Epistemological error and the illusion of phenomenological continuity

van Os and Reininghaus’ paper on the transdiagnostic “extended psychosis phenotype” attempts to present an exhaustive framework for the nosology and pathogenesis of psychiatric and especially psychotic disorders1. We have there the genes, gene‐environment interactions, an emphasis on the role of childhood trauma (resurrected after a period of skepticism about psychoanalytic theories and reluctance to ascribe independent causal role to retrospectively ascertained events), and a dimensional approach to phenotypic manifestations. There is also a theory of symptomatic dimensions and their combinations to yield a few categorical entities of “psychotic disorders”.

Since the paper aspires to break a new ground and radiates an air of recency and novelty, it seems relevant to mention that a somewhat similar dimensional approach was tried out on a sample of psychiatric inpatients already around the time of World War II2. Those hand‐made calculations, prior to factor analysis, revealed three main psychotic “dimensions”: the paranoid (positive), the heboid (disorganized) and the schizoid (negative). The dimensional approach was then pursued through the creative contributions of P. Meehl and the scales by the Chapmans3. Despite some useful information, this line of research has not resulted in a radically new understanding of mental disorders.

The multiple scales on subthreshold psychotic symptoms, applied in the studies to which the authors refer, are not a product of original research into the life‐world of psychosis but rather a reflection‐based attenuation of DSM criteria for schizophrenia – formulated at a very high chronicity level – in order to be applicable to young, first‐contact patients with a schizophrenia spectrum disorder4. Such simplification is certainly amplified by conceptual ambiguities. For example, the authors seem to use the notions of “weak expressions of positive psychotic symptoms”, “psychotic experiences”, and “psychotic symptoms” more or less equivalently, without explaining their relations or, more basically, what makes a symptom “psychotic”5.

The fundamental problem is of an epistemological kind. Since the creation of DSM‐III, the symptom is considered a thing‐like object, existing in itself, i.e. independently of other symptoms, larger Gestalts, and structures of consciousness. For example, a phenomenon of self‐reference is considered as such independently of whether it is caused by melancholic guilt feelings, insecurity after a panic attack, or a sense of being the center of the universe in incipient schizophrenia. This simplification of the concept of symptom, with a complete absence of holistic, contextual and gestaltic considerations, has contributed to a situation in which diagnoses are assigned on the basis of accidental recombination of criteria, with a neglect of differential‐diagnostic considerations6. This is well illustrated by the authors’ reference to common mental disorders like anxiety with psychotic experiences as anteceding a full‐fledged psychosis. These patients already having psychotic experiences should have not fallen into the category of “common mental disorders” in the very first place, and their anxiety may be fundamentally different from “common” anxiety.

The fundamental epistemological problem of operationalism results in a homogenization, trivialization, and non‐specificity of mental symptoms, which invites an illusion of “phenomenological continuity”. Feeling that other people stare at one, because one is the center of the universe, is not the same as feeling that others have noticed one's panic attack. In a very important work, Stanghellini et al7 demonstrated that “hallucinatory experiences” in a non‐clinical population are qualitatively different from hallucinations in schizophrenia (see also Henriksen et al8). Similarly, Schultze‐Lutter et al9 documented that self‐reported “psychotic‐like experiences” are simply uncorrelated with the clinician‐assessed “attenuated psychotic symptoms”.

Another example of the metaphysical reification of symptoms, implicit in van Os and Reininghaus’ paper, is their claim that, under pressure of traumatic experiences, hallucinations and delusions amplify each other because of increased “connectivity” of symptoms (presumably a connectivity between the networks responsible for single symptoms). We are not offered any psychological or phenomenological considerations of higher‐level interactions between the psychotic phenomena.

Blankenburg10 emphasizes that many patients with a schizophrenia spectrum disorder initially present with vague or unspecific complaints, for instance of fatigue, feeling unmotivated or having problems with occupational performance. Through a phenomenological interview, he demonstrates that these seemingly “non‐specific” complaints often are shorthand for much more “specific” ones. For example, one patient reports: “The situation is that I do not feel a genuine drive. I always come so rapidly to the dead point. My patience is then almost run out…”10. He feels “exhausted”; everything “gets on his nerves”10. While these complaints may seem “non‐specific”, the patient's further utterances testify to their embeddedness in a much more recognizable, “specific” clinical Gestalt of schizophrenia spectrum: he complaints of “lacking distance to his surroundings”, of only perceiving “the front” of things, and “of a failing approach to everyday life, to the reality of ordinary life”10. Consequently, even the simplest task is felt as a burden, requiring massive cognitive efforts on his behalf, and partly so because he is unable to take for granted what others consider obvious or self‐evident (i.e., “lack of common sense” and hyper‐reflectivity)11.

Early diagnostic assessment requires not only a superficial symptomatic screening but also an insight into the life‐world of the patient, implying considerable psychopathological knowledge. Moreover, it is unwarranted to perceive a symptom (e.g., a “psychotic experience”) in abstraction from other symptoms, larger Gestalts, and structures of consciousness. We are certainly able to build up scales trivializing symptoms into phenomenological continua, but in this move the symptoms are emptied of their clinical validity.

Josef Parnas, Mads Gram Henriksen
Faculty of Health and Medical Sciences & Faculty of Humanities, University of Copenhagen, Copenhagen, Denmark