Beyond Adaptive Preferences: A Different Model of “Hard Cases” for Feminism


Feminist critique is a kind of moral critique of things and people that contribute to the inequality or oppression of women. But an important challenge for feminists quickly arises: what should feminists say about women who participate in and endorse practices that contribute to or constitute the oppression of women?

Many feminists understand such participation via the lens of “adaptive preferences,” i.e. preferences that have been shaped by conditions of oppression to favor those conditions (e.g. Nussbaum 1999). However, this puts feminist critics in the position of critiquing the apparently free choices of adults, and it risks critiquing “legitimate but unfamiliar conceptions of the good” (Khader 2012, 302) while impugning the agency of precisely those who they are centrally concerned to protect and uplift, rendering them as “dupes of the patriarchy” (Narayan 2002). On the other hand, regarding endorsed norms or practices as legitimate risks abandoning feminist critique (since most objects of feminist critique have women who defend them).

We can recognize the rationality of women who participate in and endorse oppressive practices, and understand feminist critique as a sort of moral critique of the women and practices by adopting an understanding of gendered practices as norm-governed, suboptimal Nash equilibria for social coordination. Such a model sees participants in oppressive, gendered social practices as rational, and without better individual alternatives. But it sees feminist critique as a fundamental sort of moral critique – a critique of “altruism failures” (Kitcher 2011) that make everyone worse off than they might otherwise be.

The model also avoids commitment to specific, substantive conceptions of the good which are likely to involve question-begging assumptions in the context of cross-cultural critique. Thus, this model offers a new and normatively better explanation of what Natalie Stoljar (2013) has called “hard cases” for feminist accounts of autonomy.


7 thoughts on “Beyond Adaptive Preferences: A Different Model of “Hard Cases” for Feminism

  1. Salomone-Sehr, Jules says:

    Many thanks for the great talk, Ron (if I may). And my apologies for missing the Q&A last night (it was already late in the time zone where I’m currently located).

    When you introduce your third desideratum, you say that cases involving coercion are not the genuinely hard cases that feminist critiques have had to grapple with. One reason why these coercion-involving cases are not genuinely hard is that after all, if coerced, your conformity with the oppressive norm or practice does not require any genuine endorsement of that norm or practice. A hard case, then, seems to involve (or at least could involve) an endorsement of that sort.

    I’m extremely sympathetic to your game-theoretic account. That said, I wonder about the way in which it handles genuine endorsement. It seems to me that if I’m stuck in an oppressive coordination equilibrium, I need not embrace it; instead, I might be content with doing my bit to reap the suboptimal payoff I have access to, while wishing all along that I did not have to play my part in that oppressive practice.

    I can see how your account leaves room for instances of participation where the oppressed party endorses the oppressive practice. But could you say a bit more about how your account might not just leave room for, but also supply an explanation of, such forms of endorsement?

    Many thanks, again!

    1. Mallon, Ron says:

      It seems to me endorsement cases are the “hardest” of hard cases, and one’s where it is most plausible to postulate adaptive preferences (as opposed to the game theoretic account I suggest). While that may be the right thing to say of some cases, I think my account can explain a kind of endorsement in the same way it explains conformity. Conformity is explained by the local advantage of conforming over nonconforming (even if there were a better way to go, it’s not available to individual agents or agent-types acting unilaterally).
      Suppose someone conforms to an oppressive norm and endorses their conformity. One possibility is that they have adaptive preferences. But another (on my model) is that they (a) lack epistemic access to the possibility of a more optimal coordination point (they don’t know things could be better), or (b) judge (perhaps even correctly in some cases) that such a point is practically impossible (because of social barriers to cooperation). In such a case, they are likely to not view such an alternate possibility as “live,” and their endorsement of conformity might be best understood as sincere, because they imagine the benefits of conformity and contrast them to the relative costs they would bear if they did not conform with the oppressive norm (and did not succeed in reaching a more optimal coordination point).
      Hope that makes sense (though let me know if not). Thanks for listening to the talk and for the question. -Ron

  2. Mallon, Ron says:

    Thanks, Jonathan. Interesting and well put. The straightforward answer to your question is “yes” for the Prisoner’s Dilemma. The “not benefitting” is in contrast to cooperative , cooperate, but because that is not a Nash equilibrium, it would not be rational for either play to choose that (i.e. to cooperate, even if the other person was going to cooperate). They are rational to defect.
    However, I’ve committed to more than I wanted to, since I meant for the Prisoner’s dilemma mainly to illustrate the idea, rather than to exhaust it. If it were a “Stag’s Hunt” (in which there are two Nash equilibria: cooperate, cooperate and defect, defect), then the overall problem remains much as I described it (how to ensure cooperation), but the answer to your question might be “no.” The comparison vis-a-vis which they are clearly rational depends upon maximizing their payoff vis-a-vis non-equilibria alternatives, and the comparison with regard to which they do not benefit is with more optimal equilibria.
    To answer your final question: yes, I think it’s okay if comparisons are shifted. However, I would note that even in the Stag’s Hunt game an actor at a suboptimal Nash equilibrium (who does not benefit) is not irrational for not acting otherwise since the benefit of doing so is only available conditional upon the actions of others).
    Thank you again for your question!

