Investigating Interdisciplinary Research Discourse: the case of Global Environmental Change

Lead Research Organisation: University of Birmingham
Department Name: English Drama American and Canadian Stu

Abstract

It is generally accepted now that many real-world problems are best addressed by a number of disciplines working together rather than by individual disciplines alone. In the UK, research councils promote interdisciplinary research activity and universities in turn encourage academics to collaborate with colleagues in other disciplines.

It is not always easy, however, for researchers in different areas to cooperate, because each discipline has methods of working, expectations, value systems and ways of talking and writing that are special to that discipline and not easily shared. We believe that it is important for institutions, research councils and researchers to have a fuller understanding of what the distinctive features of discourse practices in interdisciplinary research are and of how they differ from discourse practices in conventional disciplines.

As a step toward this goal we propose investigating the discourse of a successful journal in an interdisciplinary field: Environmental Change. We will study the extent to which this field operates as a unified whole, the extent to which journal authors in the field broaden their messages to a multidisciplinary audience, and the extent to which each discipline in the field maintains a discrete identity. In order to investigate the field we will analyse the journal 'Global Environmental Change'.

We will include in our study every article published in the journal since its inception in 1990. The primary methodology we use is Corpus Linguistics, that is, using specialised software to analyse large quantities of written text. Many studies of individual disciplines, using corpus techniques, already exist, but these techniques have not yet been applied to an interdisciplinary field. We will use an approach to the description of the linguistic features of texts that will make it possible for us to cluster texts according to degrees of correlation in their linguistic profiles; this approach, developed by Douglas Biber, is called 'multidimensional analysis'. In addition, we will investigate the recurrent phraseologies of texts in this field and then cluster texts with similar phraseological profiles. Using both multidimensional and phraseological clusters, we will see whether the texts cluster following discipline boundaries. This is an innovative approach to corpus-based investigations of research discourse: instead of placing the texts into categories according to external criteria (for example, according to 'discipline'), we will group texts by text-internal features. We will also look at the citation practices (do authors tend to cite within their discipline?) and how writers and readers are represented in the texts (do writers address their readers as experts in their discipline, for example?).

We will compare the results of our investigation of the discourse of interdisciplinary research in that journal by comparing it to samples of texts taken from other journals, five representing other interdisciplinary fields and five representing specific disciplines. This will allow us to determine whether interdisciplinary research discourse is distinct in its features, and also to to see how much variation in discourse practices there may be between interdisciplinary fields and within disciplines. We will look at this across time, to see whether discourse practices change as a field becomes more established and as an interdisciplinary community develops.

To complement our analyses of texts, we will also conduct surveys and interviews of people involved in the journal publication process: editors, reviewers and authors. The data can help us to explain some of the phenomena that we observe in the texts; the influence of editorial policies, for example, or the ways that authors from different disciplines collaborate to write texts.

The work will support the research infrastructure in the UK by increasing our understanding of interdisciplinary research discourse.

Planned Impact

There are four major user-groups for this research.

1) The research councils who advise on and implement government policy in relation to research funding:

For these groups, encouragement of interdisciplinary research is a major challenge; identifying and removing any obstacles to such research, and indeed identifying how such research is successfully carried out, would facilitate interdisciplinary endeavours. Our project will not do everything that would be needed to meet this aim but it will establish ways of working with interdisciplinary discourse that will be the foundation of future studies. Our research will inform the funding councils on the following questions: how easy do researchers in different disciplines find it to build on each others' work in carrying out their own? in an interdisciplinary field, of which Environmental Change is an example, to what extent are there differences in terminology (especially words and phrases which are the same in form but different in meaning) that are likely to cause problems of understanding? are the styles of papers written from the different disciplines similar enough to escape comment or different enough to cause problems of interpretation? what kind of training might researchers in the constituent disciplines of a novel interdisciplinary field need in order to facilitate cooperation? More importantly, perhaps, we will test our methods for answering these questions so that they can be applied to other interdisciplinary fields.

2) People involved in the publication of interdisciplinary research output, directly and indirectly:

These include commercial publishers (such as Elsevier, Wiley-Blackwell, Taylor and Francis) and the large numbers of researchers that act as editors, editorial board members and reviewers. More specifically, within these operations, the findings will interest: company staff with responsibility for managing IDR journals; the editors and the editorial boards; and the staff who assist in the provision of training and writing support to authors. The findings can inform editorial and publisher policy decisions, review processes and the guidance provided to authors.

