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

Area of Science:

• Cognitive Neuroscience
• Neuropsychology
• Computational Neuroscience

Background:

• Category-specific semantic memory impairments are observed after brain damage.
• The roles of visual and functional knowledge in semantic representation remain debated.
• Existing models struggle to explain certain patterns of knowledge deficits.

Purpose of the Study:

• To propose and test a modality-specific semantic memory system.
• To investigate the differential roles of visual and functional knowledge.
• To account for specific patterns of category-specific impairments.

Main Methods:

• Development of a parallel distributed processing model of semantic memory.
• Subdivision of knowledge into visual and functional components within the model.
• Simulated lesioning of the model to assess effects on knowledge retrieval.

Main Results:

• Damage to visual semantics selectively impaired knowledge of living things.
• Damage to functional semantics selectively impaired knowledge of nonliving things.
• The model explained impaired retrieval of functional knowledge for living things and verbal probing effects.

Conclusions:

• A modality-specific semantic memory architecture can account for category-specific impairments.
• Visual and functional knowledge play distinct roles in representing living and nonliving things.
• The model provides a unified explanation for diverse semantic memory deficits.