English-medium teaching in European higher education

In the global debates on English as international lingua franca or as ‘killer language’, the adoption of English as medium of instruction in Higher Education is raising increasing concern. Plurilingualism and multilingualism are embedded in the official policies of the European Union and Council of Europe, and the Bologna Process for harmonizing Higher Education promises ‘proper provision for linguistic diversity’. But even enthusiasts acknowledge the problems of implementing such policies in the face of an inexorable increase in the use of English. This survey draws on the most recent and sometimes disparate sources in an attempt to paint a comprehensive and up-to-date picture of the spread of English-medium teaching in Europe's universities. The article sets the changes in the context of accelerating globalization and marketization, and analyses the forces which are driving the adoption of English, and some of the problems which accelerating ‘Englishization’ of European Higher Education might create.

Individual plurilingualism and societal multilingualism are the principles which underpin the language policies of both the European Union and the Council of Europe, albeit accompanied by different emphases on European identity. While Byram (2004) considers that each of the bodies both makes and implements policy, and while he reflects their view that English as a lingua franca would be politically unacceptable, inefficient for social discourse, and damaging to social identity and personal status, he also acknowledges that 'principle does not necessarily end in practice'.
The present article seeks to measure the extent to which English is progressively becoming 'the language of higher education' in Europe (Coleman 2004a). First, it describes the global context in terms of the spread of English as a lingua franca and of the marketization of higher education (HE). Secondly, it explores the reasons for English-medium teaching in Higher Education Institutions (HEIs) across Europe. Thirdly, it seeks to assess the extent of the phenomenon and its impact, and to explore possible future trends. 1

Globalization
Globalization influences both language use and the economics of HE. It is a complex phenomenon, with positive and negative social impacts, embracing economics, culture, identity, politics and technology (Block & Cameron 2002: 2-5). Globalization is characterized by the compression of time and geographical distance, the reduction of diversity through intensified trade and communication, and new social relationships marked by reduced local power and influence (Giddens 1990: 64;Waters 1995: 3;Mohammadi 1997a, cited in Dörnyei & Csizér 2002Arnett 2002: 74).

Globalization and the spread of English
The spread of English is inseparable from globalization (Hüppauf 2004).
Globalization manifests itself in the increased use of English as a second language world-wide, [and] in the corresponding decrease of importance of other languages in second language acquisition. (Gardt & Hüppauf 2004: x) The emergence of a world language is a wholly new phenomenon: consequently, 'interpretations are highly controversial; furthermore, assessments of these tendencies are often emotionally charged' (Hüppauf 2004: 4f.), as we shall indeed see. The global acceptance of English has been predicted for over 200 years (Kachru 1992a: 2), but resistance to it has grown in the last twenty, partly because of the concomitant phenomenon of language death. ■ Despite today's heightened awareness of endangered languages (Salminen 1999;Crystal 2000;Tsunoda 2003;Ostler 2005), language death is not new: the reduction in the number of distinct codes began with the settlement of hunter-gatherers into agricultural communities 11,000 years ago, intensified during the last millennium thanks to agricultural advances and urbanization, accelerated further through colonization, industrialization and oppressive language policies, and has received an additional impetus from technology, media and concentration of economic power. But the perception that the English language gains from the extinction of others has led to its being portrayed as a 'killer language' (Price 1984(Price , 2000 or as the ultimate predator Tyrannosaurus Rex (e.g. Frühhauf 1997;Swales 1997), to be resisted by a process of 'reverse language shift' (Fishman 1991(Fishman , 2000. Ever wider use of English is promoted through economic, political and strategic alliances, through scientific, technological and cultural cooperation, through mass media, through multinational corporations, through improved communications, and through the internationalization of professional and personal domains of activity (Clyne 1984(Clyne , 1995. The literature describing its growing use and influence is now vast (for example , Graddol 1997;Brutt-Griffler 2002;Crystal 2003Crystal , 2004. Stakeholders who see it as undesirable range from the large and institutional (European Commission 2003) to the individual and particular (Roy 2004).
Many critical commentators link the struggle against the expansion of English to anti-capitalist or anti-imperialist causes, and to the protection of human and community rights (Kachru 1992a, b;Phillipson 1992Phillipson , 2003Widdowson 1993;Pennycook 1994;Canagarajah 1999Canagarajah , 2002Skutnabb-Kangas 2001;Dalby 2003). Others (Nettle & Romaine 2002;Abley 2003;Graddol 2004Graddol : 1329 adopt an environmentalist position, combining language diversity, politics and ecology in equating the survival of languages with that of cultures and populations.