3) Those involved in conducting or managing IDR, within academia and in commercial research centres:

Researchers who are entering IDR environments may not be aware of the communication challenges faced in IDR; the description of the features of ID discourse will help them to understand what the rhetorical demands of writing for a broader audience are. Similarly, managers of IDR activities, whether at a local project team level, or at the research centre level, will benefit from the answers that we provide regarding training needs, and the possible obstacles to effective communication.

4) Teachers of Study Skills and English for Academic Purposes (EAP), many of whom work in service units within universities and prepare students particularly for postgraduate study and research:

A large proportion of their 'clients' are international students, who bring in substantial income to the UK as a whole, and upon whom the economic health of the Higher Education sector increasingly depends. In many cases, such students report that their work is interdisciplinary in scope and, furthermore, that they can experience difficulties meeting the demands of supervisors from different disciplines. EAP teachers make use of extant research into 'academic prose' in general and into the language and discourse of specific disciplines - this leaves them ill-prepared to respond to the needs of students undertaking interdisciplinary research. Our project will inform EAP and Study Skills teachers on the following question: what are the typical linguistic features and rhetorical practices of interdisciplinary research discourse?

Publications

10 25 50
 
Description The project used innovative methodologies to investigate the language of an interdisciplinary research journal (Global Environmental Change) over two decades, in the context of a larger set of five monodisciplinary and five further interdisciplinary journals. We wished to know how researchers write for an interdisciplinary audience, and how linguistic methods could be adapted to investigate the discourse of interdisciplinary research.

There are four key findings from the research: two relating to the nature of interdisciplinary research discourse and two relating to methodology.

The comparison through computational analysis of 53 linguistic features in the eleven selected journals demonstrated that the ID journals cannot be simply distinguished from the MD ones using linguistic features. That is, the ID journals do not form a distinctive cluster; there is no consistent interdisciplinary 'style'. One implication of this is that there may be different 'flavours' of interdisciplinary writing. There is variation between journals (tested through random forests analysis which showed that our method predicted the assignment of articles to journals with 70% accuracy) but also within journals. The consistency of journal style ranged from below 40% to above 80%.

There is evidence, however, from the corpus analysis and also from the interview data that writing for an interdisciplinary audience requires more space (longer articles) in order to make assumptions, concepts and approaches clearer for a broader audience. Writing for an interdisciplinary audience typically contains more explanation and more 'signposting'. The interdisciplinary journal writer enacts two complementary positions: how the approach they take offers a new perspective to other disciplines (divergence); and how it simultaneously contributes to a shared enterprise (convergence).

The project demonstrated that it is possible to adapt an existing methodology (Multi-Dimensional Analysis) to identify dimensions that are specific to a specialised corpus and, using those dimensions, to identify 'constellations' of papers that are distinctive in their dimensional profile. Six constellations were identified in Global Environmental Change, each representing a distinctive epistemological take on a common research challenge. Our research suggests, however, that the MDA model could be improved by adjusting the set of linguistic features that the analysis uses and we recommend this as a path for future research work.

The project also discovered and demonstrated the value of the technique known as topic modelling to explore topics not only in individual texts but also in text-parts (through an automated slicing technique). This analysis provides an overview of the content of a corpus, encapsulating the content and positioning of a large number of texts. It indicates how topics change or remain constant over time, and also how different articles may differ from or resemble each other. The method is linguistically simple but provides rich information on large collections of text, and on text structure. Our approach to topic-modelling will be of interest to a range of researchers, including those interested in text and data mining, and in corpus and computational linguistics.

The project focused on journals related to the environment, and further research needs to be conducted on interdisciplinary research in other areas. There is also room for further development of the two methodologies, using different datasets.
Exploitation Route The findings have implications for the ways that interdisciplinary research funding bids are assessed. One implication is that more space is needed for elaboration of ideas for a broad audience in funding applications, and for demonstrating how the partners complement each other; concomitantly, it is advisable to ask interdisciplinary teams to provide sections of text written for specialist audiences which would show the strength of each partner within the interdisciplinary team. We are interested in working with RCUK on recommendations for new application procedures.

The findings can be used by publishers, by interdisciplinary team managers and by writing tutors who support writers preparing to write for interdisciplinary researcher audiences so that they can understand the particularities of the rhetorical demands, and the linguistic strategies available to them for making their research accessible.

