2.2 Review Question 2:How Does Move Structure Vary...
Among the variation studies(Stage 3),possible factors causing variation in schematic structure were explored.Discipline and first language are the two most frequently examined variation factors(see Figure 5).As mentioned above,different part-genres have varied communicative purposes and schematic structures.Since the factors leading to the variation in schematic structure in each section may not be the same,the variation in each section is explored separately and reported according to the type of variation.
Figure 5 Number of studies by the factors of variation[1]
Disciplinary Variation
Studies on Introductions show evidence of disciplinary variation.When Swales developed the CARS model in 1981,he noted the schematic structure differ in the hard and soft sciences.Swales combined the rhetorical functions identified in both hard and soft sciences and made the model suitable for all disciplines.He thus aimed to develop a generalized model rather than examining the variation.The studies in Stage 2 share the focus on the generalizability of the model and aimed to test the applicability of the model in a specific discipline.Disciplinary variation was identified in studies of this stage.Lakic(1997)applied the model to Economics texts while Posteguillo(1999)examined articles in Computer Science.Lakic(1997)found that the step of Reviewing Previous Studies is obligatory while Posteguillo(1999)noted that it is optional.The contradictory results may be associated with disciplinary variation.Lakic(1997)highlighted the importance of the step by suggesting that it“establishes a solid theoretical basis”which he thought is necessary in Economics(p.413).Posteguillo(1999),however,indicates that the less occurrence of reviewing literature in Computer Science articles is possibly a result of“the relative newness of this discipline”(p.154).
A number of studies reported similar findings to those of Posteguillo(1999).The researchers also observed schematic structures unique to emerging disciplines.Samraj(2005)compared texts from two Biology sub-disciplines,comprising one emerging field and one established field.Habibi(2008)investigated texts from three Applied Linguistics sub-disciplines,including English for Specific Purposes,Psycholinguistics and sociolinguistics.Kanoksilapatham(2012)examined three Engineering sub-disciplines.The studies revealed differences in move structure between emerging and established fields.In both Applied Linguistics and Engineering,Introductions in emerging sub-discipline contain more persuasive elements.The findings from the three studies all show more occurrence of the step Claiming centrality in emerging fields.Samraj(2005)suggests that authors in the emerging fields feel the need to let the audience understand the importance of their research area.Kanoksilapatham(2012)adopted Swales’model(2004)which contains the step of Presenting positive justification.She found the step occurred three times higher in the emerging field than the established field,which is statistically significant(p＜0.001).The findings of Samraj(2005)and Kanoksilapatham(2012)show more occurrences of Announcing principal outcomes in the emerging fields,indicating that researchers in emerging field want to show the significance of their study at the outset of their research to attract more attention.The schematic structure in emerging fields across hard and soft sciences share similar features,which suggests that the variation across disciplines is more likely to be related to the distinction between emerging fields and established fields rather than the hard and soft sciences.
Three move models were developed for the Method section,including Nwogu’s model(based on Medical articles),Kanoksilapatham(2005)(Biochemistry)and Lim(2006)(Management).Both Nwogu(1997)and Kanoksilapatham(2005)have the move Describing experimental procedure in their model.It probably can be attributed to the fact that experiment is the most common data capture method in Medical Science and Biochemistry.The second move in Lim’s model Delineating procedures for measure variables is not seen in the other two models.Management research examines the characteristics of human activities which are complicated and hard to capture.Researchers usually need to break the constructs down to measurable variables.There are various data capture methods in Management,such as questionnaire,interview,experiment and obtaining data from secondary sources.Lim’s model shows that Management Methods focus on what data to capture rather than the process to capture them.Though the steps vary significantly among the three models,they all include a data sampling move,a data capturing move and a data analysis move.
The studies on the Results section do not reveal much disciplinary variation.For example,Williams(1999)and Posteguillo(1999)applied Brett’s model developed based on Sociology articles to analyse the schematic structure of the Results section of Medical and Computer Science articles and found the model suited the texts in the two disciplines.
In regard to Discussions,Holmes(1997)applied Swales’8-move model which was revised from Hopkins and Dudley-Evans’model(1988)in texts from three social science disciplines,namely History,Political Science and Sociology.Holmes found variation in both frequency and order of moves across the three sets of texts.Lim(2010)applied the commentary moves proposed by Basturkmen(2009a),who analysed texts from Applied Linguistics and Education.He found that the steps of Explaining the findings and Evaluating the findings have more occurrences in Applied Linguistics texts.The difference between the two sets of texts is statistically significant(p＜0.05).Basturkmen(2012)applied Yang and Allison’s model(2003)with her revised commentary steps in Dentistry Discussions and compared the results with her previous study in Applied Linguistics(2009).She also observed more frequent use of Explaining the findings in Applied Linguistics.A possible explanation for the higher occurrences of the step in Applied Linguistics,as Basturkmen suggested,might be that the results in humanities and social sciences emerged from“reasoning and interpretation”and therefore require more explanation while results in physical sciences which are obtained from laboratory activities are self-explanatory(p.142).This explanation,however,cannot account for the relatively lower occurrence of the step in Education which is also a discipline in social science reported in Lim’s study(2010).
