Evidence of Absence? Abstract Metrical Structure in Speech Planning

Abstract

Metrical structure (e.g., PREsent vs. preSENT) and segmental structure (e.g., dot vs. pot) are two important cues for differentiating word meanings in English. However, it is not clear whether they are independently or jointly stored in lexical representation. The current study explored this question through the lens of speech planning. In addition, we investigated whether individual differences in working memory capacity (WMC) and music perception skills moderate the effect of word pair characteristics on speech planning difficulty. To study these questions, we adapted the tongue-twister word pair repetition paradigm used by Myers and Watson (2021), and fully crossed metrical overlap, initial stress, and initial segmental overlap. We also calculated the phonological distance between the two words in each word pair. Participants repeated word pairs as quickly as possible while being audio-recorded and completed a series of standardized assessments of WMC and music perception skills. Speech planning difficulty was measured as mean word pair duration, number of correct and complete repetitions, and number of speech errors. We found that speech production was more challenging when the word pair had metrical overlap or segmental overlap, suggesting a joint representation of metrical structure and segmental structure. Speech production was also more challenging when the first syllable was unstressed. There was no moderation effect of individual differences in WMC and music perception skill on the relationship between word pair characteristics and speech planning difficulty. The implications on lexical representation, speech planning, and the connection between music processing and language processing are discussed.