Autonomy, Experience, and Therapy

The contemporary philosophical idea of autonomy has a psychological implication, to wit, that there exists a comprehensive set of ideal competences, realized in our mind/brain, that enable a person to be self-governing. Autonomy is normally accorded individuals who enjoy a certain kind of psychological functioning and, perhaps, a certain sort of psychological history (Christman 1991). We think that autonomous individuals critically evaluate their life and actions, endorse them as self-determined rather than imposed from without, and guide their own lives in accordance with the plans and values they have worked out for themselves. Such self-government has to depend on some psychological structures, but philosophers have tended to theorize about what these structures might be at a considerable distance from what the behavioral sciences can tell us. Anderson and Lux's paper is a welcome contribution to the growing literature that tries to connect issues in ethics with empirical considerations. This sort of methodological, as opposed to merely metaphysical, naturalism, is only just coming into vogue in ethics, even in moral psychology.¹ Anderson and Lux are, as far as I am aware, the first people to reflect seriously on what we can learn about autonomy from clinical experience. Theirs is a very suggestive and fruitful paper that looks at real-life cases of loss of autonomy and tries to reach conclusions about what the deficits in performance tell us about the psychological preconditions of autonomy. Their paper also opens the way for an investigation into the precise neurologic and computational underpinnings of these capacities, which is essential if we are to get a serious moral psychology underway.

I find little to quarrel with in the story that Anderson and Lux tell, but I will try to modify it slightly to bring out what I think is the important psychological point they make and explore some ramifications. I begin by claiming that they have scored a decisive point in favor of an external conception of autonomy. Second, I suggest that what Anderson and Lux have really put their finger on is less a requirement of accurate self-assessment than a requirement of sensitivity to feedback from the environment. I conclude by raising some of these issues about the relations between autonomy and assessments of psychological function when the deficit responsible is less amenable to clear cut neurologic identification.

The External Perspective on Autonomy

One way of taking the paper (although this goes beyond anything that Anderson and Lux suggest) is that there is a smooth naturalistic picture of the cognitive basis of autonomy. When people have a normally functioning neurocomputational setup in their heads, we should assume that they are autonomous, and vice versa. However, autonomy cannot simply be identifiable with a mind that is working as some scientific theory says it should, because a mind that is in that sense working correctly could, nonetheless, be directed by forces other than the agent. Servile or deluded individuals might endorse their lives even if those lives are the product of indoctrination, self-deception, or internalized oppression (Mackinnon 1997; Mele 1995). Normal mechanisms of belief fixation, for example, may not foster one's autonomy if the only context in which they are exercised is one in which many possible ways of life are simply unavailable for consideration, having been excluded from consideration by the dominant forces around one. We might, then, think of full autonomy as depending on relational rather than intrinsic properties. Perhaps autonomy requires a subject to have both normal psychological capacities and the right sort of environment in which to exercise them. I do not follow-up on this suggestion, but I mention it because it introduces two dimensions, within the subject and around the subject, where the externalist perspective on autonomy can get a grip. I discuss the former, but I do not mean to suggest that the latter is not important. A full account must deal with both. But even the psychological structures that foster autonomy cannot be identified with a mind working according to some scientific theory, as we shall see.

Many writers on autonomy argue that we must recognize appropriate, objective constraints...