Eric Raimy and Mark Koranda presented at the 92nd Annual Meeting of the Linguistic Society of America. The work titled “Phonological substance and the Successive Division Hierarchy” compares the implications of two approaches to phonological features, the ‘naive phonetic’ descriptive system used by UPSID, and that proposed by Avery and Idsardi (2001) for 213 distinct vowel inventories (Maddieson, 1984). Calculating all possible hierarchies according to the successive division algorithm (Dresher, 2009), they show that the AI system presents a smaller, and thus more tractable, hypothesis space for the language learner.
Authors: Raimy, E., Koranda, K., Kosmatka, C.