The internationalisation of higher education in two different contexts: Catalan and Estonian sociolinguistic perspectives*

ABSTRACT The topic of the internationalisation of academia has recently attracted attention from sociolinguists and language-policy scholars. In this paper, we compare two different universities from two contrasting contexts in Europe in order to find out more about their projected stance [Jaffe, A. (2009). Stance in a Corsican School: Institutional and Ideological Orders and the production of Bilingual Subjects. In A. Jaffe (Ed.), Stance: Sociolinguistic perspectives (pp. 119–145). New York, NY: Oxford University Press] and attitudes towards the different languages present in their immediate contexts. In particular, we compare the University of Tartu (Estonia) with the University of Lleida (Catalonia, Spain), analysing several key parameters. The purpose of the comparison is to contrast, from a sociolinguistic point of view, the higher education setting of two medium-sized language contexts in Europe [Vila, F. X., & Bretxa, V. (Eds.). (2015). Language policy in higher education. The case of medium-sized languages. Bristol: Multilingual Matters] with different demolinguistic and language political features. The results show that both institutions adopt a similar stance in connection to their respective national language (a protectionist attitude), but they take different approaches towards the other societal language and English. We read these differences in light of the broader historical and socio-political backgrounds, which we suggest are reflected in the microcosm of the universities here analysed.

language (a less visible and thus more negative stance towards Russian in Estonia than it is the case for Spanish in Catalonia) (see also articles by Gallego-Balsà & Cots, and Moore in this special issue).
For the purposes of the discussion, we draw on information stemming from different sources, from university policy documents to institutional webpages (we elaborate more on the methodology followed and material used below, see section on Theory and data). These parameters alone, of course, do not represent the entire language ecology of each particular university, but we believe they may be fruitfully exploited here to discuss and analyse how the higher education institutions in these two contexts adapt to the needs of this new era. One could have included information about the number of incoming and outgoing mobility students in the two universities, or the number of online programmes offered by them, but the chosen indicators, we will argue, are rich enough in order to provide a picture of the situation in the two settings that are safely comparable. Methodologically, the indicators were also sufficiently easy to access and thus analyse in a straightforward manner.

Sociolinguistic commonalities and differences between Catalonia and Estonia
From the point of view of their history, demography and sociolinguistic configuration, Catalonia and Estonia share some important points. Nevertheless, the two cases are not a frequent pair in comparative sociolinguistics, although some authors have in the past analysed in some detail their differences and similarities (e.g. Branchadell, 2011;Laitin, 1992;Shafir, 1995;Skerrett, 2010;Soler-Carbonell, 2013).
The rationale behind the comparison rests mainly on the historical, socio-political, and sociolinguistic factors. During an extended period of time in the twentieth century, the two nations found themselves dominated by authoritarian regimes (Francoism, 1939-1975, and the Soviet era, 1940-1991. In the course of those decades, the social, demographic, and sociolinguistic landscapes of the two contexts were importantly modified by the arrival of migrant workers who settled in the growing industrial areas in the urban centres of the two places (Branchadell, 2011). Moreover, those who arrived during the period were speakers of the dominant language, Spanish and Russian, respectively, and they would in most cases be able to lead their lives without an important need to become proficient in the local language. Nevertheless, and this is an important difference, the influx of newcomers in Estonia stopped and their numbers even receded in the late 1980s and 1990s. In Catalonia, in contrast, the number of residents of migrant origin has continued to grow, almost exponentially, over the past decade in particular. It is therefore possible to state that Catalonia features a socially and linguistically more heterogeneous society than Estonia, despite the fact that the latest Population and Housing Census (Statistics Estonia, 2011) has documented a total number of 157 different languages spoken presently in Estonia. In 1998, only 2.4% of the population in Catalonia were of foreign origin (i.e. from outside Spain); by 2013, that percentage had increased to 14.6% (Idescat, n.d.).
