Maggie Charles

This paper argues for an approach to LSP in which teachers and students compile their own do-it-yourself corpora using specialised texts in their area of study and teaching. I show that it is relatively easy to build a rough and ready corpus of this type and illustrate how it can be used both indirectly in the construction of materials for a German reading course and directly by EAP students taking a course on editing doctoral theses. It is suggested that both LSP teachers and students can benefit from such a corpus due its specificity and relevance to their needs. For teachers, a DIY corpus provides a means of familiarising themselves with the discourse of a specialised area and a source of authentic examples for materials production, while students particularly value the corpus as a lexico-grammatical reference resource.

Do-it-yourself (DIY) corpora can be defined as small-scale databases of electronic texts built by users for specific, limited and local purposes. Such corpora can be of great benefit to both teachers and students of English for academic purposes (EAP), who, with recent software advances, can now construct DIY corpora to their own specifications relatively easily. For teachers, who may be tasked with giving courses related to disciplinary areas in which they have little or no expertise, the specialist DIY corpus provides an opportunity to examine a body of texts that they have selected as relevant to the target course and thus enables them to familiarize themselves with the discourse of the discipline in question. Such corpus-based investigations not only facilitate teachers' understanding of disciplinary norms and practices, but also provide examples for use in class or in course materials and give frequency information on lexis and phraseology so that instructional decisions can be made on a sound evidential basis. For graduate students, who have needs that are highly specific, it is also valuable to construct DIY corpora from material within their own field. This tailor-made resource provides individual, preciselytargeted information, which can be drawn upon to answer lexicogrammatical queries both at the composing and editing stages of the writing process. This paper makes the case for the use of DIY corpora in EAP contexts, illustrating the argument with two sets of examples: first, a teacher's use of a corpus of theses in creating a course for doctoral students in materials science and second, students' uses of DIY corpora for individual writing support.

Corpora are planned collections of naturally occurring texts which are stored and accessed on a computer. They have been used in English language learning since the 1980s, particularly in the development of writing skills. With the spread of internet connectivity and the advent of web-based corpora, it has become easier for students to consult a corpus, which has led to an increase in their use independently and in class. Corpus tools consist of two elements: the corpus itself and the software to analyze it. These are often packaged together, but may also be combined by the user according to their individual needs. Unless otherwise indicated, all corpora and tools mentioned here are freely available or have some free accessibility; URLs are given in Table 1. Corpora for writing students may be divided into four types. First, large general corpora contain a representative sample of written and spoken English in several registers, including academic discourse, journalism and fiction; examples are the British National Corpus (BNC) and the Corpus of Contemporary American English (COCA) (Davies, 2008). The Web itself may also be consulted as a corpus through the WebCorp Web site. Secondly, there are specialized corpora containing only texts of a certain type; for example, the British Academic Written English Corpus (BAWE) (Nesi & Gardner, 2012) and the Michigan Corpus of Upper-Level Student Writing (MICUSP) contain texts in several different disciplines written by university students and given high grades. A wide range of other academic and professional corpora are accessible through Hong Kong Polytechnic University (HKPU). The third type, learner corpora, contain texts written in English by nonnative speakers. The International Corpus of Learner English (ICLE) (Granger, Dagneaux, Meunier, & Pacquot, 2009), composed of essays by advanced level university students, is available for purchase. The fourth option is to build a doit-yourself (DIY) corpus. At lower English levels, writers can benefit from a pedagogical corpus containing the material in their own textbook(s); students can also work on a corpus of their own writing, individually or as a group, while advanced academic writing students can compile their own discipline-specific corpora.

This study proposes a pedagogic approach which combines discourse analysis with corpus investigation and reports on students' feedback on the materials. Forty-nine international graduates evaluated the corpus work on a five-point scale. Results show that attitudes were generally very favourable: percentages agreeing with positive statements about the tasks ranged from 62% to 96%. The study examines the practical implications of using hands-on concordancing in large mixed-discipline EAP writing classes and concludes by proposing a three-stage set of goals for such classes on a cline of increasing corpus competence, from basic corpus awareness through corpus literacy to advanced corpus proficiency. 2.

