Pluricentricity in the classroom: the Serbo-Croatian language issue for foreign language teaching at higher education institutions worldwide

: The study presented in this article looks at the effects of the changes in national language policies following the break-up of the Socialist Federal Republic of Yugoslavia on teaching the Serbo-Croatian language or a “language which is simultaneously one and more than one” as a foreign language. The study explores how language ideologies and conflicting attitudes towards national standard languages, recorded both within nation-states and across nation-state borders, are understood by teachers in the context of teaching Serbo-Croatian as a foreign language. The article also examines the extent to which these understandings reflect current discussions of pluricentric languages and methods adopted for teaching pluricentric languages as foreign languages. the teaching of Serbo-Croatian.


Introduction
The process of dramatic socio-political changes and redrawing of state boundaries during the collapse of the Socialist Federal Republic of Yugoslavia in the early 1990s initiated comprehensive changes in language policy in the Serbo-Croatian speaking area. Necessitated by the identification of language with ethnicity/nation, at the time language was perceived more than ever as a symbol of national identity and a powerful tool in the nation building process. Hence, over the years, language and education policy makers in Bosnia and Herzegovina, Croatia, Serbia and Montenegro1 put exceptional effort into presenting shared cultural and linguistic practices as fragmented and imbued with insurmountable differences, changing speakers' attitudes to language from (common) language as a means of communication to (separate, national) language as identity marker.
In this article I present ways in which institutions around the world involved in teaching Serbo-Croatian, or Bosnian, Croatian, Montenegrin, Serbian (henceforth BCMS2), and teachers who teach the language as a foreign language at these institutions, understand what constitutes the language used by the speakers in the central South Slavonic space. Attempts to map these understandings have so far been fragmentary and inconclusive, focusing predominantly on the elaboration of ideological stances, without attempting to document institutional practices or attitudes of those who teach (about) the language. The main aim of my empirical study is to document a set of attitudes to the Serbo-Croatian sociolinguistic issue and the ensuing teaching practices in order to map the effects of language policies in the region in the post Yugoslav period, relating them to the still prevailing uncertainties about language naming and to the stumbling blocks in accepting the current (fragmented) status of Yugoslavia's lingua communis (Radovanović 1986). It is therefore useful to begin by situating the Serbo-Croatian language issue in the context of historically shifting ideologies and policies which have shaped the debates about the name and the status of the common language.
dialect was the most widespread in the central South Slavonic area. The process of standardising the common language, inspired by the works and ideas of the Serbian philologist Vuk Karadžić and Croatian intellectuals, members of the Illyrian movement, started with the work on the Rječnik hrvatskoga ili srpskoga jezika ['Dictionary of Croatian or Serbian'] initiated by the Yugoslav Academy of Sciences and Arts in Zagreb in the 1880s. By the end of the 19th century, Gramatika i stilistika hrvatskoga ili srpskoga jezika ['Grammar and stylistics of Croatian or Serbian'] by Tomo Maretić was published. With regards to the glottonym "Serbo-Croatian or Croato-Serbian", it was first used as a language name by the Slovene philologist Jernej Kopitar in 1836 and it gained widespread currency after the publication of Pero Budmani's Grammatica della lingua serbo-croata (illirica) ['Serbo-Croatian (Illyrian) grammar'] published in Vienna in 1867.5 Ever since the widespread emergence of national movements in the 19th century, when the idea of a common literary language for the peoples of the central South Slavonic area was first conceived, debates about the nature of such a language and about the need for one have been frequent and fierce. Debates and surveys on the relationship of the varieties of Serbo-Croatian and the role their distinctive features should play in the standardisation of the common literary language were initiated by the leading cultural journals in the region several times in the 20th century. Diverse and often opposed opinions on language corpus planning -mainly concerning orthographic issues pertaining to the choice of a phonemic or morphemic approach in defining the written norm, but also interventions in flectional morphology, opposed attitudes to borrowing from foreign languages and from other Serbo-Croatian dialects, and to the common use of the Latin and/or the Cyrillic alphabet (to name but a few of the contested aspects) -came also to be interpreted along national and ethnic lines in different intellectual and political climates.
After the Second World War, efforts to harmonise relations in the nationally heterogeneous Socialist Yugoslavia also involved the language question. The Novi Sad Agreement of 1954 was meant to bring about a compromise on the contested issue of a unified language norm. However, already by the late 1960s, based on criticisms of the solutions suggested in the new orthographic manual and dictionary of the common language, and following failed attempts to harmonise professional and scientific terminology, the notion that each national variety was a distinct language, and that every 5 In the approximately 100-year-long history of the common language, the official language names ranged from Serbo-Croatian, Croato-Serbian to Croatian and/or Serbian. Serbo-Croatian is a term which is "primarily a technical term from philology and linguistics, rather than a colloquial name routinely used by speakers" (Bugarski 2019: 106), a stance earlier elaborated by Lenček (1976). My use of the language name in this article is guided by this understanding, as well by the use of the glottonym Serbo-Croatian in the most recent and the most relevant studies on this topic (cf. Kordić 2010;Gröschel 2009). people had the right to choose the name of their language, started to gain ground, with a view to changing the status of the varieties into languages (Brozović 1970).
Prior to Socialist Yugoslavia's collapse, Serbo-Croatian was conceptualised as a language with two main national varieties: Eastern and Western. The variety used in Bosnia and Herzegovina was seen as a mixture where features of the Serbian ("Eastern") and Croatian ("Western") varieties were "combined", "neutralised" or "coexisted". Montenegrin was seen as a sub-variety of the Serbian variety.6 The unequal quantitative and functional distribution of the two main national standard varieties, and consequently greater prestige of the Eastern variety relative to other varieties, was increasingly perceived as a threat which allegedly could have led to linguistic and ethnic assimilation (Babić 1995: 31). As a result, some linguists started to question the unity of the common language more openly, as expressed in Deklaracija o nazivu i položaju hrvatskog književnog jezika ['Declaration on the name and status of the Croatian literary language'] from 1967, and insisted with increasing vehemence that the standard national varieties could function as separate standard languages.
The language issue gradually became a vehicle for not only affirming national identity but also raising awareness of ethnic boundaries. Arguments against (and also in favour of) the common language intensified during Yugoslavia's dissolution in the 1990s, diminishing even further the already weak "internal identity" of the common language.
It was during this period that national varieties of Serbo-Croatian were elevated in status to separate languages through re-standardisation. The important thing to bear in mind is that in refocusing on the new language norms, the features that prescriptivists selected might or might not have been exclusive to the re-standardised variety in question. In the same way, speakers of that variety might or might not have frequently used those features. With regard to what to call the language, the name Serbo-Croatian was abandoned on the official level. All these changes were additionally reflected in the internal realignment of national philologies and cultural institutions, which were subsequently separated through additional efforts to erase common cultural and linguistic bonds.
6 Commenting on the lack of recognition of the Bosnian and Montenegrin varieties, and the fact that the discussion revolves around the Croatian and Serbian varieties most of the time, German linguist Daniel Bunčić points out: "[i]t can be assumed that this bipolar structure was inherited from the period before 1945, and the lack of independence of the varieties of Bosnia and Herzegovina and Montenegro can be ascribed to the centralism of the First Yugoslavia" (Bunčić 2008: 96).

