University of Pennsylvania Working Papers in Linguistics


The syllabic affiliation of ambisyllabic consonants (e.g. the /m/ in limit) is unclear. Standard analyses argue for their simultaneous linkage to the preceding and following syllables (Kahn, 1976; Kenstowicz, 1994). However, others have argued that a reformulation of the crucial phonological processes in terms of foot-structure eliminates the need for syllabic representations all together (Jensen, 2000; Kiparsky, 1979). There is even a lack of consensus in the literature about the onsets and codas. It has been argued that there is little to no perceptual evidence through priming experiments for syllable structure (Schiller, Costa, & Colome 2002; Schiller, Meyer, & Levelt 1997). On the contrary, through a syllable tracking task, Nesbitt & Durvasula (2015) argue that listeners do perceive a difference between word-medial onset and codas. Furthermore, they argued that listeners treated words containing ambisyllabic consonants similar to those containing word-medial coda consonants. With so many conflicting findings, the question remains: Are syllabic representations available to speaker/listeners of American English? If so, what acoustic cues are utilized to indicate such representations for ambisyllabic consonants? For this paper, recorded speech was extracted from the Buckeye Corpus (Pitt et al 2007), and analyzed to determine the acoustic effects of word-medial consonants. We compared duration and pitch measurements of vowels preceding ambisyllabic consonants to those preceding word-medial coda and word-medial onset consonants in American English. We conclude that American English speakers have a coda representation for ambisyllabic consonants. They produce vowels preceding these consonants and word-medial coda consonants with a shorter duration and lower pitch than they do vowels preceding word-medial onset consonants.