Self and schizophrenia: current status and diagnostic implications

The notion of self‐disorders in schizophrenia emerged in contemporary psychiatry at the beginning of this millennium1. It was considered an unorthodox novelty, and neither the DSM‐IV nor the DSM‐5 contains a reference to disordered self in the schizophrenia spectrum.

However, that notion was historically co‐constitutive of the concept of schizophrenia. Bleuler2 listed experiential “ego‐disorders” among the fundamental symptoms of schizophrenia and reported patients complaining of being only “reflections of themselves”, unable to “catch up with themselves” or having “lost their individual self”. All classic texts on schizophrenia contain a reference to disordered self1. The concept of “disintegration”, widely used in psychiatry and psychoanalysis, makes only sense in the presence of some notion of self that is at stake.

The DSM‐III glossary of terms linked disturbance in the “sense of self” to schizophrenia, and the ICD‐9 definition of schizophrenia referred to disturbance of fundamental features of personality (e.g., uniqueness and autonomy), whereas in the ICD‐10 the term “personality” was removed. The disappearance of “disordered self” was perhaps linked to the simplification of fundamental symptoms into the so‐called “four A's” (autism, ambivalence, association and affect disorders) and a difficulty with conceptualizing the notion of autism.

What kind of self is disordered in the schizophrenia spectrum conditions? It is useful to follow a distinction of contemporary philosophy of mind and phenomenology between the so‐called “narrative self” and the “core self”.

The narrative self refers to features which characterize and individualize a person and which easily lend themselves to linguistic self‐description (e.g., “I have a tendency to act impulsively”) and descriptions from the third‐person perspective (“she is acting impulsively”). These features comprise biographical, characterological and cognitive characteristics and are heavily dependent on language and memory.

The notion of core self refers, instead, to the first‐person perspective which is an intrinsic structural feature of all experience and which provides us with an immediate or pre‐reflective sense of subjectivity and self‐familiarity as an “I‐me‐myself”. This can be extended to comprise a sense of temporal persistency, self‐coincidence, substantiality‐embodiment, and demarcation. All these features are never an object of ordinary experience, but provide a first‐person structure for the narrative level of experiencing oneself as, for example, “impulsive” or “suspicious”. However, these features are experientially accessible when we reflect upon the way in which our experience articulates itself.

We have previously proposed that the essential feature of schizophrenia spectrum disorders is a disturbance of the core self in its immediate relation to the world3. It is important to emphasize that we are not talking about a lack or a deficit (as in “too much or too little”) but rather of an instability or dis‐order4. This basic disturbance of self‐world relation is the generative component in the Gestalt of autism3, which “appears nowhere else in this particular fashion”2 and which imbues schizophrenia with an air of un‐understandability5.

Empirical studies1 from different groups and on different samples clearly show a selective hyper‐aggregation of disorders of core self in schizophrenia and schizotypal disorder as opposed to bipolar disorder and other psychiatric disorders. Self‐disorders typically begin in childhood or adolescence, are observed in populations at ultra‐high‐risk for psychosis, and predict subsequent schizophrenia spectrum outcome1.

Two studies have demonstrated temporal persistence and similarity of patterns of self‐disorders five years apart6. Self‐disorders are unrelated to IQ1, and preliminary data fail to show any substantial correlation with neurocognitive disorders. In sum, empirical research seems to corroborate Bleuler's idea that these phenomena are to be considered as trait features of the schizophrenia spectrum.

This structural instability of self‐world relation is the background for the development of psychotic symptoms, which in their form contain an imprint of disordered selfhood4, 7. For example, the characteristic auditory verbal hallucinations are often a progression from the state of anonymization and spatialization of thinking, where the patient's “I think” becomes transformed into “it thinks in me”. The phenomenon of thought broadcasting is a flamboyant expression of the loss of sense of demarcation. And the characteristic double‐book‐keeping involves a construction of a private world or alternative ontological framework7, 8.

The recognition of self‐disorders entails important nosological consequences. Currently, we see a decrease in the diagnosis of disorganized schizophrenia, a very uncommon use of the schizotypal diagnosis and an increasing frequency in the use of the borderline personality disorder diagnosis. This latter diagnosis is over‐inclusive and often applied to patients which would in the ICD‐9 be diagnosed with a schizophrenia spectrum condition9. It seems to us that it is nearly impossible to conceptualize a core psychopathological difference between the notion of schizotypy and the contemporary clinical application of the DSM‐5 diagnosis of borderline personality disorder9.

This diagnostic confusion is multidetermined, but mostly due to a very tolerant use of the ninth borderline disorder criterion (“transient, stress‐related paranoid ideation or severe dissociative symptoms”) and the unclarity of the borderline disorder criteria of “identity disturbance” and “feelings of emptiness”. “Feelings of emptiness” are undefined, and the identity disturbance criterion, although apparently referring to the narrative level of selfhood, is not sufficiently differentiated from disturbances of core self10. We find it crucial to sharpen the distinction between schizophrenia spectrum psychopathology (involving disturbances of both core and narrative self) and disorders of personality (which do not involve structural disturbances of the core self).

Contemporary classification is striving for simplicity and reliability, with much research being performed by for‐the‐purpose‐trained lay interviewers. The disappointment with the slow progress of pathogenetic research encourages critical voices advocating abandonment of phenotypic categories altogether. However, the story of self‐disorder research may inspire us to reconsider the phenotypic classification with a more refined psychopathological approach.

Josef Parnas1, 2, Maja Zandersen1
1Mental Health Centre Glostrup, Institute of Clinical Medicine, University Hospital of Copenhagen, Copenhagen, Denmark; 2Center for Subjectivity Research, University of Copenhagen, Copenhagen, Denmark