Can self-validating neuroenhancement be autonomous?

In moral philosophical literature, individuals whose personalities are, in some sense, replaced by new ones have mainly been discussed in psychiatric ethics and in connection with medical advance directives (see, e.g., Haji 1997; Dworkin 1993, Chap. 8). Yet the cases focused on in those debates would not appear to be similar enough to self-validating neuroenhancement to throw sufficient light on the question whether such enhancement can be autonomous. Witt (2017) touches on self-validation in connection with discussing autonomy and neuroscientific treatment. Insofar as it pertains to the topic of this article, his view will be considered below.

If self-validating neuroenhancement cannot be autonomous, it might be employed, say, to improve the mood of a person who has permanently lost her autonomy or will shortly lose it for good. In the latter case, the person might choose to undergo a neurointervention foreseen to be self-validating, while he is still autonomous. Then the person could be said to be autonomous with respect to the intervention, even though his subsequent endorsement of his new self were heteronomous. I now put such cases aside.

The technology employed in enhancing human beings could assumedly be used in improving some other beings too. Yet humans are the primary—although not necessarily the only (cf., e.g., Beauchamp and Wobber 2014)—examples of autonomous beings and I am better informed about humans than about other beings (in whose case questions about endorsement of selves might not arise at all). Accordingly, I now focus on humans. Needless to say, the idea that human capacities should sometimes be enhanced is not new. For a long time, people have used caffeine to improve alertness, chocolate to better mood, and physical exercise to develop fitness, to mention just a few examples.

Neuroenhancement of physical capacities and neuroscientific lifespan extension would appear to be different enough from neuroenhancement of mental abilities to merit their own discussion.

If his fully knowing what it is like to be the new Mike presupposes his actually acquiring the new personality, Mike’s case too would involve some initially unforeseen consequences. Yet Mike’s post-enhancement endorsement of his new personality could be taken to result from his having been altered by neuroscientific techniques, even if his case did not involve unpredicted consequences.

I now put aside the question whether enhancement is the best term to use in connection with what I named self-invalidating and indecisive neuroenhancement.

For responses to central objections to the conception of autonomy see, e.g., Beauchamp (2009) and Beauchamp and Childress (2013). Some of its critics rather want to further specify the conception or to clarify the material, social, and communicative preconditions that should be met to attain autonomy in that sense than to reject it (cf., e.g., Jennings 2016).

Accordingly, the conception of individual autonomy is also commonly seen to ground the practice of medical informed consent, in which the moral acceptability of a person’s undergoing a medical intervention is made contingent on her autonomously authorizing the intervention (see, e.g., Beauchamp 2009, p. 58; Beauchamp and Childress 2013, Chap. 4).

Conceptions about the nature of individual autonomy vary widely (see, e.g., Dworkin 1988; Garnett 2014a, b; Korsgaard 2009) enough to suggest that autonomy theorists are not always focusing on the same notion. Accordingly, noting that attempts to find a core essence for autonomy have not been successful, Anderson, for instance, argues that normative debates about criteria for autonomy are best conceived as debates about “the relative merits of various possible packages of thresholds, entitlements, regulations, values, and institutions” (Anderson 2014, p. 355), so that autonomy can mean different things in different contexts.

For discussion on other problems neuroenhancement confronts see, e.g., Jotterand and Dubljevic (2016), Levy (2007), Savulescu and Bostrom (2009), and Savulescu et al. (2011).

Someone might propose that Kelly’s endorsement of her new personality cannot be autonomous insofar as the old Kelly would not approve of the new Kelly. But why should the autonomy of the new Kelly depend on her being approved by the old Kelly? After all, that the self I had before would disapprove of me as I now am does not entail that my current personality is heteronomous (see also, e.g., Juth 2011, p. 43).

See also, e.g., Müller and Walter (2010) and Focquaert and Schermer (2015). Witt concentrates on treatment but proposes that what he says applies to enhancement as well (see Witt 2017, p. 387). Indeed, while some clear cases of both treatment and enhancement exist, the division between treatment and enhancement appears difficult to uphold in general (see, e.g., Levy 2007, Chap. 3). Witt refers to the account of autonomy presented by Beauchamp and Childress (Witt 2017, p. 387), but wants to specify it in the way proposed in the above quotation. He (2017, pp. 389–390) goes on to argue that a patient’s approval of an identity-changing intervention is an informed consent only if the approval is based on an assessment of her pre-intervention quality of life and the identity change from her pre-intervention perspective and an assessment of her post-intervention quality of life from her post-intervention perspective, provided that pre-intervention and post-intervention quality of life are weighted equally and the weight assigned to the identity change is that which the patient endorses when considering her weighing preferences from her pre-intervention perspective. If the reason for looking at the personality change a person undergoes within a neurointervention from the perspective of the person’s old self referred to here is plausible, the problem focused on in this article—whether self-validating neuroenhancement conflicts with autonomy—is solved. Accordingly, I now concentrate on that reason.

The same applies in the cases of several of the other difficulties of self-validating neuroenhancement discussed in this article. Yet, as proposed, previous discussions on the difficulties have not focused on self-validating neurointerventions.

To clarify, I am not saying that those who understand authenticity in one or more of the senses referred to above necessarily see authenticity as an element of autonomy.

Kelly’s new capacities can also improve her ability to be loyal to her essence or to create herself (see also, e.g., Levy 2011) and the new Kelly need not be less honest in her self-presentation or less self-fulfilled than the old Kelly, if these kind of things were relevant to autonomy.

Reflective endorsement of one’s personality can, of course, be understood in some other sense than the one referred to above. But then it should be clarified what the other sense is and why autonomy should presuppose reflective endorsement in that sense.

In the case of reversible self-invalidating neuroenhancement, for example, the result speaks for returning to an enhanced individual’s old self rather than for his trying to come to terms with his new self. In the case of indecisive neuroenhancement, the conclusion implies, not that indecisiveness itself is cannot be a problem from the viewpoint of autonomy, but that the significance of the indecisiveness related to that kind of enhancement rationally cannot be downplayed as not reflecting the autonomous views of the enhanced person in question.