which are present in (western) feminist discourse on women in the third world. Since I focus primarily on the Zed Press ‘Women in the Third World’ series, my comments on western feminist discourse are circumscribed by my analysis of the texts in this series.6 This is a way of focusing my critique. However, even though I am dealing with feminists who identify themselves as culturally or geographically from the ‘west’, as mentioned earlier, what I say about these presuppositions or implicit principles holds for anyone who uses these analytical strategies, whether third-world women in the west, or third-world women in the third world writing on these issues and publishing in the west. Thus, I am not making a culturalist argument about ethnocentrism; rather, I am trying to uncover how ethnocentric universalism is produced in certain analyses. As a matter of fact, my argument holds for any discourse that sets up its own authorial subjects as the implicit referent, i.e., the yardstick by which to encode and represent cultural Others. It is in this move that power is exercised in discourse. The first analytical presupposition I focus on is involved in the strategic location or situation of the category ‘women’ vis-a-vis the context of analysis. The assumption of women as an already constituted and coherent group with identical interests and desires, regardless of class, ethnic or racial location, implies a notion of gender or sexual difference or even patriarchy which can be applied universally and cross-culturally. (The context of analysis can be anything from kinship structures and the organization of labour to media representations.) The second analytical presupposition is evident on the methodological level, in the uncritical way ‘proof of universality and cross-cultural validity are provided. The third is a more specifically political presupposition, underlying the methodologies and the analytic strategies, i.e., the model of power and struggle they imply and suggest. I argue that as a result of the two modes - or, rather, frames - of analysis described above, a homogeneous notion of the oppression of women as a group is assumed, which, in turn, produces the image of an ‘average third-world woman’. This average third-world woman leads an essentially truncated life based on her feminine gender (read: sexually constrained) and being ‘third world’ (read: ignorant, poor, uneducated, tradition-bound, religious, domesticated, family-oriented, victimized, etc.). This, I suggest, is in contrast to the (implicit) self-representation of western women as educated, modern, as having control over their own bodies and sexualities, and the ‘freedom’ to make their own decisions. The distinction between western feminist re-presentation of women in the third world, and western feminist self-presentation is a distinction of the same order as that made by some Marxists between the ‘maintenance’ function of the housewife and the real ‘productive’ role of wage-labour, or the characterization by developmentalists of the third world as being engaged in the lesser production of ‘raw materials’ in contrast to the ‘real’ productive activity of the first world. These distinctions are made on the basis of the privileging of a particular group as the norm or referent. Men involved in wage-labour, first-world producers, and, I suggest, western feminists who sometimes cast third-world women in terms of ‘ourselves undressed’ (Michelle Rosaldo's term; Rosaldo, 1980: 389-412, especially 392), all construct themselves as the normative referent in such a binary analytic. ‘Women' as category of analysis, or: We are all sisters in struggle By women as a category of analysis, I am referring to the crucial presupposition that all of us of the same gender, across classes and cultures, are somehow socially constituted as a homogeneous group identifiable prior to the process of analysis. The homogeneity of women as a group is produced not on the basis of biological essentials, but rather on the basis of secondary sociological and anthropological universals. Thus, for instance, in any given piece of feminist analysis, women are characterized as a singular group on the basis of a shared oppression. What binds women together is a sociological notion of the ‘sameness’ of their oppression. It is at this point that an elision takes place between ‘women’ as a discursively constructed group and ‘women’ as material subjects of their own history.7 Thus, the discursively consensual homogeneity of ‘women’ as a group is mistaken for the historically specific material reality of groups of women. This results in an assumption of women as an always-already constituted group, one which has been labelled ‘powerless’, ‘exploited’, ‘sexually harassed’, etc., by feminist scientific, economic, legal and sociological discourses. (Notice that this is quite similar to sexist discourse labelling women as weak, emotional, having math anxiety, etc.) The focus is not on uncovering the material and ideological specificities that constitute a group of women as ‘powerless’ in a particular context. It is rather on finding a variety of cases of ‘powerless’ groups of women to prove the general point that women as a group are powerless.8 In this section I focus on five specific ways in which ‘women’ as a category of analysis is used in western feminist discourse on women in the third world to construct ‘third-world women’ as a homogeneous ‘powerless’ group often located as implicit victims of particular cultural and socio-economic systems. I have chosen to deal with a variety of writers - from Fran Hosken, who writes primarily about female genital mutilation, to writers from the Women in International Development school who write about the effect of development policies on third-world women for both western and third-world audiences. I do not intend to equate all the texts that I analyse, nor ignore their respective strengths and weaknesses. The authors I deal with write with varying degrees of care and complexity; however, the effect of the representation of third-world women in these texts is a coherent one. In these texts women are variously defined as victims of male violence (Fran Hosken); victims of the colonial process (M. Cutrufelli); victims of the Arab familial system (Juliette Minces); victims of the economic development process (B. Lindsay and the - liberal - WID school); and finally, victims of the economic basis of the Islamic code (P. Jeffery). This mode of defining women primarily in terms of their object status (the way in which they are affected or not affected by certain institutions and systems) is what characterizes this particular form of the use of ‘women’ as a category of analysis. In the context of western women writing about and studying women in the third world, such objectification (however benevolently motivated) needs to be both named and challenged. As Valerie Amos and Pratibha Parmar argue quite eloquently, ‘Feminist theories which examine our cultural practices as “feudal residues” or label us “traditional”, also portray us as politically immature women who need to be versed and schooled in the ethos of western feminism. They need to be continually challenged’ (1984: 7). Women as victims of male violence Fran Hosken, in writing about the relationship between human rights and female genital mutilation in Africa and the Middle East, bases her whole discussion and condemnation of genital mutilation on one privileged premise: the goal of genital mutilation is ‘to mutilate the sexual pleasure and satisfaction of woman’ (1981: 3-24, especially 11).9 This, in turn, leads her to claim that woman's sexuality is controlled, as is her reproductive potential. According to Hosken, ‘male sexual politics’ in Africa and around the world ‘share the same political goal: to assure female dependence and subservience by any and all means’. Physical violence against women (rape, sexual assault, excision, infibulation, etc.) is thus carried out ‘with an astonishing consensus among men in the world’ (14). Here, women are defined systematically as the victims of male control - the ‘sexually oppressed’. Although it is true that the potential of male violence against women circumscribes and elucidates their social position to a certain extent, defining women as archetypal victims freezes them into ‘objects-who-defend-themselves’, men into ‘subjects-who-perpetrate-violence’, and (every) society into a simple opposition between the powerless (read: women) and the powerful (read: men) groups of people. Male violence (if that indeed is the appropriate label) must be theorized and interpreted within specific societies, both in order to understand it better, as well as in order to effectively organize to change it.’10 Sisterhood cannot be assumed on the basis of gender; it must be forged in concrete historical and political praxis. Women as universal dependants Beverley Lindsay's conclusion to the book, Comparative Perspectives on Third World Women: The Impact of Race, Class and Sex states: ‘Dependency relationships, based upon race, sex and class, are being perpetuated through social, educational, and economic institutions. These are the linkages among Third World Women’ (1983: especially 298, 306). Here, as in other places, Lindsay implies that third-world women constitute an identifiable group purely on the basis of shared dependencies. If