The analysis and assessment of linguistic proficiency : an examination of the spoken English of advanced foreign students

• Main Author: Mendelsohn, David Julian
• Published: University of Edinburgh 1977
• Subjects: 410
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.657691

This thesis examines linguistic proficiency in the spoken English of a highly homogeneous group of Israeli undergraduates. Linguistic proficiency is defined as: "mastery of the language system together with native-like fluency". The thesis is divided into two parts: Part I: Analytic Assessment of Six Parameters of Linguistic Proficiency. Recorded data of the Israeli informants and a control group of native speakers of English is analysed and assessed on each of the following six parameters of linguistic proficiency: pronunciation at the segmental level; production error; syntactic maturity; lexical richness; intonation; word-stress. Some of the assessment procedures used are those developed by other linguists, but most have been specially developed for this study. The different analyses yield a rank order of the informants on each of the parameters. However, intonation and word-stress are found to be non-discriminating and therefore are not investigated further in Part II. Part II: Linguistic Proficiency and the Judgement of Oral Proficiency. In this part a series of experiments is administered in which thirty-five foreign-language teachers (native and non-native speaking) judge sets of data for overall ability in language. The data, unbeknown to the judges, is modified in different ways so as to control for different parameters, in order to examine the possible influence of the parameter in question on overall assessment. Conclusions. (i) The judges are shown on the whole to be able to make reliable and valid overall judgment. (ii) The linguistic proficiency parameters of segmental pronunciation and productive error (classified as "deviance-judged" parameters) are shown to be exerting a major influence on overall judgment. Lexical richness ("quality judged") is shown not to be exerting a serious influence, and the results for syntactic maturity ("quality judged") are inconclusive. (iii) A comparison of judgment by the native and non-native speaking groups of judges shows that there is little difference between the judging ability of the two groups.