Reflections on philosophy of psychiatry

In this issue of the journal, Stein et al ¹ summarize the current status of philosophy of psychiatry, focusing on several issues in the areas of nosology, psychiatric science and the mind‐body problem.

I agree with their favorable view of “soft naturalism” as an appropriate conceptualization of psychiatric diagnostic categories. “Hard” naturalism, expecting clear boundaries and essences, which applies to some entities in science (e.g., elements of the periodic table) is clearly inappropriate for psychiatric disorders. Our disorders are “fuzzy” and do not have essences.

Do psychiatric disorders exist in the world to be “discovered” or do we “create” them? Although I favor the former position, historical work suggests some skepticism. If Kraepelin went into zoology, like his older brother, would someone else have described dementia praecox and manic‐depressive illness in the same terms? Many clinicians were working, over the 19th century, to sort out the diagnostic space of mood and psychotic disorders ² . Something resembling our current views would likely have evolved without Kraepelin, but how similar would it have been?

Psychiatry has long believed that the next scientific revolution – be it brain pathology, Mendelian genes or monoamine neurotransmitters – would solve the problem of psychiatric disorders by discovering mono‐causal “essentialist” explanations. This is a regrettable, recurrent pipe‐dream of our profession. Ironically, near the beginning of the evolution of psychiatry (early 19th century), we confronted general paresis of the insane. This has been the greatest success story of our science, but left us with an unrealistic expectational set that we cannot repeat for our remaining syndromes.

With some rare exceptions, psychiatric disorders are highly multi‐causal “all the way down”. It is therefore unrealistic of us to express impatience that DSM has not yet produced an etiology‐based diagnostic system. Despite all the advances in brain imaging and molecular genetics, we still remain ignorant at a basic level about the causes of our disorders. This is not likely to change anytime soon. Indeed, advances in molecular genetics are pointing out how hard this will be, aptly called “the curse of polygenicity”.

My view of the role of values in psychiatric nosology is more restricted than that advocated by Stein et al. The most important value that enters into the DSM review process is the prioritization of potential validators. We broadly agree now that diagnostic change should, in so far as possible, be empirically based. Revised sets of criteria for a disorder have to perform better than their predecessors on real‐world data. But how to decide what kind of performance should be prioritized? Once you know what you want your diagnoses to do, it becomes more possible to at least approximate a full‐bore data‐driven approach. The current system, with many different validators of uncertain weights, is problematic ³ . The DSM Review Committees have a much more difficult job than those who evaluate randomized controlled trials for the US Food and Drug Administration, in which drug response is the only relevant variable.

I also agree with Stein et al's emphasis on the importance of pluralism in psychiatry. As a field, we have multiple legitimate scientific perspectives on our disorders, their etiology, and their treatment. The need for pluralism results from the multi‐causal nature of most of these disorders. However, I have a concern about “undisciplined” pluralism: it is at risk for degenerating into an anti‐scientific “let a 1,000 flowers bloom” perspective. To sit at the pluralist table, hard empirical work is needed. Correlations are not enough. Designs that allow causal inference are critical. Our thresholds need to be high, because the field of mental health attracts a wide variety of theorists, some of whom are little constrained by the problems of empirical evidence. I advocate for a “hard‐nosed” pluralism. I have also been long concerned with how often people enter into our field with a research agenda highly influenced by strong a priori ideological commitments. This is, I have thought, partly because of our immaturity as a field, but partly because many of our research areas touch on core assumptions about the nature of being human.

I also appreciate Stein et al's concerns on how to put together the multiple perspectives on psychiatric disorders. It is too easy for our highly specialized researchers to dig deeply into their own corner of the etiologic space of psychiatric illness for an entire career and never look up. In fact, our funding system encourages such specialization. I favor a brand of “pluralism” not discussed by Stein et al: “integrative pluralism” ⁴ . This brand suggests that, every once in a while, it is incumbent on a scientist to come out of his/her hole and spend some energy trying to integrate his/her findings with at least those of adjacent perspectives.

A comment on reduction is warranted here. I do not consider it a dirty word. Indeed, the increasing merger of molecular genetics and molecular neuroscience to uncover risk pathways from genetic variants to psychiatric disorders is among the most exciting in our field. More power (and funding) to them would be advisable. My objection is to those reductionists who argue that their approach is the only way. “Reductive hubris” has been harmful to us at several historical phases of our discipline. Certainly, all research pathways for psychiatric disorders are not equal. But psychological and social interventions have proved their value for a number of our disorders, bringing me back to the arguments above about multi‐causality.

In contrast to the above questions – which exist largely within the philosophy of science – the mind‐body problem is fundamentally a metaphysical one. While innately fascinating, the insights gained compared to the effort needed to wend one's way through the metaphysical thickets in this area have, for me, been disappointing over the years. I will therefore only make a few brief and personal comments. Philosophers ponder the mind‐body problem. Psychiatrists live it. The best metaphor I have heard to describe many psychiatrist‐patient encounters, where the good clinician has to switch back and forth between seeing his/her patients as minded and brained, is “binocularity” ⁵ . As we can see depth in the world through our two eyes, we see more deeply into our patients by seeing them through two different lens – brained and minded.

I am a non‐reductive materialist. I am not sure that this is a coherent position, but it is the best I can come up with. It means roughly that I do not, for my scientific work, assume that mind can exist independent from brain. Indeed, my mind is instantiated in my brain. But the mind‐brain system is an interactive one – although I cannot explain how that works. Mind is a level central to the psychiatric profession. The concept of a mind‐less psychiatry is oxymoronic to me. Indeed, I suggest that the tensions between mind‐based and brain‐based perspectives have been definitional for our history ⁶ . And I believe in top‐down causation. Important things “happen” at the level of mind that are sometimes of great psychiatric significance ⁷ .

In conclusion, it is heart‐warming to see a prominent psychiatric journal give space to this wide‐ranging and thoughtful essay. The nature of psychiatry is such that all researchers and clinicians bring to their work a range of philosophical assumptions. The only question is whether, at some point in their career, they take time and energy to examine some of them and decide whether any of them might need revision.