shared dependencies were all that was needed to bind us together as a group, third-world women would always be seen as an apolitical group with no subject status! Instead, if anything, it is the common context of political struggle against class, race, gender and imperialist hierarchies that may constitute third-world women as a strategic group at this historical juncture. Lindsay also states that linguistic and cultural differences exist between Vietnamese and Black American women, but ‘both groups are victims of race, sex and class’. Again, Black and Vietnamese women are characterized and defined simply in terms of their victim status. Similarly, examine statements like: ‘My analysis will start by stating that all African women are politically and economically dependent’ (Cutrufelli, 1983: especially 13). Or: ‘Nevertheless, either overtly or covertly, prostitution is still the main if not the only source of work for African women’ (Cutrufelli, 1983: 33). All African women are dependent. Prostitution is the only work option for African women as a group. Both statements are illustrative of generalizations sprinkled liberally through a recent Zed Press publication, Women of Africa: Roots of Oppression, by Maria Rosa Cutrufelli, who is described on the cover as an ‘Italian Writer, Sociologist, Marxist and Feminist’. In the 1980s is it possible to imagine writing a book entitled ‘Women of Europe: Roots of Oppression’? I am not objecting to the use of universal groupings for descriptive purposes. Women from the continent of Africa can be descriptively characterized as ‘Women of Africa’. It is when ‘women of Africa’ becomes a homogeneous sociological grouping characterized by common dependencies or powerlessness (or even strengths) that problems arise - we say too little and too much at the same time. This is because descriptive gender differences are transformed into the division between men and women. Women are constituted as a group via dependency relationships vis-a-vis men, who are implicitly held responsible for these relationships. When ‘women of Africa’ (versus ‘men of Africa’ as a group?) are seen as a group precisely because they are generally dependent and oppressed, the analysis of specific historical differences becomes impossible, because reality is always apparently structured by divisions between two mutually exclusive and jointly exhaustive groups, the victims and the oppressors. Here the sociological is substituted for the biological in order, however, to create the same - a unity of women. Thus, it is not the descriptive potential of gender difference but the privileged positioning and .explanatory potential of gender difference as the origin of oppression that I question. In using ‘women of Africa’ (as an already constituted group of oppressed peoples) as a category of analysis, Cutrufelli denies any historical specificity to the location of women as subordinate, powerful, marginal, central, or otherwise, vis-a-vis particular social and power networks. Women are taken as a unified ‘powerless’ group prior to the historical and political analysis in question. Thus, it is then merely a matter of specifying the context after the fact. ‘Women’ are now placed in the context of the family, or in the workplace, or within religious networks, almost as if these systems existed outside the relations of women with other women, and women with men. The problem with this analytical strategy is, let me repeat, that it assumes men and women are already constituted as sexual-political subjects prior to their entry into the arena of social relations. Only if we subscribe to this assumption is it possible to undertake analysis which looks at the ‘effects’ of kinship structures, colonialism, organization of labour, etc., on women, who are defined in advance as a group. The crucial point that is forgotten is that women are produced through these very relations as well as being implicated in forming these relations. As Michelle Rosaldo argues, ‘woman's place in human social life is not in any direct sense a product of the things she does (or even less, a function of what, biologically, she is) but the meaning her activities acquire through concrete social interactions’ (1980: 400). That women mother in a variety of societies is not as significant as the value attached to mothering in these societies. The distinction between the act of mothering and the status attached to it is a very important one - one that needs to be stated and analysed contextually. Married women as victims of the colonial process In Levi-Strauss's theory of kinship structures as a system of the exchange of women, what is significant is that exchange itself is not constitutive of the subordination of women; women are not subordinate because of the fact of exchange, but because of the modes of exchange instituted, and the values attached to these modes. However, in discussing the marriage ritual of the Bemba, a Zambian matrilocal, matrilineal people, Cutrufelli in Women of Africa focuses on the fact of the marital exchange of women before and after western colonization, rather than the value attached to this exchange in this particular context. This leads to her definition of Bemba women as a coherent group affected in a particular way by colonization. Here again, Bemba women are constituted rather unilaterally as the victims of western colonization. Cutrufelli cites the marriage ritual of the Bemba as a multi-stage event ‘whereby a young man becomes incorporated into his wife's family group as he takes up residence with them and gives his services in return for food and maintenance’ (1983: 43). This ritual extends over many years, and the sexual relationship varies according to the degree of the girl's physical maturity. It is only after the girl undergoes an initiation ceremony at puberty that intercourse is sanctioned, and the man acquires legal rights over the woman. This initiation ceremony is the most important act of the consecration of women's reproductive power, so that the abduction of an uninitiated girl is of no consequence, while heavy penalty is levied for the seduction of an initiated girl. Cutrufelli asserts that the effect of European colonization has changed the whole marriage system. Now the young man is entitled to take his wife away from her people in return for money. The implication is that Bemba women have now lost the protection of tribal laws. However, while it is possible to see how the structure of the traditional marriage contract (as opposed to the post-colonial marriage contract) offered women a certain amount of control over their marital relations, only an analysis of the political significance of the actual practice which privileges an initiated girl over an uninitiated one, indicating a shift in female power relations as a result of this ceremony, can provide an accurate account of whether Bemba women were indeed protected by tribal laws at all times. However, it is not possible to talk about Bemba women as a homogeneous group within the traditional marriage structure. Bemba women before the initiation are constituted within a different set of social relations compared to Bemba women after the initiation. To treat them as a unified group, characterized by the fact of their ‘exchange’ between male kin, is to deny the specificities of their daily existence, and the differential value attached to their exchange before and after their initiation. It is to treat the initiation ceremony as a ritual with no political implications or effects. It is also to assume that in merely describing the structure of the marriage contract, the situation of women is exposed. Women as a group are positioned within a given structure, but there is no attempt made to trace the effect of the marriage practice in constituting women within an obviously changing network of power relations. Thus, women are assumed to be sexual-political subjects prior to entry into kinship structures. Women and familial systems Elizabeth Cowie, in another context (1978: 49-63), points out the implications of this sort of analysis when she emphasizes the specifically political nature of kinship structures which must be analysed as ideological practices which designate men and women as father, husband, wife, mother, sister, etc. Thus, Cowie suggests, women as women are not simply located within the family. Rather, it is in the family, as an effect of kinship structures, that women as women are constructed, defined within and by the group. Thus, for instance, when Juliette Minces (1980: especially 23) cites the patriarchal family as the basis for ‘an almost identical vision of women’ that Arab and Muslim societies have, she falls into this very trap. Not only is it problematical to speak of a vision of women shared by Arab and Muslim societies, without addressing the particular historical and ideological power structures that construct such images, but to speak of the patriarchal family or the tribal kinship structure as the origin of the socio-economic status of women is again to assume that women are sexual-political subjects prior to their entry into the family. So while on the one hand women attain value or status within the family, the assumption of a singular patriarchal kinship system (common to all Arab and Muslim societies, i.e. over twenty different countries) is what apparently structures women as an