Using a corpus-based approach, this paper investigates the construction of stance in finite reporting clauses with that-clause complementation. The data are drawn from two corpora of theses in contrasting disciplines: a social science-politics-and a natural science-materials science. A network for the analysis of reporting clauses is presented which sets out the major alternatives available to academic writers and enables stance to be linked systematically to grammatical and semantic patterns of use. Quantitative and qualitative analysis of the data leads to the identification of an important, but somewhat under-researched, function of reporting clauses in academic writing: their use to report the writer's own work. Drawing on the notions of averral and attribution, the paper shows how writers can emphasize or hide their responsibility for their own propositions and thereby construct a stance which differs according to the epistemology and ideology of the discipline concerned. These reporting clauses play a key role in the construction of major claims, with greater writer visibility seen in politics than materials. However, despite the superficial objectivity and impersonality of writing in the natural sciences, it is argued that skilled exploitation of the interplay between averral and attribution allows writers to construct a stance that is both clear and pervasive.

This paper deals with the use of adverbial connectors (e.g. thus, however) in native-speaker undergraduate writing. Drawing on data from the British Academic Written Corpus (BAWE), it examines the frequency and grammatical patterns of 17 adverbials in four disciplines: Business Studies, Politics, Chemistry and Computer Science. It shows that connectors vary considerably in frequency according to discipline and genre. More detailed investigation of the adverbial thus reveals its characteristic patterns and functions. The paper argues that the teaching of text connection needs to focus both on the patterns of individual adverbials and on the functions they perform in context.

This study investigates the phraseological patterning that occurs in reporting clauses used to make references to others’ research. It examines finite reporting clauses with that-clause complement and draws upon two corpora of theses written by native speakers in contrasting disciplines: approximately 190,000 words in politics/international relations and 300,000 words in materials science. The findings show that both disciplines use significant

This paper reports on the long-term use of personal do-it-yourself corpora by students of EAP. Forty international graduate students attended a course in which they built and examined their own corpora of research articles in their field. One year after the course, they completed an email questionnaire, which asked about their corpus use in the 12 months since the end of the course. Results show that 70% of the respondents had used their corpus: 38% were regular users (once per week or more), 33% irregular users (once per month or seldom) and 30% non-users. Most users consulted the corpus for checking grammar and lexis while composing and revising and 93% of them considered that corpus use had improved their academic writing. Reasons for non-use included the small size of the corpus and its lack of reliability and convenience. Case studies of a user and a non-user are presented and highlight two other factors likely to affect take-up: the individual’s writing process and the focus of their current writing concerns. The paper discusses the reasons behind long-term use of personal corpora and some of the challenges to be overcome in extending the approach more widely.


This volume explores the interaction between two traditions of investigating written academic prose that might broadly be called 'discourse analysis' and 'corpus linguistics'. The two traditions have much in common. Both take selected examples of naturally occurring discourse as their starting point. Both attempt to identify recurring patterns in those examples. Both relate their findings to the social, intellectual or ideological contexts in which the discourse plays a role. The priorities of the two approaches do tend to diverge, ...

Using personal corpora independently: ‘too much effort’ or ‘a lovely friend’?
Maggie Charles
University of Oxford Language Centre
maggie.charles@lang.ox.ac.uk