The spread of English and the related -but not correlated -danger of extinction of perhaps 90% of existing languages over the next century is for many an ethical question addressed in value-laden, passionate, anthropomorphic and sometimes apocalyptic terms: 'languages are today being murdered . . . the real culprits . . . brutal market forces' (Skutnabb-Kangas 2001: 201). It is difficult to separate linguistic dominance from other manifestations of power and ideology, and unsurprisingly, some critics see volition or even conspiracy underpinning the spread of English, tying the commercial interest of English language teaching to Anglo-American imperialist hegemony. However, some adopt the more neutral term 'language evolution' (Mufwene 2001), while a few see an international language for crosscultural communication as 'a positive development' (Kachru 1992b: 67), and others credit 'happenstance rather than planning' (Brumfit 2004: 165), for the emergence of a putative global lingua franca. Montgomery (2004Montgomery ( : 1334 asserts that the process of globalizing English seems a rather wild and woolly affair, a cumulative effect of myriad decisions by editors, teachers, students, parents, writers, publishers, translators, officials, scholarly associations, corporations, schools, and so on, with an equally wide array of motives. Within Kachru's still widely adopted (1982;cf. Graddol 1997: 10;McArthur 2001: 3) division of speakers of English into an inner circle (native speakers), an outer circle (second language) and an expanding circle (foreign language), language shift can occur from outer to inner -for example, in successive generations of immigrants. However, shift principally happens from expanding to outer: English ceases to be a foreign language as it acquires a social and sometimes official role across more and more communities. Function determines use, so language and cultural practices are intimately connected. Thus, while languages may sometimes disappear as a result of population loss or political suppression, the process today results essentially from 'voluntary shift' (Nettle & Romaine 2002: 141f.; the term is unfortunate, since any choice is 'within a framework defined and overcast by systematic political and cultural domination'). Where language shift is 'voluntary', language death is top-down. In other words, in diglossic societies, the formal and prestigious functions are the first to be lost: hence the importance of higher education.
De Swaan (2001) addresses global language change through political economy and sociology, applying world system theory to put languages into one of four categories. The peripheral category comprises 98% of the world's six thousand or so languages, which are used by less than 10% of its population and have no written form. Around 100 central languages are used by 95% of the population: these are mostly national and official languages, languages of record (and of higher education: de Swaan 2001: 4). About a dozen supercentral languages, each spoken by over a hundred million speakers, are shared across central languages, and used for long-distance and international communication (as well as for higher education in ex-colonies). One hypercentral language is used between speakers of different supercentral languages: English.
Language exchange, de Swaan notes, is never on even terms. The majority of languages learnt over and above the mother tongue(s) will be from a higher category. Taking language as a form of collective good, he can calculate a Q-value for each language by multiplying the proportion of the entire population who use it by the proportion of multilingual speakers who use it. The higher the Q-value of a language, ■ English-medium teaching the more desirable its acquisition. Such a calculation puts in perspective Graddol's (1997Graddol's ( , 2004 projections based on populations, suggesting that even if the raw number of mother-tongue Spanish-or Arabicspeakers were to surpass that of English-speakers, these languages will not attain the desirability of English nor supplant it as lingua franca. 2 Crystal (1997: 10, cited by Alexander 2004) makes a similar point, arguing an intimate link between language dominance and cultural power, regardless of number of speakers.
Even dispassionate observers draw a clear distinction between a lingua franca, used in communications between two non-native speakers, and asymmetrical dominance, where communication takes place between one native speaker and one non-native speaker (Ammon 1991). And accepting English as a lingua franca (Huber 1997;Wright 2000;Seidlhofer 2001) does not necessarily mean siding with the imperialists. Already, non-native speakers of English outnumber native speakers in a ratio of about 5:1 (Kachru 1996: 241;Crystal 2003: 69). Once ELF (English as a Lingua Franca) has been objectively described as a variety and has lost its stigma, and once the previously assumed target of native speaker proficiency is set aside as 'unrealistic and unnecessary' (Ritzen 2004: 37), then new and less inequitable conceptions of global English and its learning and teaching become possible. Such conceptions could perfectly well embrace democracy and promote human and linguistic rights. However, Seidlhofer (2004) is herself acutely aware of the risks of putting her ELF corpus in the public domain today, lest confusion between description and prescription lead to misrepresentation.
Not all the global drivers of Englishization apply to the same extent in Europe. The countries involved are not ex-colonies; science and technology, broadcasting and economic power are more evenly spread than in other continents. And Europe is further distinguished by shared policies on language and on higher education.

Globalization and the economics of higher education
When last universities shared an international language, in the Europe of the Middle Ages and Renaissance, and lecturing or publishing in the vernacular rather than in Latin was actually censured (Nastansky 2004: 49), higher education was the reserve of a smallélite. Today HE belongs to a globalized market. In the struggle to determine university policies, a new balance has emerged between autonomous institutions, individual and col-lective governments, and the market. The United States, Australia, New Zealand and Japan are seeking to include HE in the General Agreement on Trade in Services (GATS), and European governments, struggling to retain their treasured national institutions and the cultural heritage they represent, are unlikely to resist indefinitely. Students and academics are more mobile than ever before, and competition for both is becoming fiercer. The significant correlation between national economic prosperity and rapid expansion of HE, demonstrated positively in Ireland and Finland and negatively in France and Germany, is encouraging European states to adopt the Lisbon 2000 target of 50% of young people in HE. Such expansion requires additional funding, especially since the new recruits would previously not have considered university and will need supplementary support. Yet meeting the supplementary financial demand out of general taxation is politically unpopular: hence the introduction and inexorable rise in student tuition fees in several states. But Europe is ageing: there will be fewer 18-year-olds to take up the places, and already, in many domains, in a reversal of the relationship which characterized the twentieth century, supply exceeds demand. The combination of higher individual fees, greater student mobility, and excess of supply over demand has accentuated the market character of HE: the student has become the customer. Universities are no longer institutions but brands. University rankings, modelled on North America, and which already inform student choice in the UK, Germany and other European countries, have now gone global thanks to Shanghai's Jiao . National selfinterest in attracting fee-paying international students seems likely, as I have argued elsewhere (Coleman ■ 2003, 2004b, cf. Ritzen 2004, to overtake any altruistic implementation of the Bologna Process, leaving the way free for market forces. Launching the UK Government's International Strategy for Education and Skills (also in November 2004), the Secretary of State asserted that 'the global market for international tertiary education is expected to rise from 2.1 million places in 2003 to 5.8 million by 2010', and that 'the UK needs its share'. The British Council has set a target of 850,000 international students by 2020; Shattock (2003: 48) estimated the UK share at over £1 billion a year, whilst the Secretary of State spoke in December 2004 of 'more than £4 billion'. Globally, the HE market has been growing by 7% a year since the late 1990s, with two million students (nearly 2% of the world's total) paying an estimated $30 billion in tuition fees (The Economist, 26 February 2005).

