Developmental Perspectives on Transfer in Third Language Acquisition

Sammanfattning: The aim of this thesis is to examine how learner-general developmental stages in syntax and morphology interact with a language-specific factor, the influence of—or transfer from— the language learner’s first (L1) or previously learned second (L2) language on the acquisition of a third language (L3). It thereby aims to bring together two lines of research whose main concepts—transfer and developmental stages—have often been defined as mutually exclusive and generally studied in separate lines of research. The results obtained are discussed in the light of several transfer hypotheses for L3 acquisition that focus on the initial state of language acquisition (e.g., Bardel & Falk, 2007; Flynn, Foley, & Vinnitskaya, 2004; Rothman, 2011). These hypotheses are divided into Initial L1 Transfer Hypotheses and Initial L2 Transfer Hypotheses depending on the expected source language of transfer in the acquisition of a certain structure. The results are also discussed in the light of a further transfer hypothesis that, by contrast, takes a developmental perspective: the Developmentally Moderated Transfer Hypothesis (DMTH; Håkansson, Pienemann, & Sayehli, 2002). The DMTH does not expect developmental trajectories to be modified by language transfer. Instead, a learner’s present state of development is assumed to constrain the transferability of structures from previously learned languages. In addition, a further factor which has been suggested to affect transfer—the learner’s perception of the distance between two languages (psychotypology)—is explored. Five different tasks (elicited imitation, communication task, picture-based storytelling, interview and questionnaire) were carried out with L1 Swedish L2 English speakers who were learning L3 German in a Swedish school context (n = 61). The participants’ production of morphological and syntactic structures representing developmental stages of German L2/L3 acquisition was tested quasi-longitudinally. The results suggest that learners follow general developmental trajectories for both syntax and morphology. These trajectories did not seem to be susceptible to modification by structural similarity between the participants’ L1 or L2 and their target L3. Further, no effects of the learners’ psychotypological estimates of the distance between the native and the target language were found. Hence, it is concluded that developmental trajectories constrain L1 or L2 transfer effects. Only when a learner is developmentally ready may structures from L1 or L2 transfer to L3.