Background. In a broad cognitive study of schizophrenia we investigated the relationships of verbal memory impairments with cognitive underpinnings on the one hand, and clinical symptomatology on the other. The results have been reported in previous papers. In this paper we show how all these data could be integrated into a consistent pattern of associations.

Method. Fifty schizophrenic patients underwent a cognitive battery including a verbal memory task with free recall and recognition, a source memory task, and tests of processing speed and selective attention. Ratings for positive, negative and depressive symptoms were available for 40 of the patients.

Results. A factorial analysis revealed a distinction between measures of memory efficiency and measures of memory errors. The system of memory efficiency was associated with processing speed and selective attention at the cognitive level, and with depression at the symptom level. The system of memory errors was assumed to be underlain by source-monitoring deficits. These memory errors were increased by positive symptoms and decreased by certain negative symptoms.

Conclusions. All the measures drawn from various memory tasks could be integrated into a model describing their associations with cognitive underpinnings and clinical symptomatology. This model provides a heuristic for the cognitive and pharmacological treatments of verbal memory impairments in schizophrenia, as well as for the understanding of positive symptoms.