A computational model of semantic memory impairment: modality specificity and emergent category specificity.

It is demonstrated how a modality-specific semantic memory system can account for categoryspecific impairments after brain damage. In Experiment I, the hypothesis that visual and functional knowledge play different roles in the representation of living things and nonliving things is tested and confirmed. A parallel distributed processing model of semantic memory in which knowledge is subdivided by modality into visual and functional components is described. In Experiment 2, the model is lesioned, and it is confirmed that damage to visual semantics primarily impairs knowledge of living things, and damage to functional semantics primarily impairs knowledge of nonliving things. In Experiment 3, it is demonstrated that the model accounts naturally for a finding that had appeared problematic for a modality-specific architecture, namely, impaired retrieval of functional knowledge about living things. Finally, in Experiment 4, it is shown how the model can account for a recent observation of impaired knowledge of living things only when knowledge is probed verbally. How is semantic memory organized? Two general answers to this question have been proposed. One is that semantic memory is organized by taxonomic category, such that different parts of the system represent knowledge about objects from different categories. Alternatively, semantic memory could be subdivided by modality of knowledge, such that one component is responsible for visual information about objects, another for auditory information , and so on. Patients with selective losses of knowledge after brain damage appear to provide a direct source of evidence on the organization of semantic memory. Unfortunately, this evidence yields conflicting answers. In most cases , the losses appear to be tied to specific modalities , resulting in impaired recognition of objects in just one modality (e. , visual or auditory agnosia) or in impaired manipulation of objects with specific uses, despite intact recognition of them (apraxia; e. g., a key might be pulled, rather than turned). These observations are consistent with recent neurophysiological data showing that most cortical neurons are modality-specific, even in regions that were traditionally viewed as supramodal association areas (e. , Sereno & Allman, 1991). In some cases, however, brain damage seems to cause category-specific losses of knowledge, which cut across different modalities. Specifically, there are patients who seem to have lost their knowledge of living things, and others who seem to have lost their knowledge of nonliving things. These observations suggest that the architecture of semantic memory incorporates at least