In addition, there is yet another important (and obvious) difference between the two cases, which is at the same time of particular interest for the purposes of this paper. At the end of the authoritarian regime, Estonia became an independent country in 1991, whereas Catalonia was reframed into the Spanish Autonomous Communities in the transition from Francoism towards democracy (1975). From a Catalan perspective, it is interesting to see the Estonian approach towards language political matters, particularly given the recent growth in impetus of the debate on Catalonia's political independence. In the last decade, there has been a growing discussion about the political relationship between Catalonia and Spain; the origins of this debate could be traced much further back than that, but the discussion on the Statute of Autonomy reform initiated in 2005 was a particularly important moment. Moreover, ever since 2010 the debate has speeded up and increasingly attracted the attention from many observers, particularly international ones. It can, therefore, potentially be illuminating to contrast these two cases against this background.

Theory and data
The paper's theoretical framework is informed by the concept of 'stance' (Du Bois, 2007;Jaffe, 2009). 'Stance' is a form of embodying a particular style or character and by choosing some particular forms of expression, speakers align themselves with a given personality. According to Jaffe (2009), in bilingual territories the significance of languages is determined by the particularities of the sociolinguistic context and that, for this reason, language choice is already a form of stance-taking. By selecting one language over the other, speakers position themselves in a particular way. The notion of 'stance' is thought to be applied in interactional settings, that is, at the speaker level. However, it can also be usefully applied at the institutional level and in this way analyse, for example, how higher education institutions position themselves in relation to the 'language question' in their internationalisation process.
The analysis is divided into two parts. First, we conduct a content analysis (Hsieh & Shannon, 2005;White & Marsh, 2006) of a set of institutional language policy documents in an attempt to understand what language(s) the universities claim to promote. In the analysis of the University of Tartu (UT), the following documents are considered: (1) the Language Principles of the University of Tartu  The language policy of UdL is made explicit in the document Language Policy of UdL: Towards a Multilingual Reality (UdL, 2008), which has been recently revised by the same institution as part of its new Strategic Plan 2013-2016 (UdL, 2013b) that the university has drafted as a means to becoming more international. As a consequence, the language policy published in 2008 has been replaced by the Operational Plan for Multilingualism (OPM) of the UdL 2013-2018 (UdL, 2013a), based on the 2008 document.
All the documents were obtained from the universities' webpages. In the case of the Estonian documents, they are available also in English, and this is the version used for the analysis. For Lleida, in contrast, the Catalan-language version of the documents was used; all the extracts provided below from UdL are therefore our translation.
The content analysis of these data has been conducted using a corpus analytic tool, AntConc (Anthony, 2014), which has allowed us to extract the most recurrent themes in the texts and their relevance in context. Following Björkman (2014), the number of occurrences per keyword in each of the data sets was counted, and then we manually analysed these occurrences in order to retrieve the most salient themes for each setting. The keywords used for UT were 'Estonian', 'English', 'Russian', and 'foreign languages', and for UdL 'Catalan', 'English', 'Spanish', and 'other languages' (in Catalan: 'català', 'anglès', 'castellà/ espanyol', and 'terceresllengües').
These types of documents represent different genres: a language policy document is not written for the same purposes as an institutional strategic development plan. However, given that the strategy documents from both universities contain important language-related sections and content, it was appropriate for us to include them in the analysis.
In parallel with the analysis of the policy documents, a series of other parameters that show the language use in these two contexts were considered. These parameters provide a more indirect, yet illustrative, access to the specific (language) management practices and preferred linguistic choices by the stakeholders on the ground. The parameters considered are: (1) the distribution of the languages of instruction both at the undergraduate and graduate levels; (2) the language(s) in which PhD dissertations are written; and (3) the languages in which the universities' web pages are available. Data for these three parameters were obtained directly from online sources, from the respective institutional webpages. In all of them we have striven to use the most up-to-date information that was available, either from the universities' websites or from direct contact with the administration.

