Collocability Difficulty: A UG-based Model for Stable Acquisition

Abstract

An L2 learner is distinguished from another inasmuch as he/she has a good command on formulaic language like collocations which are deemed to be very specific to native speakers of a language being part of their competence and intuition. If this is true, it follows that English collocation, for instance, will be difficult for those learning it as an L2. Thus, this study is set to provide empirical evidence for collocabilty difficulties encountered by Yemeni Arabic-speaking advanced learners of English through investigating collocation errors deeply probing their sources, consequences and the cognitive strategies employed while acquiring such collocations. Sixty essays selected randomly from 107 given to Yemeni advanced learners majoring in English as home assignments. After identifying, classifying and tabulating errors, sources were classified into four categories, viz. L1-transfer, L2-influence, mutual and unrecognized.(30%) of the errorscommitted were ascribed to L1-transfer, (54%) to L2-influence, (13%) to mutual and (3%) to unrecognized. Based on such findings, I have proposed a UG-Model for learning collocations based on subcategorization and selectional restrictions. Though Yemeni learners are involved in this study, this model could be utilized for learning collocations in similar ES/FL contexts.

Keywords: Collocability Difficulties, Error Sources, UG-Model, Subcategorization and Selectional Restrictions