oppressed group in these societies! This singular, coherent kinship system presumably influences another separate and given entity, ‘women’. Thus all women, regardless of class and cultural differences, are seen as being similarly affected by this system. Not only are all Arab and Muslim women seen to constitute a homogeneous oppressed group, but there is no discussion of the specific practices within the family which constitute women as mothers, wives, sisters, etc. Arabs and Muslims, it appears, don't change at all. Their patriarchal family is carried over from the times of the Prophet Muhammad. They exist, as it were, outside history. Women and religious ideologies A further example of the use of ‘women’ as a category of analysis is found in cross-cultural analyses which subscribe to a certain economic reductionism in describing the relationship between the economy and factors such as politics and ideology. Here, in reducing the level of comparison to the economic relations between ‘developed’ and ‘developing’ countries, the question of women is denied any specificity. Mina Modares, in a careful analysis of women and Shi'ism in Iran, focuses on this very problem when she criticizes feminist writings which treat Islam as an ideology separate from and outside social relations and practices, rather than a discourse which includes rules for economic, social and power relations within society (Modares, 1981: 62-82). Patricia Jeffery's otherwise informative work on Pirzada women in purdah (1979) considers Islamic ideology as a partial explanation for the status of women in that it provides a justification for the purdah. Here, Islamic ideology is reduced to a set of ideas whose internalization by Pirzada women contributes to the stability of the system. The primary explanation for purdah is located in the control that Pirzada men have over economic resources, and the personal security purdah gives to Pirzada women. By taking a specific version of Islam as the Islam, Jeffery attributes a singularity and coherence to it. Modares notes, ‘“Islamic Theology” then becomes imposed on a separate and given entity called “women”. A further unification is reached: Women (meaning all women), regardless of their differing positions within societies, come to be affected or not affected by Islam. These conceptions provide the right ingredients for an unproblematic possibility of a cross-cultural study of women’ (1981: 63). Marnia Lazreg makes a similar argument when she addresses the reductionism inherent in scholarship on women in the Middle East and North Africa: A ritual is established whereby the writer appeals to religion as the cause of gender inequality just as it is made the source of underdevelopment in much of modernization theory. In an uncanny way, feminist discourse on women from the Middle East and North Africa mirrors that of theologians' own interpretation of women in Islam.... The overall effect of this paradigm is to deprive women of self-presence, of being. Because women are subsumed under religion presented in fundamental terms, they are inevitably seen as evolving in nonhistorical time. They have virtually no history. Any analysis of change is therefore foreclosed.(Lazreg, 1988: 87) While Jeffery's analysis does not quite succumb to this kind of unitary notion of religion (Islam), it does collapse all ideological specificities into economic relations, and universalizes on the basis of this comparison. Women and the development process The best examples of universalization on the basis of economic reductionism can be found in the liberal ‘Women in Development’ literature. Proponents of this school seek to examine the effect of development on third-world women, sometimes from self-designated feminist perspectives. At the very least, there is an evident interest in and commitment to improving the lives of women in ‘developing’ countries. Scholars like Irene Tinker, Ester Boserup, and Perdita Huston11 have all written about the effect of development policies on women in the third world. All three women assume that ‘development’ is synonymous with ‘economic development’ or ‘economic progress’. As in the case of Minces' patriarchal family, Hosken's male sexual control, and Cutrufelli's western colonization, ‘development’ here becomes the all-time equalizer. Women are seen as being affected positively or negatively by economic development policies, and this is the basis for cross-cultural comparison. For instance, Perdita Huston states that the purpose of her study is to describe the effect of the development process on the ‘family unit and its individual members’ in Egypt, Kenya, Sudan, Tunisia, Sri Lanka and Mexico. She states that the ‘problems’ and ‘needs’ expressed by rural and urban women in these countries all centre around education and training, work and wages, access to health and other services, political participation and legal rights. Huston relates all these ‘needs’ to the lack of sensitive development policies which exclude women as a group. For her, the solution is simple: improved development policies which emphasize training for women field-workers, use women trainees and women rural development officers, encourage women's cooperatives, etc. Here, again women are assumed to be a coherent group or category prior to their entry into ‘the development process’. Huston assumes that all third-world women have similar problems and needs. Thus, they must have similar interests and goals. However, the interests of urban, middle-class, educated Egyptian housewives, to take only one instance, could surely not be seen as being the same as those of their uneducated, poor maids. Development policies do not affect both groups of women in the same way. Practices which characterize women's status and roles vary according to class. Women are constituted as women through the complex interaction between class, culture, religion and other ideological institutions and frameworks. They are not ‘women’ - a coherent group - solely on the basis of a particular economic system or policy. Such reductive cross-cultural comparisons result in the colonization of the specifics of daily existence and the complexities of political interests which women of different social classes and cultures represent and mobilize. Thus it is revealing that for Perdita Huston women in the third-world countries she writes about have ‘needs’ and ‘problems’, but few if any have ‘choices’ or the freedom to act. This is an interesting representation of women in the third world, one which is significant in suggesting a latent self-presentation of western women which bears looking at. She writes, ‘What surprised and moved me most as I listened to women in such very different cultural settings was the striking commonality - whether they were educated or illiterate, urban or rural - of their most basic values: the importance they assign to family, dignity, and service to others’ (Huston, 1979: 115). Would Huston consider such values unusual for women in the west? What is problematical, then, about this kind of use of ‘women’ as a group, as a stable category of analysis, is that it assumes an ahistorical, universal unity among women based on a generalized notion of their subordination. Instead of analytically demonstrating the production of women as socio-economic political groups within particular local contexts, this analytical move - and the presuppositions it is based on - limits the definition of the female subject to gender identity, completely bypassing social class and ethnic identities. What characterizes women as a group is their gender (sociologically not necessarily biologically defined) over and above everything else, indicating a monolithic notion of sexual difference. Because women are thus constituted as a coherent group, sexual difference becomes coterminus with female subordination, and power is automatically defined in binary terms: people who have it (read: men), and people who do not (read: women). Men exploit, women are exploited. Such simplistic formulations are both historically reductive; they are also ineffectual in designing strategies to combat oppressions. All they do is reinforce binary divisions between men and women. What would an analysis which did not do this look like? Maria Mies's work is one such example. It is an example which illustrates the strength of western feminist work on women in the third world and which, does not fall into the traps discussed above. Maria Mies's study of the lace-makers of Narsapur, India (1982), attempts to analyse carefully a substantial household industry in which ‘housewives’ produce lace doilies for consumption in the world market. Through a detailed analysis of the structure of the lace industry, production and reproduction relations, the sexual division of labour, profits and exploitation, and the overall consequences of defining women as ‘non-working housewives’ and their work as ‘leisure-time activity’, Mies demonstrates the levels of exploitation in this industry and the impact of this production system on the work and living conditions of the women involved in it. In addition, she is able to analyse the ‘ideology of the housewife’, the notion of a woman sitting in the house, as providing the necessary subjective and socio-cultural element for the creation and maintenance of a production system that contributes to the increasing