Language policy documents and strategic plans
Language policy documents represent normative official papers that, within the university context, may affect ordinary academic activities and define a framework for the use of languages. This has the potential to affect how teachers and administrative staff make use of the languages available in the sociolinguistic context and sets the scenario for students. Table 1 shows an overview of the directives at the UT and the UdL mentioned before. The table shows the number of words in each document and the hits in connection with Estonian or Catalan (the local minority languages), English, Russian or Spanish (the local majority languages), and 'foreign language(s)'.
After manual analysis and revision of the data, the following emerging themes were detected. First, we present the themes extracted from the UT documents and, second, those extracted from the UdL documents. Table 2 summarizes the different themes emerging from UT policy documents.
As regards to Russian, it is worth highlighting that there is no explicit reference to it across the three UT documents analysed. In these documents, the most prominent theme that can be identified is Theme 1. The documents are very clearly geared towards framing the protection, promotion, and development of the Estonian language in the domain of higher education. This is significantly the case in the Strategic Plans (A2015 and A2020), where almost no other theme in relation to language emerges from them. The 'protectionist' stance normally appears when explaining the need to maintain Estonian as a language of instruction in all fields. The 'promotion' stance is usually associated with giving international students and staff the opportunity to learn Estonian, as well as valuing the publication of research results in the language. Finally, the 'development' theme normally appears in connection to the advancement of Estonian as an academic language, with the appropriate terminology, textbooks, and scientific materials. Next are some sample extracts from the documents that illustrate each of these subthemes in connection to Theme 1: . The University of Tartu as a national university of Estonia bears the responsibility for solving problems faced by the society by ensuring the continuity of Estonian intellectuals and language and culture and by contributing to the development of education, research, and technology and other creative activities throughout the world (A2020). . Guarantee the provision of higher education at the Bachelor's and Master's levels in all the fields of study on the basis of Estonian-language curricula, to avoid any field being taught only in foreign languages (UT Language Principles).
In relation to English, as Table 1 indicates, there are not many instances in the documents where the language is explicitly mentioned. It is particularly absent in the strategic plans (note that A2020 does not mention it at all). However, when it does get mentioned, the themes that emerge are the following: first of all, it is referred to as a useful and In connection to Estonian Theme 1 Estonian needs to be protected, promoted and developed.
In connection to English Theme 2 English as a tool: it is important for pragmatic purposes, economic reasons, competitiveness, and internationalisation objectives. Theme 3 English has to coexist with other languages. Theme 4 The 'foreign language(s)' label as an ambiguous way to refer to English. necessary tool for different purposes (including economic competitiveness and internationalisation efforts) (Theme 2); secondly, it is also mentioned in connection to other languages, explicitly pointing out the idea that English is not enough and that it should coexist with other languages (Theme 3); and finally, it can also be implicitly referred to, but without naming it explicitly, using instead labels such as 'foreign language(s)' (Theme 4). The latter is what we refer to as 'the invisibilization English', which we will return to in the last section of the paper. Next we provide some extracts from the data by way of exemplification of these three themes.
Theme 2: . Ensure that members and partners of the university have access to information in English and communication in English in a professional capacity (A2015). . Give students the opportunity to learn English, including specialised academic English, and to study in English, to ensure that graduates have competitive English language competency for the international labour market (UT Language Principles).
Theme 3: . Offer language support in publishing research results in English and other languages (UT Language Principles). . Create the conditions to enable students to study at least one foreign language in addition to English (UT Language Principles).
Theme 4: . Develop subjects and modules in foreign languages within the curricula taught in Estonian, thus increasing the opportunities for Estonian and international students to study together (UT Language Principles). . Develop Estonian scientific language in doctoral theses written in Estonian, in the Estonian summaries of doctoral theses published in other languages, in textbooks for higher education and in the scientific literature (UT Language Principles).
Turning to the policy documents at UdL, Table 3 summarizes the different themes emerging from them, which were obtained after a manual analysis and revision of the data. Table 3. Emerging themes from UdL policy documents.
