Listener knowledge about sociolinguistic variation
Speakers realize sociolinguistic variables in predictable ways; certain grammatical and phonological properties probabilistically condition speech production. This project investigates to what extent listeners expect to hear variables realized according to those constraints. With Tyler Kendall.
Influences of social and distributional information on perceptual learning
We ask whether social and distributional information affects perceptual learning and how listeners generalize what they have learned. We examine lexically guided perceptual learning of English fricatives specifically, investigating the conditions that promote learning and generalization. With Molly Babel and Michael McAuliffe.
Characterizing variability in native and non-native speech
Non-native speech is often characterized as being more variable than native speech, yet this assumption is not well studied. We examine the production of linguistic features in Japanese by native and non-native speakers in order to understand how speakers acquire within-category variation. With Melissa Baese-Berk and Kaori Idemaru.
Factors affecting intelligibility of accented speech
Do listeners' attitudes and beliefs about an accented speaker affect their comprehension of that speaker? In particular, how do (1) beliefs about the source of a speaker's accent, (2) listeners social and personality characteristics, and (3) the various intersectional aspects of a speaker's identity, factor into listeners' evaluation of accented speech?
Social meaning throughout the lifespan
Most adults have beliefs about the languages and accents they think are smart and sophisticated, and which sound uneducated. But, how did they come to have those beliefs? How and when in development do features of language come to take on social meaning? And, how can social meanings change throughout the lifespan? With Kara Becker.
Indexical processing in bilinguals and monolinguals
In my dissertation, I examined which features of a talker's speech must be processed integrally by a listener. When listening to a bilingual speak, must a listener process what language that person is speaking along with more fundamental characteristics of that talker, such as their gender? Advised by Ann Bradlow.
Native language classification
I conducted a computational classification of transcribed spoken English into native language backgrounds using SVMs with character n-grams as feature vectors. With Janet Pierrehumbert and Hannah Rohde.
Neighborhood density effects in coda position
I examined the effects of lexical competition (namely, neighborhood density) on speech production, in a project with Matt Goldrick. Specifically, we measured pre-coda vowel duration in words with and without a minimal pair differing only in coda voicing.
Vowel plots in motion
How can we visualize vowel plots dynamically, over the course of a speech recording? Some initial explorations with one sociolinguistic interview recording can be found at the following links:
With
Anne Fabricius and
Tyler Kendall.