A Corpus-based Comparative Study On Formulaic Sequences Between Chinese And British Senior High School Students' Written English

In the past,formulaic language was regarded as a minor part in foreign language learning for the traditional viewpoint that language can be acquired by learning the grammar rules of the language and formulaic language can be neglected when considering the kernel of language.Accordingly,the role of grammar was over-emphasized in language learning.However,in the recent two decades,formulaic language has attracted a relatively high degree of research interest in cognitive linguistics,psycholinguistics as well as corpus linguistics.Alison Wray(2002)first named the formulaic aspect of a language as formulaic sequence,which was composed by more than one word and stored and retrieved as a whole unit in human brain.Formulaic sequence is often seen as a holistic language unit,which can be directly applied in language output.Recent researchers(Nattinger & DeCarrico,1992;Lewis,1993,1997)believe that formulaic sequence plays a significant part in enhancing language fluency and accuracy,saving processing effort,and achieving native-like language production,etc.They hold the belief that this kind of exemplar-based language input instead of rule-based input should be the basic input unit.Taking formulaic sequences as the starting point,the present study focuses on the authentic using of formulaic sequences in Chinese senior high school students' written English.The current study combines both quantitative and qualitative approaches to achieve two main research objectives based on two corpora.The first objective is to prove the positive role of formulaic sequences in improving writing performance.The second aim is to conduct a contrastive study on the use of formulaic sequences between Chinese learners and their native counterpart.To be more specific,the comparison was conducted among the distribution,structure characteristics and pragmatic functions of formulaic sequences between the two comparative groups.In the first place,the present study simulated a College Entrance Examination testing environment and collected participants' compositions which were finished within thirty minutes.The participants came from two province-leveled demonstrative high schools in Lanzhou,Gansu province.The present study divided the collected 369 compositions into four groups – from the fifth to the second rank according to the specified scoring standard.The effective formulaic sequences with their length from three to six-word and their occurrence frequency in the essays of the fifth rank(excellent level)were extracted with software AntConc 3.2.2 and manpower.Next,the present study put these formulaic sequences extracted from the fifth rank(excellent level)into essays of the fourth rank(good level),the third rank(moderate level)and the second rank(poor level)respectively and got their occurrence frequency respectively by using software AntConc 3.2.2.Then,the present study made three correlation analyses between the use of formulaic sequences and the corresponding scores using SPSS 20.0.During these correlation analyses,the scores of the participants were treated as dependent variable while the occurrence frequencies of the formulaic sequences in the three lower rank groups were regarded as independent variable.Finally,the results of the three correlation analyses showed that there was a significant correlation between the use of formulaic sequences and the writing scores.Thus,the first research objective is attained.In addition,in order to fulfill the second research aim,the present study took the self-built corpus and part of LOCNESS(Louvain Corpus of Native English Essays)as the comparative data source.All the three to six-word formulaic sequences were extracted from native corpus by using the same method of extracting effective formulaic sequences from learner corpus.The comparison was conducted among the distribution,structure characteristics and pragmatic functions between the two corpora.The present study adopted Altenberg's(1998)structure classification of formulaic sequences,according to which,there are three broad categories of formulaic sequences: full clauses(include independent/dependent clauses),clause constituents(include multiple/single clause constituents)and incomplete phrases.A series of Chi-square tests were applied to examine whether there existed significant differences of the occurrence frequencies among the distribution and different structure categories of formulaic sequences between the two comparative corpora.The research findings are as follows:(1)According to the results of Chi-square tests of the general distribution between the two corpora,the general distribution of formulaic sequences is rather significant.To be more specific,Chinese senior high school students use more three to six-word formulaic sequences in their written English than their native peers.But there is one thing in common,three-word formulaic sequences took the largest part in number and occurrence frequency while the number and frequency declined as the number of words constituting formulaic sequences increased in the two corpora;(2)Regarding the specific differences of structural characteristics,the present study found that among Altenberg's three broad structural types,clause constituents occupied the largest proportion in the two corpora,followed by full clauses,with incomplete clauses as the fewest type.According to the results of the Chi-square tests,there was a significant difference in independent clauses,dependent clauses and incomplete phrases between the two corpora.There was a difference in single clause constituents while there was no difference in multiple clause constituents.Although Chinese learners used more full clauses and clause constituents in written English than those of native peers,they employed less incomplete phrases in their written English.Chinese senior high school students tended to overuse and underuse formulaic sequences in their written English.To be more specific,they generally overused full clauses as well as clause constituents while underused incomplete phrases;(3)When analyzing the pragmatic functions of formulaic sequences of each category in the two corpora,the present study found that they shared more similarities than differences.Chinese learners used almost all the pragmatic functions which were similarly used by native peers.However,Chinese learners tended to use less diversified means in achieving these pragmatic functions.The structures of formulaic sequences were rather fixed,which means they memorized a great deal of formulaic sequences and when they were writing English composition,they simply retrieved what they think were useful in their writings;(4)The present study found three more pragmatic functions of single clause constituents on the basis of Altenberg's six functions of single clause constituents.They are reason expressions,result expressions and conditional expressions.Based on the findings of the present study,the thesis puts forward some suggestions on English teaching and learning and hopes that the suggestions can provide useful implications to language learners.