Acquisition of English relative clauses by L1 Tamil speakers from rural areas in Malaysia

This study examined L1 Tamil learners’ acquisition of English Relative Clauses (RCs) from the perspective of the Universal Grammar (UG) theoretical framework. It studied the capability of the L1 Tamil speakers to reset the parameter of Tamil Relative Clause into the setting of English Relative Clause. Specifically the study investigated the development of English RC acquisition among three different proficiency levels of learners. The participants’ knowledge of English RCs was measured through their performance in judging the grammaticality of English RCs in a Grammaticality Judgement Test (GJT). The participants’ potential to produce grammatical English RCs was measured by using a Sentence Combination Test (SCT). The participants of this study were L1 Tamil learners selected from two secondary schools in Banting, Selangor. A total of 145 L1 Tamil speakers took part in this study. The participants were assigned to three different proficiency levels namely the Elementary,Intermediate and Advanced level based on their scores on the Oxford Placement Test 2 (Allan, 2004). The results of the GJT were analyzed using a one way ANOVA and Bonferroni post hoc tests. Results were compared among the three proficiency levels. In the production task, the results were also analyzed based on the participants’ age. The results showed that the acquisition of English RCs among the L1 Tamil speakers improved with prolonged exposure to and increased proficiency in English. The incremental trend suggested that with a higher level of English proficiency, the participants’ ability to judge English RCs also improved, in particular the subject and direct object RCs.However, the mean scores of correct judgement of acceptability showed that the participants’ ability to judge and accept grammatical items was better than their ability to reject ungrammatical items. This indicated that the participants’ underlying knowledge of English RC was non-native like since native speakers would equally be able to correctly judge grammatical items as acceptable and ungrammatical items as unacceptable. Even the Advanced L1 Tamil speakers accepted ungrammatical sentences (up to 43.84%) which were constructed in line with the non wh-movement L1 (Tamil) RC syntactic structure, i.e. they had resorted to the non wh-movement option. Another result that supported this explanation is that some of them produced sentences with resumptive pronouns, a construction considered ungrammatical in English although the use of resumptive pronoun/noun did decrease with increased proficiency. This further strengthened the claim that the participants’ default representation of English RC is the one with no movement. Since such operations in syntax involve functional categories and associated features, whatever functional categories that have been instantiated in the learners’ L1 during the critical period (Smith and Tsimpli, 1991, Tsimpli and Smith, 1995 as cited in Hawkins and Chan,1997) and which are available in the L2 would be acquirable. Further, if two options of a parameter are available in the L1, then one of the settings would be the default, and this tends to be carried over to the L2 acquisition process and is evident in the initial stages. Thus, the findings showed that the learners’ L2 grammar are UG constrained but what they lack are functional categories involved in English RC formation, which are not accessible once they have gone past the critical period. Thus, the results of this study support the Failed Functional Features Hypothesis (FFFH).

Preview PDF FBMK 2014 4RR.pdf Download (968kB) | Preview

Actions (login required)

View Item