A Corpus Based Approach To Units Of Meaning In Chinese Learner English

The concept of "extended unit of meaning" is put forward by Firth and developed during a trend of "British contextualism" (Firth, 1957b; Halliday, 1966; Sinclair, 1991; Stubbs, 1993, etc). It is consistant with lexico-grammar, which associates syntactic structures with lexical items and is tested by many scholars using native corpus. Sinclair (1996:39) manifests through corpus study that lexical item is composed of five structural categories: an obligatory core, semantic prosody which is also obligatory, collocation, colligation, semantic preference and their interrelationships.The notion of extended unit of meaning is widely used in corpus study on the meaning of native English and many comprehensive conclusions are drawn. However, this notion is not used as much in the study of English as a L2 (second language). It is also the case in China, where most studies on the semantic characteristics of English as a L2 have their attention paid only to certain aspects of the semantic features. And few researches could come out with a systematic conclusion.Keeping this in mind, the present paper adopts a corpus-based Contrastive Interlanguage Analysis approach and applies the concept of 'extended unit of meaning' to the written English by Chinese learners. Taking the word serve as an example and using CLEC (Chinese Learner English Corpora) as a L2 corpus, this research makes a detailed analysis of the meaning of serve from all the aspects of extended unit of meaning, identifies the semantic characteristics of it in a L2, the similarities and differences of it in different corpora. With all the results identified, the current research also intends to explore the underlying reasons for the differences and their implications to cultural transmission and language learning.