PATTERNS OF LEXICAL COHESION IN A MULTIPLE-CHOICE CLOZE TEST: DIFFERENCES BETWEEN ENGLISH AS A SECOND LANGUAGE LEARNERS AND NATIVE SPEAKERS OF AMERICAN ENGLISH

The purpose of this research was to investigate patterns of lexical cohesion in a multiple-choice cloze test produced by native speakers of English and learners of English as a Second Language (ESL) to determine tendencies in the semantic patterning of lexical ties across sentence boundaries. The framework for the analysis of semantic patterns was the theoretical treatment of lexical cohesion in English by Halliday and Hasan (1976, Chap. 6); one aim of this study was to investigate the validity of this framework in a research setting involving second language learners. The research question that this study investigates is whether or not the strategies underlying the selection of a lexical alternative in a discourse context are dependent upon language proficiency.;In terms of group agreement on the most appropriate choice to be inserted in the blanks of the multiple-choice cloze test, low ability non-native speakers exhibited more variability in their response to an item than did the other two groups. The results of the analysis of the statistical association among items across sentence boundaries within each of the three groups indicated three tendencies: (a) low ability subjects tended to relate selections to other choices in the immediate discourse context (i.e., contiguous items or items in close proximity), (b) the native speaker group showed a greater perception of the interdependence in item choice than the groups of non-native speakers of English, although the exact nature of the interdependence was not determined, and (c) low ability subjects exhibited the semantic strategy of repetition and partial repetition more often than the other groups of subjects with the native speaker group showing the fewest instances of repetition. The appropriateness of the multiple-choice cloze test as a test instrument for discourse studies is discussed and the implications of the results for English as a Second Language (ESL) pedagogy are presented.;ESL learners at two second language proficiency levels as determined by the Test of English as a Foreign Language (TOEFL) from an Oriental first language background (Chinese, Japanese, and Korean) were compared with a control group of native speakers of American English on a task involving a multiple-choice cloze test.