Are the neurocognitive correlates of subtle subjective symptoms the way forward in psychiatry?

Abstract

In their thoughtful overview, Sass and Byrom (1) correlate three levels of description: the subjective experience of the patient as understood by a third person (phenomenology), cognitive models and neuroscience. In particular, they link two diverging models with each other, i.e., selected phenomenological accounts of delusions and their potential underlying neurocognitive explications. They identify correlations between the two models and crystallize neglected areas in neurocognitive accounts, filling them with phenomenological notions. Basically, they argue to look for fine grained, subjective symptoms – beyond structured interviews and DSM/ICD criteria – and to relate them to neurocognitive data. The authors’ ambition is to find a neuroscientific explanation of subjective experiences. However, neuroscience should not be a validator for phenomenology, as they are two different realms of description, which should inform each other. There is of course a fundamental “explanatory gap”. The subjective consciousness (first person perspective) has to be translated by the patient into a second person “output”, usually verbal and non-verbal; this has to be understood by the interviewer and then related to brain states (third person perspective). The mind-body relation is a yet unresolved philosophical question; therefore we can only correlate such levels of descriptions, but we cannot explain them with one another. Correlating symptoms with brain states has already been successfully achieved by a neuroimaging “symptom catching” approach, mainly for auditory hallucinations and formal thought disorder. In the beginning, cruder syndromes have been related to brain activation (2,3), while recently a more fine grained and subjective psychopathology has been applied (4,5). In the article, highly selected phenomenological ideas (models) are correlated with current neuroscience models. Since there is not much replicated data on the neurobiology of delusions, Sass and Byrom's account is rather a comparison of models. Ideally, one would have a phenomenologically derived symptom or experience and relate this to empirical data, rather than comparing models with each other. More empirically testable predictions from phenomenology would be welcome. However, this begs the question of which among the many phenomenological accounts (Sass and Byrom only discuss a few) should be translated into experiments and tested. For current neuroscientific data acquisition, a fluctuating phenomenon/symptom is usually considered, and the patient is investigated experimentally at two or more points in time, with symptom severity being correlated with cognitive measures and/or brain states. This methodological constraint vastly reduces the number of phenomena possibly investigated. A further current problem is that the conception of our patients’ inner lives has been oversimplified by the operationalized diagnostic approaches. This oversimplification has been accompanied by a reliance on methodologies that are unable to capture the delicate forms of human experience and expression (6). If the cognitive scientist wants to base experiments on phenomenological notions, he or she must be deeply familiar with the philosophical-phenomenological literature. This needs thorough reading and understanding of the ideas, which are currently marginalized. Many French and German texts have not been translated into English and are therefore inaccessible to a wider international audience. For example, Sass and Byrom discuss concepts by K. Conrad, but another important phenomenological contribution on chronic, negative symptoms stems from W. Blankenburg (7), both from Marburg. Their two main books have equally not been translated into English. The clinician should be familiar with the rich descriptive and hermeneutic tradition of psychiatry. We only see in our patients what we know. There are other levels for the explanation of experiences than cognitive psychology and neuroscience, and a deep understanding of the patient's world is interesting and therapeutic in itself. A phenomenologically informed systems neuroscience, integrating environmental, genetic, molecular and brain imaging data in the same patient across his life, will eventually explain the etiology and pathophysiology of mental disorders. It can, however, not replace an interpersonally shared understanding of the patient's inner life as a humane value and source of content in itself.