Teaching staff ’ s views about the internationalisation of higher education : The case of two bilingual communities in Spain

The process of internationalisation of higher education can be seen as fluctuating between two main discourses: economic competition and academic internationalisation (Bolsman & Miller 2008). Within the former type of discourse, internationalisation is constructed as a means to generate income, in competition with other institutions, through the provision of research and teaching services of a high quality to as many ‘clients’ as possible. From the point of view of academic internationalism, internationalisation is represented as a joint enterprise by institutions from different countries for the advancement of human knowledge and intercultural understanding. In this paper we aim to explore the views of the teaching staff of two bilingual universities in Spain in connection with the process of internationalisation of their institution, placing a special emphasis on its impact on language policy. The sample for this study was made up of 173 university teaching staff who completed a questionnaire in which they were asked to express their views on issues such as internationalisation at higher education institutions, academic mobility and their attitudes towards multilingualism. The participants belonged to two officially bilingual universities in Spain, namely the University of Lleida (UdL) and the University of the Basque Country (UBC). Variables such as the sociolinguistic context, gender, mother tongue and age were considered when scrutinising the participants’ answers with a view to drawing a picture of internationalisation which included opinions on academic mobility and multilingualism from the teaching staff’s perspective.


Introduction
The research reported in this paper is part of larger project focusing on the ambiguities and tensions between internationalisation and language policies in three universities located in the bilingual territories of the Basque Country and Catalonia, in Spain, and Wales in the United Kingdom, which, as part of a process of political devolution, are actively engaged in reversing the language shift to the majority language.The phenomenon of internationalisation of higher education is relatively recent in non-English speaking universities, and in Europe it has been accelerated in the last 20 years with the establishment in 1987 of the Erasmus mobility programme, which in the academic year 2009-2010 mobilised a total of 213,266 students and 37,776 members of staff, with an annual increase of 7.4 percent (Lifelong Learning programme, 2011).
Faced with this demand from an increasingly international and linguistically diverse student body, universities, especially those in bilingual territories, have been forced to reconsider their language policies, which in many cases had never been made explicit.This is the case, for instance, with the University of Helsinki, whose language policy document was approved in 2007 and was clearly connected with its internationalisation policy (University of Helsinki 2007: 39): The University of Helsinki Strategic Plan 2007-2009 cites internationalisation as one of five key areas for development.Preparation of the Strategic Plan required a comprehensive view and thorough analysis of the usage of different languages within the University.… For the first time, the University is assembling its views on the relationship between Finland's national languages (Finnish and Swedish), English, and other foreign languages, as well as their status within the University.
A similar process of justification of the need for and the definition of their language policy has been taking place in bilingual universities in Spain.Thus, in 2007 the University of the Basque Country (UBC) also passed its Language Policy Plan on Basque (University of the Basque Country 2007), and a year later, the University of Lleida (UdL) made explicit its language policy in the document Language Policy at UdL: Towards a Multilingual Reality (University of Lleida 2008).In the three cases there is a genuine concern on the part of the institution to make compatible the existing bilingual diversity (special emphasis being placed on the protection of the minority language) with the need for the university to form part of a globalised higher education environment through acknowledging the role of English as a third medium of instruction.
This paper is based on the premise that the success or failure of the implementation process of a particular university policy cannot be fully understood without taking into account the attitudes and practices of the members of the academic community.In this case, we want to focus on the attitudes of teaching staff towards internationalisation, including academic mobility and multilingualism, the two most obvious issues on which it has an impact.Specifically, we want to explore, in the first place, the extent to which the attitudes of the teaching staff depend on their sociolinguistic context (Basque Country or Catalonia), gender, L1 and age.In the second place, we aim at identifying the most important factors that the teaching staff associates with the notions of internationalisation, academic mobility and multilingualism in higher education.

The internationalisation of higher education
Most higher education institutions' (HEIs) representatives will agree with claims about the rising importance of internationalisation in the last 20-25 years.The Bologna Declaration in Europe (1999) and the development of specific internationalisation policies in European universities are two examples of the institutional concern about a manifested need for internationalisation.However, as Schoorman (2000) mentions, HEIs do not always clearly conceptualise what internationalisation represents for them.The oft-quoted definition of internationalisation of HEIs by Knight (2003a;cited in Koutsantoni 2006: 11) sheds some light on the academic debate: 'The process of integrating an international, intercultural or global dimension into the purpose, functions and delivery of post-secondary education'.The concept of integration in the definition suggests 'the process of infusing … the international and intercultural dimension into policies and programmes to ensure that the international dimension remains central, not marginal, and is sustainable' (Knight 2003b: 3).When it comes to implementing such integration, the rationale ranges from economic competitiveness, derived from income generation and institutional accountability, to academic internationalism, encompassing mainly curriculum internationalisation and student and staff mobility.
After analysing European HEIs' internationalisation, Wächter (2008) lists six clusters of phenomena that the term applies to: (i) (physical) mobility for students and staff, including non-degree and degree mobility; (ii) academic activity recognition across country borders, comprising degrees and other qualifications; (iii) curricular reform, whereby the content and delivery of programmes are internationalised mostly through the introduction of Englishmedium instruction (EMI) but also of comparative, regional or international studies (e.g.International Business, European Studies, etc.); (iv) transnational  education (international education, off-shore campuses, etc.); (v) branding and international positioning, aimed at recruiting international students and staff; and (vi) adoption of the entire agenda of the European Higher Education Area, in the form of the same three-cycle degree structure and common guidelines for assurance and accreditation, among others.To these six clusters, others aimed at internationalisation 'at home' could also be added (Knight 2003a;cited in Koutsantoni 2006: 11) including, besides the afore-mentioned curricular reform, the establishment of educational projects with new ethnic (immigrant) groups settled close to the university (cf.European HEIs' managerial agenda for internationalisation takes a neoliberal stance, as one of the dominant raisons-d'être of management is their increasing capacity to attract international students to the institution, especially non-EU students, who, as in the case of British universities, pay higher fees (Bolsmann & Miller 2008).According to the neo-liberal view, education is conceived for its use value, or as an 'investment in human capital which will enhance competitiveness and [will bring] rewards to the individual, corporations and the national economy' (ibid: 78).Moreover, neo-liberal education policies promote competition both internal and external to the institution for the sake of institutional efficiency, organisation (re)appraising and accountability in the workforce (Deem 2001, in Servage 2009).The notion of accountability in the workforce has an important impact on the academic staff, as the institution holds them directly responsible for the learning outcomes and degree of satisfaction of an increasingly diversified student body, including local and international students.