English and higher education
While the global status of English impels its adoption in HE, the adoption of English in HE further advances its global influence. The process might be termed the Microsoft effect: once a medium obtains a dominant market share, it becomes less and less practical to opt for another medium, and the dominance is thus enhanced. Béacco & Byram (2003: 52) opt for a different metaphor: 'a self-reinforcing upward spiral operates in favour of English as the first foreign language in almost all educational systems'. The conclusion is the same: 'English is the language of science. That is the language we have to use if we wish to prepare our students for an international career in a globalizing world' (Kruseman 2003: 7). But is English in HE 'the academic lingua franca' (van Leeuwen 2003: 20) or the 'lingua franca trap' (Breidbach 2003)?
All education is influential. The extent of foreign language learning in the curriculum at primary and secondary level, and the hierarchy of languages learnt, are highly significant. In Europe, the most recent survey finds that a majority of primary school children in ten countries are learning English (Eurydice 2005a: 44), and shows graphically that 'English is the most taught language in virtually all countries. . . . Furthermore, in both primary and secondary education its dominant position is becoming even stronger' (Eurydice 2005a: 11). At HE level, Graddol has described the global trend and its sociocultural and economic consequences: One of the most significant educational trends world-wide is the teaching of a growing number of courses in universities through the medium of English. The need to teach some subjects in English, rather than the national language, is well understood: in the sciences, for example, up-to-date text books and research articles are obtainable much more easily in one of the world languages and most readily of all in English. (Graddol 1997: 45) He suggests that English-speaking graduates may extend the language to social use, and pass it on to their children, as it becomes a marker of social privilege: 'English-medium higher education is thus one of the drivers of language shift, from L2 to L1 English-speaking status' (1997: 45). At the same time, developing countries can themselves become exporters of educational services. Also on the positive side, in formal international scientific settings, English facilitates global academic exchange, advancement of knowledge, and career advancement and mobility (Montgomery 2004(Montgomery : 1334. Although English is already the language of HE in Africa, where a recent Ph.D. thesis defined principles of English-medium course design (Coetzee 2004), the role of HE has generally received less attention in Europe until very recently. Nettle & Romaine (2002: 31) refer only in passing to the influence of Englishmedium university teaching, while its omission from Clyne's (1995) list of factors supporting the spread of English, even though Clyne recognizes that younger, educated people tend to be the drivers of linguistic shift, underlines how recent is the trend.

Drivers of the Englishization of European higher education
The reasons which have impelled individual HEIs, as well as regional or national governments, to introduce programmes and courses taught through the medium of English may be allocated across seven categories: CLIL, internationalization, student exchanges, teaching and research materials, staff mobility, graduate employability and the market in international students. It might be suggested that this rainbow of motives ranges from the ethical and pedagogical through the pragmatic to the commercial. Foreign language learning in itself is NOT the reason why institutions adopt Englishmedium teaching. The relationships between the seven drivers are considered below, although only CLIL is treated in detail.
In its diverse manifestations, CLIL (Content and Language Integrating Learning 3 ) resembles the approach which North America calls 'immersion', in which parts of the curriculum are delivered through a second or foreign language. Learners acquire the target language (TL) naturalistically by studying content through it (Swain 1996;Snow & Brinton 1997;Marsh & Hartiala 2001;Wilkinson 2004), and TL exposure is increased without a correspondingly higher demand on precious curriculum time. The many dynamic European CLIL projects seeking to achieve the double benefit of subject knowledge ■ English-medium teaching and improved target language proficiency 4 received a boost from the widely reported study (Mechelli et al. 2004) showing increased grey-matter density of the left inferior parietal cortex in early language learners, translated into headlines as 'Learning languages boosts brain'. Jäppinen (2005) reviews the claims that CLIL can also advance intercultural knowledge, understanding, communication and mobility. Less studied than immersion or bilingual education, CLIL is usually found to enhance TL proficiency without concomitant problems in cognitive development or in knowledge of other subjects. The study (Marsh, Hau & Kong 2000) which found large negative effects in non-language subjects among Hong Kong students when TL education was delayed until secondary school (cf. also Tollefson & Tsui 2004) would not meet the definition of CLIL which requires 'focus on non-language subject-matter, and focus on the language in terms of both communication and cognition' (Marsh & Laitinen 2005).
