Distinct mechanisms in the processing of English past tense morphology: A view from L2 processing*

According to the Dual-Mechanism Model (Pinker & Prince, 1988), the processing of language is based upon two fundamentally different components: a rule component and an associative memory. The aim of this study was to analyze whether the theoretical tenets of the Dual-Mechanism Model, which has originally been established as a model of L1 processing, can be extended to the processing of L2 morphological processing. For this purpose, 49 adult Turkish learners of L2 English at two distinct proficiency levels and 8 adult native speakers of English participated in an elicited English past tense production task that tested their regular and irregular past tense formation behaviors on real and nonce verbs. The results display that L2 subjects behave quite similar to L1 subjects and compute regular and irregular past tense forms in qualitatively distinct ways, as proposed by the Dual-Mechanism Model. A second question that the study tried to approach is related to whether L2 learners at low proficiency levels lack a rule-system and initially compute regular and irregular forms exclusively over the associative memory until a certain proficiency level is reached as proposed by Zobl (1998). The results of the present study show that both low and high proficiency L2 learners were able to implement the past tense rule that adds the default past tense suffix “-ed” to stem forms and did not rely on the associative memory only. The results thus suggest that the two distinct mechanisms posited by the Dual-Mechanism Model are in use from early stages of interlanguage development on.