University of Michigan, Ann Arbor
Scrutinizing Feminist Epistemology, edited by Cassandra Pinnick, Noretta Koertge, and Robert Almeder (2003), (henceforth, SFE) offers a systematic critique of feminist epistemology. It aims to show that the entire enterprise is a failure. How should a critique that purports to decisively undermine an entire field of inquiry be evaluated? I propose the following standards. First, accuracy. The critique must accurately represent the field as it stands today, paying close attention to what its actual proponents say, in context. It should not attack straw men or beat dead horses killed long ago in debates internal to the field. Second, perspective. To be illuminating, a critique should not simply deploy its presuppositions against its rivals, but make them explicit, situate them relative its rivals in a field of possibilities, explain why these are the possibilities, and why its presuppositions should be accepted rather than rivals. Third, normative consistency. The critique should live up to the same normative standards it applies to its object. Let us assess Scrutinizing Feminist Epistemology by these standards.
SFE levels four major charges against what it takes feminist epistemology to stand for. First, political correctness: feminist epistemologists hold that political criteria should be deployed to pre-empt or override scientific reasoning based on evidence, in order to suppress truths that are inconvenient to the feminist agenda, and promulgate falsehoods congenial to it. Second, tribalism: feminist epistemologists think that all women (or all feminists) do, or should, think alike. In particular, they do or ought to adopt some common “feminine” epistemic style or methodology, applicable to all fields of inquiry, held to be superior, both in generating more objective knowledge and in serving women's interests Third, self-defeating conservatism: feminist epistemology defeats its own aims in taking nonwestern, women's, or “feminine” values as an uncriticized given, even when these values underwrite sex and caste oppression. Fourth, cynicism: feminist epistemologists reject the quest for objectivity and truth as an impossibility, and regard the claim to pursue it is a mask for a power play that in practice serves the interests of white heterosexual Western men at everyone else's expense. I shall argue that all of these criticisms are based on gross misrepresentations of feminist epistemology.
(a) Political Correctness
Susan Haack, Noretta Koertge, and Robert Almeder accuse feminist epistemology of political correctness. Haack says the aim of feminist epistemology is to legitimate the idea that “feminist values should determine what theories are accepted” (12). Such politicized inquiry leads to “sham reasoners seeking only to make a case for some foregone conclusion” (15), and threatens “honest inquiry,” which she characterizes as “research not informed by political ideas at all” (16). Koertge says feminist epistemologists want “to place ideological constraints on the content of science,” (230) raising the specter of Lysenkoism (229). Almeder says that they aim “to dogmatize and indoctrinate by eliminating open discussion,” (190) and by replacing “traditional canons of evidence and argument” with the test of “whether [an idea] furthers the political interest of the oppressed” (193). Political correctness is the most serious charge against feminist epistemologists, since it accuses them not merely of error or cynicism, but of intellectual dishonesty, dogmatism, and even tyranny. What is the evidence for it?
Haack cites Helen Longino’s underdetermination argument for permitting values to influence theory choice. Longino observes that the evidence by itself underdetermines hypothesis choice; it supports a given hypothesis only in conjunction with background assumptions. In practice, at the cutting edge of research, incompatible background assumptions are available for interpreting the evidence. Since various background assumptions could be legitimately selected for any reason, no logical or methodological principles prevent scientists from choosing some on account of their congruence with their moral or political values. Longino concludes that feminists may select their background assumptions on account of their congruence with feminist values. In so selecting her background assumptions, the feminist scientist “admits political considerations as relevant constraints on reasoning” (Longino 1990, 193). This is the key passage Koertge cites to support her claim that Longino “advocate[s] curtailing the hypothesis space” by political values in the context of justification--i.e., at the point at which the truth of hypotheses is being determined (227).
Haack and Koertge suggest several worrying possibilities that Longino’s defense of feminist science might raise. First, a hypothesis might be accepted, even if the evidence suggests it is false, because it serves some political interest to believe it. Similarly, a hypothesis might be rejected, even if the evidence suggests it is true, because it serves some political interest to disbelieve it. This is the worry suggested by Haack’s analogy of feminist science with “sham reasoners” (15), and Koertge’s suggestion that feminist epistemologists might want to “make an exception” in favor of “hypotheses without an empirical basis” when they are “deemed . . . politically progressive” (231). Second, political constraints on the investigation of acceptable hypotheses might be imposed on the scientific community as a whole, thereby undermining their intellectual freedom, as well as preventing some scientists from discovering certain truths (16, 229). Third, science would be hobbled by rancorous political debates over which political values should constrain acceptable hypotheses (229-30). Fourth, the epistemic authority of science would be undermined, as people came to see that scientists were reaching conclusions on the basis of political considerations rather than the evidence (230). Finally, in cases where the evidence is not sufficient to decide between rival hypotheses, Longino’s argument would let us “choose to believe whatever theory suits our political purposes,” when the correct stance to take in this case is a suspension of judgment (12).