Internationalisation and language policy
The relationship between internationalisation and language policies in higher education has been little researched.Risager (2012) underscores three main types of responses to internationalisation in terms of their language policies.The first type involves a monolingual policy of replacing the national language by English more or less exclusively; this would be the case of some degree programmes in Scandinavian and Finnish universities, especially at postgraduate level, which are only offered through the medium of English.The second type of response consists in a bilingual policy where English is used side-byside with the national language either by offering parallel courses or programmes in the two languages or by introducing specific modules in English.In the third place, Risager mentions a trilingual policy, which is the case of bilingual territories like the Basque country or Catalonia, in which English needs to coexist with the national and the regional language, an option which requires not only a serious investment but also a complex organisation.According to Risager (ibid.), the tendency nowadays is to favour an extended or even exclusive use of English to attract foreign students, especially in Masters and PhD programmes.This tendency to connect internationalisation with the Englishisation of universities in non-English speaking countries has been pointed out by Ljosland (2005) and Coleman (2006), among others.The first author complains about the 'irresponsibility' of the Bologna Declaration in establishing the European Higher Education Area without making any provision for or even reference to the language diversity in Europe, which has led to a situation in which every individual country or university decides on the language policy to be implemented.Thus, many universities, in their fierce competition to attract international students, have opted for removing the potential language barriers by introducing entire programmes in English and not requiring international students to prove their competence in the national language in order to earn a degree.Ljosland (2005) mentions the example of universities in Norway, where, paradoxically enough, while foreign students are not asked to learn Norwegian, local students going on an exchange programme are encouraged to learn the language of their host country.For Coleman (2006) this Englishisation of HEIs is essentially the result of their being pushed into a global market to compete for financial resources linked to either research outputs or the provision of high-quality training/educational programmes.In this sense, having an 'international reputation' is becoming a must for any institution, to the point that it has become necessary even to attract local students (Kurtán 2004).
Since the internationalisation of higher education is a recent phenomenon in many non-English speaking universities, little research has yet been undertaken to delve into what its consequences may be, especially in bilingual communities in which majority and minority languages already coexist.From this point of view, Balfour's work (2007) represents a first step in this process of enquiry in that it focuses on the universities' policies to develop minority languages in South Africa and Wales.This author points out the damaging effects of governments and academic institutions without a real commitment to the promotion of minority languages beyond the political correctness of policy documents.
In order to understand the impact of a language policy on a specific academic community, we need to take into account the daily communicative practices of the members of the institution, the teaching staff constituting a very important group.Given the intellectual (as well as political) power of this sector of the academic community, it seems important to acknowledge that the success or failure of any university policy cannot be fully explained without taking into account the academic practices as well as the ideologies of the members of the academic staff.According to Tange (2010), a university policy of internationalisation can force lecturers to acquire intercultural skills with very little support from the institution; under this circumstance some of the teaching staff, with more experience and pedagogic resources, are favourable, whereas others, less experienced and with fewer pedagogic resources, feel somewhat overwhelmed by the changes in the profile of their student body and the diversity of learning styles with which they now have to cope.Erlenawati (2011) finds some degree of intra-and interpersonal tension between two approaches adopted by lecturers with international students: valuing 'sameness' vs. valuing 'difference'.The author also remarks the greater open-mindedness of humanities and social science subject teachers to adapt to different types of students by abandoning a traditional teaching method which we could characterise as teacher-centred and essentially based on an information-transfer view of education and replacing it with a more constructivist, learner-centred method, according to which the teacher's main role is to design appropriate activities for the learners to negotiate and build their knowledge autonomously.The introduction of EMI (English-medium instruction) is perhaps one of the clearest ways in which internationalisation has impacted on the professional task of teaching staff.Among the obstacles that Coleman (2006: 6-7) mentions for EMI in European universities, he includes the 'inadequate language skills and the need for training of indigenous staff and students' and the 'unwillingness of local staff to teach through English'.This is perhaps more of a problem in countries like Spain, in which the introduction of English as foreign language has a much shorter tradition than in other countries in Europe.However, Spanish teachers and students show a positive attitude towards EMI, even though teachers are slightly less enthusiastic because of their greater responsibility and lack of incentives (Dafouz et al. 2007).The reason for this general satisfaction may be the view, perhaps more prevalent in teachers than students, that English is an essential component in their careers and EMI is an opportunity for developing their communicative skills.This might help to explain the findings by Aguilar & Rodríguez (2012) that after an EMI experience, teachers were more satisfied than the students because they felt that it provides them with spoken practice.

