Working Memory Capacity and Relative Clause Attachment Preference of Persian EFL Learners: Does Segmentation Play Any Role?

Abstract

To contribute to a better understanding of first language (L1) and second language (L2) sentence processing, this study investigated the relationship between working memory capacity (WMC) and relative clause (RC) attachment preference of Persian learners of English as a foreign language . Additionally, the impact of segmentation of experimental stimuli on the participants' RC attachment preferences was explored. The participants' L1 attachment preference was also measured as baseline. Sixty-two native speakers of Persian participated in offline reading tasks in return for course credit. Results of WM test revealed no correlation between the participants' WMC and both their L1 (Persian) and L2 (English) attachment preferences. Results of the impact of the segmentation of the stimuli, on the other hand, showed when L1 and L2 experimental sentences are displayed with a break between the second determiner phrase (DP2) and the RC, readers prefer to attach the RC more to the first determiner phrase (DP1), especially in L2 (English). This finding provides support for Fodor's (1998) Implicit Prosody Hypothesis. In addition, the study found a positive correlation between L1 and L2 attachment preferences. Also, in order to explore the possible interactions between the variables under investigation, linear mixed effects model was run. Results revealed no interaction between the variables of the study. These findings might justify both interlingual and to some extent intralingual variations in attachment preferences. The findings of the study provide some implications for language teachers with regard to raising L2 learners' awareness for the prosodic aspects of language and the role of L1 transfer in L2 sentence processing.