An investigation of the integrity of semantic boundaries in schizophrenia

• Published in: Schizophrenia research Vol. 53; no. 3; pp. 187 - 198
• Main Authors: Elvevåg, B., Weickert, T., Wechsler, M., Coppola, R., Weinberger, D.R., Goldberg, T.E.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 15.01.2002
• Subjects: Adult; Association; Brain - physiopathology; Cognition Disorders - diagnosis; Female; Humans; Male; Memory Disorders - diagnosis; Middle Aged; Neuropsychological Tests; Reaction Time; Schizophrenia; Schizophrenia - physiopathology; Semantic boundaries; Semantics; Thinking; Schizophrenia; Semantic boundaries
• ISSN: 0920-9964; 1573-2509
• DOI: 10.1016/S0920-9964(01)00202-X

Organizing information and knowledge, and hence categorization, requires specifying boundaries between items, concepts and words. Over-inclusiveness in categorization may be seen as looseness of association; over-inclusive thinking was an early description of schizophrenic thinking. Recent studies suggest semantic memory problems in schizophrenia, and that thought disorder is associated with a disorganized semantic network. One such study [Psychol. Med. 24 (1994) 193], using a word categorization task, found patients slowest to respond to items semantically related to, but outside the category, whereas controls were slower responding to items sharing less features of the category (i.e. borderline). The authors suggested that there is an outward shift of semantic category boundaries in schizophrenia. In Experiment 1, we replicated Chen et al.'s (1994) methods, but did not find this qualitative difference in patients (28 patients, 26 controls). We extended this question in Experiment 2 to a more visual domain using pictures that ‘morphed’ from one entity into another and asked participants to decide when they no longer considered an item to be that item (20 patients, 25 controls). We did not find a difference between patients and controls in their sensitivity to detect boundaries of representations. These two experiments do not support the notion that thought disorder with postulated looseness of association or over-inclusive thinking is related to reduced awareness of boundaries of semantic category membership or entities, and inferentially their featural network. Despite anomalies in the semantic system in schizophrenia, we found aspects to be intact. This specificity of semantic processing is promising, suggesting that research will be informative as to how semantic memory is constructed, and thus how it can selectively break down. Moreover, this study indicates that patients do not ‘fail’ semantic tasks (e.g. priming) because of globally disorganized decision-making: here their capability to make precise distinctions between representations was intact.