pauperization of women, and keeps them totally atomized and disorganized as workers. Mies's analyses show the effect of a certain historically and culturally specific mode of patriarchal organization, an organization constructed on the basis of the definition of the lace-makers as ‘non-working housewives’ at familial, local, regional, statewide and international levels. The intricacies and the effects of particular power networks are not only emphasized; they also form the basis of Mies's analysis of how this particular group of women is situated at the centre of a hegemonic, exploitative world market. This is a good example of what careful, politically focused, local analyses can accomplish. It illustrates how the category of woman is constructed in a variety of political contexts that often exist simultaneously and overlaid on top of one another. There is no easy generalization in the direction of ‘women’ in India, or ‘women in the third world’; nor is there a reduction of the political construction of the exploitation of the lace-makers to cultural explanations about the passivity or obedience that might characterize these women and their situation. Finally, this mode of local, political analysis which generates theoretical categories from within the situation and context being analysed, also suggests corresponding effective strategies for organizing against the exploitations faced by the lace-makers. Here Narsapur women are not mere victims of the production process, because they resist, challenge, and subvert the process at various junctures. This is one instance of how Mies delineates the connections between the housewife ideology, the self-consciousness of the lace-makers and their inter-relationships as contributing to the latent resistances she perceives among the women: The persistence of the housewife ideology, the self-perception of the lace makers as petty commodity producers rather than as workers, is not only upheld by the structure of the industry as such but also by the deliberate propagation and reinforcement of reactionary patriarchal norms and institutions. Thus, most of the lace makers voiced the same opinion about the rules of purdah and seclusion in their communities which were also propagated by the lace exporters. In particular, the Kapu women said that they had never gone out of their houses, that women of their community could not do any other work than housework and lace work etc. but in spite of the fact that most of them still subscribed fully to the patriarchal norms of the gosha women, there were also contradictory elements in their consciousness. Thus, although they looked down with contempt upon women who were able to work outside the house - like the untouchable Mala and Madiga women or women of other lower castes, they could not ignore the fact that these women were earning more money precisely because they were not respectable housewives but workers. At one discussion, they even admitted that it would be better if they could also go out and do coolie work. And when they were asked whether they would be ready to come out of their houses and work in one place in some sort of a factory, they said they would do that. This shows that the purdah and housewife ideology, although still fully internalized, already had some cracks, because it has been confronted with several contradictory realities. (Mies, 1982: 157) It is only by understanding the contradictions inherent in women's location within various structures that effective political action and challenges can be devised. Mies's study goes a long way towards offering such an analysis. While there are now an increasing number of western feminist writings in this tradition,12 there is also unfortunately a large block of writing which succumbs to the cultural reductionism discussed earlier. Methodological universalisms, or: women's oppression is a global phenomenon Western feminist writings on women in the third world subscribe to a variety of methodologies to demonstrate the universal cross-cultural operation of male dominance and female exploitation. I summarize and critique three such methods below, moving from the most simple to the most complex methodologies. First, proof of universalism is provided through the use of an arithmetic method. The argument goes like this: the more the number of women who wear the veil, the more universal is the sexual segregation and control of women (Deardon, 1975: 4-5). Similarly, a large number of different, fragmented examples from a variety of countries also apparently add up to a universal fact. For instance, Muslim women in Saudi Arabia, Iran, Pakistan, India and Egypt all wear some sort of a veil. Hence, this indicates that the sexual control of women is a universal fact in those countries in which the women are veiled (Deardon, 1975: 7, 10). Fran Hosken writes: ‘Rape, forced prostitution, polygamy, genital mutilation, pornography, the beating of girls and women, purdah (segregation of women) are all violations of basic human rights’ (1981: 15). By equating purdah with rape, domestic violence, and forced prostitution, Hosken asserts its ‘sexual control’ function as the primary explanation for purdah, whatever the context. Institutions of purdah are thus denied any cultural and historical specificity and contradictions and potentially subversive aspects are totally ruled out. In both these examples, the problem is not in asserting that the practice of wearing a veil is widespread. This assertion can be made on the basis of numbers. It is a descriptive generalization. However, it is the analytic leap from the practice of veiling to an assertion of its general significance in controlling women that must be questioned. While there may be a physical similarity in the veils worn by women in Saudi Arabia and Iran, the specific meaning attached to this practice varies according to the cultural and ideological context. In addition, the symbolic space occupied by the practice of purdah may be similar in certain contexts, but this does not automatically indicate that the practices themselves have identical significance in the social realm. For example, as is well known, Iranian middle-class women veiled themselves during the 1979 revolution to indicate solidarity with their veiled working-class sisters, while in contemporary Iran mandatory Islamic laws dictate that all Iranian women wear veils. While in both these instances similar reasons might be offered for the veil (opposition to the Shah and western cultural colonization in the first case, and the true Islamicization of Iran in the second), the concrete meanings attached to Iranian women wearing the veil are clearly different in the two historical contexts. In the first case, wearing the veil is both an oppositional and revolutionary gesture on the part of Iranian middle-class women; in the second case it is a coercive, institutional mandate.13 It is on the basis of such context-specific differentiated analysis that effective political strategies can be generated. To assume that the mere practice of veiling women in a number of Muslim countries indicates the universal oppression of women through sexual segregation is not only analytically reductive, but also proves to be quite useless when it comes to the elaboration of oppositional political strategy. Second, concepts like reproduction, the sexual division of labour, the family, marriage, household, patriarchy, etc., are often used without their specification in local cultural and historical contexts. These concepts are used by feminists in providing explanations for women's subordination, apparently assuming their universal applicability. For instance, how is it possible to refer to ‘the’ sexual division of labour when the content of this division changes radically from one environment to the next, and from one historical juncture to another? At its most abstract level, it is the fact of the differential assignation of tasks according to sex that is significant; however, this is quite different from the meaning or value that the content of this sexual division of labour assumes in different contexts. In most cases the assigning of tasks on the basis of sex has an ideological origin. There is no question that a claim such as ‘women are concentrated in service-oriented occupations in a large number of countries around the world’ is descriptively valid. Descriptively, then, perhaps the existence of a similar sexual division of labour (where women work in service occupations like nursing, social work, etc., and men in other kinds of occupations) in a number of different countries can be asserted. However, the concept of the ‘sexual division of labour’ is more than just a descriptive category. It indicates the differential value placed on ‘men's work’ versus ‘women's work’. Often the mere existence of a sexual division of labour is taken to be proof of the oppression of women in various societies. This results from a confusion between and collapsing together of the descriptive and explanatory potential of the concept of the sexual division of labour. Superficially similar situations may have radically different, historically specific explanations, and cannot be treated as identical. For instance, the rise of female-headed households in middle-class America might be construed as indicating women's independence and progress, whereby women are considered