Disciplinary Differences in the Use of English in Higher Education

Disciplinary differences in the use of English in higher education: reflections on recent language policy developments Author(s): Maria Kuteeva and John Airey Source: Higher Education , May 2014, Vol. 67, No. 5 (May 2014), pp. 533-549 Published by: Springer Stable URL: https://www.jstor.org/stable/43648673

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Higher Education

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 DOI 10. 1 007/s 10734-0 1 3-9660-6

Disciplinary differences in the use of English in higher education: reflections on recent language policy developments

Maria Kuteeva * John Airey

Published online: 4 September 2013 © Springer Science+Business Media Dordrecht 2013

Abstract In post-Bologna Europe, there has been a noticeable increase in English- medium instruction. In this article we take the case of Sweden as an illustrative example of the wider disciplinary issues involved in changing the teaching language in this way. By 2008 the use of English in Swedish higher education had risen to such an extent that it had to be regulated at the governmental level and through university language policies. Such policies have attempted to provide generalised pragmatic guidelines for language use across educational programmes. In this paper we argue that such general policies fail to take into consideration fundamental disciplinary differences and their potential impact on language use. We present a theoretical argument about the knowledge structures of disciplines, relating these to the disciplinary literacy goals of educational programmes. We then illustrate our argument using data from an extensive survey carried out at a major Swedish university. We conclude that the disciplinary variation in the use of English can be seen as a product of different knowledge-making practices and educational goals. This conclusion problematises "one-size-fits-all" language policies which only deal with general features of language use and do not allow for discipline-specific adjustments.

Keywords Disciplinary differences • Disciplinary literacy • Disciplinary knowledge structures • Teaching in English • Language choice • Language policies • Parallel language use

M. Kuteeva (El) Department of English, Centre for Academic English, Stockholm University, 10691 Stockholm, Sweden e-mail: [email protected]

J. Airey Physics Education Research Group, Department of Physics and Astronomy, Uppsala University, Uppsala, Sweden

Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms 534 High Educ (2014) 67:533-549

Introduction

In post-Bologna Europe, there has been a noticeable increase in the teaching of university courses through the medium of English. This trend has been particularly strong in the north of Europe, with research singling out the Netherlands and the Nordic countries as having the highest percentages of graduate and undergraduate courses taught in English (Wächter and Maiworm 2008). Although the internationalisation driving this increase is perceived as a desirable goal (cf. Nilsson 1 999), the use of English in the Nordic countries has now risen to the extent that it has had to be regulated at the governmental level (e.g. Nordic Council of Ministers 2007; Swedish Government 2009). Sweden provides an illustrative example of this trend (Kuteeva 201 1 ), and it is here that warning voices have been raised about the marginalisation of Swedish as an academic language (e.g. Gunnarsson and Öhman 1997; Gunnarsson 2001; Teleman 1992). In response to this perceived threat, universities in Sweden have been directed to produce documents outlining their language policies in an attempt to provide pragmatic guidelines for the handling of English and Swedish across educational programmes (Swedish National Agency for Higher Education 2008). However, as a recent study of academisation and marketisation in Swedish higher education has shown, such decrees 'from above' are not always compatible with the actual practices and ideals prevailing in different disciplinary and professional areas (Ek et al. 201 1). We argue that the recent language policies in Swedish higher education may not sufficiently take into consideration fundamental differences in disciplinary practices and their potential impact on language use. Although the impact of disciplinary differences on teaching and learning has been widely discussed in educational research (e.g. Becher 1989; Becher and Trowler 2001; Lindblom-Ylännea et al. 2006; Neumann 2001 ; Neumann and Becher 2002), little research has explored this issue in relation to English-medium instruction in Europe. In this respect, Airey (2009, 2010) has pointed out that in many countries the development of academic literacy (Lea and Street 1998) actually involves two or more languages - the local language^) and English - illustrating how the balance between these disciplinary languages is contested and intimately related to the acquisition of disciplinary knowledge. In this paper we use the case of Swedish higher education to illustrate our argument that general language policies may fail to take into consideration the potential impact of fundamental disciplinary differences. To do this we first map out the current linguistic landscape in Swedish higher education. Then, in order to account for differences in the use of English as an academic language, we apply insights from Bernstein's (1996, 1999, 2000) notion of disciplinary knowledge structures that are predominantly horizontal or hierarchical. Drawing on a recent report by the Swedish Language Council (Salo 2010) and our own data, we argue that disciplinary variation in English language use is not arbitrary, but rather can be seen as a product of the different knowledge structures of the disciplines. Next, we analyse different disciplinary literacy goals of educational programmes (Airey 201 1 ), before finally illustrating our theoretical arguments using specific examples from a range of disciplines representing so- called pure hard, pure soft, applied hard, and applied soft sciences (Becher 1989). In light of our findings, we problematize university- wide language policies, which recommend parallel use of Swedish and English without allowing for discipline-specific adjustments.