But if CLIL has demonstrated benefits at secondary level, and is firmly harnessed to European ideals of multilingualism and the MT + 2 formula (mother tongue and two additional languages for all citizens including school pupils), in reality the target language adopted as the medium of teaching can often be English. In a recent compendium, D. Marsh feels the need to assert more than once that 'there is no a priori reason why English should be the main target language ' (2002: 71, cf. 70, 76, 77); yet of seventeen case studies, two are in the UK, and English figures in fourteen of the other fifteen.
CLIL has not yet been widely adopted in HE, although Marsh & Laitinen (2005) offer practical and research-informed suggestions for raising the quality of English-medium teaching in European HE which might also reduce the impact of market forces and resist the undermining of linguistic pluralism. But ironically, in Europe's English-speaking countries, in the one domain where language and content were frequently integrated and where non-Englishmedium classes were common, i.e. in Modern Languages, the move has typically been away from the teaching of content through the TL. When Parkes (1993) extolled the benefits of having the same academic linguists delivering both language and content courses, much of the teaching of Business, Cultural, Literary and Area Studies took place in the target language. The advent of modularized courses uniting in a single classroom students from a range of disciplines, some without the necessary linguistic proficiency, allied to the stark reduction in specialist language courses, has led to an increase in the proportion of the 'content' of a Modern Language degree delivered through English, while the foreign language is taught as a skill by a separate group of staff (McBride 2000;Rodgers et al. 2002;Coleman 2004c).
Virtually all HEIs recognize the imperative of internationalization, and its potential impact on modernization, on the quality of the student learning experience, on raising the cultural awareness, perspectives and skills of indigenous academic staff and students, on the attractiveness of an institution to staff and students both local and global, and on profile and prestige. Internationalization is directly linked to introduction of English-medium HE teaching (Marsh & Laitinen 2005). In Central and Eastern Europe, European integration and EU membership fuel an additional need to engage with the market economy.
However, in the studies exploring reasons for moving to English-taught courses (OECD 1995;Tella 1999;Räsänen 2000;Ammon & McConnell 2002;Maiworm & Wächter 2002;Hellekjaer & Westergaard 2003;Kurtán 2004;Lehikoinen 2004;Marsh & Laitinen 2005), the initial impetus typically emerges as participation in higher education exchange programmes. In countries whose national language(s) are little taught elsewhere, bilateral exchanges are only possible if courses are delivered through an international language, most frequently English. An opportunity to study abroad is at the same time seen as better preparing domestic students for international careers. Regrettably, such studentcentred impulses have often now been overtaken by a desire to share in the lucrative European and global markets in university students. The phrase 'international students' increasingly means not the 'organized mobility' of mutual exchanges but the 'spontaneous mobility' of fee-paying individuals. And in today's HE market, internationalization is necessary even to attract domestic students (Kurtán 2004: 131).
In either case, the recruitment of international students and international staff, which English facilitates, leads to enhanced institutional prestige, greater success in attracting research and development funding, and enhanced employability for domestic graduates. Institutional and individual self-interest thus coincide both for academic staff, whose international careers depend on a demonstrated ability to teach and publish in English, and for students whose access to a good employment track on graduation also depends heavily on their proficiency in English. Employability of graduates is, in turn, a common criterion of university rankings.
Thanks to universities' dual function as teaching and researching institutions, a powerful impact is exerted by the language of academic publication. Whilst a century ago English, French and German ■ were in near parity, share of global academic literature correlates with economic power (Ammon & McConnell 2002: 11-20), and the research which teachers cite in today's classrooms is increasingly in English, not only in sciences but across the disciplinary panoply (Hoberg 2004: 91, citing Ammon 1998. The process of supplanting other languages as the medium of academic communication was accelerated in the 1960s when U.S universities abandoned foreign language requirements for students (Ammon 2004: 164). It is being completed by the dominance of English-language online academic databases: searching for keywords 'language', 'langue', 'Sprache' and 'nyelv', for example, shows the unevenness of the battleground. Publication and teaching in English also allow academics in poorer states to improve their career prospects by becoming job-mobile.

English-medium teaching in European higher education
In claiming that 'for the first time in recorded history, ALL the known world has a shared second language of advanced education', Brumfit (2004: 166) asserts that, globally, English is already the language of higher education. In Europe, English is 'the most dominant L2 medium of instruction', with its position forecast to strengthen further (Marsh & Laitinen 2005: 2). Its academic potential was recognized a generation ago by Ferguson (1981: xvif.): English is widely used on the European continent as an international language. Frequently conferences are conducted in English (and their proceedings published in English) when only a few of the participants are native speakers.
And a contemporary survey (Schröder & Macht 1983) of 1916 German, Belgian and Finnish university students found 69.1% support for a single official European language, which 84% wanted to be English.
It is, however, in the last fifteen years that Englishmedium teaching in European universities has shown exponential growth, initially in Masters courses but increasingly also in undergraduate degrees. Although HEIs in the Netherlands and Sweden were teaching through English as early as the 1950s, and others such as Finland, Hungary and Norway had followed suit by the 1980s, the trend really takes off in the 1990s, both in Western Europe and, for rather different reasons, in Central and Eastern Europe. Once introduced, the upwards and outwards pattern of ever more Englishmedium courses across ever more disciplines is universal: 'increasingly more universities are offering programmes wholly or partly in a foreign language, almost always English' (Ritzen 2004: 33).
