Low proficiency does not mean ab initio: A methodological footnote for linguistic transfer studies

Abstract

The goal of this brief article is to highlight a specific methodological consideration pertaining to the examination of linguistic transfer in sequential language acquisition: When and how can transfer be meaningfully disentangled from issues pertaining to developmental trajectories of the target language? While this methodological issue is relevant for all transfer studies irrespective of learner type or linguistic domain of inquiry, herein we focus on a set of third language acquisition data. We examine the domain of negative quantifiers <i>nobody/nothing</i> and negative polarity items <i>anybody/anything</i> by Catalan-Spanish early bilinguals learning English as the L3 in adulthood. We offer two group analyses. The first is the superset of low beginner proficiency speakers (all participants taking part in a specially designed English course) and then a subset group (only those who were true <i>ab initio</i> L3 learners—that is, with no previous study of English). The analyses combine to show that exposure matters beyond proficiency—even when proficiency is held constant at very low levels, low proficiency L3 learners who have had some instruction/exposure to an L3 pattern differently from truly <i>ab initio</i> L3 learners. We discuss how this reality complicates isolating L3 transfer proper from effects of L3 development/acquisition and thus, by extension, to all cases of transfer such as adult and child L2.