The new multidimensional model needs to be tested and refined on other research article datasets. Diachronic MD analyses can indicate to editors and publishers how balanced the mixture of disciplines within the journal is, and this can feed into editorial policy decisions. The potential of topic modelling in analysing journal content datasets needs to be explored and exploited too, to identify the expansion and recession of research themes, of emerging concepts and techniques, and to assist in data mining projects.
Sectors Environment,Other

URL http://idrd-bham.info/idrd/
 
Description We have written, with the assistance of a professional technical writer, a booklet on 'Preparing to write for an interdisciplinary audience'. This was published on the Elsevier PublishingCampus website and has now been moved to the new Researcher Academy site (URL for document: http://tinyurl.com/y2fn77wr), and is aimed at researchers worldwide who are writing their first or second paper for an interdisciplinary audience. Researcher Academy has 8000 online visitors a day. The content of the booklet is built around the findings of the project. On 23rd March 2018, the PI, Paul Thompson, led a workshop on writing for an interdisciplinary research journal audience for thirty international early career researchers at the SPP SeaLevel Early Career Scientists (ECS) Meeting in Hamburg, based on this research. On 15th July 2019, Paul Thompson was a key trainer on a webinar for researchers worldwide (603 participants) on 'How to write for an interdisciplinary audience' organised by Elsevier Researcher Academy.
First Year Of Impact 2015
Sector Other
Impact Types Societal

 
Title Research journal MDA model 
Description We have developed a six dimension multi-dimensional analysis model, based on Biber's 1985 approach to MDA but where his framework described the dimensions of general language use, our model addresses the discourse of published research articles. This is the first model of its kind. It allows the researcher to place research articles at points on six different dimensions and then to cluster articles according to these linguistic feature dimensions. 
Type Of Material Computer model/algorithm 
Year Produced 2016 
Provided To Others? Yes  
Impact The model has been applied to our own dataset but there are no notable impacts to report as yet. 
URL http://reshare.ukdataservice.ac.uk/
 
Description Partnership with Elsevier UK 
Organisation Elsevier
Department Elsevier UK
Country United Kingdom 
Sector Private 
PI Contribution We are providing the partner with linguistic analyses of the holdings of eleven of their journals and findings that can be used in management and development of their research journals, and in providing guidance to authors and editors.
Collaborator Contribution The partners have given us full access to the holdings of eleven journals, and have worked with us to select the journals using confidential data from their database, and they have also conducted analyses of their citation data, as part of the overall research design.
Impact A booklet on 'Preparing to write for an interdisciplinary journal: Understanding the needs of your multidisciplinary audience' has been prepared for publication on Elsevier's PublishingCampus website (https://www.publishingcampus.elsevier.com/).
Start Year 2013
 
Description 2014 Interdisciplinary Conference, IAS, Durham 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact This was a presentation made to researchers, managers and administrators of interdisciplinary research, representative of the identified end-user groups for the project, from different parts of the world. My talk raised considerable interest as it explored the language used in interdisciplinary research journal articles and how this differs from writing in specialist journals, and this is a perspective that noone else at the conference had offered. The discussion continued with various members of the audience throughout the period of the conference.

After my talk, I was invited to visit the University of Warwick to talk to PhD students about the nature of writing for an interdisciplinary audience (date yet to be decided).
Year(s) Of Engagement Activity 2014
 
Description BAAL 2014 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Postgraduate students
Results and Impact A paper given at the annual British Association for Applied Linguistics conference, Warwick. The paper reported on the new multidimensional analysis framework that we have developed.
Year(s) Of Engagement Activity 2014
URL http://idrd-bham.info/idrd/
 
Description BALEAP PIM 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Postgraduate students
Results and Impact This was a talk delivered to one of our usere groups for the project - EAP lecturers. The talk was aimed at rasing awareness of the project to the user group.
Year(s) Of Engagement Activity 2014
URL http://idrd-bham.info/idrd/
 
Description Conference paper, AAAL 2017 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact A talk to applied linguistics in a conference themed around transdisciplinarity, bringing a corpus perspective to the general debate and raising awareness of the power of corpus approaches.
Year(s) Of Engagement Activity 2017
 
Description Developing Corpus Linguistics in South Asia Conference (Islamabad) 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact This was a workshop for introducing corpus linguistics to Asian researchers. The multimensional analysis model for analysing academic discourse was presented to the audience and received with great interest. One Pakistani researcher is now working at Birmingham on academic corps analysis under Dr Thompson's supervision.
Year(s) Of Engagement Activity 2015
URL http://idrd-bham.info/idrd
 
Description ESSE 2014 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Postgraduate students
Results and Impact This was a workshop bringing together researchers who take corpus-based approaches to the investigation of academic discourses. Our paper was primarily methodological.
Year(s) Of Engagement Activity 2014
URL http://idrd-bham.info/idrd/
 
Description Early Career Interdisciplinary researcher workshop 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact Paul Thompson was an invited speaker at a workshop on writing for interdisciplinary research audiences. The participants at the workshop were doctoral students and research fellows in the SPP-1889 "Regional Sea Change and Society" research network. He gave a talk and led a workshop on "Writing/presentation skills for interdisciplinary topics and audiences".
Year(s) Of Engagement Activity 2018
 
Description Final seminar on ESRC project (BIRDS) 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact The research team gave talks on all aspects of the project in a series of short presentations, and the event concluded with a 45 minute plenary discussion of the implications of the research and the methods employed.