Our knowledge of the real situation regarding English-medium teaching in European HE depends on two large and several smaller studies. The most recent major quantitative survey has been the Academic Cooperation Association Survey conducted in 2001/02 (Maiworm & Wächter 2002), including data from 1558 HEIs involved in SOCRATES-ERASMUS programmes in 19 countries where English is not a native language. It is important to note that the study documents only entire programmes delivered through English, and not the much larger number of individual courses. Maiworm & Wächter found that overall only 1% of programmes were delivered through English, but 30% of universities had at least one Englishmedium programme. The figure was 100% in both Finland and the Netherlands. The authors stress that English-medium teaching is a recent phenomenon: only 8% of the courses had existed prior to 1990, and a majority had been launched since 1998. They were typically located in larger international universities, especially those north of the Alps. The first programmes had been situated in Engineering and Business courses, and at the postgraduate level. The Bologna Process, of course, obliges HEIs to certificate foreign language competence in Masters level courses, and in most cases this naturally means English, so the emphasis on postgraduate teaching is likely to continue (cf. Marsh & Laitinen 2005).
A less widely cited study (Ammon & McConnell 2002) was undertaken earlier, in 1999/2000, following Germany's introduction of Internationale Studiengänge in 1997/98. As well as covering Germany in depth, it adopted a pragmatic methodology to survey English-taught programmes in 22 European countries. For each, it quantifies national population, school and HE students, foreign students, HEIs and English-taught programmes. The text describes types and numbers of programmes and student enrolled, start dates and rationales, overall trends and goals, and problems faced. The authors also reference earlier national studies. Once again, postgraduate teaching leads, and the first subjects to adopt English are typically Economics/ Business Administration/Management on the one hand and Engineering and Science on the other, with extension to Law, Medicine, Agriculture, Mathematics, Information Technology also widespread, and Humanities noticeably much less common. Table 1 summarizes the findings, stripped of caveats and details. Because of its methodology, the study's coverage is a little uneven, and data refer only to entire programmes of study, not individual courses.
English-medium teaching has been so widely adopted despite predictable problems: • inadequate language skills and the need for training of indigenous staff and students • ideological objections arising from a perceived threat to cultural identity and the status of the native language as a language of science ■ English-medium teaching • unwillingness of local staff to teach through English • the lack of availability on the international market of sufficient anglophone subject specialists • the inability of recruited native speaker tutors to adapt to non-native speaking students • inadequate proficiency of incoming international students in the host language • organizational problems and administrative infrastructure • lack of interest from local students • loss of confidence and failure to adapt among local students • lack of critical mass of international students • lack of cultural integration of international students • financing the teaching of international students where no fees exist • financing for international students from poorer countries where fees do exist • uniformity and availability of teaching materials • equity of assessment for native and non-native English speakers (Smith 2004)  The problems can arise from the marginalization of an HEI's linguists from the making and implementation of decisions: even if staff have an adequate command of English (and questions often remain over verification and appropriate staff development opportunities), they are unlikely to have specialist knowledge of the particular demands of university-level education through an L2, where mixed ability becomes the norm and complex content exacerbates already high cognitive processing loads (Marsh & Laitinen 2005). English-taught programmes have sometimes attracted more capable and diligent students (Hellekjaer & Wilkinson 2003;Wilkinson, personal communication).
Although there exists no comprehensive, reliable and wholly contemporary survey, national illustrations of an accelerating trend can be added to the general picture painted by the two earlier studies.
The University of Maastricht is at the forefront of English-medium teaching. English-medium has been offered since 1987, originally in a perspective of bilingualism: but already the Economics Faculty teaches exclusively through English, and it is likely that the whole institution will shortly become English-only -including the administration. The University, 35% of whose students are international, also provides a focus for discussion of the Englishization of HE at conferences, the most recent in 2003 (Wilkinson 2004) and the next in 2006. 5 In Hungary, before 1989, English had the status of a social accomplishment, and the few students seeking to improve their access to the classics of English literature undertook academic grammar-translation exercises. Less than a decade later, the need for an international lingua franca arising from the demise of the Soviet Union and integration into global markets led more than half of school pupils across Central and Eastern Europe to learn (rather than study) English, and the percentage of Hungarian students learning English rose from below 20% to over 50% (Enyedi & Medgyes 1998: 4). Dörnyei & Csizér (2002) provide a longitudinal survey of globalization, motivation and language choice in Hungary between 1993 and 1999. They show that Russian has virtually disappeared, that there is lower cultural interest in all foreign languages and that actual contact with native speakers is falling for all languages despite increased opportunities. On the two conventional axes of longterm language learning motivation, integrativeness and instrumentality (Gardner & Lambert 1972;Gardner 1985;Dörnyei 2001) ■ dropped significantly during the six-year period. Instrumental attitudes are stable for German, the traditional regional lingua franca, and up for English, but down for all other languages. Hungarian higher education typifies Central and Eastern Europe in many ways: marketization is advanced by 'the intersection of high private demand with inadequate public supply [and] teaching in English is spreading fast' (Floud 2004).