In connection to Catalan Theme 1 Catalan has to be protected and promoted. It is the own and the official language of the university and of Catalonia. In connection to Spanish Theme 2 Spanish as an international and co-official language. In connection to other languages Theme 3 Multilingualism as a tool, a goal, and a means to enhance students' mobility and economic competitiveness. In connection to English Theme 4 English has to be promoted because of economic and academic reasons. Theme 5 English has to coexist with other languages.
There are two themes that emerge from the UdL documents: first of all, the promotion and protection of Catalan (Theme 1) and secondly, the development of multilingualism, which is to an extent explicitly connected to English (Themes 3, 4, and 5). In relation to Catalan, the documents activate a discourse of 'authenticity' (Woolard, 2008) by framing it as the own language of Catalonia (llenguapròpia). They take a 'protectionist' stance by emphasising the need to safeguard the predominance of Catalan as a language of instruction and to promote it among the foreign students.
Theme 1: . The Catalan language needs to have the presence that corresponds to it as the own language of the country and of UdL (Statutes of UdL, art. 4). Promoting the use of Catalan in all the areas of the university life needs to be a priority (UdL Language Policy). . In the specific area of undergraduate and master studies, and on the basis of the current legal framework, in formal education it is necessary to continue ensuring the predominance of Catalan (UdL OPM). . Elaborate and apply a new plan of language policy of the UdL that does not affect negatively its internationalisation and simultaneously promotes a certain degree of functional plurilingualism, which will strengthen the use and the protection of the Catalan language (UdL Strategic Plan).
As for Spanish, the comparatively low number of hits across the documents hints at a lower degree of concern from the institution. Even though Spanish is recognised as one of the official languages of the territory and also as an international language, the measures mentioned to promote it are scarce. It could be interpreted that the institution's stance of indifference towards Spanish derives from the privileged position that this language has internationally and also as a majority language in Spain. It can also be read as a politically motivated move to make Spanish be less present in the policy documents, thus blurring the language from this context (Irvine & Gal, 2000).

Theme 2:
. In relation to the Spanish language, given its status as a co-official language and its widespread international interest, the UdL needs to make sure that students will have the necessary tools to develop a professionally and academically correct use of the language (UdL Language Policy).
The second most relevant aspect of the UdL language policy is its focus on the promotion of multilingualism. Multilingualism is presented by the policy documents as a means to increase the economic competitiveness of the university and the students' mobility and employability (Theme 3). It is true that the analysed documents from UdL also make use of an ambiguous category, that of 'third languages' (terceresllengües), but the number of explicit mentions of the terms 'multilingualism' and 'plurilingualism' is as equally high as the number of hits for 'third languages'. We will return to this in more detail in the discussion section of the article.
English is the most mentioned foreign language in the policy documents, framed as an important language because of economic and academic reasons (Theme 4). However, oftentimes it is mentioned in connection to a more heterogeneous form of multilingualism (Theme 5). In addition, it is interesting to note that in the Language Policy document of 2008 the number of hits in connection with English was much higher than the hits in connection with foreign languages (23 and 5 hits, respectively); however, the more recent OPM (2013) has reversed this situation. The 2013 revisited language policy contains 60 hits in connection with 'foreign languages' (including explicitly named languages, and the labels 'multilingualism', 'plurilingualism', and 'third languages') and only 18 hits refer explicitly to English. This indicates an evolution in the stance the institution adopts towards multilingualism from a more English-centred one, towards one that emphasises the coexistence between English and other languages.
Theme 3: . Promote the development of third languages, the contact with the professional context, and the mobility of international students (UdL OPM).
Theme 4: . English will be introduced as a working academic language, considering its importance academically and professionally at an international level (UdL Language Policy). . Establish a system of incentives and awards for the administrative staff with working skills in English (UdL OPM).
