Commentary on "Aristotle's Function Argument and the Concept of Mental Illness"

Chris Megone’s proposed definition of illness as a “functionally explicable failure of development,” combining as it does elements of modern evolutionary theory with the conceptual resources of Aristotelian biology, opens up a potentially rich seam of research into core health concepts. In this commentary, I want to fine-tune his (generous) references to my own work in this area on three points: (1) the overall structure of my argument; (2) the sense in which I use the term overt; and (3) the relationship I suggest between function and action (and hence between disease and illness). This fine-tuning will endorse Megone’s claim that action-based theories of the kind I have proposed are incomplete: the fact that they incorporate “growth buds” for future development is, I believe, a virtue of such theories. My fine-tuning, though, will contradict Megone’s claim that action, in the relevant sense, can be replaced by an Aristotelian notion of function. It will suggest, on the contrary, that Aristotle offers us a function-based medical model no different in principle from those that action-based theories aim to replace. Indeed, the main virtue of Aristotle’s account is that it makes explicit a hidden teleological element of meaning currently operating as a suppressed premise in modern attempts to ground a value-free account of function on evolutionary theory (Fulford forthcoming). In opening up Megone’s seam of research, then, one can learn as much from Aristotle’s errors as from his insights.

The Structure of the Argument

Megone correctly notes that one consequence of my value-based analysis of medical concepts, as set out in my Moral Theory and Medical Practice (1989), is to reverse the usual pejorative perception of the concept of mental illness. In the standard medical model, which takes the concepts of illness and disease to be at heart “scientific” and hence value free, mental illness, in being more value laden a term than physical illness, is conceptually suspect. In a value-based model, by contrast, mental illness, in being more value laden a term than physical illness, offers a window onto the meaning of medical concepts generally. Once established, this window makes possible a number of argumentative shortcuts (as in Fulford 1991, quoted by Megone). In the way the argument is developed in my book, however, the window offered by mental illness is not, as Megone suggests, a starting point but a conclusion.

The basic structure of my argument is set out in detail in chapter 1 of Moral Theory and Medical Practice. The essential point can be made in terms of the linguistic-analytic distinction between reflective definition and the actual use of concepts. In much of the literature on medical concepts, it has been assumed that, because the term mental illness is more problematic in use than the term physical illness (partly because it is more value laden), the former term must be the more obscure in meaning. Hence the debate about the “meaning of mental illness” has proceeded essentially by comparing and contrasting mental illness with what has been taken to be the relatively transparent concept of physical illness. But the debate itself, as I show, turns primarily on differences of view about the meaning not of mental illness, but of physical illness! Hence the meaning of physical illness is at the very least no less obscure than that of mental illness. Nevertheless, there is more agreement about the use of the term physical illness; so it still makes sense to start the debate from (agreed) examples of physical illness; albeit that the form of argument adopted has to be one of generalization rather than of direct comparison.

An argument by generalization from agreed examples of physical illness does not prejudge the validity of mental illness. If, in Szasz’s still-challenging phrase, mental illness is a “myth” (1960), then the generalization will fail. If, on the other hand, there is a legitimate use of the concept of mental illness, then the generalization will succeed (assuming competent analyses in...