1 Contrastive Chinese-English Academic Writing in English: Interpretations & Implications of Research Findings Sheena Gardner and Chao Han School of Humanities, Faculty of Arts and Humanities Coventry University, UK BALEAP 2017 BRISTOL ADDRESSING THE STATE OF THE UNION: WORKING TOGETHER = LEARNING TOGETHER
2 Outline The BAWE corpus and published contrastive Chinese-English L1 studies with Interpretations and Implications for teaching The Han CH-EN corpus research on Transition Markers with Interpretations and Implications for teaching Summary and Questions
3 Number of texts by level and disciplinary group6,506,995 words 2,896 texts from 2,761 assignments 1,953 written by L1 speakers of English; 245 (14%) by L1 speakers of a variety of Chinese 1,251 “distinction” and 1,402 “merit” 812 students, 72 (12%) L1 Chinese 1000+ modules & 300 degree courses BAWE Corpus Contents Year 1 Year 2 Year 3 Year 4 Arts & Humanities 255 229 160 80 Life Science 188 206 120 205 Physical Science 181 154 156 133 Social Science 216 198 170 207 Number of texts by level and disciplinary group
4 13 Genre families 5 Social Purposes Exercises ExplanationsDemonstrating knowledge and understanding Critique Essay Developing powers of independent reasoning Literature Survey, Methodology Recount, Research Report Building research skills Case Study, Design Specification, Problem Question, Proposal Preparing for professional practice Narrative Recount, Empathy Writing Writing for oneself and others
5 Summary BAWE corpus of successful university student writingCan be freely used to explore and compare the language of university disciplines, genre families and levels of study Not designed to be used for contrastive studies of different L1s… mitigated by differences across disciplines and genres Eg in Engineering vs Economics Eg design specifications vs case studies
6 Uneven numbers of texts by Chinese writers across disciplinesEngineering & Business NS – 50% CH – 30% Other NNS – 20% History & Classics NS – 90% CH – 0% Other NNS – 10%
7 Contrastive Chinese-English studies using BAWEChen, M. (2013) Overuse or underuse: A corpus study of English phrasal verb use by Chinese, British and American students. International Journal of Corpus Linguistics, 18 (1) Chen, Y.-H. & P. Baker (2010). Lexical Bundles in L1 and L2 Academic Writing. Language Learning and Technology, (14) 2, Gardner, S. (2016) A genre-instantiation approach to teaching English for Specific Academic Purposes: Student writing in Business, Economics and Engineering. Writing and Pedagogy 8 (1): DOI: /wap.v8i Lee, D. & X. Chen (2009) Making a bigger deal of smaller words: function words and other key items in research writing by Chinese learners. Journal of Second Language Writing, 18 (4) Lee, D. & X. Chen (2008) Small words, big deal: teaching the use of function words and other key items in research writing. In: Frankenberg-Garcia, A., Rkibi, T., Braga da Cruz, M., Carvalho, R., Direito, C. & Santos-Rosa, D. (eds) Proceedings of the 8th Teaching and Language Corpora Conference. Lisbon, Portugal: ISLA, pp Leedham, M (2015). Chinese Students’ Writing in English: Using visuals and lists. English Teaching in China (ETiC), 6 pp. 16–20. Leedham, M. (2015) Chinese students’ writing in English: Implications from a corpus-driven study. Routledge. Leedham, M. (2012) Combining intuition with corpus linguistic analysis: a study of marked lexical chunks in four Chinese students' undergraduate assignments. Nordic Journal of English Studies, 11(3), pp. 155–187. Leedham, M. (2012) Writing in tables and lists: A study of Chinese students' undergraduate assignments in UK universities. in Tang, R. (ed.) Academic Writing in a Second or Foreign Language: Issues and challenges facing ESL/EFL academic writers in higher education contexts. London: Continuum Nesi, H. & Moreton, E. (2012) EFL/ESL writers and the use of shell nouns. in Tang, R. (ed.) Academic Writing in a Second or Foreign Language: Issues and challenges facing ESL/EFL academic writers in higher education contexts. London: Continuum Smith, S. & Keng, N. (2013) The Acquisition of Classical Origin Words by Chinese, French, and Finnish Learners. Language Education in Asia, 4 (2), Zhao, C. (2014) Lexical Cohesion of Sino-British College Students’ EAP Writing. Theory and Practice in Language Studies, 4 (10)
8 Lexical Bundles(1) Chen and Baker (2010) Data: 3 corpora:1. BAWE-CH: 53 Chinese writers in BAWE 2. BAWE-EN: 60 ‘comparable’ English L1 writers in BAWE 3. FLOB: 80 Professional writers in FLOB FLOB – widest range of lexical bundles, C-BAWE smallest range Referential expressions (e.g. in the context of) more frequent in FLOB. Also in so far as, the degree to which, the extent to which, and to a large extent, Discourse organisers (e.g. in this essay I) more frequent in BAWE-CH and BAWE-EN than in FLOB. BAWE-CH tend to use in the long run, in the recent years, and all over the world, which are not found in the NS corpora.