Over the last two decades there have been many accounts of direct corpus use with language learners, particularly those studying academic writing in English (see Boulton 2010; Yoon 2011). To date, the majority of this work has evaluated the success of corpus consultation immediately on completion of an in-class corpus intervention. As Pérez-Paredes et al. (2013) note, fewer studies attempt to examine independent corpus use over the longer term. One notable exception is Yoon (2008), who tracks six students’ corpus use over a six month period; however, her study provides qualitative rather than quantitative data on continuing corpus use. Work by Charles (2013, in press) reports on independent corpus use by 40 students one year after their corpus course, finding that 70% of respondents were corpus users. However that study does not shed much light on the reasons behind use or non-use of the corpora.
To investigate this question further, the present paper reports on a new set of data from 72 students who responded to an on-line survey one year after using corpora on an EAP course. During this course, students built their own personal corpus from research articles in their field and used it to explore discourse functions in their discipline. The data reported here is part of the more wide-ranging survey and focuses particularly on factors that are likely to promote ongoing corpus use and those that may hinder or prevent take-up. The quantitative results are supplemented by qualitative data from interviews with two students, a corpus user and a non-user.
Respondents were first asked whether they had used their corpus at any time since the academic writing course had ended; 41 students (57%) had done so, while 31 (43%) had not. The two largest groups of users were those who were currently using their corpus (15, 37% of users) and those who had used their corpus in the past, but were not using it at the time of the survey because they were not doing any academic writing (16, 40%). Some users mentioned other factors which negatively affected their corpus use: preference for other resources, lack of time and lack of usefulness were each noted by 3 students (7%), while technical problems were important for just 1 user (2%). The findings for non-users presented a similar pattern: the biggest single reason for non-use was that respondents had not done any academic writing (11, 36% of non-users). Seven students (23%) preferred other resources and 6 (19%) did not find the corpus useful. Lack of time was cited by 4 (13%) and lack of experience by 2 (6%), while technical problems affected one student (3%).
The first conclusion that can be drawn from this data is that when students work independently, the need for writing resources such as corpora is likely to be sporadic. This has implications for the provision and content of corpus courses, as well as for their timing. Two specific factors that have a negative effect on continuing corpus use also stand out: lack of time and a preference for other resources, often perceived as quicker and more convenient to use. The importance of these factors is also underlined by student ratings of the potential disadvantages of their personal corpus: 41 students (58% of all respondents) considered lack of time and lack of convenience to be very important or important.
However the likelihood of take-up or rejection of personal corpus use also depends upon individual student concerns, as illustrated by the contrasting attitudes reported by two students. Ahmad  characterised his personal corpus as ‘like having a lovely friend with you who can advise you any time you want.’ Piotr, however, concluded that personal corpus use ‘just took too much time, too much effort’. In accounting for this difference, it is noteworthy that Ahmad’s self-reported writing needs were mainly lexico-grammatical. Thus it was worth his while to build up his corpus to over a million words and he incorporated corpus use as a proofreading tool within his pre-existing writing practices. Piotr, however, was mainly concerned with the coherence of his text. Since such issues of overall textual organisation lend themselves much less easily to corpus investigation, he saw little or no benefit in investing time and effort in corpus construction and use.
This paper presents and discusses further the data on factors affecting independent personal corpus use and draws out the implications for fostering corpus consultation among students.
References
Boulton, A. 2010. “Learning outcomes from corpus consultation”. In M. Moreno Jaén, F. Serrano Valverde and M. Calzada Pérez (eds.), Exploring new paths in language pedagogy: lexis and corpus-based language teaching. London: Equinox, 129-144.
Charles, M. 2013. “Student corpus use: giving up or keeping on?” In A. Lenko-Szymanska (ed.) TaLC10: Proceedings of the 10th international conference on teaching and language corpora. Warsaw: Institute of Applied Linguistics, University of Warsaw. Available online at http://talc10.ils.uw.edu.pl/proceedings/.
Charles, M. 2014. “Getting the corpus habit: EAP students' long-term use of personal corpora”. English for Specific Purposes 35: 30-40.
Pérez-Paredes, P., Sánchez-Tornel, M. and Alcaraz Calero, J. (2013). “Learners’ search patterns during corpus-based focus-on-form activities”. International Journal of Corpus Linguistics 17 (4): 482-515.
Yoon, C. 2011. “Concordancing in L2 writing class: an overview of research and issues”. Journal of English for Academic Purposes 10: 130-139.
Yoon, H. 2008. “More than a linguistic reference: the influence of corpus technology on L2 academic writing”. Language Learning and Technology 12 (2): 31-48.