Language attitudes towards multilingualism
In the study of language attitudes in multilingual settings, variables such as the sociolinguistic context, gender, mother tongue (L1) and age have been examined due to their influence on the formation of language attitudes (cf.Huguet 2006;Lasagabaster 2003).The sociolinguistic context variable is approached as the environment where the subject under scrutiny is situated, and how this environment shapes the subject's attitudes towards multilingualism (cf.Portolés 2011).Recent research in bilingual communities in Spain shows how the sociolinguistic context conditions respondents' linguistic attitudes to multilingualism.Lasagabaster (2005) reveals that within the Basque Autonomous Community (BAC) university students living in Basque-speaking environments (i.e.smaller towns and villages) support more favourably Basque (the minority language) than those living in Spanish-speaking contexts (i.e.bigger towns and cities), the latter showing a more congenial attitude towards Spanish (the majority language).The results of this study also bring to the fore that attitudes towards English differ according to the sociolinguistic context: students coming from Spanish-speaking contexts had a more supportive attitude towards the foreign language (FL) than those students residing in Basquespeaking areas.In this same vein, Portolés (2011) obtains similar results for the Valencian Autonomous Community.The outcomes of the study show that students from the Jaume I University, in which Spanish and Valencian are coofficial languages and Valencian has a strong presence in the street, support the minority language with the most favourable attitudes, followed by Spanish and English.In contrast, students coming from the Catholic University of Valencia, which only has Spanish as its official language and is placed in the capital city of the autonomous community, bear the most favourable attitudes towards Spanish and English, and less congenial attitudes towards Valencian (ibid.: 86-91).
As for gender, Pavlenko & Piller (2007) concede that multilingualism is a gendered practice in terms of access to valued languages, but also in terms of multilingualism as a commodified practice.Research suggests a tendency for women to show a more positive attitude towards multilingualism.In Ladegaard (1998), for instance, while the boys show great attachment to the general use of their regional dialect independently of the occasion, many of the girls, despite thinking that being able to use dialectal varieties is 'exciting' or 'nice', show a tendency towards standard Danish in specific situations.Women's more positive attitude towards multilingualism can also be seen in Bilaniuk's study (2003: 73), according to which Ukrainian women are 'more likely to pursue the benefits accorded by Russian and English than are men'.In this case, as in  Ladegaard (1998) or Gal (1998), women's more positive attitude towards multilingualism can be explained by the fact that they consider that it gives them better opportunities for social and economic promotion.Kissau's (2006: 85) grade-nine students in Ontario provide one more confirmation for Piller & Pavlenko's idea of multilingualism as a gendered practice, as boys report being less good at French because they feel that society does not consider it 'normal' to be successful in French and, therefore, they fear negative social appraisal.
Considering the effects of the variable mother tongue (L1), Portolés maintains that it is the most influential variable in language attitude formation, as speakers hold positive attitudes towards their L1 'as a sign of identity and belonging to their roots ' (2011: 44).As reported in Lasagabaster & Huguet's research on 9 European bilingual communities ( 2007), two trends may be observable: on the one hand, L1 speakers maintain a more fully supportive attitude towards their L1, be it either the majority or the minority language of the community.This conclusion is evidenced in Catalonia (Huguet 2007: 35), Galicia (Loredo et al. 2007: 58), the Basque Country (Lasagabaster 2007: 86), Friesland (Ytsma 2007: 158), Ireland (Ó Laoire 2007: 179) and Wales (Laugharne 2007: 225).On the other hand, minority language users show a non-integrating attitude towards the FL.Baker (1992: 136) describes this linguistic attitude as a bunker attitude towards bilingual practices, inasmuch as minority language bilinguals stand up for their minority language when interacting with majority language monolinguals.However, this opinion is not extensively evidenced among minority language users in the European bilingual communities examined in the volume edited by Lasagabaster & Huguet (2007).Only 4 out 9 of bilingual communities hold such a bunker attitude towards the FL: the Basque Country (Lasagabaster 2007: 86), Friesland (Ytsma 2007: 158), Ireland (Ó Laoire 2007: 179) and Wales (Laugharne 2007: 225).
We are aware of the fact that the term 'mother tongue' has been criticised in recent sociolinguistic research on the grounds that it may sometimes be vague and ambiguous, particularly in bi-and multilingual contexts such as Catalonia and the Basque Country.From this point of view, it may be claimed that the participants in our study could have felt forced to choose only one language and respond with Basque/Catalan to a question about their mother tongue, even if they in fact grew up bilingually with Basque/Catalan and Spanish.However, we decided to maintain the examination of the effect exerted by this variable for three main reasons.Firstly, because the questionnaire included the option 'Catalan/Basque and Spanish' and therefore the participants were made fully aware of the possibility of choosing this option.Secondly, because the use of the mother tongue variable is deeply rooted in the sociolinguistic surveys carried out in Spanish bilingual regions such as Catalonia, Galicia or the Basque Country.In fact, this item (mother tongue) is always included not only in official sociolinguistic surveys undertaken by the autonomous Basque/ Catalan/Galician governments, but also in many applications such as those which parents have to fill out when enrolling their children in infant or elementary schools.Thirdly, because studies carried out in bilingual contexts (those included in Lasagabaster & Huguet 2007;Portolés 2007, to name a few) recurrently confirm that the mother tongue exerts a very significant influence on language attitudes.The age variable has also been explored to determine its possible effects on linguistic opinions and judgements.Baker (1992), for instance, points out that in Wales positive linguistic attitudes towards Welsh tend to decrease with age among school students, establishing the age of 14 as the turning point.This finding corroborates Gadner & Smythe's study (1975;cited in Lasagabaster 2003), which also reports older children having less favourable attitudes towards the minority language.On the other hand, McDonough (1981, cited in Chambers 1999) concludes that students aged 11 or 12 begin to show more congenial attitudes to foreign languages, as they begin considering the advantages of FL proficiency for employability.This employability factor associated with foreign languages may be extended to include other kinds of instrumental justification.Kormos & Csizér (2008) explore FL (in this case English) learning motivation considering the respondents' attitude towards their English learning process.Three age cohorts from Budapest (Hungary) are selected: secondary school students, university students and adult students.The three age groups share a positive attitude towards learning English due to its role in today's globalised world.However, the results for language learning attitudes of the three age groups differ on the nature of the instrumental justification for language learning: the motivation for EFL learning held by the secondary-school students and university students is underpinned by classroom experience and academic pressure, that is to say, compulsory foreign language learning and accreditation in Hungary; and adult EFL learners seem to decide to learn English to improve their employability or even for leisure.