9 Interpretations & ImplicationsThey say: “BAWE-EN and BAWE-CH contain more VP-based bundles [e.g. is one of the] and discourse organizers [in this essay I] than native expert writing does, which appears to be a sign of immature writing.” “ELT publishers or practitioners [should] put more emphasis on computer-retrieved formulaic language in the curriculum and/or materials.”
10 Lexical Bundles(2) Lee and Chen (2009)Data: 3 corpora of Applied Linguistics in English: 1. CAWE: 78 L1Chinese UG Dissertations 2. EXJA: 56 ‘Expert’ Published Journal Articles 3. BAWE-L 76 L1 English linguistics assignments Items most significantly used more by Chinese (1) than others (2 and 3) were items that tend to be ignored in academic writing instruction: function words (the, can); high frequency verbs (make) connectors (besides, according to) In CAWE we see make a conclusion/ study/ research, .. In BAWE we see reach a conclusion, conduct a study ….
11 Interpretations & ImplicationsA series of ‘small words’ are examined, issues are explored, and appropriate teaching strategies suggested. “Through structured exposure to genre-relevant samples of language use, apprentices can hone their intuitions of how certain phrases are used by expert writers, and learn the alternatives by example.” (2009:293)
12 Sample Exercises Step 1: Look at examples of make in NS writing. Is there an element of negative connotation in MAKE s.o. do s.th. Step 2: Look at examples from Chinese apprentice writers. Do you think these are appropriate uses of MAKE? Step 3: Do the above cases sound like direct translations from Chinese? [Note to the Teacher: The above sentences could be translated into perfectly appropriate Chinese sentences using (ling or shi), without any negative associations, unlike English.] Step 4: Rewrite Chinese sentences using more appropriate verbs than MAKE. Step 5: Raise awareness of two groups (HELP, CAUSE s.o to do s.th.) Step 6: Group others verbs in this pattern (ENABLE, AID, FORCE, COMPEL, … ) Step 7: Improve these sentences Make the learners become aware Make the news become more interesting Make the newspapers become more popular What is the pattern here?
13 Shell Nouns Nesi and Moreton (2012) BAWE –NS BAWE -NNS Focus: Formal and Functional Patterns of the most frequent shell nouns - change, effect, fact, factor, method, process, problem, result, system Detailed analysis. E.g. NS use fact for anaphoric and cataphoric reference, NNS tend to use it anaphorically. E.g. Redundancy in because of the fact that.
14 Interpretations & ImplicationsCautious interpretation of the findings because of the lack of comparable genres and disciplines for NS and NNS Make a case for teaching shell nouns – compared to discourse markers they are not widely taught. Specific activities suggested, based on examples in the corpus, include rewriting short paragraphs using shell nouns, and asking students to identify shell nouns in their own writing. Includes student examples of shell noun use. The circuit has now been set up with strain gauges to sense the stresses in the shaft, and these mechanical outputs have been converted into electrical ones and amplified to a sufficiently high level with the noise attenuated so the signal can be read by a voltage-measuring instrument. This system now needs to be able to be read and calibrated for data presentation. (0018a, Level 2, Engineering, Cantonese speaker)
15 Keywords, Figures, ListsLeedham (2015) conducted a series of studies 1. 29 texts by 4 CH students over three levels of study in Engineering, Food Sciences, & (2 from) HLTM (management) 2. All CH and EN writing in Biology, Business, Engineering, Food Sciences Use of preferred connectors (on the other hand, besides) decreases over years of study; as does informal language (what’s more, last but not least). Keyword comparisons of L1 Chinese and L1 English based on BAWE found higher frequencies in CH of Numbers, formulae, and references to data, such as according to [or refer to] figure/appendix/equation.. Use of visuals increases over the years for CH and EN
16 Two first year biology texts
17 Two second year Economics Texts
18 Interpretations & ImplicationsCH and EN language develops naturally over the years of study CH Effective strategies: Use of visuals and lists Tutors are happy with such formatting Where has it come from? Why are Chinese students using these strategies?