Does Longino’s defense of feminist science really lead in these worrisome directions? Interpretations of others’ views should be tested for consistency with what they say directly about the issues in question, and with their other core ideas. In Longino’s case, the other core ideas are her account of the standards of evaluation for scientific theories, and her account of objectivity. With respect to standards of evaluation, Longino is an empiricist. While different scientists may stress different values (e.g., reductionism vs. ontological heterogeneity), one standard is mandatory for all scientists: empirical adequacy--“that is, truth of the observationally determinable portion of theories or models” (Longino 2001, 185–6; See also Longino 1993, 261–3; Longino 1994, 476). With respect to objectivity, Longino argues that an epistemic community's conclusions are objective, or count as knowledge, insofar as they are the product of “effective critical interactions.” Such interactions “transform the subjective into the objective, not by canonizing one subjectivity over others, but by assuring that what is ratified as knowledge has survived criticism from multiple points of view” (emphasis mine) (Longino 2001, 129). To assure that critical interactions are effective, an epistemic community must (1) establish public forums for criticism, (2) change its beliefs and theories in response to criticism, (3) in accordance with shared public standards, and (4) recognize a “tempered” equality of intellectual authority among inquirers, which allows taking their possession of cognitive virtues or vices, but disallows taking someone's social position or power, as grounds for taking someone seriously or discounting their views (Longino 2001, 128–135; Longino 1990, 76–81).
These well known aspects of Longino’s epistemology plainly rule out the worries Haack and Koertge express about the implications of Longino’s argument. The first worry, that a community might accept empirically disconfirmed hypotheses for political reasons, is ruled out by Longino’s requirement that theories be empirically adequate. Haack and Koertge might reply that the disconfirming evidence might never be gathered, or superior hypotheses never get the chance to be developed, because the community rules out their investigation from the start, for political reasons. But this second worry is ruled out by the objectivity requirements. Epistemic communities must hold themselves open to criticism. No one is entitled to dictate theory choice to others, because all enjoy tempered equality of intellectual authority. “Individuals have doxastic autonomy” (Longino 2001, 154). “A diversity of perspectives is necessary for vigorous and epistemically effective critical discourse” (Longino 2001, 131). Even among the subset of inquirers doing science as feminists, no particular articulation of feminist values is entitled to dominance over the rest. “Different feminist perspectives may be represented in theorizing” (Longino 1990, 194). This rules out the third worry, that the progress of science will be impaired by endless squabbling about what the politically correct values are. Since no one has the authority to demand that others adopt their values, each is free to choose background assumptions in accordance with their own personal values.
What about the fourth worry, that the epistemic authority of science will be undermined once people become aware that theory choice is being influenced by political considerations? Since all theories must meet the test of empirical adequacy, this worry could arise only in the case of “ties,” where rival research programs can each claim empirical successes and neither is refuted by the evidence. Longino clearly states that in such cases, "as long as both frameworks offer coherent and comprehensive accounts of the relevant data, neither can displace the other" (Longino 1990, 130). In other words, the scientific community should not consolidate a position around one or the other view. Individual scientists should remain free to further develop either theory. Longino therefore does not disagree with Haack about the proper epistemic attitude the community should take in cases of empirical “ties.” So the fifth worry is also baseless. And this provides an answer to the fourth worry. Although individual scientists may continue to pursue one or another theory because it is more congruent with their political goals, no particular theory will be incorporated into a scientific consensus, to the exclusion of its rivals, until it uniquely survives criticism from all points of view, including those grounded in opposed political perspectives. Longino specifically denies that her account of science can "grant to some form of feminism or to any other social or political program an exclusive grant to truth" (Longino 1993, 270).
What, then, does Longino mean when she says that the feminist scientist “admits political considerations as relevant constraints on reasoning”? Doesn't this mean, as Koertge thinks, that Longino “advocate[s] curtailing the hypothesis space” by political values in the context of justification? And what could it mean when Longino says we may choose background assumptions, and hence theories, in light of our values? Janet Kourany, the only contributor to SFE sympathetic to feminist epistemology, gives us a clue (210). Before scientists begin to gather evidence, they must make numerous critical choices, not only about the part of nature they will investigate, but about the questions they want answered, the terms in which they will describe that part of nature, their measuring tools and procedures for eliciting data, and so forth. They make these choices in light of their goals--some of which, Longino observes, may be political. These choices inherently constrain, in advance, the hypotheses that can be investigated. Yet such constraints are unavoidable preconditions for getting research underway. Of course, other scientists make other choices, and so are subject to different constraints. There is no question of imposing a single set of constraints on the entire body of researchers, since this would undermine the conditions of objectivity, which require pluralism. Values enter at this stage of research--what is conventionally called “the context of discovery”--not the context of justification, insofar as the latter is understood narrowly, as the point at which the truth or warrant of hypotheses is determined. Even hard-core logical positivists allowed values to influence theory “choice” at this stage--i.e., the choice of theories to develop and investigate, rather than to consolidate into the body of accepted scientific findings or shared beliefs. Koertge allows this, too. Longino’s theory therefore does not violate Koertge’s constraints on legitimate inquiry.