Theme 5: . English needs to be progressively included as a language of instruction at the undergraduate level; at a postgraduate level, plurilingualism is to be promoted (at least in Catalan, Spanish, and English) (UdL OPM). . All this does not imply the exclusion of other languages from teaching and development activities. In particular, languages with strong scientific traditions will be promoted, such as French, German, Italian, or Portuguese. Circumstances allowing, and according to the demand, languages such as Chinese or Arabic will be promoted as well, since these might also be considered of particular interest in today's Catalan society (UdL Language Policy).

The languages of instruction
Another important parameter that we want to consider in the paper is the languages of instruction present at each institution. With reference to UT, the university currently offers 62 bachelor's degree programmes, 54 master's degrees, and 36 doctoral programmes (retrieved from www.studyinestonia.ee). At the undergraduate level, there are two English-taught programmes: Business Administration (3 years) and Medicine (6 years: integrated BA and MA studies); the remaining 60 are all taught in Estonian as the medium of instruction. At the Master's level, the increase in English-taught programmes is substantial, but Estonian still dominates. Finally, at the PhD level, all the 36 programmes that the university offers are officially taught in English. Interestingly, there is no programme taught in Russian as the medium of instruction. The data presented here is in agreement with the general situation in Europe, where the majority of English-taught programmes (ETPs) can be found at the more advanced level of education (Master's and PhD). At the Bachelor's level, ETPs at UT represent only a minor fraction of all programmes, while at the Master's level there is already a substantial increase, with 100% of doctoral studies officially taught in English (see Table 4). An important caveat is in order here: this does not mean that the entire degree is taught in a given language exclusively. Even PhD studies, although advertised as fully ETPs, may contain several courses (sometimes obligatory ones) taught in Estonian. Unfortunately, UT does not have any official data on the languages of instruction subject by subject. Regardless of that, this gives us enough grounds to present an overall picture of the current situation in terms of medium of instruction at UT.
At the UdL, information regarding the languages of instruction is stated in the course programme, which is a requirement of the 'language safety principle'. This principle states that every teacher has to pre-define the language of instruction of the course that they are going to teach and the information needs to appear explicitly formulated in the course programme; this is intended to avoid potential tensions in classrooms with mixed audiences, with both national and international students. Table 5 shows the percentages of the courses offered in Spanish, Catalan, and in third languages in Bachelor's and Master's degrees. In the case of the UdL, the available data does not allow us to report the language of specific degree programmes, but rather on individual courses only. In addition, we cannot report the total number of courses in each category since the source of this information (UdL, 2015) offers only the percentage and not the total number of courses. Finally, the UdL does not offer PhD courses as part of its PhD programmes and, therefore, the only data related to the languages of the PhDs at UdL is the language in which PhDs are written (see Table 6).
Similar to the UT case, the local language, in this case Catalan, is the most widely used language of instruction at Bachelor and Master's levels followed by Spanish. At the Master's level, Spanish and foreign languages (mostly English) gain more prominence, but Catalan still is most commonly used as a language of instruction. However, the higher presence of Spanish at more advanced levels of instruction suggests that there is a greater need for it to be used as a lingua franca in that context. As we shall develop further in the discussion section, this may indicate that in the context of UdL, Spanish rather than English fulfils the role of the institution's international language. The language of PhD theses When comparing the languages in which the theses have been written in the period 2000-2012, significant differences between the UT and the UdL can be observed. One of them is that the total number of theses submitted at the UdL is half the number of theses at UT (1116 and 558, respectively). Table 6 shows the proportion of theses that have been submitted in each language during the period 2000-2012 at the UT and UdL. Although Table 6 does not show the progression of the number of theses in each language over the years, at UT, English has remained steadily at the first position, especially in the Faculties of Science and Medicine. At UdL, the tendency has also been for Spanish to be the most frequently used language over the past two decades, with English increasingly growing its presence in the more recent period. For instance, in 2012, 39% of the theses at UdL were written in English, very close to Spanish. The present situation is that there is balance between the number of theses written in English and Spanish, but Catalan is losing presence in relative terms as an academic language at that level.