Two bilingual sociolinguistic contexts
The Basque Autonomous Community (BAC) and Catalonia are two of the 17 autonomous communities that make up Spain.Both are bilingual communities in which Basque or Catalan, respectively, are official languages together with Spanish.It is important to point out that whereas Basque is a pre-Indo-Euro- pean language (Trask 1997), genetically unrelated to Spanish, Catalan is a Romance language and, therefore, it shares a large number of basic linguistic features with Spanish.
In 2011, the BAC had a total population of 2,184,606.As for language competence, 32 percent of Basque citizens are fully bilingual, 17.2 percent passive bilinguals and 50.8 percent non-Basque speakers (Basque Government 2012).In the period 1996-2006, there was a significant increment (5.4 percent) of full bilinguals, a rise which relies mainly on the educational system.The percentage of bilinguals has increased in all age groups below 50, but the highest percentage of bilinguals is found among people under 35; in fact, 60 percent of the population between 16 and 24 are nowadays fully bilingual.
According to the Institut d'Estadística de Catalunya (n.d.), in 2011 Catalonia had a population of 7,539,618.As for language competence, in 2008 95 percent of the population could understand spoken Catalan and 75 percent could speak it; 35.6 percent of the population declared Catalan as their ordinary means of expression, 45.9 percent Spanish and 12 percent both.If we compare these figures with those obtained in the previous 2003 survey (50.1 percent Catalan, 44 percent Spanish and 4.7 percent both), a significant increase in the number of people who use both Spanish and Catalan on an ordinary basis can be observed.This increase has been mainly at the expense of habitual users of Catalan.
In the Basque educational system both Basque and Spanish are taught throughout all the rungs of the educational ladder.However, the presence of these languages at pre-university education varies depending on the linguistic model of the school which parents can choose.In model A, the language of instruction is Spanish, and Basque is taught as a second language for 3-5 hours a week.In model B, both Spanish and Basque are used as languages of instruction for about half of the school time each, although this is a rather heterogeneous linguistic model.Finally, in model D Basque is the only language of instruction and Spanish is taught as a subject for 4-5 hours a week (see Zalbide & Cenoz 2008 for further information on this issue).Similarly, Basque is used as means of instruction in the three universities located in the BAC to different degrees, depending on the availability of Basque-speaking teaching staff.
The pre-university educational system in Catalonia does not contemplate a division of the students according to the presence of Catalan.The Law of Education of Catalunya (Generalitat de Catalunya 2009) establishes Catalan as the ordinary language of communication and learning for the educational system, and the schools must make a special provision to provide individual assistance to those students entering the system whose first language is Spanish and do not understand Catalan.Spanish is taught as a second language and the school must guarantee that by the end of their compulsory education, students will be fully competent in both Catalan and Spanish.In universities Catalan is the most common language of instruction.
The UBC is the only public university in the BAC, with one campus in each of the three provinces that from part of BAC (Álava/Araba, Biscay and Gipuzkoa).More than 78 percent of the students enrolled in tertiary education in the BAC complete their degrees in this university.In the academic year 2008-09, there were 47,000 students registered.Although most of the courses at the UBC are offered in Spanish, there is often a parallel offer of those same courses with Basque as medium of instruction, which in the academic year represented 77 percent of the compulsory subjects.Some courses are also taught in English and French as part of the so-called Multilingualism Programme: 126 subjects in the 2010-11 academic year.
The UdL is one of the smallest universities in the autonomous community of Catalonia.It is located in Lleida, the capital city of the westernmost province of the four territories into which Catalonia is divided.The university student body in the academic year 2008-09 comprised a total of 8,800 students, a figure which represents only 4 percent of the total university student population for Catalonia (Institut d'Estadistica de Catalunya n.d.).Based on the results of the annual survey on the use of languages at the institution (University of Lleida 2011) in the academic year 2010-11, the lecturers used Catalan in 65.6 percent of the subjects, Spanish in 28.2 percent, English in 3.2 percent, other languages in 0.7 percent; and the lecturers left the survey unanswered in 2.3 percent of the subjects.