to have chosen to be single parents, there are increasing numbers of lesbian mothers, etc. However, the recent increase in female-headed households in Latin America,14 where women might be seen to have more decision-making power, is concentrated among the poorest strata, where life choices are the most constrained economically. A similar argument can be made for the rise of female-headed families among Black and Chicana women in the US. The positive correlation between this and the level of poverty among women of colour and white working-class women in the US has now even acquired a name: the feminization of poverty. Thus, while it is possible to state that there is a rise in female-headed households in the US and in Latin America, this rise cannot be discussed as a universal indicator of women's independence, nor can it be discussed as a universal indicator of women's impoverishment. The meaning and explanation for the rise must obviously be specified according to the socio-historical context. Similarly, the existence of a sexual division of labour in most contexts cannot be sufficient explanation for the universal subjugation of women in the workforce. That the sexual division of labour does indicate a devaluation of women's work must be shown through analysis of particular local contexts. In addition, devaluation of women must also be shown through careful analysis. In other words, the ‘sexual division of labour’ and ‘women’ are not commensurate analytical categories. Concepts like the sexual division of labour can be useful only if they are generated through local, contextual analyses.15 If such concepts are assumed to be universally applicable, the resultant homogenization of class, race, religious, and daily material practices of women in the third world can create a false sense of the commonality of oppressions, interests and struggles between and amongst women globally. Beyond sisterhood there is still racism, colonialism and imperialism! Finally, some writers confuse the use of gender as a superordinate category of organizing analysis with the universalistic proof and instantiation of this category. In other words, empirical studies of gender differences are confused with the analytical organization of cross-cultural work. Beverley Brown's review (1983) of the book Nature, Culture and Gender (1980) best illustrates this point. Brown suggests that nature: culture and female:male are superordinate categories which organize and locate lesser categories (like wild/domestic and biology/technology) within their logic. These categories are universal in the sense that they organize the universe of a system of representations. This relation is totally independent of the universal substantiation of any particular category. Her critique hinges on the fact that rather than clarify the generalizability ofnature:culture::female: male as superordinate organizational categories, Nature, Culture and Gender, the book, construes the universality of this equation to lie at the level of empirical truth, which can be investigated through field-work. Thus, the usefulness of the nature:culture::female:male paradigm as a universal mode of the organization of representation within any particular socio-historical system is lost. Here, methodological universalism is assumed on the basis of the reduction of the nature:culture: :female:male analytic categories to a demand for empirical proof of its existence in different cultures. Discourses of representation are confused with material realities, and the distinction between ‘Woman’ and ‘women’ is lost. Feminist work on women in the third world which blurs this distinction (a distinction which interestingly enough is often present in certain western feminists' self-representation) eventually ends up constructing monolithic images of ‘Third World Women’ by ignoring the complex and mobile relationships between their historical materiality on the level of specific oppressions and political choices on the one hand and their general discursive representations on the other. To summarize: I have discussed three methodological moves identifiable in feminist (and other academic) cross-cultural work which seeks to uncover a universality in women's subordinate position in society. The next and final section pulls together the previous sections attempting to outline the political effects of the analytical strategies in the context of western feminist writing on women in the third world. These arguments are not against generalization as much as they are for careful, historically specific generalizations responsive to complex realities. Nor do these arguments deny the necessity of forming strategic political identities and affinities. Thus, while Indian women of different backgrounds might forge a political unity on the basis of organizing against police brutality towards women,16 an analysis of police brutality must be contextual. Strategic coalitions which construct oppositional political identities for themselves are based on generalization and provisional unities, but the analysis of these group identities cannot be based on universalistic, ahistorical categories. The subject(s) of power This last section returns to an earlier point about the inherently political nature of feminist scholarship, and attempts to clarify my point about the possibility of detecting a colonialist move in the case of a structurally unequal first/third-world relation in scholarship. The nine texts in the Zed Press ‘Women in the Third World’ series that I have discussed17 focused on the following common areas in discussing women's ‘status’ within various societies: religion, family/kinship structures, the legal system, the sexual division of labour, education and, finally, political resistance. A large number of western feminist writings on women in the third world focus on these themes. Of course, the Zed texts have varying emphases. For instance, two of the studies, Women of Palestine (1982) and Indian Women in Struggle (1980), focus explicitly on female militancy and political involvement, while Women in Arab Society (1980) deals with Arab women's legal, religious and familial status. In addition, each text evidences a variety of methodologies and degrees of care in making generalizations. Interestingly enough, however, almost all the texts assume ‘women’ as a category of analysis in the manner designated above. Clearly this is an analytical strategy which is neither limited to these Zed Press publications, nor symptomatic of Zed Press publications in general. However, in the particular texts under question, each text assumes ‘women’ have a coherent group identity within the different cultures discussed, prior to their entry into social relations. Thus, Omvedt can talk about ‘Indian Women’ while referring to a particular group of women in the State of Maharashtra, Cutrufelli about ‘Women of Africa’ and Minces about ‘Arab Women’ as if these groups of women have some sort of obvious cultural coherence, distinct from men in these societies. The ‘status’ or ‘position’ of women is assumed to be self-evident because women as an already constituted group are placed within religious, economic, familial and legal structures. However, this focus on the position of women whereby women are seen as a coherent group across contexts, regardless of class or ethnicity, structures the world in ultimately binary, dichotomous terms, where women are always seen in opposition to men, patriarchy is always necessarily male dominance, and the religious, legal, economic and familial systems are implicitly assumed to be constructed by men. Thus, both men and women are always seen as preconstituted whole populations, and relations of dominance and exploitation are also posited in terms of whole peoples - wholes coming into exploitative relations. It is only when men and women are seen as different categories or groups possessing different already constituted categories of experience, cognition and interests as groups that such a simplistic dichotomy is possible. What does this imply about the structure and functioning of power relations? The setting up of the commonality of third-world women's struggles across classes and cultures against a general notion of oppression (primarily the group in power - i.e., men) necessitates the assumption of something like what Michel Foucault calls the ‘juridicodiscursive’ model of power (1980: 134-45), the principal features of which are: ‘a negative relation’ (limit and lack); an ‘insistence on the rule’ (which forms a binary system); a ‘cycle of prohibition’; the ‘logic of censorship’; and a ‘uniformity’ of the apparatus functioning at different levels. Feminist discourse on the third world which assumes a homogeneous category - or group - called ‘women’ necessarily operates through such a setting up of originary power divisions. Power relations are structured in terms of a unilateral and undifferentiated source of power and a cumulative reaction to power. Opposition is a generalized phenomenon created as a response to power - which, in turn, is possessed by certain groups of people. The major problem with such a definition of power is that it locks all revolutionary struggles into binary structures - possessing power versus being powerless. Women are powerless, unified groups. If the struggle for a just society is seen in terms of the move from powerless