Prosodic categories in speech are acoustically multidimensional: evidence from dimension-based statistical learning

Segmental speech units such as phonemes are cued by multiple acoustic dimensions (e.g. F0 and duration), but dimensions do not carry equal perceptual weight. The relative perceptual weights of acoustic speech dimensions are not fixed but vary with context. For example, when speech is altered to create an accent in which two acoustic dimensions are correlated in a manner opposite that of long-term experience, the dimension that carries less perceptual weight is down-weighted to contribute less in category decisions. It remains unclear, however, whether this short-term reweighting is limited to segmental categorization, or if it extends to categorization of suprasegmental features which span multiple phonemes, syllables, or words, which would suggest that such "dimension-based statistical learning" is a widespread phenomenon in speech perception. Here we investigated the relative contribution of two acoustic dimensions to word emphasis. Participants categorized instances of a two-word phrase pronounced with typical covariation of fundamental frequency (F0) and duration, and in the context of an artificial accent in which F0 and duration (established in prior research on English speech as primary and secondary dimensions, respectively) covaried atypically. When categorizing accented speech, listeners rapidly down-weighted the secondary dimension (duration) while continuing to rely on the primary dimension (F0). This result indicates that listeners continually track short-term regularities across speech input and dynamically adjust the weight of acoustic evidence for suprasegmental categories. Thus, dimension-based statistical learning appears to be a widespread phenomenon in speech perception extending to both segmental and suprasegmental categorization.