Acoustic evidence for consonant cluster organization across contexts

Abstract

Listen

Translate article

C-center organization, in which onset consonant clusters exhibit temporal coordination similar to singletons, is an important theme in speech research. As an alternative to expensive and difficult articulatory research methods, Durvasula et al. (2021) developed a technique for identifying C-center timing using acoustic data alone. We replicate the English results of that study and investigate additional contexts. Following Durvasula et al. (2021), we noted the acoustic midpoints of onset consonants, took the average of these midpoints as “acoustic C-center,” and used the acoustic end of the vowel as an anchor. If C-center coordination were present, C-center-to-anchor intervals should, as compared to singleton onset-to-anchor intervals, be similar in magnitude but more stable. We recorded 18 English speakers producing /s/+nasal, stop+liquid, and /s/+stop+liquid clusters along with segmentally matched non-clusters (e.g., splay, play, lay). Diagnostics indicated C-center-like-timing in initial clusters across segment types, as well as in tautosyllabic medial clusters; the evidence of C-center timing was not found where clusters were split across a syllable boundary. These results are consistent with established models of syllable structure and support the use of acoustic diagnostics for gestural coordination.