  3. Schaffer, Jonathan says:

    Thanks Ron, I found that very insightful!

    I won’t be able to attend but wanted to ask about how your 2nd desideratum (that the women participating do not benefit) interacts with your 5th desideratum (that the women participating are acting rationally). It seems to me that (i) there is a comparison relative to which the women participating both benefit and are acting rationally, namely the comparison between the actual “both defect”-style outcome and the “only the other defects”-style outcome; and (ii) there is also a comparison relative to which the women participating both do not benefit and are not acting rationally, namely the comparison between the actual “both defect”-style outcome and the “both cooperate”-style outcome. **But is there any one comparison relative to which the women both do not benefit yet can still be said to be acting rationally?** That is, is there any one way in which your 2nd and 5th desiderata are both met? (Maybe that is ok with you. Maybe all you want to say is that there is a comparison relative to which the 2nd desideratum is met, and there is a comparison relative to which the 5th desideratum is met. In that case please take my question to be: Is it ok if these desiderata are each met but only relative to shifted comparisons?)

    1. Mallon, Ron says:

      Hi Jonathan. I replied to you at 14:42 but it wasn’t “tagged” as a reply. Thanks for the question!

  4. Martin, Laura says:

    Hi Ron,

    Thank you very much for your interesting talk! Unfortunately my time zone is 12 hours behind, so I’ll be unable to attend the Q&A, and thought I would pose a question here instead…

    I’d like to press you to say a bit more about the contrast you draw between cases where there is NO coercion (and hence, what needs to be explained is the free choice of individuals), and cases where there IS coercion. I’m not convinced there is such a clear line here and, if that is true, then the framing of the whole problem as one about how to explain the wholly uncoerced choices of individuals (in ways not to their benefit) becomes more questionable. For instance, to take the example of raunch feminism, it seems difficult to explain the women’s participation in these events without considering the effect that capitalist interests and patriarchal backlash have had on how feminism is presented in the public sphere – in magazines, TV, etc. for instance, ‘owning one’s sexuality’ is presented as ’empowering’ and ‘feminist’ – when in fact, as you point out, this tends to serve the interests of men (as well as industries that profit off women buying clothes, makeup, etc.).

    I can see why, on a very local level, we would want to say that the women’s participation is uncoerced – no-one forces them to do it, after all! But it seems we lose something very important for feminist critique if we ignore the background relations of power that allow sexual objectification of the self to be prominently cast in the media as feminist and empowering. In light of this, my question is whether (and how) you think these more distant relations of power should inform how we understand the nature of the problem – if at all!

    1. Mallon, Ron says:

      Thanks for this question, Laura. This is obviously a very thorny question since it is difficult to say what counts as coercion, unacceptable manipulation, acceptable persuasion, and so forth, even though we can perhaps think of cases that seem like clear cases of each.
      I think you are correct that I am assuming a distinction between genuine coercion and other noncoercive cases that raise questions of the autonomy of the women involved. Coercion seems philosophically easy, since we can attribute the conformity of a person with oppressive practices by appeal to their rational response to coercion. In contrast, I take it the interesting question is the extent as to whether the agency of oppressed persons might be compromised. (Again, I take it that I am following others in assuming this.)
      So, in cases in which people aren’t coerced (what Uma Narayan calls “literally coerced”), the question emerges about whether agency is compromised or some other explanation (e.g. my explanation, perhaps) might be operating. The cost of saying autonomy is compromised is just the cost of potentially failing to recognize the agency of people who have it, and of potentially justifying paternalistic policies or norms concerning such persons. And so, it seems to me like it would be (normatively) useful if there were an alternative.
      Still, I don’t deny that other sorts of explanations also have purchase for various cases, i.e. that there may be cases of adaptive preferences. And I haven’t shown (in the talk) that my model is a better empirical explanation of the cases that I described than those alternatives. (I wonder if that is the view that you are tempted by in your question.) Thank you for your question!

Leave a Reply