English language policies and practices in Swedish higher education

As is the case in other European countries, higher education in Scandinavia has experienced a number of substantial changes resulting from the Bologna process (2010). One of

â Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 535

Fig. 1 English-language PhD dissertations in 16 subject categories (by the Royal Swedish Library). Percentage of the total number in the respective subject category (adapted from Salo 2010, p. 24) the main changes has been an increase in the use of English in an attempt to attract the best international students (cf. Coleman 2006). English has been widely used as a language of research and scholarly publication in Scandinavian universities since the 1950s (Salo 2010), but its official status had become the subject of a heated debate by the late 1990s and culminated in the introduction of the Language Act (Swedish Government 2009), which established Swedish as the country's official language. The Swedish Language Council (Salo 2010) documents the increasing use of English in Sweden in some academic disciplines, particularly in the natural and exact sciences. In terms of teaching, just over 20 % of all courses and programmes now use English as a medium of instruction. However, even on courses where English is not the official medium of instruction, a high proportion of textbooks are in English. At postgraduate level English is much more prevalent: here approximately 65 % of Master's and other postgraduate programmes are taught through English. Moreover, around 87 % of all dissertations at Swedish universities are currently written in English. Interestingly, the increasing use of English in Swedish higher education is not evenly distributed across disciplines. For example, in the sciences the proportion of dissertations written in the English language is as high as 94 %, the same figure in the social sciences is 65 %, and in the humanities this figure falls to 37 % (Salo 2010, p. 7). As mentioned above, the report indicates that the frequency with which English is used to write dissertations varies greatly according to discipline (see Fig. 1). The academic subjects where English is used more frequently as a language of dissertations are, in decreasing order, science, mathematics, medicine, engineering, philosophy and psychology, economics, social sciences and linguistics. Those subjects with less than 50 per cent of dissertations written in English are (again in decreasing order): archaeology, literary studies, ethnography and social anthropology, geography and local history, history, reli- gion, art and music and theatre and film, and educational studies. As far as the language of instruction is concerned, in order to preserve the national language, the Swedish Language Council report recommends that parallel language use should also be a key goal. This concept of parallel language use is largely rooted in Scandinavian reality, although similar practices may exist in other European countries. In 2006, ministers for education and culture and other governmental representatives from the Nordic countries met to discuss language policies. This discussion resulted in Deklaration om nordisk sprâkpolitik/Declaration on a Nordic Language Policy (Nordic Council of Ministers 2007). This document contains an entire section on the parallel use of English and one or several Nordic languages in research and education. The declaration also states

£) Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms 536 High Educ (2014) 67:533-549 that both Nordic languages and English should be used as languages of science, that higher education institutions should develop language policies, including parallel language use and language instruction, and that competent bodies should coordinate translation and terminology in scientific domains. In Sweden, universities were directed to produce policy documents outlining their position on the relationship between languages (Swedish National Agency for Higher Education 2008). Following the introduction of the Language Act (Swedish Government 2009), which established the official status of Swedish as the country's main language, parallel language use has increasingly been regarded as a guiding principle for the dual use of Swedish and English in higher education. In this paper, we discuss the language policy and practices at one of the major Swedish universities which has adopted the parallel language use of Swedish and English in order to promote internationalisation and to improve its position in university rankings. This policy was introduced in 201 1 and contains entire sections on parallel language use and language awareness. It states that high language standards are essential for being nationally and internationally competitive, both in research and education. The policy underscores the importance of English for research dissemination and for preparing students to work in an international work environment. Parallel language use implies that Teachers and students have a full command of Swedish in parallel with English of high quality' (p. 2, our translation). The main goal of parallel language use is to strengthen the international dimension and at the same time to ensure the development of subject-specific terminology and disciplinary discourses in Swedish. The policy elaborates on some possible detrimental effects of using a foreign/second language such as English on the quality of education, such as domain loss and competence attrition (kapacitetsförlust (Swedish), i.e. a diminished capacity to express oneself with nuance and precision). It is recommended that English should be used for educational purposes after a careful consideration of all advantages (e.g. international dimension) and disadvantages (e.g. domain loss and capacity attrition). The section dealing with the guidelines for language use at the university states that all academic disciplines should be able to communicate in 'clear and comprehensible' (p. 3) Swedish and English, and all researchers are responsible for the development of terminology and disciplinary discourses in Swedish. As far as education is concerned, the policy recommends that, with the exception of special courses for international students, all undergraduate education should take place in Swedish, with a possibility of introducing English in some course components. At the Master's level, increased use of English is recommended, and the education can take place either in English or Swedish or in both languages in parallel. At the PhD level, the students can choose which language to use, but all theses in English should be accompanied by a Swedish summary and vice versa; theses written in other languages (e.g. French, Spanish, etc.) should be accompanied by both Swedish and English summary. Interestingly, apart from the possibility of using a third foreign language in some specific fields concerned with the study of that language and related literature or culture, no mention is made of disciplinary differences and how these may affect language use. Initially conceived by policy-makers, the notion of parallel language use does appear to have received widespread support; however, the reality is that its status is still one of an unoperationalized political slogan. To date, little thought has been given to the day-to-day practicalities of parallel language use by students and teachers (Airey 2009). In fact, Airey and Linder (2008) criticized the notion of parallel language use, pointing out that it unhelpfully focuses our attention on regulating language use in higher education rather than the more pressing agenda of formulating a set of disciplinary linguistic skills that students, lecturers and researchers need to acquire. The full implications of parallel

£) Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 537 language use remain unclear, and more research needs to be done on how this concept might be implemented in practice. According to the 2010 report, a balanced parallel language use in education is a natural and desirable development in an internationalised educational world, which involves more than the language(s) one uses to educate students. Salo (2010, 54-55) draws on the example of university physics where students can work towards bilingual scientific literacy, so that they can be empowered to do their university physics studies in both Swedish and English. Thus, the issue of disciplinary teaching and learning is raised in the report in relation to language use but this route is not explored further. The above-mentioned language policy does not sufficiently elaborate on disciplinary differences in language use either. Our paper endeavours to fill this notable gap. To what extent is it possible to develop bilingual disciplinary literacy in different domains? Is it possible to have the same goal and the same language policy across different educational programmes? The following section approaches these issues by reviewing work on disciplinary differences and language use in higher education.

