(1991, xiv). The challenge for feminists, she argues, is to disrupt cultural representations of an incommensurable sexual difference which organize "the world 'as if' women stood in a derivative (but also opposed) relation to 'man'" (1991, xiv), and, we would add, "as if" that relation were not only immutable but also more socially significant than the relations of race and class.
The instability of the category "women" and of naturalized sexual difference, as demonstrated in the works of feminist theorists, has important implications not only for political theorists, but also for more empirically oriented political scientists. Although these implications will become clearer as the debate over the category "women" continues to play itself out over the next several years, we can put forth a few preliminary observations from the vantage point of the early 1990s.
The work of feminist theorists suggests that researchers in the 1990s should be more cautious and self-reflective about their use of the category "women" than they have been in the past. Most empirical work focusing on women or gender (or even including sex as a control variable) has assumed that the category "women" has an inherent political meaning, that there is something politically relevant, most commonly "interests" (see Sapiro 1981 for a discussion of issues involved in defining women's interests), that all women share. Most empirical researchers believe that by controlling statistically for variables such as race, age, income, and education, they can isolate and measure the effects due to gender. However, applying statistical controls to isolate the effects of gender is not sufficient to deal with the problem of the category "women" as it is posed by feminist theorists. Rather, feminists who have interrogated the category "women" question whether such gender effects, free from the influence of other confounding variables, exist at all -- whether there is any commonality, any essence, or even any interests that all women share after variations in their race, age, education, income, and the like are taken into consideration. Instead, they suggest that race, class, and gender, for example, are so intertwined that all work together to shape identity (or, in this case, political identity). Women exist in an historical and cultural context, and to erase or ignore this context is to gloss over important differences among women.
What does this actually mean for those who do empirical research? First, it means that we should be less concerned with comparing women with men and more concerned with examining how different subgroups of women in different contexts behave politically. If we take the critiques of the category "women" seriously, our research is likely to become more contextual and more historical. It will make visible those women (e.g., women of color, poor women) whose experiences are often erased in the pursuit of scientific generalizations that purportedly apply to all women. Taking seriously critiques of the category "women" will likely move us in the direction of reducing our knowledge claims and our pursuit of universal truths. It will make us less likely to generalize across women (or men) of differing classes, races, cultures, nationalities, ethnicities, or generations. However, what we sacrifice in our ability to generalize (which those who theorize about the category of women would suggest has led only to false or misleading generalizations anyway) will be more than made up for in the depth and richness of our analysis and our understanding.
While empirical researchers who are sensitive to critiques of the category "women" may well become more reflective and self-conscious in their use of the term and more contextual and culturally specific in their approach to research, there are, as we noted earlier, good reasons for researchers not to abandon fully the category that has enabled their work. By using the category "women," feminist political scientists have been able to call into question some of the central assumptions and frameworks of the discipline. The concerted focus by feminists on women and women's experience has helped us as a discipline to see the biases and the blinders that characterized pre-feminist work and to improve our knowledge base by correcting for these biases and removing the blinders. In the same way, the current feminist interrogation of the category "women" may lead to research that will expand and improve our disciplinary knowledge base through greater historical and cultural specificity and more attention to heretofore neglected segments of the population. A major task for women and politics scholars in the 1990s will be to work through the many questions surrounding the category "women," perhaps finding some middle ground between uncritical acceptance and total abandonment of the category. In doing so, feminist scholars are likely to continue to pose new and important challenges to the discipline of political science.
We would like to thank Ada Finifter, Virginia Sapiro, Barbara Crow, Kathleen Casey, and an anonymous reviewer for their careful readings and constructive suggestions regarding this essay.