- Dysfunction as a Value-Free Concept:A Reply to Sadler and Agich
classification, disease, function, psychiatric nosology
In a series of recent articles (Wakefield 1992a, 1992b, 1993), I have presented the "harmful dysfunction" analysis of the medical concept of disorder. My analysis implies that "disorder" has an evaluative component ("harmful"), but that the negatively evaluated conditions that legitimately fall within medicine's conceptual domain are limited to those that satisfy an additional descriptive criterion, namely, that the harm is due to a dysfunction. "Dysfunction" is here understood as the failure of some internal mechanism to perform a natural function, and "natural function" is understood in explanatory terms as an effect that itself somehow played a role in shaping the biological mechanism that causes it. For example, we judge that enabling us to see is a natural function of the eye because we believe that it cannot just be accidental that the eye is structured so as to have that effect; thus we conclude that the effect of enabling us to see must somehow have played a role in leading to the existence and structure of the eye in our species. We now know that such shaping of biological mechanisms by their effects is due to natural selection, so the definition of "natural function" in terms of the explanatory relation between effect and cause can be given scientific substance by linking it to evolutionary theory. Note that, according to this analysis, evolutionary theory is not part of the definition of natural function—nor could it be, because functional concepts existed in biology long before the advent of evolutionary theory—but rather it is a theory of the essential processes that give rise to the phenomena that fall under functional concepts.
Sadler and Agich (1995)1 lodge four objections against the "harmful dysfunction" analysis of disorder: that my definition of "dysfunction" is in fact value-laden because value terms appear in it; that evolutionary theory itself, in the form that I rely on it, is covertly value-laden; that even if my evolutionary descriptivist analysis of "dysfunction" succeeds, it yields an analysis of "disorder" that has no practical implications for psychiatric nosology; and that my attempt to delineate a descriptive "dysfunction" component of "disorder" is driven by my relativism about values and could be dispensed with if I adopted a more sophisticated account of values. The objections are aimed at disputing my claims that "dysfunction" captures the descriptive component of [End Page 233] the meaning of "disorder" and that "dysfunction" is best understood as a value-free scientific concept tied to evolutionary theory.
Examining the objections in the order they were presented, I argue that all four of Sadler and Agich's objections to my analysis of "disorder" are invalid.
The "Value-Ladenness" Objection
Sadler and Agich claim that my definitions of "natural function" and "dysfunction" are not truly value-neutral because I "fail to define function and dysfunction independent of value terms" (Sadler and Agich 1995, 219). They note that "if a concept is truly value-neutral, value terms should neither appear nor be necessary in defining it," and claim that my analysis "does not meet this requirement" (221). Sadler and Agich offer the standard test for value-neutrality: "For a concept to be value-neutral, it should be amenable to redescription without reference to evaluative terms" (223); that is, a term is value-neutral if its definition can be formulated in a way that retains the term's meaning but contains no value terms. The only evidence presented that my analysis of "function" fails this test is that in some of my definitions of "function" I use the term benefit to describe the relevant effects. Sadler and Agich assert without argument that these value terms "are central to [Wakefield's] analysis" (224).
However, according to Sadler and Agich's own criterion, the test of value-ladenness is not whether a value term happens to occur in a description of a term's meaning (for it might appear incidentally, contingently, or superfluously) but whether the value term is necessary for capturing the intended meaning. If a value term does appear, the question then becomes...