Differences between morphological and repetition priming in auditory lexical decision: Implications for decompositional models

Publication date: Available online 19 October 2018

Source: Cortex

Author(s): Robert J. Wilder, Amy Goodwin Davies, David Embick

Abstract

A natural way of probing the effects of morphology on lexical processing is to directly compare morphological priming, for which primes and targets share a stem but are mismatched in morphological structure (e.g., frogs→frog), with outright repetition priming (e.g., frog→frog). However, work making this comparison has reported no difference between these two types of priming. Importantly, the reported non-differences have been found in the visual domain. Here, we investigate morphological (Morph) vs. repetition (Rep) priming in two auditory primed lexical decision experiments. Using the English plural suffix -/z/, we compare Rep priming with Morph priming for both singular and plural target conditions (e.g., frog/frogs→frog, frog/frogs→frogs). Overall, we find robust priming in both Rep and Morph conditions. However, for both singular and plural targets, there is consistent evidence that Rep priming is greater than Morph priming at early lags of 0 and 1 intervening items. This facilitation decreases with an increasing number of intervening items. Comparisons with phonological and semantic controls demonstrate that this pattern cannot be attributed solely to shared form or meaning. We interpret these findings in a decompositional model of morphological processing. The robust facilitation in Morph and Rep conditions is attributed to the activation of a shared stem representation. The convergence of Morph and Rep is attributed to a diminishing episodic trace related to morphological recombination.

https://ift.tt/2yLMyXO