Context-sensitive computational mechanistic explanation in cognitive neuroscience

Area of Science:

• Cognitive Neuroscience
• Neuroscience
• Computational Neuroscience

Background:

• Cognitive neuroscience seeks to explain behavior by linking organismal abilities to neural networks via computation.
• Current models face challenges due to empirical findings that contradict stable mappings between analysis levels.

Purpose of the Study:

• To review the commitments of cognitive neuroscience in mapping behavior to brain networks.
• To highlight the mismatch between theoretical models and empirical research findings.
• To propose a reorientation of methodological and theoretical approaches.

Main Methods:

• Integrative review of existing literature and empirical research findings.
• Analysis of phenomena such as context-dependent neural tuning, neural reuse, and plasticity.
• Examination of research on task context and dynamic neural parcellation.

Main Results:

• Stable mappings between organismal, computational, and neural levels are lacking.
• Neural function is context-dependent, preventing stable computational descriptions.
• Task context dynamically influences neural component organization.

Conclusions:

• Instability across analysis levels poses an epistemological challenge for cognitive neuroscience.
• A shift towards contextual explanations at all levels is necessary.
• Brain mapping efforts require methodological and theoretical reorientation to achieve mechanistic explanations of behavior.