This presentation reports on students’ use of corpora one year after they took a 6-week corpus course. The paper first describes the 6-week, 12-hour course in which students learnt how to build and use corpora to explore discourse in their own field. I then report 5 years of questionnaire data on 134 respondents who took the corpus course between 2009 and 2013 and were surveyed one year later. Participants were graduates from many different countries, studying in a wide range of disciplines. Results show that 68% of students continued to use their corpus one year after the end of the course and 57% consulted their corpus frequently (once per week or more). The vast majority of users, 92%, considered that using their corpus improved their academic writing. These data indicate that students see the corpus as a valuable long-term resource that they can access independently to support their learning. I conclude by arguing for greater use of corpora in EAP classes.

This paper reports on the long-term effects of a course on academic writing, in which students built individual ‘do-it-yourself’(DIY) corpora of research articles in their field and used them to explore the discourse of their own discipline (Lee and Swales 2006). I first give details of the 6-week, 12-hour course and then report 5 years of questionnaire data on 144 respondents who took the course between 2009 and 2013 and were surveyed one year later. Participants were graduates from many different socio-cultural backgrounds, studying in a wide range of disciplinary contexts. The aim of the study was to investigate the nature and extent of independent corpus use and the attitudes to corpus consultation that prevailed over the longer term. Results showed that 91 students (63%) continued to use their corpus one year after the end of the course and 52 users (56%) consulted their corpus frequently (once per week or more). The vast majority of users, 83 (91%), considered that using their corpus improved their academic writing. These data indicate that students see their corpus as a valuable resource that they can access independently to support their writing. The use of DIY corpora thus helps satisfy not only the immediate, but also the on-going writing needs of international students, which often continue long after the end of their EAP courses. I conclude by arguing for the greater use of DIY corpora in academic writing classes, where they offer a freely available technological resource of great benefit.


Lee, D., and J. Swales. (2006) A Corpus-Based EAP Course for NNS Doctoral Students: Moving from Available Specialized Corpora to Self-Compiled Corpora. English for Specific Purposes 25 (1), pp. 56–75.

All tooled up: Corpus-assisted editing for academic writers

Teaching and Language Corpora Conference, Giessen, July 2016

Maggie Charles

Oxford University Language Centre

This paper reports on a course which aims to teach students how to use corpus tools for editing their texts. Although much has been written about the benefits of data-driven learning, particularly for academic writing (for a review, see Yoon, 2011), less attention has been paid to the potential of individual corpus tools for addressing specific editing concerns. The present course is targeted at doctoral students who have already completed part of their thesis in draft form. After an initial session introducing corpus work, students built two corpora: 1) research articles (RAs) in their own field; 2) their own writing. The freeware AntFileConverter (Anthony, 2014) was used for converting batches of pdf files to plain text format to build the RA corpus and AntConc (Anthony, 2015) for editing purposes. Class sessions provided demonstrations of how specific tools can be used for editing, followed by individual practice in which students used the tool to edit their own writing.

The course has run nine times and evaluation data are available for 66 students (41% natural sciences; 30% social sciences; 29% humanities). All participants gave a positive answer to the question ‘Is it helpful to use your corpus and AntConc for editing?’ (79% yes definitely; 21% yes probably). Students were asked to rate the individual tools for editing purposes as very useful, useful, fairly useful, of little use or not useful. Combining the very useful and useful categories shows that, unsurprisingly, the most highly rated tool was Concordance with 95% of responses. This was followed by Clusters (82%), Collocates and Keyword List (both 74%), N-grams (70%), Context Searching (67%), Concordance Plot (63%) and Word List (59%).

While the utility of concordancers and other tools that show collocations has been discussed in the literature (e.g. Flowerdew, 2015), I argue that tools such as Keyword List, Concordance Plot and N-Grams have affordances that are particularly relevant to students who are editing texts. For example, the N-Grams tool can be used to make a list of all the 3-grams in the student’s own writing and compare it with those in their RA corpus, thereby revealing differences in phraseology. Issues concerning the content of the text can be addressed using both Keyword List and Concordance Plot. A keyword list of one section or thesis chapter compared to the rest of the text identifies the words that occur more (or less) frequently than expected. This tool can therefore reveal the most salient words in a section or chapter and thus the extent to which the writer deals adequately with the topic under discussion. Concordance Plot provides a graphic representation of the distribution of a search term throughout the corpus files. When the term chosen is central to the student’s argument, this tool can show how the content develops over the course of the whole text. The present paper discusses further the course and the affordances of the corpus tools for editing, illustrating the findings with examples of student searches.