Disciplinary differences: knowledge-making practices and disciplinary literacy

Knowledge-making practices

The significance of disciplinary differences has been a recurring theme in educational research for over 20 years (e.g. Becher 1989; Becher and Trowler 2001; Lindblom-Yl- ännea et al. 2006; Neumann 2001; Neumann and Becher 2002; Christie and Maton 201 1). Drawing on earlier work by Biglan (1973) and Kolb (1981), Becher (1989) classified disciplines according to four major categories: 'pure hard', 'pure soft', 'applied hard' and 'applied soft', depending on their epistemological characteristics. Neumann and Becher (2002: 406) characterise pure hard knowledge (e.g. physics or chemistry) as having a 'cumulative, atomistic structure, concerned with universais, simplification and a quanti- tative emphasis'. On the other hand, pure soft knowledge (e.g. history or anthropology) is 'reiterative, holistic, concerned with particulars and having a qualitative bias' (ibid). Unlike pure hard disciplines, pure soft disciplines have no clear knowledge community nor any sense of superseded knowledge. Thus, academic enquiry in such disciplines is usually of an individual and interpretative nature, valuing creativity and fluency of expression. In the more applied sciences, hard applied knowledge (e.g. engineering) is purposive and pragmatic, based on factual understanding and concerned with the mastery of the physical environment and practical applications. Soft applied knowledge (e.g. education) is also of functional, practical nature, informed primarily by soft pure disciplines; it lays emphasis on professional practice or personal growth. Although this classification is somewhat sim- plified and does not take into account possible differences within disciplines, it remains useful for mapping out some fundamental distinctions. The link between disciplinarity and academic discourse has also been explored by Bernstein (1996, 1999, 2000) in relation to disciplinary knowledge structures. Similarly to Becher and other researchers in educational science, Bernstein explains disciplinary differences through epistemological factors and shows how these differences are manifested in different types of academic discourse, which he terms 'vertical discourse':

Vertical discourse takes the form of a coherent, explicit and systematically principled structure, hierarchically organised as in the sciences, or it takes the form of a series of

Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms 538 High Educ (2014) 67:533-549

hierarchical knowledge ^structure ■ - "warring a - ^tnangles" horizontal /' knowledge / ' / physics biology ^ ^ linguistics sociology ^ < history literary studies ^ science social science humanities

Fig. 2 Knowledge structures across disciplines. Adapted from Martin (20 1 l, pp. 42-43)

specialised languages with specialised modes of interrogation and specialised criteria for the production and circulation of texts as in the social sciences and humanities. (Bernstein 1999, p. 159).

Thus, the different forms of vertical discourse are a product of the different knowledge structures of disciplines (Bernstein 1996, 1999, 2000). As illustrated in Fig. 2, at one end of the continuum lie hierarchical knowledge structures, which attempt 'to create very general propositions and theories, which integrate knowledge at lower levels' (Bernstein 1999, p. 162). At the other end, there are horizontal knowledge structures, in which knowledge is built through a series of specialised languages. Muller (2007) suggests that hierarchical knowledge structures test theories against data, whereas horizontal knowledge structures use theory to interpret texts. Thus, hierarchical knowledge structures are typical of natural sciences, in which knowledge rests upon the same foundation and is accumulated through empirical enquiry, whereas the humanities create knowledge through multiple interpretations of the same phenomena and artefacts. Wignell (2004, in Martin 2011) proposed that the social sciences can be viewed as 'warring triangles', in the sense that these disciplines often model themselves on sciences but have different theoretical foundations; some theories may become dominant and marginalise others, but a single theory seldom takes over the whole discipline. As far as language is concerned, the sciences usually have an agreed set of terminology which has a direct correspondence to the concepts used by the discipline. The sciences also tend to make use of established methods and procedures for conducting research, which is reflected in their disciplinary discourses, both at the macro-level (the schematic and rhetorical structure of academic genres) and the micro-level (lexicogrammatical features, formulaic language). On the other hand, most disciplines in the humanities, e.g. literary and cultural studies or history, tend to use different perspectives to interpret texts, artefacts, performances, events, and so forth. For example, Martin et al. (2010) offer an account of how different paradigms in the study of history impact disciplinary discourse. Hood (201 1 ) examines how disciplinary knowledge-making practices are reflected in the way that sources are projected in academic texts across disciplines, including the natural sciences, social sciences and humanities. All in all, the discourse in the humanities is characterised by a stronger authorial presence compared to the sciences. Language serves as the means to construct knowledge and is therefore used in a more flexible and creative manner. To sum up, language is shaping and is being shaped by disciplinary practices and epistemologies across a wide range of speciali- sations, from sciences to arts and humanities (e.g. Hyland 2000). Thus, considering the role that language plays in the construction of disciplinary knowledge, we can predict that disciplinary differences will affect the use of English as an

Ô Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 539

Most use of English Least use of English

Natural sciences Social sciences Humanities

Fig. 3 Predicted use of English across disciplines academic language with the natural sciences having the most use of English and the humanities having the least use of English. This relationship is illustrated in Fig. 3. Our predicted cline is clearly visible in the data from the 2010 report on the use of English in PhD theses presented at Swedish universities (Fig. 1). Interestingly, this trend is not limited to written discourse, and a similar cline can be observed in the proportion of undergraduate and Master's lectures given in English across the four faculties at a major Swedish university, whose language policy was discussed in the previous section (Fig. 4). Again, the use of English decreases from sciences to social sciences to law and the humanities (which, in Swedish higher education would include disciplines that in the English-speaking countries would often be placed at faculties of Arts, e.g. literature, history, theatre and cinema studies, and so forth). At this point, three issues with our claims about disciplinary differences in language use should be pointed out. First, Bernstein's classification contemplates exclusively 'pure' rather than 'applied' disciplines (i.e. what Bernstein termed singulars). Next, Bernstein addresses disciplinary knowledge structures rather than specific educational goals in different kinds of disciplinary contexts. Finally, although our work emphasises differences across disciplines, we do accept that there are, in fact, important differences within disciplines (e.g. Becher and Trowler 2001). The following section will discuss what constitutes disciplinary literacy.

