Henrik Walter

Research Division of Mind and Brain, Department of Psychiatry and Psychotherapy, Charité Universitaetsmedizin Berlin, Berlin, Germany

2013 Sep 5.


Modern psychiatry has taken a lot of effort to make the description of psychopathology reliable by introducing standardized ways of exploring, describing and rating psychopathological patterns over time.

In America, psychiatric disorders are diagnosed using the DSM-IV (published 1994), the Diagnostic Statistics and Manuals of Mental Disorders, the official handbook of the American Psychiatric Association (APA), sometimes referred to as the “bible” of psychiatry. According to DSM-IV mental disorders are diagnosed by carefully checking if subjects fulfill a certain number of psychopathological criteria for a certain amount of time. DSM-IV is agnostic on etiopathogenesis, i.e., the causal genesis of disorders, but rather has put emphasis of establishing a reliable, intersubjective schema for diagnoses on the psychopathological level. But what about validity, i.e., what is measured or rather intended to be measured with DSM-criteria? What kind of things are mental disorders? Kendler et al. (2011a) have distinguished four types of kinds that mental disorders could be.

Essentialist kinds are based on an essence, e.g., an underlying cause, from which the defining features (the typical symptoms) do arise. Although this theory fits to some cases like progressive paralysis in syphilis or Mendelian defects in cholesterol metabolisms, it is now widely acknowledged that this model neither fits most chronic diseases as atherosclerosis, hypertension or autoimmune disease, nor psychiatric disorders. Rather, it is generally accepted that psychiatric disorders arise from a multitude of causes that are probabilistically related to signs and symptoms. Even in cases, where family and twin studies unambiguously have demonstrated that most of the variance is explained by genetics factors (e.g., up to 80% in schizophrenia) there is no single gene causing this disorder. Recently discovered risk variants explain only a tiny portion of variance, usually less than 1%, although, using imaging genetics, they can be shown to have much stronger effects on the brain level (Walter et al., 2011).

A second approach is to understand psychiatric disorders as socially constructed kinds which are brought about solely by the human activity of describing and classifying but not by an underlying structure independent of human constructs. Although this still is a popular thesis in the camps of cultural relativists and anti-psychiatry, this theory is rarely taken serious today. Instead, it is now widely acknowledged that cultural influences and social factors play important roles in the expression of symptoms, e.g., in the content of delusions. But it is also clear that for certain prototypical diseases (e.g., schizophrenia, bipolar disorder, depression, and some anxiety disorders) there are invariant patterns in experience and behavior despite eminent cultural differences. Many people think that what matters most is how we handle mental problems. So maybe psychiatric disorders are best understood as practical kinds. This approach holds that psychiatric disorders do not carve nature at its joints but just are those kinds which are most useful for certain purposes, ranging from medical ones (diagnoses and treatment) to sociological or even political ones (this is the point of departure of anti-psychiatry). This model is grounded in pragmatist philosophy and instrumentalism and has some appeal. In fact, the philosophy of DSM is very close to this approach with its agnostic and atheoretical framework. Although the practical kind of view avoids metaphysical discussion (like: What is schizophrenia really?), it gives us no advice as how classifications should be build and goes against many realistic intuitions that are the basis of successful applications not only in medicine. Instead, Kendler et al. (2011a) argue for a concept that is based on a model originating in the philosophy of biology, dealing with the problem how species are classified and on recent developments in the theory of neuroscience: the concept of mechanistic property clusters (MPC). According to this view, the items to be classified rest on properties that need not to be shared by all members of a class, rather they should be understood as a cluster within an abstract space of features or properties in a multidimensional space. Some of those features may be more essence like, some more practical. Importantly, the MPC-view encourages the thought that there are robust explanatory structures to be discovered underlying psychiatric disorders.

These explanatory multidimensional structures (genes, cell receptors, neural systems, psychological states, environmental inputs, social-cultural variables) are interacting in a complex and intertwined way, are sometimes fuzzy, but nevertheless stable. Importantly, it cannot simply be read from the superficial structure of items if they belong to the same kind. Rather, their membership is explained by the causal mechanisms that regularly ensure that their properties are instantiated together (a historical account). The interaction typically is inter-level, but can also be on the level on symptoms, thus mutually re-enforcing the pattern, e.g., in depression insomnia predisposes to tiredness and guilt predisposes to suicidal ideation. As MPC kinds are defined in part by the mechanisms that underlie and sustain them, the reductionist intuitions of old wave biological psychiatry are partially satisfied. However, the kind cannot be fully explained and thus understood if inter-level interactions, which are often hidden to the subject as well as to the external observer, are not taken into account. For example, it has been empirically shown that subjective explanations for depressive episodes by patients do not correlate with objective risk factors for depression (Kendler et al., 2011b) – a finding that makes it likely that explanations based on just a selection of levels (subjective experience and remembered behavioral events) do not explain depression well. The same can be said for simplified biological models of depression as a neurotransmitter deficit that ignores many of the other levels. Although the MPC-model does not tell us in advance what the relevant causal mechanisms are, it is consistent with the new biological wave in psychiatry which we will now characterize by describing a controversy around the introduction of DSM-5.

%d bloggers like this: