Second Language Acquisition vs. Grammaticalization: The Status of the functional category 'have' in the Second Language Acquisition of English Temporal Adjunct Clauses by Japanese speaking learners

Ms Miki Hyun Kyung Bong

Saturday, 8 May 2004; 7:30-9:30pm

Keynes Hall, King's College

This study reports a study investigating the status of the English auxiliary 'have' in the second language acquisition (SLA) of English Temporal adjunct clauses by Japanese-speaking learners. English has the [have + verb perfect form] construction which is interpreted as 'experience', 'completion', 'continuity', or 'resultative' of the event. In contrast, Japanese language has no such a functional category as auxiliary 'have' (namely, have-less language). Instead, Japanese uses the iru 'be' with -te perfective morpheme construction (V-te-iru) to convey the resultative or continuous of the event.
Recent second language (L2) research within the generative framework on the syntactic establishment of formal functional features (to account for the prevalent failure, or the poverty of output in SLA) by L2 learners diverges into two broad approaches: L2 learners have (1) the same syntactic representation with a full feature specification, but with Mapping Problem (Lardiere 2000 inter alios); (2) different syntactic representation from native speakers due to the absence of parameterised formal (uninterpretable) features (Hawkins & Chan 1997; Hawkins 2003; among others).
The present study provides robust evidence that L2 learners have a different representation of such parameterised functional categories as `auxiliary have' from English native speakers (regardless of its interpretability?). Judgement data in the study indicate that for example, Japanese learners erroneously judge such sentences as 'John will go to the park after he has eaten breakfast' as unacceptable, and correct 'have+verb perfect' form to a simple present form of the verb 'eats'. Japanese learners seem to have established a novel functional category which is in conformity neither with English `have' nor with Japanese `be': meanings and functions of the presumably newly established functional category are different from L1 and from L2.
In this talk, it is attempted to provide a modified representational deficit hypothesis, referring to the aforementioned results and the general results from SLA in English TACs by Japanese speakers. The results suggest that when an overt exponent of a particular functional category is absent in L1 but present in L2, SLA involves establishment of a new functional category, which is similar to a diachronic phenomenon of grammaticalization (see Roberts and Roussou 2003 on syntactic change, and Millet 1912), and which leads to assume the different representation in L2 grammar. This SLA grammaticalization process in turn seems to suggest that all functional features (UG lexicon) should be fully accessed by L2 learners and support the claim that all functional features are present in all languages although may not all be realized. To conclude, L2 interlanguage grammar with respect to the underlying syntactic representation differs from the target language. To be more precise, the realization/attraction property of functional heads may be different due to grammaticalization or due to the principle of economy in SLA. The application of such processes is only relevant to the lexicon (by means of construction), not to the computation.