A Corpus-based Contrastive Study Of English Downtoners From The Perspective Of Extended Units Of Meaning With Almost As An Example

This paper takes one of the English downtoners—almost as an example to conduct a comparative investigation on English downtoners used by Chinese EFL learners and native speakers in their oral language from the perspective of extended units of meaning (Sinclair 2004). To be more specific, it exams the differences between these two groups of people concerning the use of English downtoners from four specific aspects— collocation, colligation, semantic preference and semantic prosody.Contrastive Interlanguage Analysis (CIA) (Granger 1996; 1998) is the method used in this research. Two corpora (COCA & SECCL) are used to provide authentic oral language materials containing almost. The research findings show that native speakers and Chinese EFL learners differ in frequency, collocation, colligation, semantic preference and semantic prosody of the English downtoner almost. In terms of frequency, Chinese EFL learners underuse almost in their speaking. As for collocation which is reflected by collocates, only four of the collocates (doubled/doubles, drowned/drown, every, forgot/forget) overlap with each other in COCA and SECCL, which indicates that the collocations of almost used by Chinese EFL learners have significant differences from those of native speakers. As for the four mentioned simultaneously retrieved collocates, although Chinese EFL learners have mastered some of their typical collocation patterns with almost, their use of collocation patterns are much limited compared with those of native speakers. Besides, there are cases of misuse. For example, in the collocation pattern of almost every+noun (plural form), Chinese EFL learners tend to add nouns of plural form after every. However, native speakers prefer to add nouns of singular forms after it. Another example is "almost forget", native speakers tend to use "almost forgot" instead of "almost forget". As regards colligation, most of the colligation types shared by both groups of people are used less frequently by Chinese EFL learners such as RBIN, RB_CD_NN/NNS etc. And only three of them are used more often by Chinese EFL learners. They are RB_VBZ/VBD/VBP, Be_RB_VBN and RB_DT_NNS. Besides, some other colligation types used by Chinese EFL learners are not used by native speakers such as RB_MD, RB_PRP/NNP_VB etc. And in contrast, some of them are used by native speakers but not by Chinese EFL learners such as RB_DT_NN, RB_VBZ_IN, RB_CD etc. As for semantic preference, native speakers tend to use almost to modify adjectives expressing "a kind of ability or disability", "a state of mind or the character of a person". Also it can modify adjectives expressing "a lack of something" and "having the same features of another person or thing". Additionally, they tend to use almost to modify adverbs whose meanings are related to the notion of "time" and verbs related to "number". However, Chinese EFL learners prefer to use almost to modify verbs meaning "causing damage to something", "shedding tears" and "coming to an end". Moreover, they use almost to modify nouns expressing "certain time and certain place." The semantic preference of almost between these two groups of people are mainly different. Though they have similarity in using almost to collocate words indicating "time", the collocates of almost in COCA are mainly adverbs while those in SECCL are mainly nouns. As regard semantic prosody, native speakers tend to use almost in positive prosody more often, but this is not the case for Chinese EFL learners, they prefer to use it in neutral and negative contexts.The significance of the research exist in its implications in language teaching and learning, the applicability of extended units of meaning in the study of vocabulary, textbook design and dictionary compiling. Firstly, corpus can be a useful tool in language teaching and learning because of the large amount of authentic language materials collected in it. Additionally, Chinese EFL learners can learn to make word choices according to their extended units of meaning, taking collocation, colligation, semantic preference and semantic prosody into consideration. Secondly, the application of extended units of meaning provide researchers with a more comprehensive perspective to study vocabulary because it realizes the co-selection of lexicon, syntax, meaning and pragmatic functions (Zhang 2004). Thirdly, the design of the textbook had better try to increase students' awareness of extended units of meaning of words by providing explanation of words from this perspective and by designing some related exercises. Last but not least, dictionary compilers can adopt the theoretical frame of extended units of meaning to explain a word in a comprehensive way so that learners can not only know the meaning of the word but also the way to use the word.