19 Studies so far: Focus on Comparisons of Others Lexical bundlesSmall words Shell nouns Visuals: Figures, Tables, Lists Lexical cohesion Genre instantiation approach Comparisons of BAWE NS & Other CH (Chen & Baker; Lee & Chen, Zhou) BAWE NS & BAWE NNS (Nesi & Moreton) BAWE NS & BAWE CH in 5 disciplines/ 4 learners (Leedham) BAWE NS & BAWE CH in Business, Economics & Engineering (Gardner) Others BAWE as reference corpus (Li & Wharton) BAWE as source of vocabulary (Smith & Keng)
20 Findings so far: Similar findings to previous contrastive studies with CH writers: less hedging than EN narrower range of connectives favouring ‘on the other hand’, ‘besides’ redundancies ‘we can see’ unusual collocations with ‘make’ ‘do’ idiosyncratic use ‘in the light of’, ‘in one word’ Helpful where the findings inform teaching strategies for small words shell noun functions use of figures In most cases the academic writing is ‘surprisingly similar’
21 Useful language in BAWE: due to the in MRs
22 Han CH-EN Corpus CH texts from L1 CH OSA in BAWEEN texts from L1 EN UKA in BAWE --- Matched by level, genre and discipline
23 across 9 genre families: across 4 levels:Texts of the Han CH-EN Corpus are across 4 disciplinary groups: across 14 disciplines: across 9 genre families: across 4 levels:
24 Paragraph & sentence length CH 25 Text length increases by level, CH 26 Transition Markers-ComparisonCH EN P-value per 1000 words however 1.440 1.911* 0.047 but 0.937 0.907 0.873 while 0.553* 0.127 0.000 on the other hand 0.206* 0.037 whereas 0.189* 0.069 0.024 meanwhile 0.087 0.004 0.054 in contrast 0.070* 0.012 0.025 similarly 0.057 0.066 0.787 on the contrary 0.043 0.017 0.311 at the same time 0.039 0.008 0.134 likewise 0.031 0.115 rather 0.028 0.034 0.737 conversely 0.022 0.035 0.490 by contrast 0.007 0.684 in the same way 0.010 0.002 0.456 correspondingly 0.320 whilst 0.075* 0.001 27 Transitions markers -ConsequenceCH EN P-value per 1000 words because 0.844 0.888 0.813 therefore 0.842 1.049 0.318 so 0.789 0.740 0.783 although 0.636 0.613 0.871 since 0.628* 0.331 0.048 thus 0.601* 0.340 0.023 though 0.201 0.105 0.253 nevertheless 0.179* 0.045 0.007 hence 0.166 0.238 0.347 while 0.104 0.062 0.352 as a result 0.101 0.049 consequently 0.109 0.907 yet 0.077 0.160 0.115 even though 0.025 0.064 0.111 nonetheless 0.005 0.144 as a consequence 0.018 0.003 0.286 thereby 0.015 accordingly 0.014 0.058 0.243 still 0.016 0.393 whilst 0.279* 0.000 Transitions markers -Consequence 28 Transitions markers -AdditionCH EN P-value per 1000 words and 1.962* 1.186 0.032 additionally 0.567 0.224 0.247 furthermore 0.340 0.257 0.343 also 0.207 0.097 0.064 in addition 0.202* 0.047 0.005 moreover 0.123 0.056 0.095 besides 0.070 0.017 0.138 again 0.041 0.085 0.180 further 0.004 0.024 0.187 29 1. however (EN>CH sig diff) --however by DisciplineP-value AbsFreq per 1000 per 1000 Law 47 3.13 52 3.47 0.751 Business 36 1.75 44 2.02 0.591 Food Science 21 1.23 23 1.52 0.673 Biology 20 0.93 29 1.56 0.183 Engineering 13 0.49 71 1.65* 0.014 30 1. the positon of however Sentence Initial: CH = EN e.g.Thus, the Dutch were able to replace the Portuguese’s monopoly position in trade with Asia, as well as developing trade with Africa and Americas. However, its trade position was weakened as the Dutch became involved in wars with Britain (CH1ESECO-0071a). Within a clause: EN > CH e.g. The monetarist approach however, emphasizes the role of expectations in distinguishing the long-run Phillips curve and the Short-run Phillips curve (EN2ESECO-399b). After a semi-colon: CH > EN e.g. It is said they wanted to understand if the mood state was related to gamble; however, researchers only assessed the anxiety/arousal levels (CH4CRPSY-0104b). 31 However used as a conjunction: EN > CH e.g.The ones with 0.5 and 0.8 dispersal rate maintained slightly above the initial population size, however this could not be recognized strikingly when dispersal rate was 0.2. 