Skip to main content

Marc Garellek

UCLA Linguistics

Monday, February 25th at 1:30 p.m., Marc Garellek will give a colloquium in the UCSD Linguistics Department, in AP&M 4301.

The role of prosody in the production and distribution of glottal stops

Glottal stops are very common sounds, occurring in the segmental inventory of about half of the world’s languages, and optionally before word-initial vowels in many others (“word-initial glottalization”). Nonetheless, much about the production and distribution of glottal stops remains unclear. Remarkably, we still do not know whether glottal stops are only glottal vs. involve supraglottal constriction, nor why glottal stops occur optionally before word-initial vowels in so many languages.

To address whether glottal stops are truly glottal sounds, I use high-speed imaging of the larynx to determine whether supraglottal articulations are involved in producing onset and coda glottal stops, and if so, how often. Results show that supraglottal constriction is indeed common, but it reaches its peak only in the middle of the glottal stop’s closure. This suggests that supraglottal constriction serves to reinforce glottal closure, but is not necessary for vocal fold adduction. Because I find that glottal closure occurs more frequently in strong phrasal environments, I argue that reinforcement serves as a type of prosodic strengthening.

To determine when and why glottal stops occur before word-initial vowels, mixed-effects logistic regression modeling is used to predict the occurrence of word-initial glottal stops in an English corpus. The results indicate that phrasal prominence and phrasing are overwhelmingly the most important factors in predicting glottal stop occurrence. Moreover, prominent word-initial vowels that are not preceded by a full glottal stop [ʔ] nonetheless show acoustic correlates of glottal constriction, whereas non-prominent phrase-initial vowels do not. These findings are subsequently confirmed articulatorily using electroglottography, and then extended to Spanish. Based on the results, I propose and motivate a prominence-driven theory of word-initial glottalization: word-initial glottalization marks prominence, but whether there is a full stop closure depends on phrasal position. Predictions of this theory for cross-language and cross-speaker variability in glottal stop occurrence will be discussed.