Age of acquisition and the completeness of phonological representations

Abstract

The age of acquisition effect(faster recognition and production of earlierlearnt than later learnt words) is a robustfinding in both picture naming and written wordrecognition and naming. One possibleexplanation of this effect is the PhonologicalCompleteness Hypothesis of Brown and Watson[(1987) Memory & Cognition 15: 208–216], which proposes that early acquired wordsare recognised and produced faster than lateacquired words because they have lessfragmented phonological representations. Thoughoften cited, this hypothesis has never beentested experimentally. The present study setout to test this hypothesis using aphonological segmentation task. If earlyacquired words are stored in a more completeform, then adult participants should be slowerto segment early words than late acquiredwords. In addition, if the AoA effect is aconsequence of the quality of an individual'sphonological representations then there shouldbe a clear relationship between phonologicalskill (as measured by the phonologicalsegmentation task) and the magnitude of the AoAeffect size. In order to assess therelationships between phonological skill andthe AoA effect in adults, participants werealso given a word and nonword naming task. Theword naming task manipulated AoA andspelling-sound consistency. The results of thesegmentation task failed to provide any supportfor Brown and Watson's (1987) phonologicalcompleteness hypothesis. Phonological skill wasfound to predict the size of the AoA effect inthe word naming task, but not the size of theAoA effect in the segmentation task.