A study investigating embodied accounts of semantic memory assessed navigation differences between congenitally blind and sighted individuals using a property listing task with concrete and abstract concepts. The researchers computed semantic entropy, an embedding-based natural language processing metric that captures the predictability of retrieval, to analyze these patterns.
- Generalized linear mixed models revealed distinct navigation patterns across groups based on visual experience.
- Sighted individuals showed higher entropy for abstract than concrete concepts.
- Blind participants did not show this pattern but instead exhibited higher entropy for visually salient concrete concepts, such as penguin.
These results underscore the role of visual experience in the organization and dynamic navigation of semantic memory.