Matters may seem otherwise, because feminist epistemologists work with a richer conception of both the context of discovery and the context of justification than many other philosophers of science do. In standard accounts of the discovery/justification distinction, the only issues discussed in the context of discovery are the selection of topics to investigate and the causal origins of hypotheses. Then the account jumps to the context of justification, where the data have already been gathered, and the main issue is the bearing of evidence on the truth of hypotheses. Feminist epistemologists frequently focus on what's missing from the context of discovery in the standard accounts--the innumerable, frequently value-laden choices concerning research design, methodology, and model construction that constrain the subsequent possibilities of discovery (Anderson 1995b; Anderson 2004; Harding 1993, 56–7, 70; Longino 1990, 83–102). Feminist epistemologists have also sometimes used the term “context of justification” to refer to more than the process of determining the truth or warrant of theories. Theories are evaluated with respect to all of the goals for which they were constructed, including their utility in facilitating successful practice, and their relevance for answering particular, often value-laden, questions (Anderson 1995a; Anderson 1995b). If we construe the context of justification narrowly, to the consideration of truth or warrant alone, then feminist empiricists have strictly adhered to the orthodox view that the political desirability of reaching certain conclusions cannot determine their truth or warranted assertability--i.e., their entitlement to be incorporated into the body of accepted scientific claims.
This is not news. Feminist empiricists and their close allies have painstakingly explained these points many times before, often in direct response to critics of feminist epistemology who have accused them of “political correctness” (Anderson 1995b; Lacey 1999; Lloyd 1995; Lloyd 1997; Tiles 1987). The critics of feminist epistemology, including those represented in SFE, have shown zero uptake of these careful replies to their criticisms. They simply recycle the same accusations over and over again, on the basis of innuendo, superficial misreadings of texts, and quotations out of context. Haack’s first article in SFE, for example, is reprinted unrevised, although I had replied to it in detail years ago (Anderson 1995b). As I have shown in the case of Longino’s theory, which provides the central point of departure for feminist empiricist discussions of values in science, the critics’ accusations of political correctness cannot withstand even the most basic tests of credibility. To read them again in SFE reminds me of how evolutionary theorists must feel when they encounter the latest creationist assault: they are trotting out that old, long-refuted objection from the Second Law of Thermodynamics again? How many times must one refute such accusations, before one is entitled to ignore them, on the ground that the accusers have failed to live up to the terms of reasoned debate?
This question arises with most force with respect to Almeder’s charges of political correctness. Unlike Haack and Koertge, Almeder does not cite any argument of any feminist epistemologist to substantiate his charge. Instead, he draws up a bill of indictment concerning the practice of feminist academics in general and challenges them to prove their innocence. Men, he says, “are regularly prohibited. . .from teaching in women's studies programs” because, “being men, their thinking is infected with the bias of the male oppressor” (191). Women's studies courses, he alleges, will not assign the work of critics of feminist epistemology such as Susan Haack, or people he calls “conservative equity feminists.” And so on (190-1). The terms of Almeder’s challenge are unacceptable. He has set himself up as prosecutor, judge, and jury, exempting himself from any obligation to produce evidence for his charges, since by his rules, feminists are guilty until proven innocent. This is not academic discourse. It is an inquisition. I shall discuss the implications of Almeder’s conduct below. For now it is enough to observe that, having removed himself from academic discourse, he has not presented his challenges in a form entitled to normal academic response.
Susan Haack, Noretta Koertge, Robert Klee, and Cassandra Pinnick accuse feminist epistemologists of tribalism--insisting that all women, or feminists, do or ought to think alike, in conformity with sexist stereotypes about feminine cognition. Koertge, for example, says that feminist philosophy of science was “strongly influenced by women's studies” which “greatly emphasized models of ‘women's ways of knowing’ and ‘tended to affirm various stereotypes about female mentality and ‘valorize’ them” (47). Supposedly, feminist epistemologists claim that thinking in stereotypically feminine ways does or will make people both better scientists and better able to advance the interests of the oppressed. The contributors to SFE voice several objections to this position. First, Haack complains that feminist epistemologists fail “to appreciate each woman's individuality” when they echo “old, sexist stereotypes” of feminine mentality and encourage women academics to follow them (250). Second, Klee suggests that feminist epistemologists claim that “increasing the amount of feminism-informed inquiry throughout all of science must . . . lead to greater accuracy in any and all domains of inquiry” (37). He argues that such global claims about the scope of feminism-informed inquiry are indefensible. At best, such inquiry could be expected to improve theories with sociopolitical content, not theories in the natural sciences. Third, Pinnick argues that Sandra Harding lacks data to support her principle contention of standpoint epistemology, which Pinnick claims is that “feminists, as marginalized persons or as a marginalized social group, do science better than nonmarginalized persons” (24).
