Formal Generative Phonology
Abstract: This paper outlines a program for the study of phonology as a branch of cognitive science. Building on the legacy of classical generative phonology and biolinguistics, it provides a theoretical framework that strictly differentiates phonological competence from aspects of articulation, acoustics and perception. We argue that phonological competence is to be characterized as a formal — that is, explicit, logically precise, and substance-free — manipulation of abstract symbols. We propose that a productive way to execute this program is to adopt a model called Logical Phonology, where phonological competence is described and explained by a system that maps between phonological data structures like strings via rules constructed from basic set-theoretic operations. We show the merits of this model by applying it to Turkish, Hungarian and English data. The remote and complex relationship between phonological competence and speech is elucidated by Cognitive Phonetics, which proposes that the outputs of phonology are transduced via two algorithms into temporally distributed neuromuscular activities. Taken together, Logical Phonology and Cognitive Phonetics aim to ex-plain the nature of what is loosely referred to in the literature as ‘the externalization of language’ and to delineate its components.
Volenec, Veno & Reiss, Charles. 2020. “Formal Generative Phonology”. Manuscript available in Radical: A Journal of Phonology.