I am happy to welcome Rosa Terlazzo as the next featured philosop-her.  She is an Assistant Professor of Philosophy at Kansas State University, and specializes in social and political philosophy. Her current research concerns adaptive preferences, children, autonomy, and political liberalism. Rosa’s work has been published in Ethical Theory and Moral Practice.

Her post follows.

–MK

Children, Adults, and Adaptive Preferences

Rosa Terlazzo

Many thanks to Meena for starting this blog, and for letting me be a part of it! In this post, I’m going to discuss a problem at the intersection of two of my areas of interest: adaptive preferences and non-ideal theory. Roughly, the problem is this: while theorizing about adaptive preferences rightly belongs to non-ideal theory, even the best current work on the problem remains idealized in one unacknowledged and unacceptable way: it considers only the needs and interests of adults, and fails to acknowledge the ways in which those of children differ. This is clearly a problem that must be addressed in the adaptive preferences literature, but I’ll suggest that it also provides some broad guidance for the way that ideal and non-ideal theory ought to be done more generally.

It should be clear, I think, that philosophical work on adaptive preferences should be part of non-ideal theory. Broadly, adaptive preferences are those preferences that have been in some way deformed by unjust or otherwise morally unacceptable circumstances, so that those circumstances come to be in some sense preferred despite the harm that they do to the person who prefers them. Although accounts vary, the mechanism of this deformation is generally taken to have something to do with aspiration: persons end up aspiring to the impoverished circumstances that they can actually hope to achieve, precisely because they cannot hope to achieve any better. But whatever the mechanism, as long as the circumstances preferred are harmful, unjust, and actual, it is non-ideal theory that must deal with adaptive preferences. There are various accounts of ideal and non-ideal theory available in the literature, but for my purposes in this post, I can assume Rawls’s original characterization: If ideal theory is concerned with the principles of justice that would govern a perfectly just society, then ideal theory will say nothing about adaptive preferences, since the circumstances that lead to them will be absent in any society that could plausibly count as perfectly just. Any philosophical work on adaptive preferences, then, would seem to fit squarely into the realm of non-ideal theory.

Much of the work actually done on adaptive preferences, however, is nevertheless surprisingly idealized in one important way. Just as Rawls developed his theory of justice as fairness in abstraction from real-world considerations of disability and chronic illness, current work on adaptive preferences abstracts away from an even more omnipresent set of considerations: those involving children. This is a crucial failing for adaptive preference theorists, I think, because any account of adaptive preferences that goes beyond simple conceptual analysis ought to make recommendations for public policy; and indeed, virtually all of those writing on adaptive preferences are equally concerned with the question of what adaptive preferences are and the question of how they may permissibly be responded to by the state. But focusing these discussions only on the needs of adults yields unacceptable public policy prescriptions, since it ignores the interests of children. Let’s see how this is so.

Considering some examples will be useful. Imagine Simone, a wealthy woman who engages in socially sanctioned beauty practices that involve significant discomfort and cost on her part, and Elisa, a poor woman who eats only after her husband has had his fill, even though this means that she often goes hungry. Classically, discussions of appropriate ways of responding to such adaptive preferences have centered around the provision of respect for persons’ autonomy. While adaptive preferences demand attention because they are taken to support unjust or otherwise bad norms and social conditions, many of those writing on adaptive preferences recognize that reasonable persons can disagree about the good, and that some reasonable persons are accordingly likely to disagree with any proposed set of conditions and norms taken to be unjust or bad. It is not completely implausible to think either that Simone might see her beauty practices as a kind of art well worth suffering for, or that Elisa might see extreme self-sacrifice as the highest good. Other theorists argue that persons taken to have adaptive preferences are better understood as making difficult compromises in impossible circumstances, and that how best to navigate these circumstances must be determined by the person forced to do so, not by a third party who lacks her knowledge and need not live with the consequences of her actions. In this case, Simone may rightly take her beauty practices to significantly increase her chances of finding a romantic partner, and Elisa might allow her husband to eat his fill because he engages in physically demanding labor that supports the family financially. Both of these considerations, in conjunction with standard liberal presumptions against paternalism, make many theorists hesitant to recommend using coercive measures to encourage persons with suspected adaptive preferences to act in accordance with “better” or “more just” norms. Instead, what I take to be the best proposals in the literature recommend responses to adaptive preferences that rely on deliberative engagement directed by persons’ own convictions and conceptions of the good. Such proposals respect the autonomy of those taken to have adaptive preferences.

