"L1 to L2 Transfer in Bilingual Sentence Processing¡¨

ABSTRACT
This study examines the influence of L1-to-L2 transfer on L2 sentence processing. Two experiments test English sentence comprehension in Spanish-English bilinguals (SEBs), Chinese-English bilinguals (CEBs), and English native speakers (ENSs) with a region-by-region, self-paced reading task.

Experiment 1 investigates the processing of English sentences containing possessive pronouns. Unlike in English and Chinese, Spanish possessive pronouns agree in number with their sister noun. We manipulated (a) the number feature of the possessive pronoun (plural/singular) and (b) whether this feature matched that of its sister noun (match/mismatch) in sentences like The actress(es) / called / her (their) hairdresser(s) / for an appointment. If possessive pronoun agreement properties transfer, SEBs (but not CEBs or ENSs) should have prolonged reading times (RTs) at the possessive pronoun+noun region when the number of the pronoun and its sister noun mismatch
(e.g. her hairdressers and their hairdresser). Preliminary results reveal no evidence of transfer (only an additive effect of plurality at the critical region for all three subject groups).

Experiment 2 examines the comprehension of English sentences with human and non-human direct objects (DOs). Unlike in English and Chinese, human DOs in Spanish require an extra (clitic) marker that is not required for nonhuman DOs. Test sentences varied with respect to humanness (human, nonhuman) of the NP as a direct object (DO) compared to as an object of the preposition (OP) (in baseline conditions) in sentences like The artist / saw (was next to) / the dancer (the painting) / at the trendy night club. If this L1 marking requirement transfers, SEBs (but not CEBs or ENSs) (a) should have longer RTs in the critical region for sentences with human DOs than for sentences with nonhuman DOs and (b) should show no such difference between sentences with human OPs and nonhuman OPs. Preliminary results are consistent with these hypotheses: SEBs have longer RTs in the critical region for sentences with human DOs than for sentences with nonhuman DOs. CEBs and ENSs show no RT differences in this region. These results suggest that SEBs may transfer their expectation for a clitic marker before a human DO.

Taken together, these results suggest the influence of some aspects of L1 in the comprehension of sentences in L2. The complete set of data and a full discussion of the differential transfer effects found in this study will be presented.