Ewa Grzeszczak
Affiliation: Jagiellonian University
Category: Philosophy
Keywords: natural kinds, Homeostatic Property Clusters (HPC), philosophy of psychiatry
Date: Wednesday 3rd of September
Time: 16:00
Location: GSSR Plenary Hall (268)
View the full session: Psychotherapy & Psychiatry
The question of natural kinds, in its broadest and classical form, asks about categories that are non-arbitrarily defined. In addition to the objectivity of the categories, natural kinds are also thought to provide explanation and prediction, which are key features of scientific inquiry. That is why the presence of natural kinds in a field is often tied to its status as scientific, so the question of natural kinds in psychiatry is strongly connected to the debate on psychiatry's medical status and its alignment with the medical model. A currently popular conception of natural kinds in psychiatry is the theory of Homeostatic Property Clusters (HPC), which understands natural kinds as clusters of properties sustained by mechanisms (first proposed by Boyd, e.g., 1991, 1999; for applications to psychiatry, see e.g., Samuels 2009, Beebee & Sabbarton-Leary 2010, Kendler, Zachar & Craver 2011, Tsou 2016). One reason for its popularity is that HPC allows for the inclusion of many properties from different levels – including neurobiological and psychological – within a single cluster. However, this undeniable advantage of multi-levelness has a flip side: the relationship between neurobiological and psychological properties must be clarified, and the definitions of key concepts crucial for HPC, such as causality and mechanism, must be adjusted accordingly. I aim to show that such adjustments result in HPC applying to psychiatry only in a weakened version, which has significant drawbacks. Firstly, applying HPC to psychiatry only allows for mental disorder to form a whole in a trivial, deflationary sense, and thus fails to provide substantial explanations of it as a unified neurobiological-psychological phenomenon. Secondly, applying HPC to psychiatry forces a global commitment to a weak theory of natural kinds, which seems to contradict the motivations behind the search for natural kinds in psychiatry. My argument is based on demonstrating that psychological properties are incompatible with the requirements of a strongly understood HPC. I first discuss how the psychological level requires giving up a productive understanding of causality in favor of interventionism. This shift, in turn, raises the question of how interventionist HPC relates to the role and definition of mechanism. To address this question, I distinguish between a substantive mechanistic explanation and a deflationary mechanistic explanation. The former increases understanding of phenomena in terms of 'how' and can serve as the basis for epistemically substantive explanations. The latter gives up understanding and epistemically substantive explanation in favor of projectibility limited to the mere co-occurrence of properties (so-called bare projectibilism). I argue that although a substantive mechanistic explanation may, in general, be compatible with interventionism, the specific relationship between the psychological and neurobiological levels makes this compatibility untenable. In other words, I claim that only the deflationary mechanistic explanation is possible between the neurobiological and psychological levels in the HPC cluster of mental disorder. What are the implications of this? It raises the central question of whether applying HPC to psychiatry is worth doing. First, if the causal link between biological and psychological properties can only be deflationary rather than substantive, this means that the explanatory value of natural kinds in psychiatry is diminished. After all, one of the reasons we care about natural kinds is that they allow us to categorize – dividing the world into members and non-members of a given natural kind – and then use these categories to provide explanations. However, if the psychological level is merely "attached" to the rest through a deflationary mechanism, then we cannot offer interesting, substantive explanations about the whole cluster, the whole category. Substantive explanations do indeed exist within some parts of the cluster, but due to this deflationary "gap" between biological and psychological levels, we cannot extend those substantive explanations to the whole cluster. Second, an interpretation of HPC that accepts deflationary mechanisms necessitates a broader acceptance of weakly understood natural kinds not only within psychiatry but across other domains as well. If we agree that a connection through a deflationary mechanism is sufficient to classify something as an HPC kind, then we significantly lower the threshold for recognizing natural kinds. If a substantive mechanism is no longer required to secure HPC status, then HPC kinds become exceedingly common in the world – any two or more properties that merely co-vary due to intervention would suffice to constitute a natural kind. Thus, applying HPC to psychiatry entails accepting a specific theory of natural kinds – namely, one that postulates an abundance of kinds and undermines the special status of those posited by sciences such as biology or medicine (for example, a theory akin to that proposed by Khalidi 2013). Importantly, a significant part of the motivation for natural kinds in psychiatry was the conviction that the presence of natural kinds ensures the scientific status of the discipline – directly linking psychiatry to medicine. However, if the application of HPC to psychiatry is only possible at the price of accepting that natural kinds no longer guarantee scientific distinctiveness, a serious doubt arises as to whether the whole endeavor was worthwhile. In summary, the application of HPC to psychiatry allows the thesis of natural kinds in psychiatry to be defended. However, at the same time, both the broader significance of natural kinds and their presence in psychiatry become less impactful – they do not provide substantive explanations of categories and are no longer confined to scientific domains. I believe that, for at least some philosophers advocating for the application of HPC to psychiatry, this may represent an unnoticed and too high a price to pay.