Would Genetically Engineering a Child’s Preferences Diminish Her Autonomy?

Abstract


 
 
 Proponents of “moral enhancement” argue that we should harness emerging technologies such as genome editing to instill preferences for performing morally good actions over bad ones, and they suggest that this would be worthwhile even if it made performing bad actions psychologically impossible. Critics object that such a gain in moral behavior would not be worth the resulting loss to human autonomy. This debate has remained confined within the context of moral preferences. Genetic engineering for non-moral preferences (GENP) – such as enjoying music or athletics – however, has not been discussed. Since preferences – what an agent likes, values, or tends to desire – form the core of who she is and what she pursues in life, and since autonomy requires that an agent be her own person and pursue her own life projects, her autonomy might seem to be diminished if her preferences were selected before birth by a third party. Plausible as this may seem, we disagree. In this paper, we argue that parents could select for a wide variety of substantive (i.e. not merely insignificant) non-moral preferences without compromising their child’s autonomy, provided that certain criteria that we propose are satisfied, e.g. the selected preferences must not be such as to inhibit the agent’s capacity for rational deliberation, which is a key component of the conception of autonomy we employ. We then respond to two objections: (a) GENP would be inherently incompatible with autonomy, regardless of whether it meets any such criteria, and (b) even if GENP would not be inherently incompatible with autonomy, people might still regard it as an alienating influence and might regard their selected preferences as non-autonomous. We then argue that this second objection would be less forceful according to a more “externalist” rather than “internalist” conception of autonomy.