The chapter “Should Psychiatry Be Precise?” challenges the Precision Medicine Initiative (PMI) and the Research Domains Criteria (RDoC) initiative that have been advanced by the National Institute of Health. The chapter includes valuable points about the challenges of applying reduction across levels of analysis, the value of nosological revision, and the potential pitfalls of using big data. But the overall argument constructed in the chapter is a straw man that does not reflect well the intentions of those who designed these initiatives nor the understanding and aims of scientists now engaged in research under the aegis of these initiatives. A more constructive approach might focus on specification of tractable questions about causality in mental health research, and aspects of subjective experience that are currently under-represented in biological psychiatry.

TThe twenty-first century paradigm of precision medicine can be characterized by its joint commitments to (1) the revision of traditional nosological systems (2) the utilization of transformative new methods of data collection and analysis (“big data”), and (3) the employment of scientific methods able to reduce complex phenomenological, behavioral, and physiological signs and symptoms to underlying biomechanisms. This chapter assesses the value of these commitments for psychiatry, and concludes that they are collectively neither necessary nor sufficient for progress in the explanation of, and intervention upon, mental disorders. Each holds promise and has proved transformative in some areas of psychiatric research and practice, but their appropriateness is better assessed independently and circumstantially. Meanwhile, the value of other traditional psychiatric commitments – such as to the principled demarcation of the pathological from the normal, and the prioritization of research that has clinical application – should not be abandoned amidst the current vogue for precision.

Psychiatric diagnosis and research is hampered by problems in nosological classification. Recent development has seen the suggestion of dimensionally-based classification systems like the Research Domain Criteria (RDoC) and cognitive ontology (CO), with the latter being developed by Bilder in his chapter. I here discuss some usually tacit or implicit presuppositions of CO concerning brain and mind. I conclude that, despite shifting from an entity-based approach (as in the DSM) to a dimensional approach, the cognitive ontology project still encounters the problem of connecting neuronal changes to psychopathological symptoms and, more generally, brain and mind.