Gouskova and Gallagher 2020

Gouskova, Maria and Gillian Gallagher. Inducing nonlocal constraints from baseline phonotactics. 2020. Natural Language and Linguistic Theory 38:1, pp. 77-116.
[pdf]

Nonlocal phonological patterns such as vowel harmony and long-distance consonant assimilation and dissimilation motivate representations that include only the interacting segments–projections. We present an implemented computational learner that induces projections based on phonotactic properties of a language that are observable without nonlocal representations. The learner builds on the base grammar induced by the MaxEnt Phonotactic Learner (Hayes and Wilson 2008). Our model searches this baseline grammar for constraints that suggest nonlocal interactions, capitalizing on the observations that (a) nonlocal interactions can be seen in trigrams if the language has simple syllable structure, and (b) nonlocally interacting segments define a natural class. We show that this model discovers nonlocal restrictions on laryngeal consonants in corpora of Quechua and Aymara, and vowel co-occurrence restrictions in Shona.

The code for the learner, the datasets, and the simulation results are available here.