The question what mental disorders are lies at the heart of the philosophy of psychiatry. In search of a valid taxonomy of mental disorders, it is a question that needs a proper answer. In recent work, Kenneth Kendler et al. (2011) have put forward the “mechanistic property cluster” (MPC) model of mental disorder. On this view, mental disorders are mechanistically mediated clusters of multi-level (bio-psycho-social) properties. Kendler et al. present the MPC-model as a non-reductionist form of realism – realist because it tries to account for mental disorders in terms of the causal structure of the natural world, non-reductionist because it views mental disorders as clusters of multi-level properties. For the project of psychiatric nosology, such non-reductionist realism would be a great step forward and indeed preferable to pragmatist and constructionist models of mental disorder.

To critically assess the MPC-model in light of arguments against realism about mental disorders presented in the philosophical literature.

To achieve a proper understanding of the ontology of mental disorders that can inform future psychiatric nosology.

Literature study and conceptual analysis.

Despite appearances, the MPC-view fails to take into account the various (societal, practical, scientific) values that determine the delineation of mental disorders. It ultimately faces philosophical problems similar to those of more reductionist forms of realism.

The MPC-model fails as a realist model of mental disorders. Its non-reductionism, however, is an important contribution to theories of explanation in psychiatry.

The authors have not supplied their declaration of competing interest.