Disentangling objects′ contextual associations from perceptual and conceptual attributes using time-resolved neural decoding

The preprint operates within the established paradigm of cognitive neuroscience, specifically focusing on visual object recognition and its neural underpinnings. It adheres to the dominant assumption that object knowledge is represented in a distributed and dynamic manner in the brain, with perceptual and conceptual features playing distinct roles.

This preprint investigates the temporal dynamics of object processing in the brain, using representational similarity analysis (RSA) to correlate EEG data with behavioral models of object similarity. The study builds upon existing research that has demonstrated the distinct roles of perceptual and conceptual features in object recognition. The findings largely confirm and extend previous work, showing a robust perceptual-to-conceptual cascade in neural responses to object images, regardless of presentation modality (image vs. word). The authors also explore the role of contextual features, but find little evidence for their unique contribution to neural object representations beyond what is already captured by conceptual features. This finding, while unexpected given previous literature, is thoroughly discussed and potential explanations are offered. The preprint’s focus on refining existing models and addressing specific questions within the established framework places it firmly within the category of normal science. There is no evidence of challenging or attempting to overturn the existing paradigm. However, the unexpected lack of unique contextual coding might lead to future research that could qualify as “model drift” if it consistently demonstrates limitations in the current understanding of contextual object knowledge.