References
Anthony, Laurence. 2014. AntConc (Version 3.4.3) [Computer Software]. Tokyo, Japan: Waseda University. http://www.laurenceanthony.net/
Anthony, Laurence. 2015. AntFileConverter (Version 1.2.0) [Computer Software]. Tokyo, Japan: Waseda University. http://www.laurenceanthony.net/
Flowerdew, Lynne. 2015. Using corpus-based research and online academic corpora to inform writing of the discussion section of a thesis. Journal of English for Academic Purposes 20. 58–68.
Yoon, Choongil. 2011. Concordancing in L2 writing class: An overview of research and issues. Journal of English for Academic Purposes 10. 130–139.

This thesis uses a corpus-based approach to investigate the construction of stance in two contrasting disciplines: a social science - politics; and a natural science - materials science. The study draws upon two corpora, each of which consists of eight theses written by native speakers in the discipline concerned. Investigations of the following grammatical features are carried out: interpersonal and textual adverbs; adjectives in introductory it patterns; nouns followed by a that-clause complement; nouns preceded by sentence initial deictic This; and
reporting verbs followed by a that-clause complement. In each investigation, the quantitative data from concordance lines are supplemented by consideration of contextual information in
order to arrive at a qualitative interpretation of the findings. It is shown that writers make use of the grammatical features in order to construct the stance of an insider in the discipline, of one who is familiar with disciplinary norms and competent to engage in disciplinary practices.
The thesis also compares and contrasts both the extent to which politics and materials deploy the grammatical features and the functions they are used to perform in the two corpora. This comparison reveals how specific differences in the use of the grammatical features can be linked to differences in ideology and epistemology between the two disciplines. The study concludes by discussing and evaluating the use of corpus based methods in research on stance.

This study proposes a pedagogic approach which combines discourse analysis with corpus investigation and reports on students' feedback on the materials. Forty-nine international graduates evaluated the corpus work on a five-point scale. Results show that attitudes were generally very favourable: percentages agreeing with positive statements about the tasks ranged from 62% to 96%. The study examines the practical implications of using hands-on concordancing in large mixed-discipline EAP writing classes and concludes by proposing a three-stage set of goals for such classes on a cline of increasing corpus competence, from basic corpus awareness through corpus literacy to advanced corpus proficiency. 2.

This paper investigates using data-driven learning (DDL) as a means of stimulating greater lexicogrammatical knowledge and reading speed among lower proficiency learners in an extensive reading program. For 16 weekly 90-minute sessions, an experimental group (12 students) used DDL materials created from a corpus developed from the Oxford Bookworms Graded Readers, while a control group (10 students) had no DDL input. Both classes were required to read a minimum of 200,000 words during the course. An embedded-experiment design (Edmonds & Kennedy, 2017) was adopted consisting of both qualitative and quantitative forms of investigation. Quantitative data from the Vocabulary Levels Test by Nation and Beglar (2007) and a C-test (Klein-Braley & Raatz, 1984) constructed from an upper-level Bookworms reader found statistically significant lexicogrammatical improvements for both groups, but greater improvement took place within the control group. Qualitative data derived from a repertory grid...

Do-it-yourself (DIY) corpora can be defined as small-scale databases of electronic texts built by users for specific, limited and local purposes. Such corpora can be of great benefit to both teachers and students of English for academic purposes (EAP), who, with recent software advances, can now construct DIY corpora to their own specifications relatively easily. For teachers, who may be tasked with giving courses related to disciplinary areas in which they have little or no expertise, the specialist DIY corpus provides an opportunity to examine a body of texts that they have selected as relevant to the target course and thus enables them to familiarize themselves with the discourse of the discipline in question. Such corpus-based investigations not only facilitate teachers’ understanding of disciplinary norms and practices, but also provide examples for use in class or in course materials and give frequency information on lexis and phraseology so that instructional decisions can be...