The research questions
Based on the previous review of the literature, this study revolves around two main research questions: 1.How do independent variables such as the sociolinguistic context, gender, L1 and age affect teaching staff's views on internationalisation, academic mobility and multilingualism at university? 2. In the teaching staff's opinion, what are the most important factors when it comes to internationalisation, academic mobility and multilingualism at university?The participants were 173 teaching staff members, 40 percent of them working at the University of Lleida (UdL) and 60 percent at the University of the Basque Country (UBC).The data were gathered among lecturers from 18 different faculties (ranging from Law through Engineering to Medicine), 51.4 percent of them being male and 44.5 percent female (4 percent did not specify their gender).
As for age, 68.4 percent of the subjects were in the 25-49 age range, whereas 31.6 percent were 50 or over.Those who had either Basque or Catalan (the coofficial language) as L1 were 31.2percent, while the majority had Spanish as L1 (52 percent).The remaining 16.8 percent had other L1s (Basque and Spanish, Catalan and Spanish, etc.), which will not be considered in this study due to the limited number of respondents in each of these groups.

The instrument
The data were gathered through a questionnaire that was based on three main factors: a) the initial goals of a European project on the internationalisation of bilingual universities in Spain and Wales; b) the review of the literature; c) the issues raised by focus groups carried out both in the UdL (Catalonia) and the UBC (Basque Country) in which university teaching staff were asked to share their thoughts about the internationalisation of their universities and the role to be played by academic mobility and multilingualism in the new global higher education space.
The questionnaire consisted of five sections.The first section was aimed at obtaining background information such as the participants' sociolinguistic context, gender, L1 and age.The second section dealt with issues related to the internationalisation of university and academic mobility.The third section focused on multilingualism at tertiary level, whereas the fourth section requested information about their multilingual practices.The final section was made up of open-ended questions.In this paper we will focus on the first three sections of the questionnaire, which included the 29 items under analysis.The questionnaire included a five-point Likert scale in which option 1 indicated strong disagreement with the item concerned, whereas option 5 showed strong agreement.Thus, the higher the mean, the more positive the respondents were about the statement put forward by the researchers.The questionnaire was designed to contain multiple items focusing on each of the content areas under scrutiny, namely internationalisation, academic mobility and multilingualism, since the internal consistency reliability of a given scale depends on both the number of items that make up the scale and the internal consistency of such items (Dörnyei 2007: 206).Internal consistency reliability was measured by means of the Chronbach Alpha coefficient.The reliability coefficients were satisfactory in the case of the three scales considered in this study, as can be seen in table 1.The internationalisation scale consisted of 10 items and the reliability coefficient was above 0.77, the mobility scale was made up of 9 items and the coefficient was even higher (0.811), and the multilingualism scale included 10 items and obtained an Alpha coefficient of 0.735.We can therefore conclude that the three scales are reliable and adequate to be analysed with our two research questions in mind.These results led us to calculate an average mean for each of the scales by adding the means obtained in each of the items and dividing the result by the number of items included in each scale.This allowed us to have an average mean for the internationalisation, academic mobility and multilingualism scales, which will also be examined in the results section together with the 29 individual items.

Table 1: Reliability analysis (Chronbach Alpha).
Internationalisation α = 0.771 (10 items) Item 1.There is a lot of English language used in the classes.Item 2. There is a lot of English language used when doing research.Item 3. The university promotes several foreign languages as well as the 2 co-official languages.Item 4. The university has many international students who complete all or part of their studies in our university.Item 5.There are many different nationalities studying at the university.Item 6.The university has a high international ranking amongst universities.Item 7. All university bodies are willing to use English.Item 8.All clerical and technical support staff can speak Basque/Catalan, Spanish and at least one foreign language.Item 9.The university promotes itself in a variety of ways to attract international students (e.g.website, student recruitment events, printed publicity, etc.).Item 10.Students are required to have knowledge of English at the end of their studies.
Mobility α = 0.811 (9 items) Item 11.Many home students study abroad on exchange programmes (e.g.Erasmus).Item 12.Many members of the teaching staff spend time working in universities abroad.Item 13.The university has a substantial presence of teaching staff from other countries.Item 14.The university offers a wide range of foreign language courses.Multilingualism α = 0.735 (10 items) Item 20.The university should provide more opportunities to learn foreign languages for both staff, students, and administration personnel.Item 21.Foreign language courses should be compulsory for all students in university.Item 22.Using a foreign language to teach a module in a non-language subject (e.g.Mathematics) is not necessary (reversed).Item 23.Students should be required to take a certain number of modules taught in English.Item 24.Knowing English well enables me to make the most of my university life.Item 25.The university should require students to be competent in English at the end of their studies.Item 26.The university should require students to be competent in two foreign languages at the end of their studies.Item 27.I am interested in being trained in English to deliver part of my lectures in English in the future.Item 28.I am interested in being trained in a foreign language other than English to deliver part of my lectures in this language in the future.Item 29.Foreign languages should play a more important role at university.

