A diagnosis of adaptive preferences is licensed only when preferences are directed toward something suboptimal. We don't, however, make decisions about what is subotimal from Plato's heaven (or from our armchair). We observe, we evaluate, and -- crucially -- we listen to testimony. The problem with the adaptive preference model is that it allows us to dismiss certain kinds of testimony as irrational or misleading. And so the adaptive preference model can quickly become a way of defending the moral status quo. (135)
Insofar as it justifies in discounting first-person testimony, the adaptive preference model can be a way of preventing this kind of progress. Misapplied, the adaptive preference model can simply entrench pre-existing biases. Elizabeth Barnes (2016) The Minority Body: A Theory of Disability (137)
Yesterday's Impression (recall) was inspired by passages in which Barnes criticizes certain applications of an adaptive preference model to discount testimony. But that post otherwise ignored her account of adaptive preferences which is the subject of today's Digression. As the quoted passages reveal, Barnes appeals to a notion of moral progress. This appeal to moral progress (and an accompanying presupposed moral community between Barnes and her sympathetic reader) is quite important in her overall argument, and in a future post I will be critical of the role it plays in her treatment.
But I don't think the idea of moral progress is central to her critique of certain uses of adaptive preferences model. Nothing substantial (as opposed to rhetorical) is lost if we rewrite the last quoted sentence from p. 135 as follows, And so the adaptive preference model can quickly become a way of defending the status quo as understood by expert. And the first sentence from p. 137 can be rewritten as follows, Insofar as it justifies in discounting first-person testimony, the adaptive preference model can be a way of preventing moral understanding of victims of testimonial injustice. This is not to deny that moral progress plays an important role in Barnes's presentation--to motivate some of her arguments, she appeals to her readers willingness to grant that prior discounting of the testimony and preferences of women and gays as adaptive was obviously wrong (134-6).+
For, in fact, Barnes cleverly applies the growing insights of the literature on (Fricker's) testimonial injustice and (Dotson's) testimonial smothering to undermine over-reliance of adaptive preference model (especially in the capability literature inspired by Nussbaum and Sen) to ignore first person testimony of disabled people.* (I am assuming here some familiarity with these concepts.) As she writes, "whether being disabled is somehow bad or suboptimal is precisely what's up for debate." (133)+ To be sure, Barnes allows that there are occasions one can still diagnose adaptive preferences in others, especially when there are fundamental conflicts between what a person says about herself and other factors of her well-being (but the discussion is a bit brief, see 140-1).
To understand the historical significance of Barnes's critical treatment of the misapplication of the adaptive preferences model it is worth briefly reminding ourselves of the rule it played in Elster's (1983) Sour Grapes, which formulated the modern conceptual apparatus of adaptive preferences (recall). (Barnes also mentions Sour Grapes.) He did so, in part, to combat certain perverse implications of utilitarianism (see chapters III.3-4 especially). What interests me here is that in Elster's account, one role for the invocation of adaptive preferences is to combat a certain form of status quo bias when one takes (historically conditioned) given wants or subjective preferences for granted (say in one's utility calculations in evaluating some policy or another). By giving a principled justification for discounting (some) existing preferences, the theorist who relies on the adaptive preferences model can propose normative improvements.
As an aside, this is not to deny that there is not some status quo bias built in Elster's approach. Elster takes political orders and regimes largely for granted and, while he allows changing of wants by “rational and public discussions,” (140 referring back to I.5 and its discussion of the political “transformation of preferences rather than aggregation” of preferences, (35)) his notion of “collective rationality” (42) has no space for the radical, even violent, transformation of institutions. Moreover, he embraces a further status quo bias by insisting that, while acknowledging that intuitions may be “culturally relative,” we “must require of a theory of justice that it does not strongly violate our ethical intuitions in particular cases” (137—emphasis added; to be sure, Elster is less optimistic than Rawls about possibility of consensus). As Michael Della Rocca taught me, this reliance on the method of reflective equilibrium has considerable status quo bias built into it.
By contrast, Barnes shows that there are two predictable risk of (moral and economic) experts relying on an adaptive preference models. First, and this is the one she emphasizes, is that (a) they put themselves in a situation of defending an imperfect status quo (that they themselves understand as moral) while, simultaneously, (b) discounting evidence that could correct the status quo. That is to say, the adaptive preference model does not generate methodological and conceptual incentives to self-correct. Second, and this is strongly implied by her treatment, the down-side risks of (let's stipulate, well-intentioned) application of the adaptive preference model is primarily born by often vulnerable member of society.
Given that the research ethics that I propose in light of methodological analytical egalitarianism is especially critical of approaches where the down-side risk of use/application by non-expert vulnerable, one might expect me to be very critical of the adaptive preferences model (as another danger of expert over-confidence in their own model [justness, etc.]). Strikingly, the adaptive preference model was, in part, designed to address the needs of the most vulnerable in the context of the history of oppression.
Not unlike Barnes, I don't think experts should never rely on an adaptive preference model. But experts and policymakers that rely on it should also be alert for the two foreseeable risks in the approach. As noted above, Barnes supplies us with a further test in application: do the preferences expressed by particular subjects cohere with other features of the subjects' lives. (For example, if a victim of abuse says they're flourishing, but otherwise has lots of symptoms of depression.) This is an empirical question and requires considerable and careful contextual judgment(s). I would add, in addition, corrections to the very theory of adaptive preferences to account for the foreseeable and known cognitive and status quo biases of experts and policymakers alike. But that's for another time.
+ Barnes does not offer any actual example of somebody invoking adaptive preference theory in doing so.
*Barnes's position resonates interestingly with work done by Serene Khadar on adaptive preferences.