Explaining frequency of verb morphology in early L2 speech

Abstract

In speech, early L2 learners of English have been observed to supply forms of copula be more frequently than auxiliary be, and both more frequently than affixal regular past –ed and 3rd person singular present tense –s in contexts where morphological marking is required for native speakers. Early learners also use a construction not found in input: be + bare V (e.g. I’m read), allow constructions involving be to have a range of meanings not found in target English, and rarely overgeneralise –ed and –s to inappropriate contexts. The present study considers the kind of mental representation that L2 learners must have that would lead to the observed performance. A ‘nativist’ account is proposed. It is argued that the mental grammars of early L2 learners are organised in the same way as the grammars of native speakers, this following necessarily from the architecture of the language faculty. They differ minimally in the nature of their Vocabulary entries for verb morphology. This difference correlates with an early under-determination of syntactic representations where ‘uninterpretable’ syntactic features are absent from syntactic expressions. Evidence from a sentence completion task conducted with low proficiency speakers whose L1s are Chinese and Spanish is used to test this claim.