In Norway, O'Driscoll (2004) cites evidence that electrical engineering, computer studies and fisheries have been taught through English for 15 years, while Ljosland (2003Ljosland ( , 2004cf. Eik-Nes 2004) has argued that the dominant status of English in HE as a marker of professional success and scientific progress is driving language shift in other prestigious domains, so that it risks displacing Norwegian as the High Variety. Airey (2004) cites evidence that a similar process is under way in Sweden, and a rare example of deliberate reduction in English-medium teaching to maintain academic Swedish. Finland has the largest share of HE in English outside English-speaking countries, is spoken of as 'Little England' in CLIL circles, and is now second choice for students who have failed to obtain an exchange with the UK (Lehikoinen 2004: 46).
In Turkey, as in much of East and Central Europe, private sector HE has stepped into the gap between supply and demand left by under-funded and slowreacting state institutions. A list of private universities with at least one Faculty teaching through English would include Atalim University, Baskent University, Bilkent University and Cankaya University (all in Ankara); Bahcesehir University, Bilgi University, Koc University and Sabanci University (all in Istanbul); Izmir University of Economics and others. In the public sector, Ankara has ODTU (Orta Dogu Teknik University), METU (Middle East Technical University in English), Bogazaci University and Hacatape University; Istanbul has ITU (Istanbul Teknik University) and Yildiz Technical University; there is the Ataturk University in Eskisehir, and the list is far from complete 6 (cf. Fields & Markoc 2004).
In Cyprus, all private and most public vocational tertiary education has been in English for several decades, and although the University of Cyprus, founded in 1992, adopted Greek and Turkish as teaching languages, it has had to adopt 'flexibility measures' to allow English-medium teaching (Karyolemou 2004).
It is several years since concern was raised about English-medium teaching in German universities (Ammon 1998). Today, 'in their attempt to keep their educational programs internationally attractive, universities in Germany have begun offering entire degree programs in English' (Gardt & Hüppauf 2004: xi). Erling & Hilgendorf (2004) trace federal policy measures in support of HE Englishization, and provide a fascinating case study of the official and unofficial expansion of English use in one university. Despite internal press criticism (Krischke 2004), the policies have been crowned with success: a 15 million euro advertising campaign has achieved record recruitment of international students, up 63% in six years: with over 163,000 international students, Germany has overtaken France and is now the third largest global recruiter. Internationalization is still predominantly at postgraduate level, however, with only 600 of 15,800 courses at Bachelor and Masters level taught through English, involving fewer than 3% of students, according to the latest survey (Nastansky 2004). Meyer (2004) acknowledges that English now dominates the communication of research findings across the natural sciences in Germany, and extends to domains such as Economics and Linguistics; but while holding out the hope that Geisteswissenschaften will not go the same way as Naturwissenschaften, that the Humanities' embedding in national traditions will impede their transmission through English, he has to recognize that the new International University of Bremen, 'which uses English as the language of instruction and regards itself as a model for higher education in Germany' also offers Humanities programmes, albeit divorced from German history and culture (Meyer 2004: 76).
Even in France, where foreign policy for so long fought to maintain French as an alternative lingua franca before adopting a protective multilingualist stance, there is now acceptance of the need for more English-medium university courses, and even acceptance of English as lingua franca: 'La langue commune de l'Europe ne peut pasêtre autre, dans un schéma réaliste, que l'anglais. C'est le langage technique de la communication, la langue de travail commune' (Alfred Mahadavy, conseiller du commerce extérieur de la France, janvier 2004). EduFrance, <www.edufrance.fr>, exists to recruit international students.
'Danish universities are increasingly expected to run like businesses, to profile and market themselves competitively. One symptom of this is an increasing use of English', writes Phillipson (2003: 47). Of how many other European countries could the same be said? The Baltic Republics, with an insistence on the national language in nationally accredited HEIs, form an honourable exception, although even there English-language HEIs can be found, including business schools in Riga (Latvia) and Klaipeda (Lithuania) (Hogan-Brun & Ramoniené 2004; Hogan-Brun personal communication).
Such snapshots cannot amount to a full survey, but currently, although there is universal anecdotal recognition of an accelerating trend towards Englishmedium teaching, no comprehensive and up-to-date ■ English-medium teaching data is available. The Eurydice national educational surveys, <www.eurydice.org>, contain some examples, but no overall pattern. Van Leeuwen (2004) has proposed a graphic model on which university profiles could be plotted, comprising five dimensions (mother tongue(s) and language(s) of instruction, administration, environment and the labour market).
A survey undertaken in 2004 under the aegis of the European Language Council confirms that universities in countries whose national language is not widely learnt elsewhere lead the Englishization process, and that mobility programmes may be helping to accelerate it rather than promoting reciprocal multilingualism (Marsh & Laitinen 2005). There seems no doubt that 'English has become the normal medium of instruction in higher education for many countries -including several where the language has no official status' (Crystal 2004: 37), but comprehensive statistical data is still lacking.