to powerful for women as a group, and this is the implication in feminist discourse which structures sexual difference in terms of the division between the sexes, then the new society would be structurally identical to the existing organization of power relations, constituting itself as a simple inversion of what exists. If relations of domination and exploitation are defined in terms of binary divisions - groups which dominate and groups which are dominated - surely the implication is that the accession to power of women as a group is sufficient to dismantle the existing organization of relations? But women as a group are not in some sense essentially superior or infallible. The crux of the problem lies in that initial assumption of women as a homogeneous group or category (‘the oppressed’), a familiar assumption in western radical and liberal feminisms.’18 What happens when this assumption of ‘women as a oppressed group’ is situated in the context of western feminist writing about third-world women? It is here that I locate the colonialist move. By contrasting the representation of women in the third world with what I referred to earlier as western feminisms' self-presentation in the same context, we see how western feminists alone become the true ‘subjects’ of this counter-history. Third-world women, on the other hand, never rise above the debilitating generality of their ‘object’ status. While radical and liberal feminist assumptions of women as a sex class might elucidate (however inadequately) the autonomy of particular women's struggles in the west, the application of the notion of women as a homogeneous category to women in the third world colonizes and appropriates the pluralities of the simultaneous location of different groups of women in social class and ethnic frameworks; in doing so it ultimately robs them of their historical and political agency. Similarly, many Zed Press authors, who ground themselves in the basic analytic strategies of traditional Marxism, also implicitly create a ‘unity’ of women by substituting ‘women's activity’ for ‘labour’ as the primary theoretical determinant of women's situation. Here again, women are constituted as a coherent group not on the basis of ‘natural’ qualities or needs, but on the basis of the sociological ‘unity’ of their role in domestic production and wage labour.19 In other words, western feminist discourse, by assuming women as a coherent, already constituted group which is placed in kinship, legal and other structures, defines third-world women as subjects outside of social relations, instead of looking at the way women are constituted as women through these very structures. Legal, economic, religious and familial structures are treated as phenomena to be judged by western standards. It is here that ethnocentric universality comes into play. When these structures are defined as ‘underdeveloped’ or ‘developing’ and women are placed within these structures, an implicit image of the ‘average third-world woman’ is produced. This is the transformation of the (implicitly western) ‘oppressed woman’ into the ‘oppressed third-world woman’. While the category of ‘oppressed woman’ is generated through an exclusive focus on gender difference ‘the oppressed third-world woman’ category has an additional attribute - the ‘third-world difference’! The ‘third-world difference’ includes a paternalistic attitude towards women in the third world.20 Since discussions of the various themes I identified earlier (e.g., kinship, education, religion, etc.) are conducted in the context of the relative ‘underdevelopment’ of the third world (which is nothing less than unjustifiably confusing development with the separate path taken by the west in its development, as well as ignoring the unidirectionality of the first/third-world power relationship), third-world women as a group or category are automatically and necessarily defined as: religious (read ‘not progressive’), family oriented (read ‘traditional’), legal minors (read ‘they-are-still-not-conscious-of-their-rights’), illiterate (read ‘ignorant’), domestic (read ‘backward’) and sometimes revolutionary (read ‘their-country-is-in-a-state-of-war; they-must-fight!’). This is how the ‘third-world difference’ is produced. When the category of ‘sexually oppressed women’ is located within particular systems in the third world which are defined on a scale which is normed through Eurocentric assumptions, not only are third-world women defined in a particular way prior to their entry into social relations, but since no connections are made between first- and third-world power shifts, it reinforces the assumption that people in the third world just have not evolved to the extent that the west has. This mode of feminist analysis, by homogenizing and systematizing the experiences of different groups of women, erases all marginal and resistant modes of experiences.21 It is significant that none of the texts I reviewed in the Zed Press series focuses on lesbian politics or the politics of ethnic and religious marginal organizations in third-world women's groups. Resistance can thus only be defined as cumulatively reactive, not as something inherent in the operation of power. If power, as Michel Foucault has argued recently, can really be understood only in the context of resistance,22 this misconceptualization of power is both analytically as well as strategically problematical. It limits theoretical analysis as well as reinforcing western cultural imperialism. For in the context of a first/third-world balance of power, feminist analyses which perpetrate and sustain the hegemony of the idea of the superiority of the west produce a corresponding set of universal images of the ‘third-world woman’, images like the veiled woman, the powerful mother, the chaste virgin, the obedient wife, etc. These images exist in universal ahistorical splendour, setting in motion a colonialist discourse which exercises a very specific power in defining, coding and maintaining existing first/third-world connections. To conclude, then, let me suggest some disconcerting similarities between the typically authorizing signature of such western feminist writings on women in the third world, and the authorizing signature of the project of humanism in general - humanism as a western ideological and political project which involves the necessary recuperation of the ‘East’ and ‘Woman’ as Others. Many contemporary thinkers like Foucault, Derrida, Kristeva, Deleuze, and Said have written at length about the underlying anthropomorphism and ethnocentrism which constitutes a hegemonic humanistic problematic that repeatedly confirms and legitimates (western) Man's centrality.23 Feminist theorists like Luce Irigaray, Sarah Kofman, Helene Cixous, and others have also written about the recuperation and absence of woman/women within western humanism.24 The focus of the work of all these thinkers can be stated simply as an uncovering of the political interests that underlie the binary logic of humanistic discourse and ideology whereby, as a valuable recent essay puts it, ‘the first (majority) term (Identity, Universality, Culture, Disinterestedness, Truth, Sanity, Justice, etc.), which is, in fact, secondary and derivative (a construction), is privileged over and colonizes the second (minority) term (difference, temporality, anarchy, error, interestedness, insanity, deviance, etc.), which is in fact, primary and originative’ (Spanos, 1984). In other words, it is only in so far as ‘Woman/Women’ and ‘the East’ are defined as Others, or as peripheral, that (western) Man/Humanism can represent him/itself as the centre. It is not the centre that determines the periphery, but the periphery that, in its boundedness, determines the centre. Just as feminists like Kristeva, Cixous, Irigaray and others deconstruct the latent anthropomorphism in western discourse, I have suggested a parallel strategy in this article in uncovering a latent ethnocentrism in particular feminist writings on women in the third world.25 As discussed earlier, a comparison between western feminist self-presentation and western feminist re-presentation of women in the third world yields significant results. Universal images of ‘the third-world woman’ (the veiled woman, chaste virgin, etc.), images constructed from adding the ‘third-world difference’ to ‘sexual difference’, are predicated on (and hence obviously bring into sharper focus) assumptions about western women as secular, liberated, and having control over their own lives. This is not to suggest that western women are secular and liberated and have control over their own lives. I am referring to a discursive self-presentation, not necessarily to material reality. If this were a material reality there would be no need for feminist political struggle in the west. Similarly, only from the vantage point of the west is it possible to define the ‘third world’ as underdeveloped and economically dependent. Without the overdetermined discourse that creates the third world, there would be no (singular and privileged) first world. Without the ‘third-world woman’, the particular self-presentation of western women mentioned above would be problematical. I am suggesting, in effect, that the one enables and sustains the other. This is not to say that the signature of western feminist writings on the third world has the same authority as the project of western humanism. However, in the context of the hegemony of the western scholarly establishment in the production and dissemination of texts, and in the contest of the legitimating imperative of humanistic and scientific discourse, the definition of ‘the third-world woman’ as a monolith might well tie into the larger economic and ideological praxis of ‘disinterested’ scientific inquiry and pluralism which are the surface manifestations of a latent economic and cultural colonization of the ‘non-western’ world. It is time to move beyond the ideological framework in which even Marx found it possible to say: They cannot represent themselves; they must be represented. This is a revised version of an essay originally published in Boundary 2, Vol. XII, No. 3/Vol. XIII, No. 1 (Spring/Fall 1984). Acknowledgements This paper would not have been possible without S. P. Mohanty's challenging and careful critical reading. I would also like to thank Biddy Martin for our numerous discussions about feminist theory and politics. They both helped me think through and sharpen some of the arguments in this paper. Notes 1 See especially the essays in Moraga and Anzaldua (1983); Smith (1983); Joseph and Lewis (1981) and Moraga (1984). 