Refining the Diagnostic Criteria for Schizophrenia: An Infinite Task

Moritz et al¹ present an informative picture of the vicissitudes of first rank symptoms throughout the evolution of DSM from II-5. Moreover, they suggest a Schneider-inspired redefinition of auditory verbal hallucinations with a relaxation of the demand that such hallucinations have a clear perceptual quality. Reading this thought-provoking paper leads us to several questions, which are most likely interdependent: What is schizophrenia? Do we have any data on the relative validity of the different DSM versions of the definition of schizophrenia? And if so, what does this validiity imply? Is it family history, genomic, prognosis, treatment response, research utility. . . . To our knowledge, such studies concerning specifically the DSM definitions do not exist, but there are poly-diagnostic studies summarized by Jansson and Parnas² in 2007. No univocal conclusion can be drawn from these studies about the superiority of one system vs the others. In 1998, Maj³ published a critical analysis of the DSM-IV criteria for schizophrenia, an analysis that may be also applicable to other definitions of DSM, including DSM-5. His contribution is very instructive in that he demonstrates the drawbacks of each criterion and concludes that schizophrenia is mainly an exclusion diagnosis (ie, nonorganic, nonaffective, etc.). The DSM definitions lack a statement on the quiddity or whatness of schizophrenia. All founders of the schizophrenia concept had a certain shared sense of the entire Gestalt of the illness, characterized by a profound alteration of mental life, which they tried to convey through their writings. Gestalt is a German word for figure or form implying in psychopathology a certain configuration of symptoms, signs, behaviors, life-history, social, interpersonal, and occupational functioning. The Gestalt is not something ephemeral but emerges from the clinical reality of single features. Such a whole is evidently more than a sum of its phenomenal aspects and can therefore not be reduced to the latter. The Gestalt confers on a single symptom its diagnostic specificity. FRS can only be assessed in the context of the entire Gestalt of the illness, which necessitates a narrative conversational interview allowing for faith full self-descriptions. We can mention a patient of ours, who replied no to the questions concerning external influence, but later during the conversation mentioned that he had a difficulty biking in the city because emanations from electrical cables compromised his decision making in the traffic. Most of the literature evaluates the presence of FRS in different populations by using structured questionnaires. Schneider⁴ himself never tested the frequency of his symptoms by letting a research assistant perform structured telephone interviews with clinical and nonclinical populations. By the way, he probably never dreamed about performing a structured clinical interview. Schneider and his contemporaries would never look at a symptom as an isolated spatiotemporal phenomenon objectively describable. The symptom in isolation would not make any sense. A symptom had to be evaluated upon its context and the state of mind. However, the notion of state of mind requires some form of theoretical grasp of the life of consciousness or as it was termed at Schneider’s era “the whole personality.” Thus, Schneider prepared his ready to use list of symptoms for general practitioners whereas he certainly expected that the psychiatrist was familiar with a phenomenological description of mental life. He probably assumed that the psychiatrists were familiar with his mentor Karl Jaspers’ treaties of General Psychopathology.⁵ The phenomena such as thought bloc, thought intrusion, thought interference, and thought aloud were considered as FRS upon a narrative, which indicated a background of consciousness with a characteristically disturbed structure. If the patient reports loud thinking being localized to the right part of the brain, we would ascribe a specificity to the symptom because it is associated with a profound alteration of the structure of conscious life (spatialization). The delusional elaboration of such phenomena (ie, the so-called secondary delusions) was frequent, but not necessary for their FRS status. According to Jaspers and Schneider, the specificity of delusions in schizophrenia (the so-called primary delusions) was determined by their revelatory and noninferential character. Schneider included only delusional perception as a FRS because of its clinical recognizability. However, the notion of primary delusion did not enter into the DSM.

The DSM-III to 5 operate in another epistemological framework. Symptoms are considered as objectively determinable in isolation with a strong behavioristic bias. The successive definitions of DSM attempt to improve the single symptoms and their combination in order to arrive at the high specificity and reliability of diagnosis. It seems to be believed that adding 1 symptom, removing another and modifying the third would lead to an overall better definition without any attention to the gestaltic essence of the disorder. Allocating a diagnosis on the basis of a meticulous weighing of single criteria provides a comforting, but illusory sense of scientific exactitude and may miss the target.

Moritz et al¹ suggest a timely modification of the future definition of auditory verbal hallucinations. They quote several studies showing that hallucinations in schizophrenia do not always exhibit a truly perceptual character. We fully endorse this suggestion, which takes into account the clinical reality of hallucinatory experience. However, in our view, the authors seem to fall into the trap of believing that a symptom may be more adequately defined by sculpting its features as if it was an isolated quasi-objective phenomenon. We think that the essence of hallucinations in schizophrenia is linked to self-fragmentation. A part of the subject becomes an object. Thus, the patient senses some of these fragments as another presence. This alien presence may become a hallucination or a passivity experience.⁶ One of our patients said that he no longer heard voices but they “were still there.”⁷

Moritz et al¹ propose the definition of hallucinations as involving a complementarity between the degree of sensoriality and the delusional conviction (lack of insight). They make several appeals to the metacognitive deficits. Given the metacognitive biases even a weak “sensory irritation” becomes psychotically elaborated to the status of hallucination. However, introducing the notion of sensory irritation appears to us as a move from a descriptive to a physiological level. “Sensory irritation” is a causal hypothesis rather than a clinical datum.

We agree with the authors’ emphasis of the lack of insight in assessing hallucinations. However, we have another perspective on this phenomenon. Many, if not the majority of patients with schizophrenia operate under the condition of the so-called double bookkeeping.⁸ It means that the patient for the most part can separate the shared social reality from his private immanent psychotic world. This private immanent sphere is often felt by the patient as an extension of supra-natural dimension hidden to other people. A patient may agree with the clinician about the unreality of his hallucinations while fully preserving his conviction of the ultimate truth of his hallucinatory experiences. This truth has nothing to do with the objective shared world, but is an incontestable first personal affection. Thus, this conviction is not a product of a faulty inference or other metacognitive deficit, but arises in the midst of the patient’s subjectivity as a revelation. The phenomenon of double bookkeeping makes the assessment of the insight much more complicated and requires again a grasping of the patient’s experience as an aspect of the Gestalt. The contribution of Moritz et al¹ brings forth important issues in the diagnostic demarcation of schizophrenia, which deserves further conceptual discussion and adequate empirical research.