The transfer of discourse level writing skills from xhosa L1 to English L2

Abstract

This research aims to contribute to the improvement of writing in English as a second language in South African schools. It is based on transfer theory: what is known in one language may transfer to another. Much of the transfer of structural aspects of languages as different as Xhosa and English is likely to be negative. This research focuses instead on aspects of writing at discourse level in the expectation that a positive transfer of learning will take place at that level. The main hypothesis is that certain discourse level writing skills transfer to a second language (English) if they have been taught in the mother tongue (Xhosa) but not in the second language. The skills in question are • using topic sentences appropriately to introduce a paragraph • writing suitable support sentences in the rest of the paragraph • achieving paragraph unity in relation to the topic sentence • using linking words and other cohesive devices effectively. The writing corpus was obtained from 66 Grade 8 learners in the researcher’s school. Thirty three of the learners were in the Experimental Group in 2005 and 33 different learners were in the Control Group in 2006. The difference between the groups was that the Experimental group were taught the writing skills in Xhosa but not in English, whereas the Control group were taught the same skills in both subjects. Both groups were taught Xhosa and English by the researcher himself. The corpus comprises a total of just under 80 000 words of composition writing, half of which was written at the beginning of the academic year and the other half at the end. Comparisons were made to determine how much learning had taken place individually and by the different groups and subjected to statistical analysis to measure significance. The findings provide persuasive evidence of a transfer of learning. The impressive amount of learning that occurred in the first place was also rewarding, providing proof, as it did, that learners are capable of responding to greater demands than we might assume.