LINGUIST List 19.1546

Tue May 13 2008

Diss: Cognitive Sci/Phonetics/Phonology/Psycholing: VanDam: 'Plasti...'

Editor for this issue: Evelyn Richter <>

        1.    Mark VanDam, Plasticity of Phonological Categories

Message 1: Plasticity of Phonological Categories
Date: 13-May-2008
From: Mark VanDam <>
Subject: Plasticity of Phonological Categories
E-mail this message to a friend

Institution: Indiana University Program: Department of Linguistics Dissertation Status: Completed Degree Date: 2007

Author: Mark VanDam

Dissertation Title: Plasticity of Phonological Categories

Dissertation URL:

Linguistic Field(s): Cognitive Science                             Phonetics                             Phonology                             Psycholinguistics
Dissertation Director:
Robert F Port Kenneth de Jong Tracy Alan Hall David B Pisoni Stuart M Davis
Dissertation Abstract:

The traditional model for language assumes a level of competence thatrepresents words composed of distinctive features that are invariant unitsdrawn from a small, universally specified set. The present work tests twoof these assumptions, representational invariance and the assumed small setsize of these features.

This study trained naive listeners on artificially lengthened voice-onsettimes in specific target words and tested whether pretraining andposttraining perceptual boundary locations differed. Results indicatesubjects were successfully trained on the target words, but this effect didnot generalize to other, structurally similar words. In addition, robustfrequency and lexical effects are reported, but do not apparently interactwith training. As predicted, frequency effects suggest subjects preferencefor high-frequency words over low-frequency words, but, somewhatsurprisingly, lexical effects suggest a preference for non-words over words.

Results undermine foundational assumptions of the traditional model,specifically the assumptions of representational invariance and a small setsize. To account for these results, an exemplar memory model is discussedin which no principled restrictions are imposed on memory for language. Theexemplar models discussed are able to account for the facts reported here.