Jesse Zymet

Department of Linguistics
3125 Campbell Hall, UCLA
Los Angeles, California


Welcome! I’m a Ph.D. candidate in the Department of Linguistics at UCLA. My primary research develops and addresses computational models of phonology and its interfaces (lexicon, morphology, phonetics, processing, syntax), and phonological learning; and tests predictions using experimental methods such as nonce probe and artificial language studies.

I also bring fieldwork and corpus data from lesser-studied languages to bear on theories of phonology and its interfaces.

Recent updates:


Lexical propensity in phonology: corpus and experimental studies

The subject of my dissertation. By lexical propensity, I mean the gradient propensity for a particular morpheme to undergo a variable phonological process. I'm currently exploring lexical propensity in corpus data on French liaison, Slovenian palatalization, Japanese rendaku, and Spanish dediphthongization -- see here for early results. Experiment results so far suggest that speakers are able to learn the relative propensities of individual morphemes to undergo a variable process -- see here. Recent models treat grammar and the vocabulary in the same component, with exceptionality encoded as lexically specific constraints. I'm exploring a MaxEnt account of lexical variation that keeps the grammar and the vocabulary separate, with the latter encoded as a random effect. More to come.

Distance-based decay in long-distance phonological processes: probabilistic models for Malagasy, Georgian, Latin, English, and Hungarian

Based on my MA thesis. Long-distance phonological processes can exhibit distance-based decay, whereby application rate decreases as distance between the trigger and target increases. This paper gives a comprehensive treatment of the effect within Maximum Entropy Harmonic Grammar, fitting a single model to corpus data on five languages. The decay effect is here established as general phenomenon, arising in harmony and dissimilation processes regulating both vowel pairs and consonants pairs. Georgian in particular reveals that liquid dissimilation can persist out to six degrees of distance, decaying nevertheless. Distance is shown to be best measured in syllables rather than in segments or moras. A crosslinguistically fixed scale adjusting a single constraint weight fits the decay effect across all surveyed languages; by comparison, a constraint pair regulating local and nonlocal application consistently underpredicts gradience, whereas a family of distance-specific constraints is found to be unmotivated in light of scaling. The results support a theory in which the learner utilizes a restrictive scale for a single constraint.

Substantive bias and final (de)voicing: an artificial language study
Glewwe, Zymet, Adams, Jacobson, Yates, Zeng, Daland

While final devoicing is natural and typologically common, final voicing is rare or unattested. We conducted an artificial grammar learning experiment that tested for substantive (i.e, phonetic naturalness) bias in the case of final (de)voicing while controlling for formal complexity. There were three training patterns, Devoicing, Voicing, and Change (a more complex final voicing/devoicing exchange rule). In test, a 2AFC task tested participants’ learning of their training pattern. Voicing was learned better than Devoicing and Change, which did not differ. This result is inconsistent with the substantive bias hypothesis, which predicts better learning of final devoicing than final voicing. Thus we found no evidence supporting substantive bias, in line with Moreton & Pater (2012).

Lookahead in phonology

My colleague Jeff Adler and I found lookahead effects in the languages we were studying, coming to roughly the same conclusions (1, 2, 3). We decided to coauthor a paper, pooling together cases of lookahead across a diverse breadth of systems displaying stress assignment, reduplication, assimilation, and more. Optimality Theory, which has full lookahead, captures our cases, while Harmonic Serialism, which has no lookahead, crashes in the same way for each. We give a unified explanation for why OT succeeds and HS fails for our systems, which are merely superficially different: to capture each, entire procedures -- entire sequences of processes applying to the input -- must be compared to best satisfy constraints in the language. Our findings suggest that grammatical models, even if they apply changes one at a time, must have lookahead capability.

My fieldwork on Maragoli focused on its rich possessive agreement system, which displays lookahead: the order in which reduplication and hiatus repair apply is conditioned by the well-formedness of the fully formed reduplicant. The system is capturable in OT, but is challenging for HS.

Contradictory markedness preferences across morphological domains and their implications for morphophonological learning

Recent research suggests that learners prefer for phonological restrictions to match across morphological domains. Recent corpus investigations quantified the degree of matching and found that a strong restriction in one domain is accompanied by a weaker, matching tendency in another (Martin 2011, Chong 2016). In this paper, a corpus analysis of Malagasy finds that backness dissimilation applies consistently to the passive imperative suffix, but is entirely uncorroborated by phonotactics, which displays a weak harmony tendency. Though recent learning models (Martin 2011) enforce a degree of matching, these data suggest that such models must accommodate morphologically specific restrictions that lack matching tendencies elsewhere.


A crosslinguistic investigation of lexical propensity in phonology
SCAMP'17 poster: pdf

Contradictory markedness preferences across morphological domains and their implications for morphophonological learning
WCCFL35 talk (/BLS43 talk): pdf

Irreducible parallelism in phonology (with Jeff Adler, UCSC)
NELS47 talk: pdf
AMP'16 poster: pdf
SCAMP'16 poster: pdf

On the relationship between sonority and glottal vibration (Risdal, Aly, Chong, Keating, Zymet)
CUNY'16 talk: pdf
LabPhon'16 poster: pdf

What factors contribute to the sonority of a segment? We analyzed recordings of voiced segment types (obstruents, nasals, liquids, glides, vowels), obtaining the following primary results: strength and degree of voicing differs depending on segment type, with weaker and breathier voicing being associated with segments having a tighter restriction (as in obstruents); moreover, measures of voicing strength (strength of excitation) and breathiness (closed quotient) correlate with the sonority hierarchy (obstruents < approximants < vowels).

A case for parallelism: reduplication and repair in Maragoli
LSA'16 talk (/MFM32 talk): pdf

Distance-based decay in long-distance phonological processes
WCCFL32 talk (/UCSC phlunch talk): pdf
ABC↔C poster: see here, which is part of the broader ABC↔C conference archive.

An investigation into apparent sublexical coordination in English (with Clara Sherley-Appel, UCSC & Google)
LSA'14 talk: pdf

We present data on coordinated affix constructions (CACs): constructions such as pre- and post-operative. We argue that CACs are not merely coordinations of affixes, but rather coordinations of phrases that underwent subsequent right-node raising ([pre-operative] and [post-operative] → [[pre-t] and [post-t]] operative).

Updated on 09/20/17.