Results
In order to use parametric procedures, the data need to be normally distributed, which is why the Kolmogorov-Smirnov statistic was performed to check this issue.The results of the test showed that the data were not normally distributed and, therefore, non-parametric tests (Mann-Whitney U test) had to be carried out.
Our first research question was focused on the effect that different individual variables could have on the teaching staff's perceptions on the three aforementioned scales.When the variable sociolinguistic context was considered (table 2), some differences emerged.In the area of internationalisation, when the teaching staff were asked about their opinion on the presence of English at  university, those from the UdL were significantly more favourable than their UBC counterparts to three statements: students should have knowledge of English at the end of their studies (item 10), students need to take modules in English (item 23) and students need to be competent in English (item 25).However, the UBC teaching staff harboured a significantly more positive stance when asked about the possibility of students having modules delivered (item 17) and the three university bodies having language courses (item 20) taught in languages other than the two co-official ones.Despite these significant differences in 5 out of the 29 items included in the questionnaire, the variable sociolinguistic context (Catalonia versus BAC) did not show any significant difference in the mean for the three main categories under scrutiny, namely internationalisation, academic mobility and multilingualism.As far as gender was concerned (see table 3), there was hardly any difference in the internationalisation and academic mobility scales, since female  teachers were only significantly more positive towards item 6 (the university has a high international ranking amongst universities).Conversely, in the multilingualism scale the female participants showed more positive attitudes towards as many as three of the items and considered that foreign language courses should be compulsory for all students (item 21), that students should be competent not only in English (item 25), but should also have knowledge of two foreign languages (item 26) by the end of their studies.The difference for the average-mean multilingualism scale was also significant, which leads us to conclude that female teachers are more favourable to the idea that there is a need to boost a multilingual language policy at university than male teachers.The teaching staff's mother tongue did not have an effect on the mean obtained for any of the three scales (see table 4), as no differences were observed among those whose mother tongue was the minority language (Basque or Catalan) and those who had the majority language (Spanish) as their L1.However, the L1 = Spanish teachers were more positive towards four of the individual items: they were more favourable to the presence of English in the classes of an international university (item 1), and regarded as more important (i) that many home students study abroad (item 11), (ii) the option for students to follow modules in languages other than the co-officials (item 17), and (iii) the existence of many exchange programmes with foreign universi-  ties (item 18).The last three items were encompassed in the academic mobility scale.
In the case of the independent variable age, the participants were divided into two groups: those who were 49 or younger, and those who were 50 or over.This variable did not exert any significant influence on the respondents' views on the internationalisation, academic mobility and multilingualism scales (table 5).In fact, only two of the 29 items in the questionnaire were found to be significantly different, as the younger teachers (M = 4.28) were more favourable to the use of English in the research carried out at an international university than the older generation (M = 3.94), whereas the older teachers (M = 4.47) were more inclined to foster an open spirit and a universal mindset than their younger counterparts (M = 4.13).
The second research question was focused on the factors that were considered to be more significant when it comes to internationalisation, academic mobility and multilingualism at university.As for internationalisation (table 6), the results revealed that the presence of many international students was the most valued item (M = 4.21), followed by the use of English in research (M = 4.17), and the presence of different nationalities (M = 4.06).These three items obtained a mean higher than 4, which clearly indicates how important they were considered to be by the teaching staff.The item dealing with the attraction .79 The university has many international students who complete all or part of their studies in our university 2.
4.17 1.02There is a lot of English language used when doing research 3.

4.06
.80 There are many different nationalities studying at the university 4.
3.96 .83The university promotes itself in a variety of ways to attract international students (e.g.website, student recruitment events, printed publicity, etc.) 5.
3.96 1.01 Students are required to have knowledge of English at the end of their studies 6.
3.90 1.06There is a lot of English language used in the classes 7.
3.74 1.02All university bodies are willing to use English 8.
3.74 1.21The university promotes several foreign languages as well as the 2 co-official languages 9.
3.71 1.05All clerical and technical support staff can speak Basque/Catalan, Spanish and at least one foreign language 10.
3.66 The university has a high international ranking amongst universities 1.08 of international students came fourth (M = 3.96), together with that related to the requirement of knowledge of English by the end of students' studies (M = 3.96), and were followed by the use of English in class (M = 3.90).The items less widely supported had to do with the clerical and technical support staff being able to speak a foreign language (M = 3.71) and the university having a high international ranking (M = 3.66).Thus, it can be concluded that the teaching staff highly valued a university that aims to attract international students  4.17 1.00 The university website includes information in a foreign language apart from the two co-official languages 7.
4.04 .80The university has a substantial presence of teaching staff from other countries 8.
3.90 1.08 Students have the option to follow modules in languages other than the two co-official languages 9.
3.73 The university offers a wide range of foreign language courses 1.14 and in which the presence of English is remarkable both in teaching and research.Nevertheless, the knowledge of the current lingua franca should mainly be required for teachers and students, the administration personnel playing second fiddle in this respect.The presence of different foreign languages is also deemed (3.74) important, but not as much as English.
In the case of academic mobility (table 7), the most outstanding result has to do with the high means reflected in almost all the items included in this scale, as seven items out of nine showed a mean higher than 4. The items tackling the existence of many exchange programmes (M = 4.37) and of resources to make adequate provision for international students (M = 4.34), the participation of home students in exchange programmes (M = 4.26), the promotion of a universal mindset (M = 4.24), the work of the teaching staff in universities abroad (M = 4.22), information in a foreign language on the website (M = 4.17) and a substantial presence of foreign staff (M = 4.04) were highly regarded.Although still valued, the items related to the offer of modules (M = 3.90) and courses (M = 3.73) in different foreign languages happened to be the least supported ones.
The ten items included in the multilingualism scale were not as highly rated as those in the previous two scales (table 8).In fact, only one of the ten items revealed a mean higher than 4, specifically the one that stated that there 3.84 1.20 I am interested in being trained in English to deliver part of my lectures in English in the future 5.
3.78 1.09 Using a foreign language to teach a module in a non-language subject (e.g.Mathematics) is not necessary (reversed) 6.
3.62 1.11The university should require students to be competent in English at the end of their studies 7.
3.53 1.13 Foreign language courses should be compulsory for all students in university 8.
3.42 1,16 Students should be required to take a certain number of modules taught in English 9.
2.69 1.22 I am interested in being trained in a foreign language other than English to deliver part of my lectures in this language in the future 10.
2.64 The university should require students to be competent in two 1.07 foreign languages at the end of their studies should be more opportunities to learn foreign languages for the three university bodies (M = 4.25).English appeared as the language that will enable them to make the most of university life (M = 3.99), as underpinned also by their positive attitude towards being trained to deliver part of their lectures in English in the future (3.84).In teachers' opinion, foreign languages should have a more relevant role (M = 3.85) and should also be used to teach non-language subjects (M = 3.78).
This scale comprised the five items that had the lowest means amongst the 29 items under examination.Teachers seemed not to support so vigorously the demand on students to be competent in English by the end of their studies (M = 3.62), to take foreign language courses (3.53) and to complete modules taught in English (M = 3.42).Especially low are the means in the items concerning their interest in being trained in a language other than English (M = 2.69) and the requirement on students to be competent in two foreign languages (M = 2.64).
Once the three multi-item scales were considered, the academic mobility scale happened to be the one most widely endorsed by the teaching staff, followed by the internationalisation of the university.However, the multilingual language policy turned out to garner less support, especially when languages other than English and two foreign languages were considered in the equation.
According to the participants, the presence of foreign languages should be increased, but this does not mean that the requirement on students to be competent in two of them should be generalised.In addition, the majority of them were not interested in being trained to teach in a foreign language other than English.

