Semantic Analysis of Speech in Patients with Schizophrenia: Toward the Identification of Psychodiagnostic Markers

Relevance. Speech disturbances in schizophrenia are among the primary diagnostic criteria listed both in the International Classification of Diseases (ICD) and in the Diagnostic and Statistical Manual of Mental Disorders (DSM) of the American Psychiatric Association. However, clear and objective indicators for their assessment have not yet been developed. The study of semantic characteristics of speech in schizophrenia may contribute to broadening and objectifying the psychodiagnostics of speech impairments.

Intention. To identify semantic features of speech in patients with schizophrenia that distinguish them from mentally healthy individuals, patients with bipolar affective disorder, and personality disorder.

Methodology. Sample: 183 participants – 56 patients with schizophrenia, 30 healthy individuals, 51 patients with bipolar disorder, and 46 patients with personality disorder. Object – speech of patients with schizophrenia, bipolar affective disorder, personality disorder and healthy participants. Subject – semantic characteristics of their speech.

Methods: Biographical (speech recording and writing), content analysis, and statistical analysis.

Results and discussion. The written speech of patients with schizophrenia was impersonal, thematically sparse, and reflected the patient’s passivity. In contrast, their oral speech was more active, briefly describing the patient’s behaviour and the behaviour and internal experiences of others. The speech of schizophrenia patients was semantically impoverished, impersonal, and passive, whereas the speech of mentally healthy individuals, as well as the written speech of patients with personality disorder and bipolar affective disorder, described goal-directed actions of the speaker, various agents, and their internal experiences. Written speech proved to be more psychodiagnostically significant than oral speech.

Conclusion. Compared to mentally healthy individuals, both oral and written speech in schizophrenia patients lacked a semantic core, an active agent, a subject of speech, and descriptions of others’ experiences. The same semantic features were observed when comparing the written speech of schizophrenia patients with that of patients with bipolar disorder and personality disorder. These findings may serve as semantic indicators of schizophrenia patients’ speech and form the basis of a psychodiagnostic model for speech impairments in schizophrenia. The development of such a model could involve machine learning and artificial intelligence technologies.