Impact of English-medium teaching in Europe
One heading in a recent and much-publicized paper on the future of English (Graddol 2004(Graddol : 1329 predicts 'The end of Modern Languages'. If this may appear exaggerated, there is nonetheless wide recognition of the problems posed in Europe by the increase in use of English in higher education (Brock-Utne 2001;Phillipson 2001;Béacco & Byram 2003;Marsh & Laitinen 2005), as indeed by the quasi-universal adoption of English as the first foreign language in school curricula. The unmanaged expansion of English certainly threatens to undermine the policy of both the Council of Europe and the European Union (see Holdsworth 2004 for a recent summary). Like Byram, Chambers (2004) predominantly addresses policy rather than practice in contrasting the many European initiatives with the few coherent national policies, and accepting too that continued linguistic diversity requires more than international policies and utopianism. The Bologna Process guarantees 'proper provision for linguistic diversity', and a new popular booklet seeks to convey to EU citizens the Many tongues, one family message. 7 Yet the European Year of Languages in 2001, promoting plurality and linguistic diversity, and demonstrably successful in many respects (see <http://europa.eu.int/comm/education/policies/ lang/awareness/year2001_en.html>) is still seen by sceptics as a failed attempt to mask the failure of multilingual policies and the reality of English dominance. With pan-European ideals dented by referendum results in France and the Netherlands, it may prove increasingly hard to persuade autonomous HEIs that they have a social duty to moderate their institutional imperatives in the interest of national or supra-national goals related to linguistic diversity. The Bologna Process itself may contribute to the problem: the Diploma Supplement describing graduates' acquired knowledge and competences, a key element of Bologna along with the two-cycle structure, the European Credit Transfer System (ECTS) and quality evaluation, is 'generally issued in English or in the language of instruction and English' (Eurydice 2005b: 27).
Mobility, in the form of study abroad, has been a central element of European policy and its implementation for two decades, re-asserted in the 2003 Action Plan (European Commission 2003) and reiterated in the 2003 Berlin Declaration, which stresses 'the necessity of ensuring a substantial period of study abroad in joint degree programmes as well as proper provision for linguistic diversity and language learning' (ibid.). For even longer, the principal learning outcome of student residence abroad has improved foreign language proficiency (Stern 1964: 91;Coleman 1997Coleman , 1998Coleman : 182, 2004dTeichler 1997). In the United Kingdom, recruitment to language degree programmes is already in a steep decline (Moys 1998;Nuffield Languages Inquiry 2000;Kelly & Jones 2003;Coleman 2004c), which changes to the school core curriculum will further accentuate (CILT 2003(CILT , 2004. The proportion of UK students opting for European student exchanges is also falling, for linguistic and financial reasons, and in response to rising interest in work placements, and to the perceived high quality and employability benefits of anglophone universities in the US and elsewhere (Sussex Centre for Migration Research 2004;cf. Ammon 2004: 7). It is therefore a matter of concern when, as the author has recently experienced at student recruitment events in UK universities, study abroad is touted as a selling point, but with the added attraction that students do not need to study a foreign language, even for European placements. On the contrary, the increasing adoption of English-medium teaching in European HE will arguably reduce the attraction and certainly the benefits of study abroad for native English-speaking students, stripping away exposure to genres and registers of the target language which they will not encounter elsewhere. At present, in university exchange programmes, because of compulsory assessments, the classroom is the one place where British and Irish students are obliged to engage with the foreign language and its attendant cultures: replace the target language with English in lectures and practical classes, and many more will be able to spend their period abroad as tourists, skimming the surface of their host country, without the deeper involvement that will bring maturity and intercultural awareness. It may not just be to native speakers of other languages that the Englishization ■ of European HE means an impoverished learning experience.
The likely impact of English-medium teaching on student exchanges has indeed been recognized: At some stage, the EU will have to face the mismatch between its support for diversity within the foreign language curriculum (for example in the Lingua and Socrates programmes) and the reality of a dominant lingua franca'. (Wright 2001: 52) EU exchange programmes have helped over one million young people to study abroad over the past three decades, although in any given year fewer than 1% of Europe's 40 million students do so. But globally, 2.12 million students were studying abroad in 2003, and the figure is set to rise to 7.2 million by 2025 (Sussex Centre for Migration Research 2004). The opposing impact of student mobility on languages in higher education may be even more profound.
While Smith argues that, for the individual, the gains of study through an additional language outweigh the losses, there are risks of language attrition and loss of cultural identity (2004: 87), to which Sercu adds, on the basis of substantial evidence, 'a decrease in the quality of teaching and the students' overall learning results and an increase in study/ teaching load ' (2004: 547). Adopting English requires a switch from a focus on the national system and culture to an international focus, which can entail the loss of some programmes, and complex adaptation of others such as Medicine or Law where the principles are global but the practical implementation is local (Ritzen 2004). International recruitment of anglophone academics might further spread the influence of North American models of HE.
There will also be an impact on academic communication itself: it is argued (Block & Cameron 2002) that 'effective communication' is not a neutral and objective term but can disguise the dominance of a particular genre and style, namely American academic English. Block here and elsewhere (Block 2003) takes the argument further, suggesting that in Second Language Acquisition (SLA) research too, a dominant North American model centred on 'task' and 'negotiation of meaning' narrows the meaning of 'communication' by ignoring social context and other non-psycholinguistic elements including face and identity. Academic discourse, like any other discourse, is culturally bound, and translation into English implies more than merely linguistic change.
