Abstract

The harmful dysfunction (HD) analysis of "disorder" holds that disorders are harmful failures of "designed" (that is, naturally selected) functions. Murphy and Woolfolk (2000) present a series of proposed counterexamples to the HD analysis to support their claim that it fails to provide a necessary condition for disorder. They argue that disorder can exist where there is no failed function, as in failed spandrels and inflamed vestigial organs, and that there can be disorders when everything is working as designed, as in environment--design mismatches and disorders acquired through normal learning processes. They also argue that the HD analysis suffers from methodological problems, including value-ladenness of dysfunction judgments and unwarranted assumptions about the nature of internal mechanisms and their functions. In this paper, each of these objections is critically evaluated. Reanalysis of the proposed counterexamples is argued to reveal strong support for the HD analysis. For example, failed spandrels are considered disorders when and only when they imply failures of designed functions; dysfunctions of vestigial organs involve failures of function at the tissue level, not the organ level; mechanism-environment mismatches are not considered disorders; and conditions acquired through normal learning processes can involve dysfunctions and are considered disordered only when they do. Murphy and Woolfolk's methodological objections are found to be based on a misinterpretation of the HD analysis. It is concluded that Murphy and Woolfolk fail to offer a cogent objection to the HD analysis.