The universities' webpages
The universities' webpages can also be taken as an illustrative site to look for implicit views on each institution's language repertoire and their tacit attitudes toward them. In both cases, the webpage is available in three languages: Estonian, English, and Russian at UT, and Catalan, Spanish, and English at UdL, in that order. The fact that the three languages are placed in that order (at least in the default site, in Estonian and Catalan respectively) is already indicative of the type of language hierarchy that each institution projects publicly.
More interestingly, if we look at how much information there is available in each language, we obtain yet another insight into the university's projected multilingualism. Clicking at the button 'site map' we can compare how many items there are for every language. In the case of UT, it appears quite clearly that a stronger emphasis is placed on Estonian, the language in which one can obtain the most information (with 504 items); next we have English, with 320 items, and finally Russian, with just 41 items (www.ut.ee; last accessed 14 January 2015). In contrast, UdL seems to offer a very similar amount of information in all three languages, Catalan, Spanish, and English. In this case, however, this is more difficult to calculate. Clicking at 'site map' in each of the three languages, one is directed to the same site in each respective language with 18 items. One then has to check manually if all the information does feature in all the three languages. In 10 out of the 18 items, the information is available in the three languages, in four items it is only available in Catalan, two items are only accessible in Catalan and Spanish, and one item is only available in Catalan and English (www.udl.cat;last accessed 14 January 2015). This is indicative of a wish to project a more balanced degree of multilingualism, in comparison to the approach taken by UT, which wishes to project itself online more decidedly as an Estonian university. Indeed, UdL's website offers a 'translate' button using Google translate technology, which enables the user to have an approximate translation of the website into 52 different languages.

Discussion and conclusions: different routes towards internationalization objectives
The aim of this paper was to analyse and contrast several features of the language ecology of two universities from two different countries in Europe in order to detect potential tensions and ambiguities in relation to the different languages at stake. More specifically, we wanted to show the institutional language attitudes and stances taken in relation to: (1) the territories' national languages (respectively, Catalan and Estonian); (2) the territories' other societal languages (respectively, Spanish and Russian); and (3) English. Taking the results from the analysis of the policy documents, we can see that both universities explicitly note the need to protect, promote, and develop the national language of their context (Catalan and Estonian, respectively). In Tartu, the university sees it as an institutional duty and obligation to help preserve the Estonian language and culture in the field of higher education and academic research, and to develop the appropriate terminology for the language in all fields of science. Similarly, UdL also formulates the need to promote Catalan, a language that is to be encouraged among members of the international community at the university (something that UT also states in relation to Estonian). With regards to the other societal language, however, the policy documents show an important difference that we can also witness in several of the other indicators we have considered in the analysis. In short, Spanish is more present in the context of the UdL, it is explicitly mentioned in these documents as a co-official language of the region, and a language of international relevance. Spanish is also used as a language of instruction, particularly at the graduate level, PhD theses are commonly written in it, and the university's website is available in Spanish to a significant degree as well. In contrast, Russian is almost absent from UT's language ecology: it is rarely used as a language of instruction, very few theses are written in it and the university's website is only minimally available in this language. With reference to the universities' websites, the analysis also indicates the particular kind of linguistic hierarchies in place in the two settings: in the case of UdL, its webpage is available in Catalan, Spanish, and English, in that order, whereas in the case of UT, it is available in Estonian, English, and Russian.
These different hierarchies point out that UdL and UT have a different approach towards English as well. In that sense, it is certainly true that the policy documents of both universities express the following: first of all, the need to promote the acquisition of English by all its members (students, academics, and administrative staff) considering its relevance in terms of international competitiveness. Secondly, the analysed documents also indicate the idea that English is not enough and that the teaching of other languages should be promoted too. However, Tartu seems to be a bit vaguer with reference to English in that sense, and in several instances the language is hidden behind the 'foreign language(s)' label, without making it explicit which these foreign languages are. In the policy documents of UdL, in contrast, the situation is somewhat different, with English being more explicitly mentioned in them, and more often than not, appearing next to the idea that multilingualism has to be promoted amongst university members. In addition, named languages are more likely to be explicitly mentioned, and so the invisibilization of English, although it occurs to a given extent, it is not as noticeable as in the case of UT.
