Abstract

The DSM classification system has long been criticized for insufficiently prioritizing scientific accuracy by inserting social values. Naturalists imply that the DSM should define disorders in terms of biological markers, producing a value-free classification. This thesis argues that the naturalist critique imposes inappropriate goals; while scientific accuracy matters, the primary goal of the DSM is to serve as a practical diagnostic tool that promotes public well-being.
To support this normative claim, I draw on the argument from inductive risk, which argues that value judgments are necessary when evidence interpretation is uncertain. I demonstrate that psychiatric classification faces significant underdetermination at two distinct layers: the “ontological layer” concerning the nature of disorders, and the “operational layer” concerning diagnostic criteria.
Using the consolidation of Autism Spectrum Disorder (ASD) and the removal of the bereavement exclusion in Major Depressive Disorder (MDD), I show how empirical data underdetermine classification decisions. I argue that experts must manage this uncertainty by weighing the non-epistemic consequences of error. Specifically, social values regarding resource allocation justify a preference for false negatives in ASD diagnosis, while clinical urgency justifies a preference for false positives in MDD. I conclude that social values should, in addition to empirical considerations, influence how the DSM defines psychiatric disorders.