Introduction

The study of third language acquisition (henceforth L3A) from a generative perspective is a quite recent trend. Yet it has provided important insights into the language learning process, particularly the roles of first language (L1) and second language (L2) in the acquisition of a third language (L3). Although nearly all L3 studies in the existing literature (e.g. Bohnacker, 2006; Leung, 1998, 2003, 2006; Lozano, 2002a, 2002b; Vinnitskaya et al., 2003) point towards full access to Universal Grammar (UG) and crosslinguistic transfer, the findings differ as to the source of crosslinguistic transfer in L3A. Leung (1998) proposes the interlanguage transfer hypothesis, according to which transfer from interlanguages may occur when L2 and L3 are typologically close. In her later work, Leung (2003) further claims that the L3 initial state is the steady state of a previously acquired language that is typologically closest to L3. In the similar vein, Vinnitskaya et al. (2003) contend that language acquisition is cumulative, and that all prior language experience can be either neutral or enhancing in subsequent language acquisition. Lozano (2002a, 2002b), on the other hand, argues that L1 can cause persistent fossilization if L1 features do not match L2/L3. In a more recent study, Bohnacker (2006) looks at L1 Swedish, L2 English and L3 German learners in the acquisition of the verb second (V2) property. Both Swedish and German are V2 languages; English is non-V2. Bohnacker (2006) finds evidence for L2 transfer alongside L1 transfer in the L3A of V2 property.

In order to further attest the role of L1 and L2 in L3A, the present study sets out to investigate the acquisition of objects of L3 Norwegian by Chinese-speaking learners who have been advanced learners of L2 English. The issue under investigation falls into the area of L3 syntax, where relatively little research has been conducted (see Leung, 2007, for a review).