All proceedings
Enter a document #:
Enter search terms:

Info for readers Info for authors Info for editors Info for libraries Order form Shopping cart

Bookmark and Share Paper 3524

Learning Underlying Representations and Input-Strictly-Local Functions
Wenyue Hua, Adam Jardine, and Huteng Dai
143-151 (complete paper or proceedings contents)


This paper proposes a learning algorithm that infers underlying representations and phonological processes from a morphophonological paradigm, based on the hypothesis that phonological processes belong to computationally restrictive classes of functions. Specifically, the algorithm is successful because it assumes that the phonological grammar is a simplex input strictly local function -- that is, an input strictly local function (Chandlee 2014) that only makes a single change. While this is a significant restriction, it can be shown that this procedure can learn underlying forms and a grammar for a simple process of assimilation, dissimilation, epenthesis, or deletion. More importantly, it illustrates that restrictive computational principles, combined with major principles in phonological analysis, allow for significant progress in understanding how phonological grammars and underlying representations are learned.

Published in

Proceedings of the 37th West Coast Conference on Formal Linguistics
edited by D. K. E. Reisinger and Marianne Huijsmans
Table of contents
Printed edition: $375.00