This paper investigates using data-driven learning (DDL) as a means of stimulating greater lexicogrammatical knowledge and reading speed among lower proficiency learners in an extensive reading program. For 16 weekly 90-minute sessions, an experimental group (12 students) used DDL materials created from a corpus developed from the Oxford Bookworms Graded Readers, while a control group (10 students) had no DDL input. Both classes were required to read a minimum of 200,000 words during the course. An embedded-experiment design (Edmonds & Kennedy, 2017) was adopted consisting of both qualitative and quantitative forms of investigation. Quantitative data from the Vocabulary Levels Test by Nation and Beglar (2007) and a C-test (Klein-Braley & Raatz, 1984) constructed from an upper-level Bookworms reader found statistically significant lexicogrammatical improvements for both groups, but greater improvement took place within the control group. Qualitative data derived from a repertory grid...

Aarts’ definition highlights the semantic elements that characterise this logical relation, focusing particularly on the mismatch between the information given in the two clauses. Thus, in example (1) taken from the corpus described in Section 2, the expectation implied in the although clause is that an estimate would lead to inaccuracy, but the information in the matrix clause shows that this expectation is not fulfilled:

This paper argues for an approach to LSP in which teachers and students compile their own do-it-yourself corpora using specialised texts in their area of study and teaching. I show that it is relatively easy to build a rough and ready corpus of this type and illustrate how it can be used both indirectly in the construction of  materials forGerman reading course and directly by EAP students taking a course  on editing doctoral theses. It is suggested that both LSP teachers and students can benefit from such a corpus due its specificity and relevance to their needs. For teachers, a DIY corpus provides a means of familiarising themselves with the  discourse of a specialised area and a source of authentic examples for materials production, while students particularly value the corpusa lexico-grammatical reference resource.

1. Acknowledgements 2. List of Contributors 3. Editors' foreword 4. Introduction: Corpora, grammar, and discourse analysis: Recent trends, current challenges (by Groom, Nicholas) 5. Chapter 1. Pattern grammar and transitivity analysis (by Thompson, Geoff) 6. Chapter 2. Using COBUILD grammar patterns for a large-scale analysis of verb-argument constructions: Exploring corpus data and speaker knowledge (by Romer, Ute) 7. Chapter 3. "Hugh's across all that": Some changing uses of prepositions (by Francis, Gill) 8. Chapter 4. The textual functions of lexis (by Stubbs, Michael) 9. Chapter 5. Examining associations between lexis and textual position in hard news stories, or according to a study by... (by Hoey, Michael) 10. Chapter 6. I mean I only really wanted to dry me towels because ...: Organisational frameworks across modes, registers, and genres (by Warren, Martin) 11. Chapter 7. Probably most important of all: Importance markers in academic and popular history articles (by Bondi, Marina) 12. Chapter 8. Chatting in the academy: Informality in spoken academic discourse (by Buttery, Paula) 13. Chapter 9. General extenders in learner language (by Aijmer, Karin) 14. Chapter 10. Language description and language learning: The pedagogic corpus and learners as researchers (by Willis, Dave) 15. Chapter 11. "What we contrarians already know": Individual and communal aspects of attitudinal identity (by Bednarek, Monika) 16. Chapter 12. Does Britain need any more foreign doctors? Inter-analyst consistency and corpus-assisted (critical) discourse analysis (by Baker, Paul) 17. Publications by Susan Hunston 18. Author Index 19. Subject Index

Do-it-yourself (DIY) corpora can be defined as small-scale databases of electronic texts built by users for specific, limited and local purposes. Such corpora can be of great benefit to both teachers and students of English for academic purposes (EAP), who, with recent software advances, can now construct DIY corpora to their own specifications relatively easily. For teachers, who may be tasked with giving courses related to disciplinary areas in which they have little or no expertise, the specialist DIY corpus provides an opportunity to examine a body of texts that they have selected as relevant to the target course and thus enables them to familiarize themselves with the discourse of the discipline in question. Such corpus-based investigations not only facilitate teachers’ understanding of disciplinary norms and practices, but also provide examples for use in class or in course materials and give frequency information on lexis and phraseology so that instructional decisions can be...