2 Terms like ‘third’ and ‘first’ world are very problematical both in suggesting over-simplified similarities between and amongst countries labelled ‘third’ or ‘first’ world, as well as implicitly reinforcing existing economic, cultural, and ideological hierarchies. I use the term ‘third world’ with full awareness of its problems, only because this is the terminology available to us at the moment. The use of quotation marks is meant to suggest a continuous questioning of the designation ‘third world’. Even when I do not use quotation marks, I mean to use the term critically. 3 I am indebted to Teresa de Lauretis for this particular formulation of the project of feminist theorizing. See especially her introduction to de Lauretis (1984); see also Sylvia Wynter, ‘The Politics of Domination’, unpublished manuscript. 4 This argument is similar to Homi Bhabha's (1983) definition of colonial discourse as strategically creating a space for a subject peoples through the production of knowledges and the exercise of power. The full quote reads: ‘[colonial discourse is] an apparatus of power... an apparatus that turns on the recognition and disavowal of racial/cultural/historical differences. Its predominant strategic function is the creation of a space for a “subject peoples” through the production of knowledges in terms of which surveillance is exercised and a complex form of pleasure/unpleasure is incited. It [i.e. colonial discourse] seeks authorization for its strategies by the production of knowledges by colonizer and colonized which are stereotypical but antithetically evaluated.’ 5 A number of documents and reports on the UN International Conferences on Women, Mexico City 1975, and Copenhagen 1980, as well as the 1976 Wellesley Conference on Women and Development attest to this. Nawal el Saadawi, Fatima Mernissi and Mallica Vajarathon in ‘A Critical Look At The Wellesley Conference’ (Quest, IV:2, Winter 1978, pp. 101-7), characterize this conference as ‘American-planned and organized’, situating third world participants as passive audiences. They focus especially on the lack of self-consciousness of western women's implication in the effects of imperialism and racism in their assumption of an ‘international sisterhood’. Amos and Parmar (1984) characterize Euro-American feminism which seeks to establish itself as the only legitimate feminism as ‘imperial’. 6 The Zed Press ‘Women in the Third World’ series is unique in its conception. I choose to focus on it because it is the only contemporary series of books I have found which assumes that ‘women in the Third World’ is a legitimate and separate subject of study and research. Since 1985, when this essay was first written, numerous new titles have appeared in the Zed ‘Women in the Third World’ series. Thus, I suspect that Zed has come to occupy a rather privileged position in the dissemination and construction of discourses by and about third-world women. A number of the books in this series are excellent, especially those which deal directly with women's resistance struggles. In addition, Zed Press consistently publishes progressive, feminist, anti-racist and anti-imperialist texts. However, a number of texts written by feminist sociologists, anthropologists, and journalists are symptomatic of the kind of western feminist work on women in the third world that concerns me. Thus, an analysis of a few of these particular texts in this series can serve as a representative point of entry into the discourse I am attempting to locate and define. My focus on these texts is therefore an attempt at an internal critique: I simply expect and demand more from this series. Needless to say, progressive publishing houses also carry their own authorizing signatures. 7 Elsewhere I have discussed this particular point in detail in a critique of Robin Morgan's construction of ‘women's herstory’ in her introduction to Sisterhood is Global: The International Women's Movement Anthology (1984) (see Mohanty) ‘Feminist Encounters’ pp. 30-44, especially pp. 35-7). 8 My analysis in this section of the paper has been influenced by Felicity Eldhom, Olivia Harris and Kate Young's excellent discussions (Eldhom, Harris and Young, 1977). They examine the use of the concepts of ‘reproduction’ and the ‘sexual division of labour’ in anthropological work on women, suggesting the inevitable pull towards universals inherent in the use of these categories to determine ‘women's position’. 9 Another example of this kind of analysis is Mary Daly's Gyn/Ecology. Daly's assumption in this text, that women as a group are sexually victimized, leads to her very problematic comparison between the attitudes towards women witches and healers in the west, Chinese footbinding, and the genital mutilation of women in Africa. According to Daly, women in Europe, China, and Africa constitute a homogeneous group as victims of male power. Not only does this label (sexual victims) eradicate the specific historical realities which lead to and perpetuate practices like witch-hunting and genital mutilation, but it also obliterates the differences, complexities and heterogeneities of the lives of, for example, women of different classes, religions and nations in Africa. As Audre Lorde pointed out, women in Africa share a long tradition of healers and goddesses that perhaps binds them together more appropriately than their victim status. However, both Daly and Lorde fall prey to universalistic assumptions about ‘African women’ (both negative and positive). What matters is the complex, historical range of power differences, commonalities and resistances that exist among women in Africa which construct African women as ‘subjects’ of their own politics. See Daly (1978: 107-312) Lorde in Moraga and Anzaldua (1983). 10 See Eldhom, Harris and Young (1977) for a good discussion of the necessity to theorize male violence within specific societal frameworks, rather than to assume it as a universal fact. 11 These views can also be found in differing degrees in collections like: Wellesley Editorial Committee, ed., Women and National Development: The Complexities of Change Chicago: University of Chicago Press 1977, and Signs, Special Issue, ‘Development and the Sexual Division of Labor’, 7.2, (Winter 1981). For an excellent introduction to WID issues see ISIS, Women in Development: A Resource Guide for Organization and Action Philadelphia: New Society Publishers, 1984. For a politically focused discussion of feminism and development and the stakes for poor third-world women, see Sen and Grown, (1987). 12 See essays by Vanessa Maher, Diane Elson and Ruth Pearson, and Maila Stevens in Young, Walkowitz and McCullagh (1981); and essays by Vivian Mota and Michelle Mattelart in Nash and Safa (1980). For examples of excellent self-conscious work by feminists writing about women in their own historical and geographical locations, see Lazreg (1988) on Algerian women; Gayatri Chakravorty Spivak's ‘A Literary Representation of the Subaltern: A Woman's Text from the Third World’, in Spivak (1987), and Lata Mani's essay, ‘Contentious Traditions: The Debate on SATI in Colonial India’, Cultural Critique No. 7, Fall 1987, pp. 119-56. 13 See Tabari (1980) for a detailed discussion of these instances. 14 Olivia Harris in Harris (1983: 4-7). Other MRG reports include Deardon (1975) and Jahan (1980). 15 See Eldhom, Harris and Young (1977) for an excellent discussion of this. 16 See Kishwar and Vanita (1984) for a discussion of this aspect of Indian women's struggles. 17 List of Zed Press Publications: Patricia Jeffery, Frogs in a Well: Indian Women in Purdah, 1979; Latin American and Caribbean Women's Collective, Slaves of Slaves: The Challenge of Latin American Women, 1980; Gale Omvedt, We Shall Smash this Prison: Indian Women in Struggle, 1980; Juliette Minces, The House of Obedience: Women in Arab Society, 1980; Bobby Siu, Women of China: Imperialism and Women's Resistance 1900-1949, 1981; Ingela Bendt and James Downing, We Shall Return: Women of Palestine, 1982; Maria Rosa Cutrufelli, Women of Africa: Roots of Oppression, 1983; Maria Mies, The Lace Makers of Narsapur: Indian Housewives Produce for the World Market, 1983; Miranda Davis, ed., Third World/Second Sex: Women's Struggles and National Liberation, 1983. 