Fueled by the latest revision of the Diagnostic and Statistical Manual of Mental Disorders (APA 2013, DSM-5), psychiatric classification and the research based upon it have received an enormous amount of critical attention lately. Jeffrey Poland’s and Serife Tekin’s edited volume brings together different perspectives from philosophy and mental health research on these controversies. Surely, a certain amount of debate about a classification of mental disorders is expectable due to the complexity of the subject matter, as well as the impacts on science and clinical practice. Yet, the editors compellingly argue in the introduction that the current situation marks a more foundational problem. They interpret the abundance of criticisms of the DSM in terms of a paradigm in crisis, indicated by a multitude of different approaches, discussions about basic metaphysical and methodological assumptions, and the development of alternatives such as the Research Domain Criteria Project (RDoC) of the National Institute of Mental Health. The application of this Kuhnian framework sets the ground for a valuable collection of 12 articles that intertwines philosophical perspectives with more concrete problems of research on psychopathology. The chronology of the chapters (more or less) follows a path from general to more specific issues.

To start with, Edouard Machery addresses the epistemological question of how to determine whether psychiatric syndromes are taxons rather than tails of distributions of traits found throughout the general population. After disentangling common misconceptions about this difference (e.g., construing it as being about the question of biological causes), he compares informal approaches (such as clinical judgment) with more formal, statistical ones (such as cluster analysis or MAXCOV). While arguing that formal approaches have a greater potential to distinguish taxons from tails, he also points out the need to pay close attention to their underlying ontological assumptions.

Robyn Bluhm traces part of the current crisis back to tensions between two very influential contemporary strands: biological psychiatry and evidence-based medicine (or evidence-based psychiatry, EBP). Whereas EBP stands in a tradition of medical empiricism emphasizing careful observation and description, biological psychiatry embodies a more rationalist perspective that aims at causal explanation. One virtue of her article is that it applies this general distinction between empiricism and rationalism to the RDoC. She argues that the RDoC’s aim to understand mechanisms of psychopathologies should be integrated with a more empiricist outlook in order not to lose the connection to clinical relevance.

The RDoC is also discussed in detail in the following two contributions. Ginger A. Hoffmann and Peter Zachar start with a helpful distinction between “therapeutic impact” and “reflective impact” and examine how the RDoC fares regarding these two. First, therapeutic impact is assumed to benefit from an improvement in validity (a lack of validity often being considered the DSM’s fatal flaw). Importantly, the authors point out a decisive difference between the DSM’s ultimate aim of etio-pathophysiological validity of disease entities and the RDoC’s focus on factorial validity of dimensional constructs. They then argue that the path from factorial validity to etio-pathophysiological validity (and ultimately therapeutic impact) is beset with a number of difficulties: The more fine-grained the RDoC constructs become (in order to achieve factorial validity), the more difficult it gets to connect these lower-level insights to the higher-level phenomena dealt with in clinical practice and therapy. Yet, this translational difficulty could be beneficial regarding reflective impact (i.e., how a diagnosis affects patients’ self-perceptions), since it would prevent harmful internalizations of diagnostic labels. This question of how and whether the RDoC could alter patients’ self-understandings is certainly a very interesting one deserving further exploration.

In a careful examination of the applicability of the Kuhnian framework to mental health research, Claire Pouncey contradicts the common characterization of the move from DSM to RDoC as a paradigm shift. Rather than a Kuhnian revolution, she analyses the shift as a further development within a period of normal science perpetuating shared methodological and theoretical assumptions.

The following four chapters each focus on a different approach to answer the current crisis. Jeffrey Poland and Michael Frank start by pointing out central problems with the DSM and argue that we do witness a Kuhnian crisis that is characterized by debates on fundamental issues and the exploration of alternatives. Regarding the task to develop a better alternative, they provide a helpful identification of different kinds of challenges that are part of such a task: ideological/conceptual challenges, methodological challenges, clinical challenges, and transitional challenges. They then outline how computational psychiatry, by an application of computational cognitive neuroscience, could face a number of those challenges. In doing so, they make core substantive and methodological assumptions of the computational approach explicit.

Aaron Kostko and John Bickle present personalized psychiatry, which takes into account the details and context of individual cases, as another response to the crisis. At the beginning, they proclaim a tension between the requirement of personalized psychiatry to pay attention to particulars and the quest for scientific legitimacy. Such legitimacy, according to the authors, depends on achieving generalizable knowledge on causal mechanisms. Unfortunately, it is left undiscussed why generalizable mechanistic evidence should be taken as criterion of scientific legitimacy. A lack of such results may be unfortunate without necessarily threatening the scientific legitimacy of research that employs established, well-controlled methods, for instance. In any case, they interpret the named tension as due to a question about the relevance of different causal factors and, by extension, the adequacy of different accounts of causation. Based on a number of case-studies, they compare the potential of Woodward’s interventionist account and Silva, Landreth, and Bickles’ metascientific account of causation to serve a scientifically legitimate personalized psychiatry; that is, to integrate different causal variables from different levels.

Kelso Cratsley discusses how a shift towards mechanistic explanation could be beneficial with regards to the aim of a sound etiological classification. He understands mechanistic explanation as being “committed to the identification of components making up the working of the mind, characterized at complementary levels of explanation” (163). After carefully developing a broad concept of mechanistic explanation and situating it within the current discussion, he identifies a variety of challenges for an account of mechanistic explanation special to psychiatry, such as an instability of surface features of mental disorders, the causal relevance of environmental factors, or the complex organization of underlying systems. Cratsley argues that these challenges can be met by a sufficiently broad account of mechanism and mechanistic explanation. Once again, the integration of different levels is demonstrated to be a key challenge in current mental health research.