Disciplinary literacy

From an educational perspective, the goal of any degree programme is the fostering of disciplinary literacy, which is defined as 'the ability to appropriately participate in the communicative practices of a discipline' (Airey 201 1). Here, disciplinary literacy involves more than simply reading and writing the disciplines; knowledge construction, negotiation and dissemination using a wide range of semiotic resources are included in the term (cf. Norris and Philips (2003) discussion of the fundamental and derived senses of scientific literacy). Disciplinary literacy will, in fact, be made up of three separate but interrelated literacies - academic, societal, and vocational, - and each of these has a local and an international form (Airey 2011). English is most likely to be used in the international forms, whilst the local language(s) will be used in local contexts. Thus, for the purposes of this paper academic disciplinary literacy may be seen as comparable to Bernstein's vertical discourse (1999). 1 Societal disciplinary literacy refers to a popular science communication of the discipline and can be thought of in terms of the communication of 'what every citizen should know' about the discipline, as in outreach publications (cf. scientific literacy). Vocational disciplinary literacy refers to communication of the discipline in the world of work and in professional settings. The relationship between the three kinds of

1 Note that Bernstein (1999, p. 159) defined horizontal discourse as entailing "... a set of strategies which are local, segmentally organised, context specific and dependent, for maximising encounters with persons and habitats." Thus, academic disciplinary literacy clearly refers to appropriate participation in both vertical and horizonal discourse - what Gee (2005) calls pulling off a big "D" Discourse. The reason for referring only to vertical discourse here is our educational interest in the link between vertical discourse, disciplinary knowledge structures and English language use.

£) Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms 540 High Educ (2014) 67:533-549

Fig. 4 Lectures in English across faculties. Adapted from Bolton and Kuteeva (2012) literacy is represented diagrammatically in terms of the Disciplinary Literacy Triangle (Fig. 5). Unlike Becher (1989), who envisages a 'pure/applied' divide, Airey (201 1) argues that whilst the educational emphasis placed by different disciplines on these three literacies will vary, an element of all three will be present for all academic disciplines. For example, some pure disciplines such as theoretical physics or pure mathematics would be located closer to the 'academy' apex, although both of these would also have 'workplace' and 'society' dimensions. On the other hand, nursing or teacher training would be located closer to the 'workplace' apex but clearly involve societal and academic dimensions. Traditional academic disciplines such as history would also have a strong societal orientation and would often be drawing upon local resources (e.g. archives) and communicate their knowledge to the local community, thus moving between the 'academy' and 'society'. Returning to the variation in the use of English as a language of doctoral dissertations in Sweden, we need to consider the use of English in some disciplines that were not included

Fig. 5 Airey 's (201 1) Society disciplinary literacy triangle. Disciplines are located at some point within the triangle according to the emphasis placed on communicative competence within each of the three sites

The academy The workplace

Ô Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 541 in Bernstein's description of disciplinary knowledge structures (1999). One notable case is that of Education, which is placed at the far end of the continuum for its low use of English in PhD dissertations (Fig. 1). Due to its applied nature, this discipline does not feature in Bernstein's overview of disciplinary knowledge structures. However, referring to the disciplinary literacy triangle (Fig. 5), the low usage of English in Education can be explained by the vocational orientation of this discipline, which justifies a higher use of the local language. The other two notable exceptions in Fig. 1 include mathematics and philosophy, both of which are characterised by a high use of English. The role of language in the construction of knowledge in these two disciplines is very different: mathematics has its own symbolic code, whilst philosophy relies exclusively on linguistic resources and argumentation. At the same time, these two disciplines are primarily based on logic and therefore have rather unique epistemologies, neither of which fits into Bernstein's classification of disciplinary knowledge structures (as acknowledged by Bernstein himself, 1999, 164; on the uniqueness of the discourse and epistemology of mathematics see also O'Halloran 2010, and McGrath and Kuteeva 2012). Moreover, these two 'pure', traditional disciplines have a strong 'cosmopolitan' (Becher 1989) orientation, which may account for a relatively higher use of English in order to reach the international academic community. Thus, this section has shown that differences in disciplinary knowledge structures (Bernstein 1999) and disciplinary literacies (Airey 2011) largely account for variation in the use of English as an academic language, both in the case of PhD dissertations and in the case of English-medium instruction in higher education. The following section will illustrate the theoretical arguments presented above with specific examples from a range of disciplines representing different kinds of knowledge structures and different disciplinary literacy goals in relation to language policies and practices at a major Swedish university.

Attitudes towards the use of English across disciplines

A recent survey of 668 academic staff and 4,524 students conducted at a major Swedish university has confirmed that the use of English as an academic language varies sub- stantially across disciplines (for further details of the survey see Bolton and Kuteeva 2012). In the context of the present discussion, it is noteworthy that we find a clear connection between the type of knowledge structures (hierarchical versus horizontal) and the likeli- hood of language switch to English across the four faculties (as shown in Fig. 4). The survey largely confirmed that in the sciences English use is a pragmatic reality for both lecturers and students. In the humanities and social sciences, English is often used as an additional or auxiliary language in parallel with Swedish or another foreign language. In this context, the use of spoken and written English was perceived as problematic when it created major obstacles for learning and the effective communication and discussion of academic ideas and perspectives. The survey was conducted in 2009, i.e. 2 years before the language policy was introduced, and the results were presented to the senior management of the university in 2010. In this paper we make use of open comments by survey participants concerning their views on the use of English in educational and research settings. The excerpts presented below are drawn from these additional comments by academic staff at the surveyed university. The anonymity of the respondents is ensured by omitting all their personal details. At the same time, the academic rank and departmental affiliation are given in parenthesis in order to provide details concerning the disciplinary context and the respondent's position

£) Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms 542 High Educ (2014) 67:533-549

within the discipline. We provide our own translations of any comments written in Swedish. In order to demonstrate and to contrast disciplinary differences in the attitudes towards the use of English in academic settings, our analysis will proceed from the natural sciences to the humanities to the social sciences. Throughout the analysis, we will be referring to Bernstein's (1999) description of knowledge structures and Airey's (2011) disciplinary literacy triangle. At the same time, wherever relevant, we will be referring to the university language policy mentioned above. In the survey data scientists generally perceive English as a useful tool in the dissemination of research and a pragmatic reality in educational contexts. Both senior and junior scientists stress the importance of terminology in their disciplinary discourse. Such attitudes are in line with Bernstein's concept of hierarchical knowledge structures and the agreed 'language' that scientific communities develop. The following four excerpts illustrate the point:

(1) Subject-wise, English dominates my subject and there are frequently difficulties to find reasonable translations to many subject-related words. Thus, one ends up using a kind of "Swenglish" that doesn't benefit anyone. I say better stick to English and prepare the students for reality. (Professor, Department of Meteorology) (2) In our field of science, English is the dominant language, and there's not so much to do about that... I would have difficulties having a scientific discussion in Swedish, since many terms are lacking a Swedish translation. (PhD student, Department of Biochem- istry and Biophysics) (3) English IS the scientific language in my field. We ... should rather be happy that it is English, and not Chinese, or some other language that is hard to learn. (Professor, Department of Genetics, Microbiology and Toxicology) (4) English is the main language in computer science. Swedish and other languages never really had a realistic chance in this field. (Lecturer, Department of Computer and System Sciences)

While all four comments (1-4) clearly show that English is the dominant language in the respondents' disciplines, comments 1 and 2 also mention the problem of subject- specific terminology in Swedish. New and fast-developing disciplines such as Computer Science also rely heavily on English terminology (comment 4). Thus, the problem of domain loss in Swedish, which was given particular attention in the language policy, appears to be relevant in this context. At the same time, the dominance of English seems to be justified by an international dimension of research and education (e.g. comment 1). Apart from the international dimension of scientific research, the workplace angle (Airey 2011) is also important in educational settings, particularly in the applied sciences, as illustrated in this comment:

(5) In geosciences it is very important for students to be taught in English. If they want to continue in academia or industry English is essential because they may need to work abroad and will certainly need to give presentations in English in different countries. (Researcher, Department of Geology and Geochemistry)

Thus, comments 1-4 largely focus on language use for education and research, whereas comment 5 raises the question of effective communication in professional settings. According to comment 5, English is essential for students whether they choose to stay in academia or to work in the industry because in either case they will work in an international environment.

Ô Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 543

In contrast to the sciences where English seems to be a natural choice and pragmatic reality, humanities do not view language in neutral terms. As this informant put it:

(6) As far as English as a medium of instruction is concerned there are important differences between natural science and humanities/social sciences. In the humanities language cannot be viewed as a neutral carrier of meaning. The question of the position of English and its possible dominance must be posed in a different manner when it comes to natural science. (Senior Lecturer, Department of Journalism and Media Studies)

In addition to the importance of language as the main means of knowledge construction, the national orientation of some disciplines does not allow much space for the use of English:

(7) My impression is that humanities in Sweden still have a strong national character and scholars seem to be very conservative in the exclusive preference for the use of Swedish language in their field ... In my Department no course in English is available, neither at undergraduate nor at postgraduate level . . . the option to write a dissertation in English is of course given, but it is often discouraged by the advisors. . . At any rate, dissertations rarely concern topics of international interest. (Researcher, Department of Art History)

Thus, language choice is closely connected to disciplinary practices, and in the case of excerpt 7, Swedish dominates the field at all levels of education and in research. This practice seems to be in obvious contradiction with the language policy discussed in previous sections, but on the other hand, introducing more English in this case may be controversial. Comment 7 was made by a guest researcher, and local academic staff in that department seem to have their own reasons for keeping Swedish the main academic language. At the same time, although the humanists often object to English-medium instruction (e.g. comment 7), they still resort to English for research dissemination purposes. This is where they face far higher challenges compared to natural scientists, due to higher demands placed on language as a means of knowledge construction:

(8) To reach the international context, English is a must [...] However ... humanists ... place much higher demands on language formulations than in the sciences. (Researcher, Department of German)

This comment reinforces the importance of language in the construction of knowledge and refers to the additional difficulties experienced by the humanities scholars when they write for international audiences in English. However, as mentioned in the language policy, it is not only English that is being used as an academic lingua franca in some disciplines. From Bernstein's perspective, it is natural that the same discipline in the humanities develops different 'languages' to suit different kinds of disciplinary knowledge structures and to offer different interpretations of the same phenomena. Therefore, other academic languages than English play an important role, as the use of a different language may bring in a new epistemological angle. For example, as shown in comment 9, using a different language brings in a change in perspective, a new interpretation, and thereby fosters knowledge construction. This comment also underscores the importance of multilingualism in international research. Over the past decades, multilingualism in academia has become less common, and this informant regrets this loss:

£) Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms 544 High Educ (2014) 67:533-549