It is our interpretation that English functions more as an international language in the case of UT than in the case of UdL. Indeed, it has a more substantial presence among PhD dissertations, in UT's website, and it is increasingly introduced as a language of instruction, especially at the graduate level. This, we believe, reinforces the imagined potential threat that English poses to Estonian, not only among university stakeholders, but also within the society at large (Liiv & Laasi, 2006), and this is probably why policy-makers might prefer to use labels such as 'foreign language(s)' in order to avoid potential heated public debates. Indeed, Estonian occupies a very central position in the matrix of what being an Estonian means, that is, it is legitimized by its 'authenticity' value (Soler-Carbonell, 2013); discussions about language loss and linguistic homogenization frequently feature in the public fora, and English is more often than not seen as a potential threat, in that sense.
In Lleida, in contrast, the situation is different vis-à-vis English because in fact, the language that primarily functions as the institution's international language is Spanish (Llurda, 2013). Spanish is significantly present in all the indicators we have considered in the analysis: as a language of instruction, in PhD theses, and in the university's website. It is indeed an important tool in terms of UdL's efforts towards internationalisation objectives and marketing strategies, and many international students who travel to Lleida do so with the expectation of being able to practice their Spanish (Gallego-Balsà, 2014; also Gallego-Balsà & Cots in this special issue).
In addition, in parallel to the position of Spanish, the more open discussion about English and multilingual issues in the policy documents at UdL may reflect what has been detected as a trend towards a relaxation of the 'authentic' value of the Catalan language (Soler-Carbonell, 2013). Some authors go as far as to suggest a possible 'de-ethnicization' of language choice in Catalonia (Pujolar & González, 2013), among other reasons provoked by the fact that in recent years, the language has incorporated a significant number of 'new speakers' (O'Rourke, Pujolar, & Ramallo, 2015), who may value Catalan in non-essentialist terms.
All in all, we have analysed two different universities from two very different contexts adopting a similar stance towards their respective national language, but different attitudes towards the other societal language, respectively, and English. We believe it is important to read these differences in light of each university's language ecology and the recent historical developments. In Estonia, back in the 1990s, society experienced deep changes and transformations, which had an impact in the sociolinguistic makeup of the country. In particular, the shift from Russian to English for international purposes started to take place back then, inscribed in what has been called the country's 'return to the Western world' (Kasekamp, 2010, p. 172). Nowadays Estonians still know Russian, particularly the older generations, but English has replaced it as the country's first foreign language. This is clear in UT's language ecology, as we have seen from the data presented above. In Catalonia, in contrast, Catalan and Spanish coexist in a more intense manner, and both are important societal languages. However, in the context of Lleida in particular, Catalan enjoys a firmer position in the most immediate surroundings, although Spanish is also very much present at all levels, particularly in the media and in state institutions. Knowledge of English, on the other hand, is not widely spread among the local population and this is why UdL adopts a clearer stance towards incorporating it as a necessary tool for its internationalisation strategy and for the benefit of its students and staff.
The present study shows that institutions from different contexts may adopt different stances towards their linguistic repertoire and in particular towards incorporating English. This reflects the different historical and socio-political evolutions of each territory, which generates potentially different roles for all the languages at play within each specific ecology. In other words, and as Vila (2015, p. 5) has pointed out, 'medium-sized languages are extremely sensitive to the legal and political factors of their nation states', which explains why Catalan and Estonian are reserved a prominent position within the university context of UdL and UT respectively (especially as languages of instruction). It also explains why Spanish has more weight at UdL as opposed to the minimal presence Russian has at UT, and the same can be said about English and its larger impact in Tartu in contrast to Lleida. This, we believe, is evidence that in their way towards becoming more international, universities all over the world, and particularly in Europe, make strategic choices based on political, economic, and historical factors.