In contrast with the charges of political correctness, here one can find feminist texts that advance tribalist theories (and even more that offhandedly invoke untheorized tribalist claims). Carol Gilligan (1982), for example, is famous for arguing that women have a special “feminine” mode of moral reasoning, based on considerations of care rather than justice. (However, few have noticed that she thinks this mode of moral reasoning, like men’s “justice” mode, is immature, or that she advances a common ideal of moral reasoning for men and women alike). Nancy Hartsock (1983), in her original statement of feminist standpoint epistemology, argued that “feminine” cognitive styles, in their affirmation of relationality and life-affirming goals, and repudiation of dichotomous, oppositional thinking and masculine focus on death and domination, provide a superior standpoint from which to envision possibilities for overcoming oppression and building a better society. (However, as with Gilligan, Hartsock made no claim that such styles were globally superior for all domains, including physics and chemistry).
It is telling that the most unequivocal statements of tribalism in feminist epistemology date from its infancy more than 20 years ago. In failing to trace criticisms of such tribalist claims by other feminist epistemologists, and replies by advocates of standpoint epistemology, SFE paints a misleading picture of the current state of play of tribalist ideas in women's studies. Since SFE appears to be written by and addressed to readers ignorant of this history, it is worth recounting some of it here. Three debates internal to women's studies can put SFE’s tribalist picture into perspective: the equality/difference debates, the feminist methodology debates, and the debates over feminist standpoint epistemology.
Contrary to Koertge’s view that women's studies was dominated by advocates of a special female difference, from the start this view was hotly contested by feminist advocates of women's equality. Equality feminists argued that the empirical basis for asserting feminine differences in cognition was weak (Fausto-Sterling 1985; Tavris 1992). Moreover, the valorization of stereotypical feminine traits in practice encouraged women to remain in the confining settings in which these traits found their most congenial context (Tong 1993). This appears to be the view of the advocates of SFE, especially Haack (250). However, the claim of cognitive “equality” or “sameness” with men was also found to be defective, in uncritically accepting the androcentric valorization of stereotypically masculine mental qualities such as emotional detachment. Some feminist epistemologists argued that the cognitive potential of stereotypically feminine cognitive styles, such as those emphasizing emotion, should be taken more seriously, but detached from claims about sex differences in the expression of such styles (Jaggar 1989). Such a project would be naturalistic in temper, and hence averse to overambitious global claims about the epistemic worth of gender-symbolized cognitive traits. Rather, it would focus on the contingent epistemic advantages such traits may have in particular contexts of inquiry--for instance, moral inquiry (Jaggar 2000). The upshot of these debates within feminist epistemology, as in women's studies generally, was not simply to affirm the “difference” pole, but to seek a way beyond both “difference” and “equality.”
A similar dynamic played out in the feminist methodology debates. In the 1970s and early 1980s, feminist researchers considered the question of whether there was a single distinctive feminist methodology applicable across the human sciences. Some researchers, objecting to the homogenization of women's experience entailed by quantitative research, argued that true feminist research ought to stick to qualitative descriptions drawn from their subjects’ own reports (Mies 1983). They argued that the feminist commitment to valorize women's experience, after centuries of silencing and neglect, meant that researchers should accept at face value what women report about their experiences (Stanley and Wise 1983). Again, such claims were strongly disputed by other feminist researchers. For example, Jayaratne and Stewart (1991) cogently argued that whether qualitative or quantitative methods should be used depended on the question being asked. Greaves and Wylie et al (1995) argued that quantitative, structured information gathering, which uses categories devised by the researcher rather than the individual subject of study, could help women. For example, by showing victims of domestic abuse that other women have had similar experiences, it could relieve their sense of isolation. Sandra Harding (1987a, 1) made a decisive case for methodological pluralism, arguing “against the idea of a distinctive feminist method of research.” She also warned against “loyalty to gender”--that is, uncritically valuing supposedly feminine cognitive styles, since femininity itself is defined by and plays a supporting role in the unjust gender system that feminist want to change (Harding 1993, 59). Far from advocating or practicing a monolithic feminine or feminist methodology, contemporary feminist research is strongly pluralist, in both theory and practice. Numerous books and anthologies on feminist methodology reflect this pluralism (Burt and Code 1995; Fonow and Cook 1991; Harding 1987; Nielsen 1990; Reinharz 1992).