Bears Don’t Always Mess With Beers: Limits on Generalization of Statistical Learning in Speech

Perception changes rapidly and implicitly as a function of passive exposure to speech that samples different acoustic distributions. Past research has shown that this statistical learning generalizes across talkers and, to some extent, new items but these studies involved listeners’ active engagement in processing statistics-bearing stimuli. In this study, we manipulated the relationship between voice onset time (VOT) and fundamental frequency (F0) to establish distributional regularities either aligned with American English or reversed the regularity to create a subtle foreign accent. We then tested whether statistical learning across passive exposure to these distributions generalized to new items not experienced in the exposure stimuli. Experiment 1 showed statistical learning across passive exposure but no generalization of learning when exposure and test items shared the same phoneme (bear/pear beer/pier) or when they differed in phoneme but shared distributional regularities across VOT and F0 dimensions (deer/tear beer/pier). Experiment 2 showed generalization to stimuli that shared the statistic-bearing phoneme (bear/pear beer/pier), but only when the response set included tokens from the exposure and test stimuli. Moreover, statistical learning transferred to influence the subtle acoustics of listeners’ own speech productions but did not generalize to new stimuli. These findings reveal that passive exposure is sufficient to support statistical learning in perception and its generalization, but task demands modulate this dynamic. Moreover, generalization in perception is not always accompanied by generalization in production, demonstrating that production is not simply a mirror of the perceptual system.