The teaching of English for Academic Purposes has for many years been a mainstay of universities in the United Kingdom and in English-speaking countries around the world. Despite some surviving distinctions between British, North American and Australian usage, any modern coursebook shows that there is broad consensus on what constitutes a good oral presentation or a good essay: see, for example, Lowes, Peters & Turner (2004). What becomes of the famous these-antithèse-synthèse of a French academic essay? The distinctive approaches to lecturing in Italian, Spanish or German universities are currently part of the benefit of student exchanges, making young people question the narrowly ethnocentric, monocultural perspectives which too many of them take abroad, and a key element in developing their intercultural competence, their recognition that cultural norms are relative and not absolute, socially constructed and not given. How will they learn to look at their own culture in a new light if Anglo-American norms dominate a newly homogenized European academic discourse? If the only language on the PowerPoint slides is English?
Research into how languages are taught and learnt is not central to the present discussion, but it is worth noting in passing that a good deal of published SLA research, especially in North America, shares an unspoken assumption that the target language is English. Because of its global status, and because of the huge numbers of foreign students, with native languages other than English, who attend North American colleges and universities each year, there is more SLA research on the learning of English than on all other target languages put together. It is not just in Japan that 'foreign language' is synonymous with English (Kubota 2002: 19). The danger is that the findings of research into the acquisition of English are assumed to have universal validity, when this is not the case. Motivation, for example, is one of the key factors in language learning, and while motivation as a research field has moved beyond purely social psychology since Gardner and his team first identified integrative and instrumental orientations (Gardner & Lambert 1972), the fundamental integrative/instrumental opposition is among the factors which distinguish learners of foreign languages from learners of English as a global lingua franca (Coleman 1996: 130f.;Dörnyei & Csizér 2002: 453).

Conclusion: a future of universal diglossia
English is clearly the dominant foreign language used in teaching at institutes of higher education in the EU countries. . . . English is generally perceived to be the dominant language of teaching for the future. (Ammon & McConnell 2002: 7) On a broad level, the Englishization of European HE represents an extension of the global threat to minority languages. Higher education, itself now a marketized and globalized commodity, is a prime driver of language shift, embracing as it does the young, mobile, educatedélite, the leaders of social change.
What is striking to a modern linguist seeking insight into the possible outcome of language shift towards English is the consensus across different commentators. The constant which emerges from ■ English-medium teaching recent publications on globalization and language, despite their being written from divergent theoretical and disciplinary perspectives, is that we all seem to be heading for a bilingual and bicultural identity. Ultimately, the world will become diglossic, with one language for local communication, culture and expression of identity, and another -English -for wider and more formal communication, especially in writing.
The psychologist Arnett (2002: 777) believes that '[m]ost people in the world now develop a bicultural identity, in which part of their identity is rooted in their local culture while another part stems from an awareness of their relation to the global culture'. Graddol (2004Graddol ( : 1330 asserts that a 'major impact [of English] will be in creating new generations of bilingual and multilingual speakers across the world'. Wright (2004: 249) confirms that populations increasingly adopt one language for identity and culture, and another language for utilitarian communication: 'many might come to choose this differentiated bilingualism'. The new Englishes of India, many African nations, and excolonies across the globe may become dissimilar as identity-bearing vernaculars always have, but the pressures of intelligibility and the influence of global media are likely to ensure the emergence and maintenance of a standard variety of English as a lingua franca (Crystal 2004). Nettle & Romaine (2002: 190) identify the process of functional differentiation as already well advanced: In today's global village, however, increasing bilingualism in a metropolitan language, particularly English, is making the majority of the world's languages in effect minority languages. Even small languages such as Icelandic with its 100,000 speakers, and larger national languages such as Swedish and Hebrew, substantially protected by national boundaries and institutions, exist in a diglossic relationship with English at the highest levels of international communication.
Adopting an established cliché and a positive spineveryone should embrace the gain of bilingualism (2002: 190) -they conclude that 'we must think locally but act globally: local languages for expressing identities and global languages for communicating beyond local levels and expressing our identities as citizens of the world ' (2002: 197). The political economist de Swaan concurs that the world's population is heading for bilingualism, with one local identity language and another transactional global language, 'much as one might take a plane to a remote destination and ride one's bike on nearby trips ' (2001: ix).
If indeed the world's peoples use one or more native languages for local and cultural communication where their personal identity is engaged, and another for international, formal, practical communication, then it seems inevitable that English, in some form, will definitively become the language of higher education. Yet, supposing a single variety of lingua franca English were ever to emerge, it would be closer to American than to British English. While US English remains 'immune' to other Englishes, it is abundantly clear that the phonology, lexis, syntax and orthography of British English are all too permeable to transatlantic influence (cf. McArthur 2001). But it will probably differ considerably from both.
There is already evidence that students in Englishspeaking countries on SOCRATES-ERASMUS exchanges socialize more with other foreign students than with native speakers, and can better understand other non-native speakers than local students. Academics too who travel, or who deal with students of English, adopt different varieties in their professional and private lives. As ELF diverges further from standard varieties in the UK, the US, Ireland or Australia, these countries too could become diglossic, and native speaker English become a sociolinguistically marked variety, no longer automatically acceptable in international contexts. Then the predominance of international academics with a range of native tongues other than English may well diversify even academic discourse away from today's ubiquitously delocalized Anglo-American standard. And the world may see the emergence of a more democratic model of English as lingua franca prefigured by Seidlhofer's work and powerfully