18 For succinct discussion of western radical and liberal feminisms, see Eisenstein (1983) and Eisenstein (1981). 19 See Haraway (1985: 65-108, especially 76). 20 Amos and Parmar (1984: 9) describe the cultural stereotypes present in Euro-American feminist thought: ‘The image is of the passive Asian woman subject to oppressive practices within the Asian family, with an emphasis on wanting to “help” Asian women liberate themselves from their role. Or there is the strong, dominant Afro-Caribbean woman, who despite her “strength” is exploited by the “sexism” which is seen as being a strong feature in relationships between Afro-Caribbean men and women’. These images illustrate the extent to which paternalism is an essential element of feminist thinking which incorporates the above stereotypes, a paternalism which can lead to the definition of priorities for women of colour by Euro-American feminists. 21 I discuss the question of theorizing experience in my ‘Feminist Encounters’ (1987), and in an essay co-authored with Biddy Martin in de Lauretis (1986). 22 This is one of Foucault's central points in his reconceptualization of the strategies and workings of power networks. See Foucault (1978 and 1980). 23 Foucault (1978 and 1980); Derrida (1974); Kristeva (1980); Said (1978), and Deleuze and Guattari (1977). 24 Irigaray (1981); Cixous (1981). For a good discussion of Sarah Kofman's work, see Berg (1982: 11-20). 25 For an argument which demands a new conception of humanism in work on third-world women, see Lazreg (1988). While Lazreg's position might appear to be diametrically opposed to mine, I see it as a provocative and potentially positive extension of some of the implications that follow from my arguments. In criticizing the feminist rejection of humanism in the name of ‘essential Man’, Lazreg points to what she calls an ‘essentialism of difference’ within these very feminist projects. She asks: ‘To what extent can western feminism dispense with an ethics of responsibility when writing about “different” women? The point is neither to subsume other women under one's own experience nor to uphold a separate truth for them. Rather, it is to allow them to be while recognizing that what they are is just as meaningful, valid, and comprehensible as what “we” are . . Indeed, when feminists essentially deny other women the humanity they claim for themselves, they dispense with any ethical constraint. They engage in the act of splitting the social universe into “us” and “them,” “subjects” and “objects”’(99-100). This essay by Lazreg and an essay by S. P. Mohanty entitled ‘Us and Them: On the Philosophical Bases of Political Criticism’, forthcoming in The Yale Journal of Criticism in March 1989 (Vol. 2, No. 2), suggest positive directions for self-conscious cross-cultural analyses, analyses which move beyond the deconstructive to a fundamentally productive mode in designating overlapping areas for cross-cultural comparison. The latter essay calls not for a ‘humanism’ but for a reconsideration of the question of the ‘human’ in a post-humanist context. It argues that (1) there is no necessary ‘incompatibility between the deconstruction of western humanism’ and such ‘a positive elaboration’ of the human; and moreover that (2) such an elaboration is essential if contemporary political-critical discourse is to avoid the incoherences and weaknesses of a relativist position. References ABDEL-MALEK, Anouar (1981) Social Dialectics: Nation and Revolution Albany: State University of New York Press. AMIN, Samir (1977) Imperialism and Unequal Development New York: Monthly Review Press. AMOS, Valerie and PARMAR, Pratibha (1984) ‘Challenging Imperial Feminism’, Feminist Review No. 17. BARAN, Paul A. (1962) The Political Economy of Growth New York: Monthly Review Press. BERG, Elizabeth (1982) The Third Woman’, Diacritics, Summer. BHABHA, Homi (1983) ‘The Other Question - The Stereotype and Colonial Discourse’ Screen 24:6, p. 23. BOSERUP, Ester (1970) Women's Role in Economic Development New York: St Martin's Press; London: Allen & Unwin. BROWN, Beverly (1983) ‘Displacing the Difference - Review, Nature, Culture and Gender’, m/f No. 8. CIXOUS, Helene (1981) ‘The Laugh of the Medusa’ in MARKS and DE COURTIVRON (1981). COWIE, Elizabeth (1978) ‘Woman as Sign’, m/f No. 1. CUTRUFELLI, Maria Rosa (1983) Women of Africa: Roots of Oppression London: Zed Press. DALY, Mary (1978) Gyn/Ecology: The Metaethics of Radical Feminism Boston: Beacon Press. DE LAURETIS, Teresa (1984) Alice Doesn't: Feminism, Semiotics, Cinema Bloomington: Indiana University Press. DE LAURETIS, Teresa (1986) editor Feminist Studies/Critical Studies Bloomington: Indiana University Press. DEARDON, Ann (1975) editor Arab Women London: Minority Rights Group Report No. 27. DELEUZE, Giles and GUATTARI, Felix (1977) Anti-Oedipus: Capitalism and Schizophrenia New York: Viking. DERRIDA, Jacques (1974) Of Grammatology Baltimore: Johns Hopkins University Press. EISENSTEIN, Hester (1983) Contemporary Feminist Thought Boston: G. K. Hall & Co. EISENSTEIN, Zillah (1981) The Radical Future of Liberal Feminism New York: Longman. ELDHOM, Felicity, HARRIS, Olivia and YOUNG, Kate (1977) ‘Conceptualising Women’, Critique of Anthropology ‘Women's Issue’ No. 3. FOUCAULT, Michel (1978) History of Sexuality Volume One New York: Random House. FOUCAULT, Michel (1980) Power/Knowledge New York: Pantheon. GUNDER-FRANK, Andre (1967) Capitalism and Underdevelopment in Latin America New York: Monthly Review Press. HARAWAY, Donna (1985) ‘A Manifesto for Cyborgs: Science, Technology and Socialist Feminism in the 1980s’, Socialist Review No. 80. HARRIS, Olivia (1983a) ‘Latin American Women-An Overview’ in HARRIS (1983) HARRIS, Olivia (1983b) editor Latin American Women London: Minority Rights Group Report No. 57. HOSKEN, Fran (1981) ‘Female Genital Mutilation and Human Rights’, Feminist Issues 1:3. HUSTON, Perdita (1979) Third World Women Speak Out New York: Praeger. IRIGARAY, Luce (1981) This Sex Which Is Not One’ and ‘When the Goods Get Together’ in MARKS and DE COURTIVRON (1981). JAHAN, Rounaq (1980) editor Women in Asia London: Minority Rights Group Report No. 45. JEFFERY, Patricia (1979) Frogs in a Well: Indian Women in Purdah London: Zed Press. JOSEPH, Gloria and LEWIS, Jill (1981) Common Differences: Conflicts in Black and White Feminist Perspectives Boston: Beacon Press. KISHWAR, Madhu and VANITA, Ruth (1984) In Search of Answers: Indian Women's Voices from Manushi London: Zed Press. KRISTEVA, Julia (1980) Desire in Language New York: Columbia University Press. LAZREG, Marnia (1988) ‘Feminism and Difference: The Perils of Writing as a Woman on Women in Algeria’, Feminist Issues 14:1. LINDSAY, Beverley (1983) editor Comparative Perspectives of Third World Women: The Impact of Race, Sex and Class New York: Praeger. LORDE, Audre (1983) ‘An Open Letter to Mary Daly’ in MORAGA and ANZALDUA (1983). MARKS, Elaine and DE COURTIVRON, Isobel (1981) editors New French Feminisms New York: Schoken Books. MIES, Maria (1982) The Lace Makers of Narsapur: Indian Housewives Produce for the World Market London: Zed Press. MINCES, Juliette (1980) The House of Obedience: Women in Arab Society London: Zed Press. MODARES, Mina (1981) ‘Women and Shi'ism in Iran’, m/f Nos. 5 and 6. MOHANTY, Chandra and MARTIN, Biddy (1986) ‘Feminist Politics: What's Home Got to Do With It?’ in DE LAURETIS (1986). MOHANTY, Chandra (1987) ‘Feminist Encounters: Locating the Politics of Experience’, Copyright 1, ‘Fin de Siecle 2000’. MORAGA, Cherrie and ANZALDUA, Gloria (1983) editors This Bridge Called My Back: Writings by Radical Women of Color New York: Kitchen Table Press. MORAGA, Cherrie (1984) Loving in the War Years Boston: South End Press. MORGAN, Robin (1984) editor Sisterhood is Global: The International Women's Movement Anthology New York: Anchor Press/Doubleday; Harmondsworth: Penguin. NASH, June and SAFA, Helen I. (1980) editors Sex and Class in Latin America: Women's Perspectives on Politics, Economics and the Family in the Third World Massachusetts: Bergin & Garvey. ROZALDO, M. Z. (1980) ‘The Use and Abuse of Anthropology: Reflections on Feminism and Cross-Cultural Understanding’, Signs 5:3. SAID, Edward (1978) Orientalism New York: Random House. SEN, Sita and GROWN, Caren (1987) Development Crises and Alternative Visions: Third World Women's Perspectives New York: Monthly Review Press. SMITH, Barbara (1983) editor Home Girls: A Black Feminist Anthology New York: Kitchen Table Press. SPANOS, William V. (1984) ‘Boundary 2 and the Polity of Interest: Humanism, the “Center Elsewhere”, and Power’, Boundary 2, Vol. XII, No. 3/Vol. XIII, No. 1, Spring/Fall. SPIVAK, Gayatri Chakravorty (1987) In Other Worlds: Essays in Cultural Politics London and New York: Methuen. STRATHERN, Marilyn and McCORMACK, Carol (1980) editors Nature, Culture and Gender Cambridge: Cambridge University Press. TABARI, Azar (1980) ‘The Enigma of the Veiled Iranian Women’, Feminist Review No. 5. TINKER, Irene and BRAMSEN, Michelle Bo (1972) editors Women and World Development Washington DC: Overseas Development Council. YOUNG, Kate, WALKOWITZ, Carol and McCULLAGH, Roslyn (1981) editors Of Marriage and the Market: Women's Subordination in International Perspective London: CSE Books.