Conclusions
The research reported in this paper is based on the assumption that the teaching staff plays an important part in the success or failure of the internationalisation policy of a university, and therefore it is important to explore how they define the notion of 'international university' and what their attitudes towards the two most visible aspects of internationalisation in higher education are: mobility and multilingualism.With this in mind, we have taken into account four main variables which research has demonstrated to be susceptible of having an impact on the teaching staff's attitudes: sociolinguistic context, gender, L1, and age.
In the case of the sociolinguistic context, our goal was to see whether there was a significant difference between the UBC and the UdL teaching staff.The results of the analysis show that the UdL staff tends to show a slightly higher degree of agreement with those statements in which English is presented as a 'required' third academic language (items 10, 23, 25), while their Basque counterparts see it more as an 'option' or 'opportunity ' (items 17, 20).This difference may be connected with the way in which bilingualism is administered in each of the institutions.Thus, at the UBC when a course is offered in English, there is usually an alternative group of the same course which is taught in Spanish, while at the UdL this is not the case and courses are almost always offered in only one of the three languages.Therefore, we seem to be confronted with two views of multilingualism, 'forced multilingualism' vs. 'optional multilingualism', which is strongly connected with the bilingual situation of the institution and, by extension, the social context.Therefore whereas at the different levels of the Basque educational system (from primary to university) students (or parents) can always opt out of Basque as the medium of instruction because there is always the Spanish alternative, in Catalonia the possibility of opting out of Catalan is not available, as Catalan has been defined as the language of education and it also represents the official language of the UdL.We could say that the staff shows the same tendency to consider either the minority language as an option (in the case of the UBC) or as an academic requirement (in the case of the UdL) in their attitudes towards the introduction of English as a third academic language.
As for the gender variable, the results obtained in this study are consistent with previous research (Ladegaard 1998;Bilaniuk 2003;Kissau 2006) in that women tend to show a more positive attitude towards multilingualism than men.Indeed, the difference for the average-mean multilingualism scale shows that the female teaching staff are more favourable to the idea that there is a need to boost a multilingual language policy at university than male teachers and therefore they show greater agreement not only with the introduction of English as a requirement, but also of a second foreign language.The significant difference for the average-mean of the multilingual scale seems to be symptomatic of a gender-based difference towards multilingualism also at university level.
The L1 variable, that is whether the respondent's mother tongue is the minority language (Basque or Catalan) or the majority language (Spanish), only had an impact on 4 (items 1, 11, 17 and 18) out of 29 items.This fact showed that L1 = Spanish teachers were more in favour of the introduction of English and mobility.From this we could interpret that minority language speakers are more likely to adopt Baker's bunker attitude (1992) to protect their social spaces in which they can protect the minority language from the 'invasion' of other languages, especially in those spaces which have been hard to 'conquer' from the point of view of the prestige of the minority language, such as the university.Nevertheless, we must be cautious about this interpretation since it is based only on the two items that make reference to the presence of English as a medium of instruction (1 and 17) and, furthermore, there are no significant differences between the means of each of the three clusters: internationalisation, mobility and multilingualism The analysis of the fourth variable, age, did not yield any significant differences in the views of the teaching staff except for two items.Although in these items both groups show a very positive attitude towards the presence of English in their research activity and on the promotion of internationalism as characteristics of an international university, we find that the younger teachers are slightly more favourable towards the importance of English in research, whereas the older teachers tend to agree more with the goal of promoting internationalism.We would need to investigate further whether these small, although significant, differences can be attributed to a possible generation gap in the academic staff, according to which younger members could be more supportive of a 'commodity' view of languages and education, whereas older members would be more in favour of a what Bolsman & Miller (2008)  a discourse of 'academic internationalism', for which the knowledge of languages and student mobility should contribute mainly to improving future international understanding.This question could probably be answered in further research through the use of qualitative instruments (such as interviews or group discussions) to gather the necessary data.
As for the most influential factors in the definition of internationalisation, academic mobility and multilingualism at university, we have found that for the first concept student mobility and the presence of English in teaching and research are the most significant ones.The teachers' lower degree of enthusiasm about the promotion of other 'foreign languages' could be interpreted as a sort of 'skewed multilingualism' of the respondents in the sense that it places foreign languages in an asymmetrical position, with English occupying a clearly outstanding position in front of the others.These results may be put down to the role that English plays nowadays as the main lingua franca in tertiary education (Coleman 2006;Doiz, Lasagabaster and Sierra 2013;Wilkinson 2004) and which leads this language to overshadow other foreign languages.This overwhelming hegemony of English would explain the participants' reluctance to make stark demands on a second foreign language, despite their general support in favour of a multilingual policy at university.It is also worth pointing out that English is highly regarded, but when students are required to be competent (in contrast to having knowledge) and to take modules taught in this language the support wanes.
Since academic mobility appears as the most important factor for the characterisation of an international university, it seems logical that most of the items included under mobility have very high means.Interestingly enough, the two items which show the lowest means (even though they are still high) and the highest standard deviations are those which ask respondents about the presence of other foreign languages apart from English.We again interpreted this as evidence for what we have defined above as 'skewed multilingualism'.
The items under the third scale, multilingualism, did not receive as high a rating as the first two, with the exception of one item in which the teaching staff demanded that the institution provide the means for members of the academic community to improve their foreign language learning.We suggest that the explanation for the relatively lower means of the rest of the items might lie in the fact that most of them represent a direct or indirect (when the item focuses on the students) challenge for the respondent in that they are somehow forced to be competent in foreign languages.This interpretation becomes even more evident when we look at the especially low means obtained in the items concerning their interest in being trained in a language other than English (M = 2.69) and the requirement on students to be competent in two foreign lan- guages (M = 2.64).Therefore, multilingualism may be experienced as a state of tension on the part of the teaching staff surveyed between a 'what should be' and a 'what it is' attitude.This, in our opinion, raises the issue about the reception that 'mother tongue plus 2 foreign languages' recommendation by the European Union may have in bilingual communities, in which their citizens are already coping with three languages in their ordinary lives and may find it difficult to accept the inclusion of a fourth one, even in highly educated sectors of society.Further research should focus on whether it is the hegemony of English as the current lingua franca at university or the recommendation to be competent in four languages that may hinder the implementation of multilingual language policies that include languages other than English in bilingual universities such as the UdL and the UBC.applied linguistics, focusing on applied discourse analysis, foreign language teaching and learning, multilingualism, and intercultural competence.He is the director of a research group on internationalisation and multilingualism in higher education, including researchers from three different universities and supervises several Ph.D. dissertation projects.He has published 7 books (as author or editor) and more than 60 articles and chapters.

