The Role of Individual Differences in the Acceptability of Island Violations in Native and Non-native Speakers

Abstract

This study examines the acquisition of syntactic island constraints on wh-movement in English by native speakers of Najdi Arabic to test whether it is possible for second language learners (L2) to acquire syntactic constraints that are not present in their first language (L1). According to the Full Transfer/Full Access Hypothesis (Schwartz & Sprouse, 1996), L2 properties are potentially acquirable by adult L2 learners regardless of L1. However, according to the Interpretability Hypothesis (Tsimpli & Dimitrakopoulou, 2007), adult L2 learners cannot acquire uninterpretable features in the L2 if those features were not selected in the L1 during the critical period. The study tested 82 English native speakers and 72 Arabic learners of English, using a grammaticality judgment task. The results showed that Arabic learners, like English native speakers, were sensitive to syntactic island constraints on wh-movement as reflected in their lower acceptability judgments of ungrammatical island violation sentences (e.g., *what does the worker worry if the boss leaves__?), supporting the Full Transfer/ Full Access Hypothesis. This study also investigates the source of island effects that cause low acceptability judgments of ungrammatical island violation sentences. Under grammatical syntactic theories, island effects are due to violations of syntactic constraints that prohibit wh-extraction from islands. Under the resource-limitation theory (Kluender & Kutas, 1993; Hofmeister & Sag, 2010), however, island effects are due to processing difficulty because islands are complex and require additional processing resources that are beyond the capacity of most native speakers. To tease apart these contrasting theories of island effects, the present study, like Sprouse et al. (2012), focused on individual differences in processing resources, which play a crucial role in sensitivity to island effects under the resource-limitation theory but not under grammatical theories of island effects. Specifically, this study tests the relationship between working-memory capacity and sensitivity to island effects by using two measures for each individual, a measure of working-memory capacity (i.e., the operation span scores) and a measure of sensitivity to island effects (i.e., the DD scores). Neither English native speakers nor learners provided evidence of a relationship between operation span scores, which measure working-memory capacity, and DD scores, which measure sensitivity to island effects, contrary to the prediction of the resource-limitation theory. These results suggest that island effects are not driven by limited processing resources and are more likely due to syntactic constraints.