my_face

news

I recently presented a poster at PLC 42 on Gender and (Syntactic) Acceptability Judgments, which you can view here.

about

I am a Visiting Assistant Professor of Linguistics at Haverford College, as well as a computational lexicographer for the NIH. My main interests are in adverbs, mathematical models of language, computational syntax, and typology. If you have thoughts on any of these topics, feel free to send me an email!

education

  • Ph.D. Linguistics, University of Delaware (Feb. 2018)
    •       M.A. Linguistics, University of Delaware (Jan. 2013)
  • B.A. Linguistics, B.A. Mathematics, University of Southern California (2007-2011)

teaching

  • Current - Spring 2018:
    • Instructor for LING 101 (Intro to Linguistics) and LING 114 (Intro to Semantics), Haverford College
  • Previous courses taught:
    • Intro to Cognitive Science, Intro to Linguistics - Honors, Language and Gender, Intro to Syntax
  • Previous courses TA'd:
    • Intro to Biological Anthropology, Intro to Anthropology of Health, Psycholinguistics
  • ongoing research projects

    • Adverb ordering:
      • My dissertation focuses on the grammaticality of various orderings of multiple adverbs, including in wh- and focus constructions. The punchline is that adverbs are very different from adjectives, and have relatively free ordering (more free than Cinque 1999 predicts), but with a melange of syntactic and semantic constraints that make certain orders universally less acceptable.
    • The subsequential nature of dissimilation:
      • Phonology is computationally less complex than the regular region in the Chomsky hierarchy, even in long-distance dissimilations. I show that dissimilation patterns are all within the subsequential class of relations, and provide FST (finite state transducers) for each of the patterns in Suzuki and Bennett's dissimilation typologies.
    • Gender and acceptability judg(e)ments:
      • Amazon Mechanical Turk is a great way to collect grammaticality data from large populations, but large-scale data analysis shows that respondents who identify as female rate test sentences significantly higher than respondents who identify as male. Where does this hyper-acceptance come from, and how pervasive is it?

    links