As a result of the semianr, we decided to offer to speak to RCUK staff at a semianr in Swindon (at Polaris House)
Year(s) Of Engagement Activity 2015
URL http://idrd-bham.info/idrd/
 
Description Flagstaff, USA 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Postgraduate students
Results and Impact This was a presentation given at the annual American Association for Corpus Linguistics in Flagstaff, AZ. The conference was organised by Professor Biber, our consultant. The visit was important for the opportunity to discuss with Biber and his colleagues, as well as to present interim findings from our research.
Year(s) Of Engagement Activity 2014
URL http://idrd-bham.info/idrd/
 
Description Genetics and Biomedical text mining workshop 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact Paul Thompson was an invited speaker at this workshop on hypothesis generating for Genetics and Biomedical text mining, held at Lancaster University on 8th January 2019. The audience was made up of researchers from a variety of disciplines, many with an interest in computer text analysis that can lead to the generation of novel hypotheses in interdisciplinary research.
Year(s) Of Engagement Activity 2019
URL http://wp.lancs.ac.uk/btm/hg2btm/
 
Description ICAME 2014 Conference 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact Susan Hunston delivered a plenary talk and Paul Thompson gave a regular talk to participants at the ICAME conference in Nottingham (primarily corpus linguists from around the world).
Year(s) Of Engagement Activity 2014
URL http://idrd-bham.info/idrd/
 
Description Interim Seminar (BIRDIS) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact This one day interim seminar at which the project team talked about the methodology, context and results of the research to date was given to an audience of representatives from funding councils (ESRC, EPSRC), interdisciplinary researchers and EAP tutors (the user groups identified by the project), and it resulted in wide-ranging discussion of the nature of interdisciplinary research and of the differences between writing for a discipline-specific journal and writing for an interdisciplinary journal.

I was invited to give a talk to applied linguistics PhD students and staff about the project at the University of Nottingham , October 2014.
Year(s) Of Engagement Activity 2014
URL http://paulslals.org.uk/IDRD/?page_id=154
 
Description International Conference Presentation (invited keynote address), Languages for Specific Purposes conference, Bergen, June 2017 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact This was a talk on specialised discourses delivered to an international audience of LSP researchers, practitioners and translators. The talk was followed by a lively question and answer session, and requests afterwards for copies of our publications.
Year(s) Of Engagement Activity 2017
URL http://idrd-bham.info/idrd/?page_id=1332
 
Description Invited keynote talks, 2017 CLAVIER international conference, Bari, Italy, November 2017, Hunston and Thompson 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact Both invited keynote talks were given to an audience of teachers, researchers, translators and interpreters working in the area of languages for specific purposes, and they engendered lively discussion and requests for copies of our publications.
Year(s) Of Engagement Activity 2017
URL http://idrd-bham.info/idrd/?page_id=1332
 
Description Invited research seminar talk (Exeter) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Postgraduate students
Results and Impact This was an invited seminar talk to the School of Education, Exeter University. About forty staff members and postgradaute students attended, and after the seminar, we discussed the project and also a planned funding proposal that Dr Durrant was working on, and which I gave advice on. His bid was successful.
Year(s) Of Engagement Activity 2014
URL http://idrd-bham.info/idrd/
 
Description RCUK seminar, Swindon 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact This was a presentation of our research to interested RCUK staff at Polaris House in Swindon, at a lunchtime seminar. There were about 40 participants and there was discussion afterwards about the implications that the research has for the procedures for applying for interdisciplnary project funding and for reviewing such research activities.
Year(s) Of Engagement Activity 2015
URL http://idrd-bham.info/idrd/
 
Description Researching interdisciplinary discourse workshop 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Other audiences
Results and Impact Susan Hunston led workshop for staff and postgraduate students in applied linguistics at the University of Huddersfield, with a talk that was followed by questions and discussion.
Year(s) Of Engagement Activity 2019
 
Description Sinclair Seminar, Birmingham 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Results and Impact Public talk given to academics, students and EAP practitioners, which led to discussions of the methodology used afterwards.
Year(s) Of Engagement Activity 2015
URL http://www.idrd-bham.info/idrd/