A Corpus-Based Study Of "VP+NP Patterns In High School Students’ Writing And Teaching Strategies

Co-selection,as the central mechanism of language description(Sinclair,1996),includes multiple categories:co-selection of words and words,of words and grammar,of pattern(a combination of grammar and meaning)and meaning.Within the framework of co-selection,Sinclair puts forward the model of extended unit of meaning,which in turn makes the theory operational.It consists of five components:node,collocation,colligation,semantic preference,and semantic prosody.Semantic preference refers to the shared semantic features of collocates,or the semantic group to which they belong;while semantic prosody is the attitudinal meaning or pragmatic function of discourse,revealing the communicative purpose of the speaker.The model of extended unit of meaning is a systematic and centralized interpretation of co-selection theory and can serve as an important descriptive mechanism in corpus linguistics.The meaning of a single verb often needs to be realized in collocations(Sinclair,1991).VP+NP pattern consists of a verb phrase followed by a noun phrase.It has many diversified sub-types,such as V+N(refers to v-n collocations),V+N+C structure,V+S structure(refers to clause)and so on.On one hand,a large number of studies have shown that learners tend to misuse and overuse some high-frequency sub-types of VP+NP patterns,like V+N collocations.On the other hand,the study of VP+NP patterns in previous research mostly focuses on misuse by learners.Meanwhile,the studies on the features of this pattern used by second language learners are limited to college learners.However,very few studies explore the usage of VP+NP English patterns used by primary and secondary school learners,especially within the framework of the co-selection theory.In view of this,this study takes VP+NP pattern as the research object,which aims to reveal the features of and the factors for VP+NP patterns in English writings by high school English(CHSE)learners and develop an improved teaching model for VP+NP patterns teaching in high school.The corpora used in this study includes LOCNESS(Louvain Corpus of Native English Essays)and WECCHSS(Written-English Corpus of Chinese High School Students)(a self-built corpus).Within the framework of co-selection,this study compares and analyzes the VP+NP patterns that occur in native corpus and self-built corpus..The research questions to be addressed are as follows:(ⅰ)Compared with native English learners,what is the characteristic of colligations of VP+NP patterns in CHSE learners’ writings?(ⅱ)Compared with native English learners,what is the characteristic of semantic preference and semantic prosody of VP+NP patterns in CHSE learners’ writings?(ⅲ)How can the above findings be employed to improve the teaching model of VP+NP patterns?This study employs the corpus-based method.WECCHSS is used as the observation corpus,and LOCNESS,which is a written English corpus of native learners,is selected as the reference corpus.Three high-frequency verbs,namely "MAKE","FIND",and "HAVE",are selected for observation.For each verb,100 concordances are randomly selected from WECCHSS and LOCNESS respectively.The main findings of this study are as follows:First,although CHSE learners have grasped some knowledge of colligations of VP+NP patterns,they differ from native speakers in terms of frequency of use and richness of collocates.CHSE learners tend to overuse delexicalized V+N collocations and the collocates they are familiar with.Besides,they create some unidiomatic sequences.While native speakers tend to use more diversified collocates.What’s more,CHSE learners tend to use V+S rather than V+N+C.In addition,"FIND+it+NP+to+VP" is a unique pattern only used by CHSE learners.Second,CHSE learners also tend to use specific patterns where target verbs co-occur with collocates of certain semantic features in order to express specific meanings and functions.However,they are considerably different from native English speakers.For example,Chinese learners tend to use delexicalized V+N collocations and "FIND+S"where NPs or noun phrases are of "positive act" or "positive results" semantic preference to express an attitude of "suggestion" "encouragement" or "approval".While native English speakers tend to use these patterns to construct semantic prosody of "neutral" or"disapproval".For another example,native English speakers are prone to use "MAKE"causative structure in which "NPs" are "inhuman" and "VPs/APs" are of "negative"semantic preference,building semantic prosody of "negative" and "disapproval".While Chinese students tend to use "human NPs" and "VPs/APs" of "positive" semantic preference,expressing an attitude of "approval".Third,the improved teaching strategy is divided into three steps.Based on the observation and summary of data,teachers can use corpus for teaching.The data of this study shows that Chinese learners tend to underuse and misuse many word sequences,such as MAKE causative structure and HAVE+NP collocation.First,teachers present relevant word sequences through real corpus data so that students can promote their perception and implicit learning through a lot of exposure and immersion.Second,according to the findings of this study,the characteristics of native speakers’ usage are explained to the learners by means of comparative presentation.Third,through the design of corpus-based situational dialogue,English-Chinese translation and other exercises,students’ use of relevant patterns can be promoted,deepening their understanding of the meaning and attitude of these patterns.This research has certain value for exploring the characteristics of CHSE learners’ VP+NP patterns,and for analyzing and solving the learners’ learning difficulties.First,learners may have deviations in patterns,and these deviations can be revealed through corpus research;secondly,these deviations may not be solved through explicit learning or teaching,but can be solved through corpus methods and authentic data.This study also has certain limitations.On one hand,the self-built WECCHSS is relatively small compared with large corpus.On the other hand,the description of semantic prosody of VP+NP patterns is made manually,which is influenced by subjective judgment to a certain extent.