Rachel Cooper focuses on an additional dimension of the current crisis: the loss of trust in psychiatric research by patients. She locates this loss in a wider crisis of public trust in science, fuelled by knowledge about industry-funding and resulting bias. While this is not a crisis in the Kuhnian sense — an accumulation of anomalies the current paradigm cannot successfully deal with — it is nonetheless a very important problem, especially in an applied science such as psychiatry. As part of a solution, Cooper proposes an increase in user-led research (or citizen science). She argues that such research can restore trust and be epistemically beneficial, since it is able to examine research questions the industry is not interested in funding, and it can provide a helpful diversity of perspectives. She also addresses obvious worries about the reliability of such research and co-optation by industry. Applied to classification, she points out that different classifications facilitate different types of research by being more or less accessible to amateurs with limited resources; in this respect at least, the DSM classification may be more helpful then the RDoC.

The last four chapters focus on more specific issues. Richard P. Bentall argues that the familiar biomedical conception of schizophrenia shows the strengths of a reigning paradigm, as the concept endures despite accumulating counter-evidence. In particular, he sets out to expose six widespread myths about schizophrenia as at least questionable in the light of current data: the conviction that schizophrenia is a reliable concept (1); the idea of a clear distinction between schizophrenia and healthy functioning (2) as well other psychiatric disorders (3); the conception of schizophrenia as a predominantly genetic disease (4); the causal irrelevance of environmental factors (5), and the understanding of schizophrenia as a brain disease (6). It would have been interesting to relate this discussion more explicitly to the DSM-5’s shift to a schizophrenia-spectrum and its potential to overcome at least some of these worries.

Serife Tekin also starts her chapter by asserting a crisis in DSM-led research that stems at least in part from problems with the DSM’s diagnostic criteria. For example, phenomenology-oriented critics argue that the DSM’s criteria for schizophrenia fail to represent the clinical reality of the disorder. In particular, this holds for distinctive anomalies in self-experience frequently reported by patients (i.e., a loss of presence in one’s first-person experience). As Tekin points out, this symptom concerns an immediate, noninferential self-experience^subjecthood, not a reflective self-experience^objecthood. In an examination of a phenomenology-neuroscience partnership in schizophrenia research (PNP), which aims to integrate these very different approaches to mental disorders, she shows that there is a problem of “wandering terminology”: The anomalies of self-experience linked to neurological findings by PNP refer to self-experience^objecthood rather than self-experience^subjecthood. Her chapter clearly demonstrates the difficulties in integrating different approaches and paradigms, as well as the importance to pay close attention to the respective details and context of concepts and methods.

Questioning the applicability of certain diagnostic categories to children, Harold Kincaid starts by making a convincing argument for a case-by-case assessment of diagnostic validity, instead of rejecting the DSM classification on the whole. He conceptualizes the validity of diagnoses as their ability to pick out objective groupings in nature, which can be supported by a convergence of different independent measures and the embedding of categories in a nomological net. He then argues that the diagnostic categories of Bipolar Disorder and Major Depressive Disorder fare pretty well in this hindsight — applied to adults. By contrast, the application of these diagnoses to young children not only lacks evidential support from taxometric studies, but also displays many instances of scientific misconduct: Empirical results are blatantly misinterpreted, claims are based on unpublished data, or diagnostic criteria are changed as fits. It is not a far stretch to conclude from this that social factors and financial interests play a significant role here.

The volume finishes with a thought-provoking chapter by Owen Flanagan and George Graham. The authors criticize the DSM for medicalizing normal behavior due to a questionable presumed link of sanity and truth. They argue that mental states such as illusions, delusions, or hallucinations are not in general pathological, but can even play supportive roles for mental health — depending on their content. For instance, they claim people with an overly optimistic assessment of their own talents to be generally “happier and nicer” (one is left wondering about empirical evidence especially for the latter point). Relatedly, they consider religious beliefs as epistemically similar to other delusional beliefs– the important distinctive feature here being whether such beliefs are subjectively and interpersonally harmful. “Metaphysical” hallucinations, they argue, are also not generally unhealthy — even healthy people sometimes seek to make such experiences by inducing them via drugs or meditation due to their suspected positive impact. However, such self-induced hallucinations may not be comparable to the involuntary experience of hallucinations in psychosis. Moreover, the DSM in general assumes mental disorders to involve distress and disability, not just the present of certain mental states — a fact it would have been valuable to discuss a bit further here.

In general, the volume provides a worthwhile and inspiring reading for philosophers as well as mental health researchers. Its strong point is the intertwining of concrete issues in mental health research and classification with a more general philosophical analysis enabled by the application of a Kuhnian framework. The resulting picture though seems to point beyond this framework, insofar as it repeatedly shows the difficulties in integrating different levels of explanation, different approaches to certain phenomena, and the multiplicity of new research programs arising. It thereby puts forward the question of whether the emergence of a new unified paradigm is really possible and expectable, or whether it is time to embrace a more pluralistic approach to research and classification in psychiatry.

 

 

References

American Psychiatric Association (APA) (2013). Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition: DSM-5. Washington, DC: APA.

 

© 2017 Anke Büter

 

Anke Büter, Institut für Philosophie, Leibniz Universität Hannover