First Language Variation
First language variation studies are also known as contrastive genre studies.Thirty-three studies in the key word map examine variation between texts written in different languages or in English by native/near-native and non-native speakers of English(see Figure 5).A number of studies reported variation between English and other languages in the schematic structure of RA Introductions.The languages under investigation comprise Hungarian(Árvay &Tankó,2004),Persian(Mahzari &Maftoon,2007;Rashidi &Ghaffarpour,2010),Brazilian Portuguese(Hirano,2009),Chinese(Loi &Evans,2010)and Thai(Kanoksilapatham,2007).First language variation studies often control the possible disciplinary variation by collecting articles from the same discipline.Although the socio-cultural context of the contrasted languages varies significantly,the RA Introductions in non-English languages share many similar features with English Introductions.Across the studies differences were identified in Move 2 Establishing a niche and Move 3 Occupying a niche.It was reported in several contrastive studies that non-English writers incline to make a research space by raising questions rather than by indicating a gap or counter-claiming(Árvay &Tankó,2004;Loi &Evans,2010;Mahzari &Maftoon,2007).It was found that non-English speaking researchers avoid evaluating and criticizing others’works.Researchers suggest that the face-saving principle of the cultures under examination is responsible for L2 writers’reluctance in taking a critical stance(Hirano,2009;Loi &Evans,2010).However,Sheldon(2011)notes that the research context in non-English academic communities is less competitive and thus researchers do not have to criticize others to make a research space.
Another difference between English and non-English Introductions lies in Move 3 Occupying the niche.Compared with English Introductions,the texts written in other languages have fewer occurrences of Move 3 steps,such as Introducing research question,Claiming significance of the study and Announcing principal findings.The fewer occurrences of Move 3 steps shows that non-English authors prefer to write in an indirect manner and imply meanings.Some important information is left unarticulated for readers to figure out themselves(Duszak,1994;Fakhri,2004).Reader/writer responsibility is used to account for this variation(Loi &Evans,2010;Sheldon,2011).Readers of non-English languages take more responsibility in the communication than English readers.Non-English readers are expected to be more actively involved in interpreting the meaning of writing.In the reader-responsible texts,authors give more context information and write in an indirect way.Readers are expected to figure out the implied meaning on their own.English is considered a writer-responsible language in which writers state their intentions clearly.The research is therefore announced in an overt manner in Move 3.The frequent use of Indicating RA structure in English writing is regarded as being reader-friendly,which is a feature of writer-responsible texts(Sheldon,2011)compared with fewer occurrences of the step in non-English texts(Duszak,1994;Fakhri,2004).
For Discussions,Amirian and Tavakoli(2009)examined the difference in schematic structure between English and Persian Discussions by adopting Swales’8-move model(1990)and found differences in the move Un/expected outcome,Explanation and Exemplification.They accounted for the variation using socio-cultural factors.
Variation over Time
Li and Ge(2009)reported statistical significant difference in the step Presenting background information(Nwogu,1997)between the articles published in two time periods.They observed that the step became obligatory in the more recent set of texts,indicating that medical researchers placing more importance on presenting a clear picture at the beginning of their report.Sheldon(2011)also noticed changes in schematic structure over time.He compared his findings in Spanish L2 writing with that from a study conducted more than a decade ago(Burgess,1997)and found that Spanish writers use more Move 2 Establishing a niche in their English writing and their writing became more similar with the native English speaker articles,suggesting L2 writings were increasingly influenced by English writing style.
For the method section,Li and Ge(2009)investigated the evolution of move structure over time.Their study revealed a significant difference in the frequency of the step Describing data-analysis procedures.The step became obligatory in the texts published in 2000-2004 period,compared with only 40% occurrence in the 1985-1989 sub-corpus.Moreover,there was a significant rise in length of the step.The rise in frequency and length was explained as a result of the growing complexity of data analysis method and authors’increased awareness of using data analysis procedures to prove the reliability and validity of their research outcome.
For the result section,variation across time periods was identified in Li and Ge’s study(2009).They applied Nwogu’s model and found significant difference in the step Indicating non-consistent observations.It is less used in the more recent set of texts.Li and Ge explained the avoidance as a strategy to get more chance for the paper being published.It may be the result of the increasing belief that papers reporting positive results are more likely being accepted.Árvay and Tankó(2004)examined the difference between the Introductions in theoretical and empirical studies and found no significant difference.
Variation across Research Approaches
Lim(2010)explored the variation in schematic structure among qualitative,quantitative and mixed method studies.No statistical significant difference was found across the three groups.
Apart from the variation mentioned above,researchers have also explored genre differences between novice and experienced writers.For example,Basturkmen(2009)investigated the variation in the move of Commenting on Results between 10 Master’s Theses and 10 research articles.Parkinson(2011)compared the way of building arguments in the Discussion section of 30 RA in Physics and 185 lab reports written by undergraduate students in Physics.Dressen-Hammouda(2014)examined genre features of the writing of six academics in Geology over ten years after obtaining their PhD.
In sum,the majority of variation studies reviewed reported variation to some extent.Disciplinary variation was revealed in the schematic structure of Introduction,Method and Discussion.The variation was influenced by different factors in different sections:the emerging or established status of the disciplines under examination influences the variation in the Introduction section;the common research method used in different disciplines lead to the variation in the Method section;on the other hand,the factor underpinning the disciplinary variation in the Discussion section is unclear.There was not enough evidence to demonstrate the existence of disciplinary variation in the Result section.Variation in schematic structure across languages was found in contrastive research on different languages.The findings from these studies showed that English articles were written in a more direct,reader-friendly manner and it was the writer’s responsibility to convey information clearly rather than leave it for readers to figure it out.The evolutionary change in schematic structure over time was also revealed.
Based on the findings of the variation studies reviewed above,the researcher,when designing the corpus for the present study,used discipline,time and research approach as controlling variables.That is,the data were collected from one single discipline(i.e.Applied Linguistics),and mainly from three controlled years(The number of Master’s Theses submitted in the three years in the New Zealand community is not sufficient for the present study so we have to collect data from a wider time span.).In addition,possible variation of research approach and gender of author is also explored.
【注释】
[1]The total number exceeds the total number of variation study because some studies explore more than one factor at the same time.