Course announcement: Phonology 4 / Seminar in Phonology

LING 635/735
Current topics in phonology: a computational approach
Tuesday, 2:35-5:25

Instructor: Morgan Sonderegger

Course description
This course will address several topics of current interest in phonology, united by the theme of variability in sound systems, using a hands-on approach.  Students will first learn to program in Python, with a focus on tools needed to extract information from corpora which can be used to test research questions about phonological variability — though these tools are useful for experimental and theoretical studies more generally.*  We will then cover several topics related to variability, for two weeks each, including (preliminary list):

  • Sources of variability
    Explanations which have been proposed for the structure of phonologicalvariability, such as Steriade’s influential P-map hypothesis, which links the perceptibility of phonological contrasts to their likelihood of being used in a language.
  • Variability in the lexicon
    Within a given language, some unattested/attested sound sequences are judged worse/better than others by native speakers, and certain sound sequences are heavily overrepresented across the lexicon (the most famous example being co-occurrence asymmetries among consonants in Arabic).  Much recent work explores how to account for such “probabilistic phonotactics”  in terms of some type(s) of similarity between segments (e.g. perceptual distinctiveness, number of shared features).
  • Variability in realization
    Phonological variation — any situation where the same underlying morpheme can be realized as different surface forms in a given environment — has gained extensive attention in phonological theory over the past 15 years. Phenomena such as English t/d deletion (e.g. realization of “went” with or without the final [t]) are increasingly, though not uncontroversially, seen as part of phonology proper, rather than simply “phonetic implementation”.
  • Variability in grammar:
    Classic optimality theory (OT) can only account for categorical phonologicalpatterns.  The increasing interest in gradient patterns (such as probabilistic phonotactics and phonological variation) in phonology has gone hand-in-hand with the adoption of theoretical frameworks which can account for both categorical and gradient patterns, most notably Maximum Entropy grammars and Stochastic OT.

For each topic, we will alternate theoretical and practical weeks: in the first week we will discuss 1-2 key papers and formulate research questions which build on them; in the second week (and a subsequent homework assignment),  we will write and deploy Python scripts to test these questions, by extracting relevant data from corpora or running simulations.  For example, after reading Frisch et al.’s influential paper on gradient consonant co-occurrence patterns in Arabic, which explains them in terms of similarity between segments, we might write scripts to extract consonant co-occurrence data from a pronunciation lexicon of a different language, and see whether Frisch et al’s account works for it as well.

Because of the hands-on nature of the course, most evaluation will be via frequent homework assignments, which will combine programming and brief write-ups. There may also be a short final paper for students in Phonology 4, which can either build on one of the homework assignments or continue an existing research project.

Comments are closed.

Blog authors are solely responsible for the content of the blogs listed in the directory. Neither the content of these blogs, nor the links to other web sites, are screened, approved, reviewed or endorsed by McGill University. The text and other material on these blogs are the opinion of the specific author and are not statements of advice, opinion, or information of McGill.