The Likeness Argument: Reminders, Roles, and Reasons for Use

I WOULD LIKE TO respond to the four commentaries in turn. In each case I have started by setting out what I think the commentaries are claiming; in doing so, I may reveal that I have misunderstood or misconstrued, and I apologize where this is the case. My responses in many cases are provisional: the commentaries have given me much to think about. Also, my responses are selective—there are many points not touched upon here that deserve consideration. Finally, the order of my responses is arbitrary.

Response to Tyreman

Steve Tyreman (2003) wants to accept my conclusion, but he has other—and he claims better—reasons than I have.

We agree that you cannot show that so-called mental illnesses really are illnesses by means of the likeness argument. His reasons can be understood through his positive thesis about the nature of illness. Illness, he says, is a sort of experience of the failure of agency. For patients, the experience in question is of the loss of "an aspect of their agency as defined by the activities they normally take for granted" (p. 270). What this experience is due to is not well defined, and the presence or absence of the experience is not to be judged by the presence or absence of disease—at least as disease is understood on the medical model. Attempting to judge the reality or otherwise of illness (mental or otherwise) on the basis of the medical model is precisely where the likeness argument, Tyreman argues, goes wrong.

My main objection to Tyreman is that his positive thesis would legitimate the use of the likeness argument. ¹ Tyreman defines a notion of illness which meets Boorse's criteria for being generic: it shows no "obvious partiality to body over mind" (Boorse 1982, 31; see also, in relation to malady, Culver and Gert 1982, 65). The defining features of illness could be used to frame the issue about the existence of mental illness in these terms: "are human conditions such as schizophrenia characterized by this experience?" My claim is that it is framing the question about the reality of mental illness in this way that is the problem, as once we ask "is an alcoholic experiencing the failure of agency?" my strong objection to the likeness argument applies. Accounts of alcoholism seem available in which a failure of agency is not present, and describing it in illness terms then seems to presume what it is supposed to prove.

What may surprise Tyreman in this response is the application of my arguments against the likeness argument to his concept illness. What I think my doing so shows is that I do not see the likeness argument as attempting to work, or failing, in the way he does. It may seek to employ criteria for physical disease in categorizing schizophrenia or alcoholism, but it is equally at home, and equally fallacious, when it seeks to employ criteria for generic notions of illness to categorize schizophrenia or alcoholism.

Response to Gipps

Richard Gipps (2003) thinks that, in construing the likeness argument the way I have, I have missed some things. Whereas I have been inclined to see it as seeking to infer conclusions from agreed premises, and failing for the reasons I give in my strong objection, Gipps thinks it may be seen as reminding us of something. Boorse, for example, appears to argue that conditions such as schizophrenia are dysfunctions like physical diseases and hence are of the same kind. But what he is actually doing, according to Gipps, is to draw our attention to intuitions about health and illness that are constitutive of our grasp of others' states of mind. For example, when faced with someone with a delusion, Gipps says we have the pretheoretical idea (perhaps it is some form of reaction?) that this person has "something seriously wrong with their mind" (p. 257).

Gipps argues that what this means is that the...