A new approach to discourse analysis in psychiatry, applied to a schizophrenic patient's speech

Progress in the science of data analysis and computer technology has led to the development of advanced methods for investigating structure discourse in the psychiatric field, where language constitutes a useful investigative and therapeutic tool. The purpose of this study was to present and use a computer-assisted method of discourse analysis (Alceste-software) to analyse the schizophrenic subject's oral contributions regularly collected for 3 months.

The method used consisted of modelling the main word distribution in spoken recordings pooled together and identifying the repetitive language patterns most frequently used by the speaker.

Four main kinds of discourse emerged from the pool of schizophrenic's speech samples, on specific topics without any lack of ability to organize the material, but the technique analysis showed that the main kinds of discourse were interspersed with unexpected 'language satellites' consisting of a secondary short and specific discourse which was also well planned but had no relevance to the main discourse making for a lack of cohesion in the speech samples. This method allows us direct access to the inner experience of the patient. The technique highlighted a very poor pre-syntax linked to the choice of words and a tendency to make pronoun errors, possibly reflecting some confusion between the patient herself and others, mainly her mother, especially in the discourse about childhood.

This method of discourse analysis made it possible to investigate various language disturbances at the same time and at different levels. It is particularly adapted for analysing the schizophrenic's speech. The data obtained were consistent with the assumption that schizophrenia involves 'thought disorders': these ones giving rise to the language impairments.