Thought Disorder in Schizophrenia: A Reappraisal of Older Formulations and an Overview of Some Recent Studies

Abstract

Introduction. Formal thought disorder, as manifested in disorganised speech, is a cardinal symptom of schizophrenia. Method. In this review we examine several older explanations of thought disorder and consider newer cognitive accounts. Results. Older formulations often were not mechanistic and simply redescribed the phenomenon under study. Direct comparisons of thought-disordered schizophrenic patients and jargon aphasic patients indicated that the groups were not similar, as they differed on a number of language variables, including naming and comprehension. Later attentional models also had limitations. Correlations between thought disorder and ''attentional'' measures were often only moderate and the tests or scales used to assess attention or thought disorder may not have accurately measured these constructs. Recent working memory models have lacked convergent validity in that patients with frontal lobe damage and working memory problems do not exhibit thought disorder and that single word association and priming studies elicit abnormalities in schizophrenic patients, findings at odds with the view that thought disorder is a discourse planning problem. In our own series of studies we found compelling evidence that thought disorder may be related to various impairments in semantic processing. These impairments were especially evident on semantic fluency tasks. In fact it was the latter, rather than working memory or attentional tests, that predicted thought disorder. Moreover, we were able to provide a mechanistic account of thought disorder based on qualitative abnormalities in semantic priming paradigms. Conclusions. Patients with thought disorder displayed lack of priming for highly associated within-category words, suggesting that they could not access them in a timely and semantically relevant manner. The literature also has implications for understanding cognitive architecture. Although vocabulary and naming abilities were relatively normal in schizophrenic patients, cognitive maps of semantic space, as determined by multidimensional scaling analyses, indicated disorganisation among semantic entities. These results imply that the size of the lexicon is dissociable from the organisation of the lexicon. In schizophrenia in particular it appears that the latter, rather than the former, is more perturbed.