Relative clauses and conjunctive adjuncts in Syrian University student writing in English

• Main Author: Fakhra, Amani
• Published: University of Warwick 2009
• Subjects: 410; PE English
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.511552

Initial investigations into English essays written by Syrian university students triangulated Syrian and British teachers’ evaluations of the essays and the lexico-grammatical features they identified as affecting the overall quality of writing, with text analyses of the sources, types and frequency of all grammatical errors. Following this, and a review of relevant literature, the thesis presents an in-depth study of relative clauses and conjunctive adjuncts as under-researched features in Arabic speaking university student writing that can enrich their writing syntactically and semantically. The relative clause (RC) analysis shows that the 'full' form RC occurred much more frequently than the 'reduced' form, and that confusion between these two forms was a prominent source of student error. 'Pronoun retention' errors indicating L1 interference were among the most frequent RC errors – as most studies of RC use by Arab learners find. Moreover, RC constructions with 'head noun' (or antecedent) in the non-subject position and 'gap' (or relativized NP/sentence) in the subject position were dominant, while other, and more complex, construction types were much less common. This supports the AHH and PDH hypotheses on the frequency/difficulty hierarchy of RC types. Conjunctive adjunct analysis reveals that 'additive' conjunctive adjuncts were more frequent, followed by 'causals'. Despite its informality, the resultive conjunctive adjunct 'so' was used most repeatedly, followed by 'also', 'but', and 'and'. Causal conjunctive adjuncts were most frequently misused, though in general conjunctive adjunct misuse is not a major weakness. Contrastive analysis between the L2 (Syrian) and an equivalent L1 (British) corpus of literature essays revealed no significant difference between the total frequencies of RCs, 'full' RCs and 'non-subject-subject' RCs. In contrast, the total frequencies of conjunctive adjuncts in the two corpora were significantly different, with the L2 corpus containing almost twice as many conjunctive adjuncts as the L1 corpus, particularly causals and additives, this latter category being most frequent in both corpora. The British students' employment of relative clause types and conjunctive expressions was generally more diverse than that of the Syrian students. Pedagogical implications conclude this thesis.