(9) In my view the academic system has to be at least bilingual... We have to teach and publish in Swedish as well as English and preferably also in a third or a fourth language. Being from a small country and mastering several languages can potentially constitute a distinct competitive advantage. The research and academic discussion in cultures "blessed" with an international language tend to be somewhat isolated within their languages. If you have a reasonable knowledge of several languages you can act as a bridge between these isolates. (Professor, Department of Economic History)

As we can see, this comment also problematizes the perceived advantage of native speakers of English as far as international research is concerned. This problem is particularly common in the humanities, which are often rooted in local traditions, whereas natural sciences tend to be more "cosmopolitan" (Becher 1989) in their approach. Thus, compared to sciences, language choice in the humanities is much less straight-forward and is more dependent on subject matter and local academic traditions. Finally, the social sciences present a more heterogeneous picture with some disciplines closer to the sciences both in the way they construct knowledge and perceive the role that language plays in this process, and others similar to the humanities in underscoring the importance of formulations and stylistic features in disciplinary discourse. The following two excerpts illustrate the latter attitude, particularly in relation to classroom discussions (excerpt 10) and PhD theses (excerpt 11):

(10) Within humanities and social sciences, where 'meaning' is important it is almost impossible to express oneself clearly in any other language than in one's native language. At the master classes the level of the discussions can sometimes be very low. The reason for this is to large extent language problems. (Senior Lecturer, School of Business) (11) We need a lingua franca for research... But I do think that the pressure to use English, for example in PhD theses, does sometimes result in lower quality research - perhaps this is most obvious in disciplines where language is an important part of the research process. In my subject, sociology, we use language to describe and analyse social reality. You have to be extremely competent in a foreign language to be able to do that with the same depth and analytical precision as you would in your first language. (Senior Lecturer, Sociology)

Both comments 10 and 1 1 underscore the crucial role of language in knowledge construction, when "meaning is important" (10) and which requires "depth and analytical precision" (11). This challenge affects both spoken and written discourse. However, the language switch does not only present challenges; according to some academics, it can also bring opportunities. In the excerpt below, using a foreign language is seen as providing an opportunity to gain new insights and knowledge by being creative:

(12) For postgraduates, it is important to write simultaneously in Swedish and English, for them to learn in the process of "translation". Here I don't mean direct translations (lexical) but in terms of meaning-making . . . When working in a second language, you are able to treat the language in a more 'disrespectful', creative and innovative manner, which helps you theorize in more interesting ways. (Senior lecturer, Department of International Education)

On the other hand, other social sciences, such as economics, are closer to the 'hard' sciences and hierarchical knowledge structures, both in their methods and in attitudes towards the use of English:

£) Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 545

(13) There are some obvious problems with the English dominance for all of us who are not native, but those problems are not sufficient arguments for trying to reduce the importance of English in academic communication. (Professor, Department of Economics)

In this case, the use of English is again seen as a pragmatic reality, and this attitude is similar to those expressed by scientists in comments 1-4. At the same time, if a more humanistic discipline, such as social anthropology, has an international dimension both in the academy and in the workplace, it is also reflected in attitudes towards English. For example, the author of the excerpt below is very positive about the use of English in social anthropology, despite the fact that this discipline also places a great deal of importance on language formulations and style:

(14) The use of English and the proficiency of native speakers may seem unfair to many, but this is life - internationalization demands a common language, and English is it, for now at least... In my department all staff are very competent in English. Swedish students communicate well in English, but are seldom confident enough to use it for papers or exams. I believe they would be more attractive on the job market if they acquired this confidence. . . (Researcher, Department of Social anthropology)

This kind of openness towards the international job market is similar to the one expressed by a scientist in comment 5. Despite reporting some difficulties in English language use (comparable to those mentioned in comment 11), the author of comment 14 sees the value of English for both research and education. From the comments collected from academic staff working in the sciences, humanities, and social sciences, we can see that the use of English varies significantly across different disciplines and is closely related to disciplinary knowledge-making practices and the placement of the discipline between the academy, the society, and the workplace. It is also clear that the parallel language use of Swedish and English recommended by the university language policy is not unproblematic and cannot be implemented in the same way across different disciplines. Potential dangers of domain loss, related to subject-specific terminology in the local language, seem to affect the sciences more than any other disciplines. At the same time, social scientists and humanists often mention that they feel inadequate when writing or speaking in a foreign language such as English.

Discussion and conclusion

The analysis of attitudes expressed by academic staff towards the use of English in education and research problematizes university-wide language policies, which recommend parallel use of Swedish and English without allowing for discipline- specific adjustments. For example, the language policy introduced at the university where the present research was conducted recognises the importance of high language standards in national and international contexts and promotes the use of both local and foreign languages in research and education. However, the parallel use of Swedish and English across all disciplines remains a desirable goal, and it is not clear how it should be implemented in such different disciplines as biochemistry and art history, for example. While all academic disciplines should be able to communicate in 'clear and comprehensible' Swedish and English, the linguistic reality across faculties and individual departments is not the same. Although

â Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms 546 High Educ (2014) 67:533-549

