Related Experiment Videos
What does semantic tiling of the cortex tell us about semantics?
1University of Glasgow, UK.
Neuropsychologia
|April 12, 2017
Summary
Current semantic mapping in neuroscience shows brain areas active in conceptual processing but misses the underlying mechanisms. Understanding cognition requires exploring the brain's algorithmic level, not just activity traces.
Area of Science:
- Cognitive Neuroscience
- Computational Neuroscience
- Cognitive Science
Background:
- Voxel-wise modeling suggests semantic maps are distributed across the cortex.
- This approach identifies active brain areas but not the processing mechanisms.
- It overlooks Marr's algorithmic level, crucial for understanding cognitive implementation.
Purpose of the Study:
- To investigate the representation and processing mechanisms underlying conceptual abilities.
- To bridge the gap between observed neural activity and the mechanisms of cognition.
- To emphasize the importance of the algorithmic level in neuroscience.
Main Methods:
- Review of existing research in cognitive science and neuroscience.
- Analysis of mechanisms for feature, frame, grounded, abstract, and linguistic representations.
- Consideration of knowledge-based inference, concept composition, and conceptual flexibility.
Main Results:
- Established knowledge includes feature/frame representations, various representation types, inference, composition, and flexibility.
- Semantic mapping provides activity traces but not mechanistic explanations.
- Cortical activity mapping alone is insufficient for understanding conceptual processing.
Conclusions:
- Understanding conceptual processing requires elucidating underlying neural mechanisms, not just mapping activity.
- Process models from cognitive science are vital for specifying intervening mechanisms.
- Neuroscience must establish algorithmic-level explanations to achieve its fundamental goals.