The problem is that children are also subject to adaptive preferences – yet the considerations discussed above do not apply to them. First, there is rightly no general presumption against paternalism when it comes to children. There is of course disagreement about how paternalistic powers ought to be allocated, but there is broad agreement that third parties may generally permissibly act for the good of children even when the children in question do not endorse those actions. Second, it is clear that children’s disagreement about the good is rarely reasonable at early ages, and their views about their own good do not, accordingly, require the same kind of respect owed to adults. Indeed, it seems to me that it is here that paternalism is most clearly justified: paternalism may be aimed not only at preventing immediate harm to children, but at ensuring that their convictions about their own good eventually becomes reasonable. That is, it can help them to develop the skills that they need to make their preferences more a product of their informed and considered choices, and less a product of their limiting circumstances. We cannot simply assume that what adults who have suffered adaptive preferences take to be best for themselves is what will be best for their children. While Simone and Elisa might have good reasons for the choices that they make, their daughters may not be best served by following in their footsteps. Neither woman may have had the chance to consider an initial course of action that included a broader set of attractive possibilities – but their daughters still might. In other words, the considerations relevant to children point not simply towards respecting autonomy in the case of adaptive preferences, but towards promoting it. With the right emphasis on promotion of autonomy, it may be possible to prevent the formation of adaptive preferences in the first place.

These differences between adults and children may seem obvious, but unless we make them explicit in discussions of adaptive preferences, we will be unable to consider their implications for the moral permissibility of policies aimed at responding to adaptive preferences. While I lack the space to respond to these questions here, I offer the following as an initial set of questions raised by recognition of the differences between children and adults:

1. If we can meaningfully respect the autonomy of adults who already have adaptive preferences, is there really reason to prevent children from forming adaptive preferences in the first place?
2. If there is reason to prevent adaptive preference formation in children in the first place, is it permissible to use coercive means to do so? At what point does it stop being permissible to do so, and on what grounds?
3. Given that the adults in a child’s community are a significant component of the circumstances in which that child finds herself, might conflicts arise between promoting the autonomy of children and respecting the autonomy of adults? If such conflicts arise, how should they be adjudicated?

Clearly, none of these questions can be answered without an extended discussion of the ways in which children’s needs and rights differ from those of adults. But just as clearly, these questions must be asked in order for responses to adaptive preferences to be morally justifiable. On at least some plausible answers to the above questions, responses to adaptive preferences will require something very different from (or at least additional to) the deliberative engagement supported by an exclusive focus on the needs of adults.

Before closing, let’s briefly return to the implications that all of this may have for the general way in which non-ideal theory is done. In the literature on non-ideal theory, there is a debate about whether ideal theory is required as a kind of blue-print or aim for the kinds of prescriptions that non-ideal theory should make. Some, like Rawls, hold that it is, arguing that we cannot know if we are moving towards justice without having at least a general idea of what perfect justice looks like. Others, like Amartya Sen, hold that it is does not, since a general conception of perfect justice is not required in order to identify clear injustices in the world and to imagine more just alternatives to them. My discussion of adaptive preferences in this post cannot begin to answer this question. It should, however, highlight one case in which otherwise careful and high-quality work in non-ideal theory can fail to offer permissible and comprehensive action-guidance because it fails to recognize the way in which it idealizes away from another deeply relevant consideration in the non-ideal world. Minimally, the case of adaptive preferences should serve as a cautionary tale for other non-ideal theorists who may inadvertently do the same. More maximally, it provides support for an argument (compatible with and in many ways similar to one made by Ingrid Robeyns) that some considerations will be so central to most plausible theories of justice that no work in either ideal or non-ideal theory may permissibly idealize away from them without explicitly addressing their relationship to justice. Considerations of children’s unique needs, pervasive as they are, will likely fall near the top of this list.