language use is largely regulated according to level of instruction (i.e. undergraduate versus Master's and PhD), disciplinary differences are not sufficiently considered. We argue that allowing space for disciplinary adjustments in language policies is necessary. In line with Bernstein's arguments, we have demonstrated that language - in our case English as an additional language - is used in different ways and for different purposes depending on the nature and knowledge-making practices of the academic discipline. In the natural sciences, for instance, new knowledge is built on the foundations of previous knowledge in a hierarchical manner, whereas in the humanities knowledge is constructed through new interpretations of phenomena, facts, texts, and artefacts (in effect creating multiple 'languages' or 'Ls'; see also Martin's adaptation of Bernstein 2000). Thus, in the sciences knowledge is seen as cumulative and resting on the same foundations and, as a result, sciences build an agreed 'language' and specialist terminology (see e.g. comments 1 and 2). Humanities, on the other hand, often create their own discourses through stylistic, syntactic and lexicogrammatical features (comments 7, 8, 11, and 12). It is therefore easy to understand why the switch between languages may be viewed as relatively unproblematic in the sciences, since this is often "simply" a question of translating terminology. By the same token it is also easy to appreciate why this argument is less acceptable in the humanities where such translation would also involve translation of stylistic, syntactic and lexicogrammatical features. This 'translation' (see excerpt 12) is a much less straightfor- ward process, often involving cultural references and other paralinguistic factors related to disciplinary knowledge-making practices. Bernstein (1999) grouped the social sciences along with humanities as having horizontal knowledge structures; however, our data only partially confirms this view. In our study, social sciences appear to vary with some indeed resembling the humanities whilst others are seemingly closer to natural sciences, often with alternative pyramids of knowledge concerning the same phenomena (referred to as 'warring triangles' (Wignell 2007). However, Bernstein's (1999) classification of hierarchical knowledge structures only contemplates the so-called 'pure' sciences, and therefore does not account entirely for the observed variation in the use of English across academic disciplines. This is where Airey's disciplinary literacy triangle (201 1) has proven to be a useful tool for interpreting language use and language attitudes. Unlike Becher' s (e.g. 1989) earlier classifications, Airey suggests three different angles of literacy in any kind of academic discipline, each with a local and international dimension. Thus, the international versus local dimension of the academy, the society or the workplace affects language uses within any given discipline and educational programme (see e.g. comments 5, 7, and 14). Returning to the question of language policy for higher education, our study points towards a relative ease of parallel language use in the natural sciences, in which the main challenge is to develop and maintain specialist terminology in both languages. The university language policy warns against the danger of domain loss across all disciplines and urges all researchers to contribute to the development of subject-specific terminology in the local language. However, the dominance of English and lack of Swedish terms was only mentioned by scientists (e.g. comments 1 and 2). Humanists and social scientists, on the other hand, were less confident about using a non-native language (e.g. comments 7, 9, 10, 1 1 ). Although English is a "must" (comment 8) for research dissemination in the humanities, it is almost necessary to develop two kinds of academic literacy, one in the local language and one in English (this challenge concerns primarily local disciplines with their own academic traditions). The social sciences may resemble either of the two extremes with new disciplines, such as economics, being dependent on English discourse

£) Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 547 norms, and others, such as anthropology, drawing on different academic traditions and discourses. Overall, we find strong agreement between our predicted attitudes towards language use and the actual state of affairs as reflected in the above-cited excerpts, the figures presented in the Swedish Language Council report (Salo 2010), and the survey data (Bolton and Kuteeva 2012). Our conclusion is that the disciplinary variation in English language use is not arbitrary, but rather can be seen as a product of the different knowledge-making practices of the disciplines. Thus, any "one-size-fits-all" language policy for higher education will be detached from the actual practices. This conclusion problematizes university-wide language policies, which we therefore suggest should primarily deal with general features of language use, whilst allowing room for discipline-specific adjustments as deemed necessary. Our findings have implications for policymakers in European higher education, as more and more universities resort to English as a medium of instruction for local and international students. In addition, our study provides some useful insights for teachers of English for academic purposes (EAP), since they also need to deal with a heterogeneous student population. For example in many universities across the UK, EAP instruction has been aimed at developing some form of generic academic literacy (Lea and Street 1998). However, Gimenez (2012) shows how differences in disciplinary epistemologies - even in the closely related disciplines of nursing and midwifery - lead to quite different disciplinary discourses. Such disciplinary-specific literacies also need to be considered in order to provide students with appropriate language skills for their university studies.

References

Airey, J. (2009). Science, Language and literacy. Case studies of learning in Swedish University Physics. Acta Universitatis Upsaliensis. Uppsala Dissertations from the Faculty of Science and Technology 81. Uppsala. Retrieved 2009-04-27, from http://publications.uu.se/theses/abstract.xsql?dbid=9547. Airey, J. (2010). The ability of students to explain science concepts in two languages. Hermes- Journal of Language and Communication Studies, 45, 35-49. Airey, J. (201 1). The disciplinary literacy discussion matrix: A heuristic tool for initiating collaboration in higher education. Across the disciplines, 8(3). http://wac.colostate.edu/atd/clil/airey.cfm (last accessed 08/29/2013). Airey, J., & Linder, C. (2008). Bilingual scientific literacy? The use of English in Swedish university science programmes. Nordic Journal of English Studies, 7(3), 145-161. Becher, T. (1989). Academic tribes and territories: Intellectual enquiry and the cultures of disciplines. Milton Keynes: SRHE/Open University Press. Becher, T., & Trowler, P. (2001). Academic tribes and territories. Buckingham: SRHE/Open Univeristy Press. Bernstein, B. (1996). Pedagogy, symbolic control and identity: Theory, research, critique. London: Taylor and Francis. Bernstein, B. (1999). Vertical and horizontal discourse: An essay. British Journal of Sociology of Education, 20(2), 157-173. Bernstein, B. (2000). Pedagogy, symbolic control and identity (Revised ed.). New York and Oxford: Rowman and Littlefield Publishers. Biglan, A. (1973). The characteristics of subject matter in different academic areas. Journal of Applied Psychology, 57(3), 195-203. Bologna Process. (2010). Bologna Process. Towards the European higher education area. http://ec. europa. eu/education/higher-education/bologna_en.htm (last accessed 07/02/2013). Bolton, K., & Kuteeva, M. (2012). English as an academic language at a Swedish university: parallel language use and the "threat" of English. Journal of Multilingual and Multicultural Development, 33, 429-447.

