UGA Linguistics Faculty and Graduate Students present at the 178th Meeting of the Acoustical Society of America

University of Georgia Linguistics Faculty and Graduate Students will be presenting at the 178th Meeting of the Acoustical Society of America, in San Diego, California, December 2-6, 2019.

Dr. Margaret Renwick with Yanming Shi and Fred Maier of UGA's Institute for Artificial Intelligence: "Improved Vowel Labeling for Prenasal Merger Using Customized Forced Alignment."

Abstract: Forced alignment is a popular technique for gaining phone-level audio transcriptions, but the pronunciation dictionaries used by it are typically based on standard varieties of US English, leading to errorful outputs for non-standard varieties. We employ a customized pronunciation dictionary with the Montreal Forced Aligner to increase labeling accuracy of the prenasal merger (a.k.a. pin-pen merger) in Southern US English. We allow the aligner to choose between IH (/ɪ/) and EH (/ɛ/) in words where the merger is expected, rather than enforcing a standard, unmerged pronunciation. We examine the tokens reclassified from EH to IH when using the new dictionary, and we use formant values to study the acoustic separation (measured by Pillai scores and Euclidean distances between centroids) between vowel formant clusters. When applied to the Digital Archive of Southern Speech (DASS), we find that the modification increases the separation between the prenasal allophones of IH and EH, and also that the proportion of prenasal EH tokens reclassified to IH is correlated with the original degree of separation between prenasal IH and EH for each DASS speaker. K-means clustering is also used to show the modification yields more accurate phonetic transcriptions, measured by increased precision and recall.

Lisa Lipani with Michael Olsen and Rachel Olsen: "Voice Onset Time Variation in Natural Southern Speech."

Abstract: Sociophonetic research has traditionally emphasized vowels and only recently begun to examine consonants. The emerging literature on consonants has found systematic regional variation in consonant production (Jacewicz et al. 2009; Eddington & Turner 2017). These studies have focused on present-day speakers producing lab-recorded speech and thus provide little historical or naturalistic insight to consonant variation. One way that the production of stop consonants varies is in voice onset time (VOT). In this study we utilize the Digital Archive of Southern Speech (DASS) (Kretzschmar et al. 2013), a collection of 64 sociolinguistic interviews recorded between 1968 and 1983, to explore consonant variation in an unscripted historical setting. DASS was force-aligned using the Montreal Forced Aligner (McAuliffe et al. 2017), and the VOT of pre-vocalic, word-initial stop consonants in three-minute audio clips from each speaker was measured using AutoVOT (Keshet et al. 2014). VOT was normalized by dividing duration by speaking rate, and this normalized measurement was included as the dependent variable in a mixed effects model. As expected, stress, voicelessness, and a dorsal place of articulation were significant predictors. Our preliminary analysis reveals significant regional and age differences, adding evidence for speaker-specific variation of VOT as a result of sociophonetic variables.