Individual differences in speech categorization and perceptual cue reliance across phonological contrasts

Individual listeners differ in how they map acoustic cues onto phonological categories, yet it remains unclear which aspects of categorization generalize across contrasts and how these relate to cue use. This study examines whether differences in speech categorization indices—categorization gradiency (slope) and consistency (trial-by-trial stability)—are linked to listeners’ reliance on acoustic cues and whether these indices remain stable across distinct phonological contrasts. English-speaking adults (n=61) completed two-alternative forced-choice cue-weighting tasks and visual analog scaling tasks for a segmental contrast (stop voicing) and a lexical stress contrast. The results provide the first demonstration that VAS can effectively capture categorization patterns for English lexical stress, extending its application to suprasegmental contrasts. Across both contrasts, listeners with higher categorization consistency showed greater integrative use of multiple acoustic cues. Cross-contrast comparisons revealed that categorization consistency in one contrast predicted consistency in the other, whereas gradiency did not. These suggest that categorization consistency is a stable, trait-like property that is closely linked to how listeners coordinate relevant acoustic cues comprehensively. Together, these findings highlight VAS as a methodological tool for probing individual differences in speech categorization across phonological domains.