Ö Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms 548 High Educ (2014) 67:533-549

Christie, F., & Maton, K. (2011). Disciplinarity: Functional linguistic and sociological perspectives. London: Continuum. Coleman, J. (2006). English-medium teaching in European higher education. Language teaching, 39, 1-14. Ek, A.-C, Ideland, M., Jönsson, S., & Malmberg, C. (2011). The tension between marketisation and academisation in higher education. Studies in Higher Education, doi: 10.1080/03075079.201 1.619656. Gee, J. P. (2005). An introduction to discourse analysis : Theory and method (2nd ed.). New York: Routledge. Gimenez, J. (2012). Disciplinary epistemologies, generic attributes and undergraduate academic writing in nursing and midwifery. Higher Education, 63, 401-419. Gunnarsson, B. L. (2001). Swedish, English, French or German - The Language Situation at Swedish Universities. In U. Ammon (Ed.), The dominance of English as a language of science. Effects on other languages and language communities (pp. 229-316). Berlin, NY: Mouton de Gruyter. Gunnarsson, B.-L., & Öhman, K. (1997). Det internationaliserade universitetet. En studie av bruket av engelska och andra främmande sprâk vid Uppsala universitet. Text- och facksprâksforskning 16. Uppsala: Institutionen för nordiska sprâk, Uppsala universitet. Hood, S. (2011). Writing discipline: Comparing inscriptions of knowledge and knowers in academic writing. In F. Christie & K. Maton (Eds.), Disciplinarity (pp. 106-128). London: Continuum. Hyland, K. (2000). Disciplinary discourses : Social interactions in academic writing. Ann Arbor: University of Michigan Press. Kolb, D. (1981). Learning styles and disciplinary differences. In A. Chickering (Ed.), The modern American college : Responding to the new realities of diverse students and a changing society (pp. 232-255). London: Jossey-Bass. Kuteeva, M. (2011). Teaching and learning in English in parallel-language and ELF settings: Debates, concerns and realities in higher education. Ibérica, 22, 5-12. Lea, M., & Street, B. (1998). Student writing in higher education: An academic literacies approach. Studies in Higher Education, 23(2), 157-172. Lmdblom-Ylännea, S., Tngwell, K., Nevgia, A., & Ashwinc, P. (2006). How approaches to teaching are affected by discipline and teaching context. Studies in Higher Education, 5/(3), 285-298. Martin, J. R. (201 1). Bridging troubled waters: Interdisciplinarity and what makes it stick. In F. Christie & K. Maton (Eds.), Disciplinarity (pp. 35-61). London: Continuum. Martin, J., Maton, K., & Matruglio, E. (2010). Historical cosmologies: Epistemology and axiology in Australian secondary school history discourse. Signos, 43, 433-463. McGrath, L., & Kuteeva, M. (2012). Stance and engagement in pure mathematics research articles: Linking discourse features to disciplinary practices. English for Specific Purposes, 3/(3), 161-173. Muller, J. (2007). On splitting hairs: Hierarchy, knowledge and the school curriculum. In F. Christie & J. Martin (Eds.), Language, knowledge and pedagogy : Functional linguistic and sociological perspectives (pp. 64-86). London: Continuum. Neumann, R. (2001). Disciplinary differences and university teaching. Studies in Higher Education, 26(2), 135-146. Neumann, R., & Becher, T. (2002). Teaching and learning in their disciplinary contexts: A conceptual analysis. Studies in Higher Education, 27(49), 405-417. Nilsson, B. (1999). Internationalisation at home - theory and praxis. EAIE ( European Association for International Education) Forum, Spring 1999: unpaginated. Nordic Council of Ministers. (2007). Deklaration om nordisk sprákpolitik 2006 [Declaration on a Nordic language policy]. Copenhagen: Nordic Council of Ministers, http://www.nordicom.gu.se/common/ publ_pdf/280_nordisk%20sprakpolitik.pdf (last accessed on 07/02/2013). Norris, S. P., & Philips, L. M. (2003). How literacy in its fundamental sense is central to scientific literacy. Science Education, 87(2), 224-240. O'Halloran, K. (2010). The Semantic hyperspace: Accumulating mathematical knowledge across semiotic resources and modalities. In F. Christie & K. Maton (Eds.), Disciplinarity : Functional linguistic and sociological perspectives (pp. 217-236). London: Continuum. Salo, L. (2010). Engelska eller svenska? En kartläggning av sprâksituationen inom högre utbildning och forskning. Stockholm: Spràkrádet. Swedish Government (2009). Sprâklag [The Language Act]. SFS 2009:600. http://62.95.69.3/SFSdoc/09/ 090600.PDF (last accessed on 07/02/2013). Swedish National Agency for Higher Education. (2008). En högskola i världen - internationaliseringen for kvalitet [Swedish universities and globalization - internationalization for quality]. Stockholm: Högskoleverket. Teleman, U. (1992). Det svenska rikssprákets utsikter i ett integrerat Europa [The outlook for Swedish in an integrated Europei. Sprâkvârd, 1992, 7-16.

Ô Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms High Educ (2014) 67:533-549 549

Wächter, B. & Mai worm, F. (2008). English-Taught Programmes in European Higher Education. The Picture in 2007. ACA Papers on Cooperation in Education. Bonn: Lemmens. Wignell, P. (2007). Vertical and horizontal discourse and the social sciences. In F. Christie & J. R. Martin (Eds.), Language , knowledge and pedagogy: Functional linguistic and sociological perspectives (pp. 184-204). London: Cassell.

Ô Springer

This content downloaded from 86.59.13.237 on Fri, 11 Jun 2021 08:36:41 UTC All use subject to https://about.jstor.org/terms