Abstract
This thesis explores island constraints in the L2 English of Norwegian learners, in connection to issues of transfer and learnability. Despite claims of universal treatment, current research has found cross-linguistic differences; English rejects all island constraint violations, while Norwegian accepts some of the same violations. Under standard views on transfer, the insensitivity to island constraints in the L1 Norwegian speakers’ grammar would predict a corresponding (if slightly reduced) insensitivity in the L2 English grammar. Accordingly, this thesis aims to uncover whether the participants’ L2 grammar display parametric settings equivalent to the suggested L1 grammar, i.e., whether island constraints and island insensitivity are subject to transfer. The research focused on three syntactic constructions: embedded questions, relative clauses and subject phrases. Additionally, the thesis examines whether the universal account of island constraints can be maintained despite cross-linguistic differences.
The L1 Norwegian and L2 English grammar of an experimental group and the L1 English grammar of a control group were examined through acceptability judgment tests. The tests were developed in accordance with the factorial design (Sprouse, 2007). Analysis of the judgments resulted in inconclusive findings regarding learnability issues in SLA. However, based on the experiments’ results, previous research and current theories on cross-linguistic influence, I argue for indications of transfer in the participants’ L2 grammar, aligning with the FT/FA-model (Schwartz & Sprouse, 1994, 1996). Additionally, the results align with previous findings suggesting an extended complementizer domain in Norwegian (e.g., Nyvad, Christensen, & Vikner, 2015), which enables a universal account of syntactic islands, despite cross-linguistic differences.
