This is a repository containing notes, code, figures, and the paper for LIN429 (Research in Phonology) which I took in Fall of 2025 at the University of Toronto, and taught by Dr. Peter Jurgec.
Long-distance harmony harmony is a rare, theoretically challenging phonological process (Hansson, 2001). It occurs most frequently in consonant harmony, where segments may need to agree over an arbitrary number of segments (including morpheme boundaries). In order to accommodate for it, rule-based theories require strangely defined environments, and autosegmental theories require careful argumentation (Jurgec, 2011). Optimality theory yields more success, though it struggles in situations where rules interact opaquely (McCarthy, 1999).
Opaque long-distance harmony compounds these factors. No popular theory is well suited to handle opaque long-distance harmony. This creates a vacuum, and encourages new tools and theories to gain strong intuition and make predictions.
I will start by describing Chumash, an isolate language indigenous to California, and argue that previous analyses (Poser, 1982; Poser, 1993; Hansson, 2001; McCarthy, 2007) of its phonology are incorrect. Instead, after the correction in Heinz & Idsardi (2010), Chumash actually displays an example of opaque long-distance harmony, caused by fed anterior sibilant harmony counterfeeding dissimilation on focus.
Next I apply the theory of output multi-tier strictly local (OMTSL) maps, first described in Burness & Mcmullin (2019), to succinctly explain Chumash's opaque sibilant harmony. I then show that the previous standard analysis of Chumash was not just incorrect, but never even possible under the OMTSL framework.
In fact, this obsolete analysis belongs to a wider class of opaque patterns that OMTSL predicts not to exist in languages with long-distance harmony. I discuss the ramifications of this typology, other languages it may apply to, and other patterns that may warrant investigation.
See the full bibliography file here, or in the paper here.
To compile paper.pdf, you need to use pdflatex. This also depends on a
linquiry2.bst. You may need to change some image paths in the file as well.
To compile poster.pdf, you need to use luatex. This also depends on this
format, which
you will need to download additional files from.
The slideshow should be compilable directly using Typst.