The near-term goal of my research is a clearer understanding of socioindexical cues in the speech signal: how they are processed, how they are stored, and how they can enhance or frustrate perception. This is part of a larger project to investigate the usefulness of variation (coarticulatory, allophonic, socioindexical, indexical, etc.) )both to speakers and hearers.
Linguistics, I believe, currently stands in the midst of a sea change regarding the importance of variation. Interest in variation is by no means new; sociolinguistics has been a productive field for over 40 years. However, the importance of variation has only recently begun to reach into areas like speech perception and formal phonology which hitherto considered it to be noise —mere accidents of performance. By moving variation from the periphery of usage and into grammar, this change has the potential to reshape our core notions of what it means to study 'language' and what native speakers know when they know a language.
On this page I describe a selection of my projects, completed or ongoing, in pursuit of this goal.
My dissertation investigated the influence of listeners' socioindexical expectations on speech perception. A number of studies have shown that listener expectations can be manipulated to alter or degrade perception or to create perceptual illusions. My research is unique in assessing the extent to which socioindexical expectations can enhance the transcription of speech (in much the same way as semantic priming). In a matched-guise transcription task, I manipulated listeners' beliefs about the race of a speaker and asked them to transcribe accented speech in noise. Listeners to a Chinese-accented voice do, in fact, make fewer transcription errors when they believe their speaker to be Asian than listeners who believe the speaker to be Caucasian or, in a control condition, have no racial indications to work with. This effect is equally strong for listeners with detailed experience of Chinese-accented speech and for listeners with only limited or stereotypical knowledge of the accent. Listeners in this task appear to draw on `Chinese' or `Asian' social category knowledge to constrain the space of possible phonetic variants and simplify the task of perception.
Presentation given at LSA 2012 in Portland (SVG format, should display in browser)
In ongoing collaboration with Anna Babel of The Ohio State University we are exploring listeners' covarying awareness of social categories and the phonological representations underlying Spanish and Quechua-dominant speech in a Bolivian crossroads. In an initial field experiment, we manipulated whether listeners believed the speaker in an AXB task to be from the Quechua-dominant west of the country or from the Spanish-dominant east. In an unusual twist for this type of experiment, we ran this study using a within-subjects matched guise design in which listeners were presented with both guises in subsequent blocks with no change in voice. Listeners who are initially told the speaker is Quechua-dominant perform essentially at chance on the listening task and continue to perform at chance on the task when switched to the Spanish-dominant guise. However, when these same stimuli are presented with the guises reversed very clear category boundaries for the vowels in each continuum emerge and are retained across the switch from Spanish to Quechua guise conditions. In contrast to these empirical results, interviews give every indication that participants believe the guise switch --commenting on everything from the pronunciation to the relative education level of the new speaker.
With the support of a departmental research grant, David Medeiros and I have been investigating the role of talker expectations during the production of tongue twisters.
In eye tracking work with Patrice Speeter Beddor, Julie Boland, Andries Coetzee, and Anthony Brasher we investigated the time course of listeners' use of nasal coarticulation. A portion of this work was presented at the 2009 Acoustical Society of America meeting in San Antonio ( ASA eye tracking poster ) and an updated verion at LabPhon 12. This manuscript has been conditionally accepted by the Journal of the Acoustical Society of America and will be linked here when published.
I've recently completed my qualifying research project and achieved candidacy. That work, entitled, “Aerodynamic Modeling of Coarticulation for Unit-Selection Speech Synthesis” has been presented at the Acoustical Society of America 2009 October meeting in San Antonio ( McGowan ASA 2009 poster ), a Michigan Linguistics departmental colloquium, and the 2010 meeting of the Linguistic Society of America ( McGowan LSA 2010 poster ).
In the summer of 2008 I had the tremendous pleasure to work with Professor Sarah G. Thomason and the elders of the Salish — Pend d'Oreille Culture Committee. This project, an ultrasound investigation of Montana Salish articulation, is in progress.