Bionotes
Guzman Mancho-Barés has been a lecturer in the Department of English and Linguistics of the University of Lleida (Catalonia, Spain) since 2009/10, where he teaches English language and Business English.He has focused his research mostly on computer assisted language learning, English for Specific Purposes, and multilingualism at higher education.He has co-edited a multi-authored volume on the use of wikis in Higher Education in Spain; he has also published on Business English needs analysis; and he has researched on the lecturers' and students' attitudes towards teaching/learning through the medium of English at Catalan universities.
Stier 2002; Wächter 2003 on the Internationalization at Home concept; Chan & Dimmock 2008 on the translocalist model of internationalised HEIs).

The study 4.1 The sample
Brought to you by | CUNY Graduate School and University Center Authenticated | david.lasagabaster@ehu.esauthor's copy Download Date | 11/6/13 5:21 PM 4

Table 6 :
Most important factors in the internationalization of the university.

Table 7 :
Most important factors when it comes to academic mobility.

Table 8 :
Brought to you by | CUNY Graduate School and University Center Authenticated | david.lasagabaster@ehu.esauthor's copy Download Date | 11/6/13 5:21 PM Most important factors when it comes to multilingualism at university.
Brought to you by | CUNY Graduate School and University Center Authenticated | david.lasagabaster@ehu.esauthor's copy Download Date | 11/6/13 5:21 PM David Lasagabaster is associate professor at the University of the Basque Country (UPV/EHU), Spain.He has published on CLIL (Content and Language Integrated Learning), English-medium instruction at university, attitudes and motivation, and multilingualism.He has published widely in international journals, books and edited books.Among others, he has co-edited Multilingualism in European bilingual contexts: Language use and attitudes(Multilingual Matters,  2007); CLIL in Spain: Implementation, results and teacher training (Cambridge Scholars Publishing, 2010); and English-medium instruction at university: Global challenges (Multilingual Matters, 2013.Since 2008 he has been a member of the executive committee of the International Association for Language Awareness.Josep M. Cots is a professor in the Department of English and Linguistics of the University of Lleida (Catalonia, Spain), where he teaches English language and applied linguistics.He has carried out most of his research in the field of Brought to you by | CUNY Graduate School and University Center Authenticated | david.lasagabaster@ehu.esauthor's copy Download Date | 11/6/13 5:21 PM