Language and cognition revisited: Labels alter post-sensory processing in visual and auditory categorization

How does language influence human categorization? Linguistic labels may affect categorization by either enhancing sensory encoding or modulating post-sensory processing. To dissociate these accounts, this study investigated whether label effects arise during sensory encoding or post-sensory processing, and whether they are modality-specific or modality-general. Using visual and auditory categorization tasks with a priming paradigm, we combined hierarchical drift–diffusion modeling (HDDM), electroencephalogram (EEG) decoding, and neurally-informed modeling to compare label effects against pseudoword baseline. Our results revealed consistent patterns across visual and auditory tasks. Behaviorally, congruent labels facilitated faster and more accurate responses in both visual and auditory tasks, while incongruent labels impaired accuracy in visual categorization. HDDM revealed that labels reduced response boundaries across tasks, with congruent labels shortening non-decision time and incongruent labels prolonging it. Critically, EEG analyses demonstrated that labels selectively amplified late neural components (post-sensory processing) without affecting early sensory encoding. Neurally informed modeling further confirmed that changes in response boundary arise from the enhanced efficiency of the post-sensory process. These findings support a modality-general, post-sensory account of label effects in categorization, aligning with the label-as-marker framework, where labels act as supervisory signals guiding categorical decisions rather than perceptual features. The study advances understanding of language-cognition interactions by delineating the temporal dynamics of label influences on categorization.