The influence of L1 syntax on L2 agreement processing: Evidence from the selection of possessive pronouns/adjectives

Foucart, A. 1 , Santesteban, M. 2 , Pickering, M. 1 & Branigan, H. 1

1 University of Edinburgh
2 University of the Basque Country

Previous cross-linguistic studies examining agreement have suggested that L2 processing is affected by L1 syntax. (Foucart & Frenck-Mestre, in press; Nicol and Greth, 2003; Tockowitz & MacWhinney, 2005). To further investigate this question, we examined the performance of English native speakers and Spanish-, French- and Greek-English late bilinguals (medium-high proficiency) in the production of possessive pronouns and adjectives. In English, the possessive adjective and pronoun agree with the possessor in gender in the singular (i.e., his/her, his/hers), but are not gender marked in the plural (i.e., their, theirs). Greek follows the same pattern as English. In Spanish, in contrast, the possessive adjective agrees with the possessee in number but not in gender (i.e., number: su/sus, for both masculine and feminine); the possessive pronoun agrees in both number and gender with the possessee (i.e., number: suyos/suyas, gender: suyo/suya, for masculine and feminine respectively). In French, the possessive adjective and pronoun agree with the possessee in both gender and number (i.e., son/sa/ses, le sien/ la sienne/les leur).
In two experiments, we manipulated gender agreement matching/mismatching in a picture naming task involving the use of possessive-adjectives (e.g., the ballerina chases her/*his son, Experiment 1) and possessive-pronouns (e.g., the ballerina says the daughter is her/*his, Experiment 2). Results revealed that, when the possessee did not linearly intervene between the possessor and the adjective, only the groups of bilinguals produced errors, with all the groups showing similar attraction effects (Experiment 1). In contrast, when the possessee intervened between the possessor and the pronoun (Experiment 2), Spanish- and French-English bilinguals produced more errors than English native speakers and Greek-English bilinguals. This ‘possessee gender attraction’ effect found in L2 speakers whose L1 agreement rules differ from English suggests that possessive-pronouns selection is affected by L1 syntax during L2 language production.