32 Interpretations and Implicationstransfer from Chinese (initial position) Where CH use visuals(more tables, figures, formulae), EN use words (more ‘however’s) English writers may lack grammar lessons (used however as conj.) 33 2. whereas (CH>EN sig diff)--whereas used as however in CH e.g. Whereas, to what extent their needs affect organisation productivity and how to deal with informal social power are not explicitly mentioned by Mayo. (CH4ESBUS-0124a) Interpretations and Implications: CH students might be confused about the use of however and whereas; 34 3. thus (CH>EN sig diff) --used as a conjunction in both subcorporae.g. Workers are more likely to form their own groups, thus team work in companies has much broader influence than that implied by formal labour-management ( Johnson 1949 ). (CH4ESBUS-0124a) The strain varies inversely with the square of the thickness, thus we can say that this is the most influential parameter. (EN2MRENG-0243a ) Thus adv. As a result of something just mentioned (OLDAE 2014: 835) Interpretations and Implications: for CH, “所以” is one equivalent of “thus”, which is used to link two clauses in a sentence—language transfer; for EN, because of the lack of grammar lessons; 35 4. contrast while (CH>EN sig diff.)e.g. To conclude, we can say that Britain succeeded in making the transition into " modern economic growth " while the Dutch did not. (CH1ESECO-0071a) It is interesting that the control sample was also translucent while the unpasteurized control sample was still cloudy. (EN2MRFS-6004d) 36 4. contrast while used as but in CHIn commodity-capitalist society, exchange-process seems to be dominant, while in fact, production plays a more decisive role in determining the magnitude of value. (CH4ESSOC-0319a) 37 4. contrast while-- four positions in CHIt can be easily seen that language use was identified as a very important factor and the difference is that males thought the language used at work was the second important factor just less than the age influence, while females thought a lot of overall language use. (CH4ESLIN-6058a) To conclude , we can say that Britain succeeded in making the transition into " modern economic growth " while the Dutch did not. (CH1ESECO-0071a) The former emphasizes the importance of the state intervention in economic development. While the latter claims that the less state intervention can make the national economy more competitive. (CH4ESPOL-0257d) Both of the studies depended mainly on participant observation and unstructured interviews to gather information. Goffman went "undercover" to be the director's assistant in the mental hospital; while Porter pretended his research topic was something other than racism, in order to ensure that the interviewees were not "pretending" when they were interviewed. (CH4ESSOC-0350c) 38 4. contrast while-- two positions in ENIt is interesting that the control sample was also translucent while the unpasteurized control sample was still cloudy. (EN2MRFS-6004d) In the lenis consonants , it is thought that , due to the air resistance caused by voicing , the frication is less intense, while the lack of voice in the fortis allows more air to be available during the period of frication. (EN4ESLIN-6044a) 39 5. whilst (EN>CH sig diff)Interpretations and Implications: EN favours whilst regarded by some as archaic 40 While vs Whilst While CH > EN Whilst EN > CHEN whilst > EN while whilst used most in Sociology Essays E.g. Whilst it holds true that there is a link between chronic illness and age, the way in which congenital abnormalities and illnesses acquired in childhood can be worked into a notion of 'biographical disruption' is uncertain. (EN4ESSOC-0405a) 41 Interpretations and ImplicationsChinese students may have language transfer, e.g. prefer to use initial-position however; e.g. thus to link two clauses; CH students might be confused about some items, like however vs. whereas; while vs. but; English writers may lack grammar lessons, e.g. use however and thus as conjunctions to link two clauses; The influence of teaching and text books for both CH and EN students warrant investigation; The nature of English may be changing through EN and CH ‘non-standard’ uses. 42 Concluding Remarks Student writing is different from published research Writing in [science] courses where Chinese students study is different from writing in [arts] courses Findings from studies in single disciplines can be checked in BAWE (e.g. ‘small words’) Examples in textbooks can be checked in BAWE (e.g. by comparison is rare, yet is presented alongside however, which is frequent in Oxford EAP B2 (de Chazal & McCarter, 2012) 5. IELTS prep may help CH writers in using figures and graphs; may hinder CH writers in producing longer, wordier texts 6. Most contrastive studies recommend ‘C-R’ activities and provide examples of teaching activities. 7. We welcome more analysis of CH writing in Han CH-EN and BAWE corpus, with more examples of teaching activities. 43 Using BAWE (all links and references at www.coventry.ac.uk/BAWE)Oxford Text Archive: the BAWE corpus is available to researchers at SketchEngine free access for anyone