Background and theoretical considerations
The interest in exploring different approaches to teaching a language "that no longer is", "is one and more than one", or a language that "disintegrated", "dissolved" and "broke up", evolved through my own teaching practice. I am a teacher of Serbo-Croatian or BCMS as a foreign language at the School of Slavonic and East European Studies (SSEES) at University College London (UCL). When I started teaching at UCL in the early 2000s, the language name in the language modules was Serbian/Croatian or Serbian and Croatian. At the time the Chair of Serbo-Croatian Studies at SSEES was Dr Celia Hawkesworth. In her account of Serbo-Croatian language teaching at British universities Hawkesworth (2004: 277) notes: the language continued to be called 'Serbo-Croat' in all our teaching and official materials until late 1990s, i.e. sometime after this became a far from neutral label. The present compromise is to call the language 'Serbian and Croatian' while courses involving literature include also Bosnian. This represents our endeavour, as an academic institution, to reflect what we still see as the linguistic, rather than the political, facts. In our teaching practice, it is quite clear that we view the language, whatever its name, as one linguistic entity.
In line with the stance elaborated by Hawkesworth, the institutional approach continues to be inclusive of literatures, cultures and linguistic idioms across Bosnia and Herzegovina, Croatia, Montenegro and Serbia.
Both alternative language names, Serbo-Croatian, and BC(M)S, are currently used by some linguists and academic institutions with a view to designating one linguistic system. The continuing use of the glottonym Serbo-Croatian, as has been argued (Kordić 2010: 134-136), reflects a positioning which insists on the continuity with a scholarly tradition which has maintained that the common language of the peoples in the central South Slavonic space, based on the Štokavian dialect, is one single linguistic unit with several national varieties (cf. Raecke 1996;Lehfeldt 1996;Šipka 2003;Kordić 2001;Gröschel 2009;Corbett/Browne 2009[1987Kordić 2010;Bugarski 2012). Equally, the other designation, Bosnian/Croatian/Serbian (henceforth BCS), has been gradually gaining acceptance among foreign scholars and a great number of academic institutions outside the Yugoslav successor states over the past twenty years. Many such universities have retained Serbo-Croatian studies as a single subject area in their Slavonic departments (cf. Bugarski/Hawkesworth 2004: 235-282;Greenberg 2008: 181), but have changed the language name to BCS in order to acknowledge the political realignment along with the change in attitudes of the speakers of Serbo-Croatian. This stance reflects the understanding of sociolinguistic change according to which Serbo-Croatian can be described as unicité sur le plan linguistique ['singularity on the linguistic level'] but multiplicité sur le plan politique ['plurality on the political level'] (Thomas 1994: 257), or as "one language linguistically but three languages politically" (Bugarski 1997: 7), as put forward by linguists who refused to accept the outright elimination of the concept of a common language. Hence, in the domain of presenting and teaching the language to foreign students, BCS seems to manifest the view that "language is simultaneously one and more than one: that is, there is a single common core, and there are separate standardized representations of this core, each bearing the name of a national/ethnic group" (Alexander 2005: 212).7 On the other hand, in the domain of official language policy in Bosnia and Herzegovina, Croatia, Montenegro and Serbia, the fragmentation of the Serbo-Croatian language has been entirely accepted and even constitutionally enshrined. The changes in the constitutional stipulations regulating language policy in the successor states were followed by the "necessary" changes in the national curricula. In school curricula ethnic/national languages and local cultures have been given objective existence and have been predominantly represented as separated from the neighbouring languages and cultures (cf. Gustavsson 2009).
Likewise, in the context of teaching Serbo-Croatian as a foreign language, educational institutions in the successor states chose to abandon the use of the glottonym Serbo-Croatian and to redefine their practices predominantly in accordance with the new political agendas of their governments, insisting on the treatment of their national languages and literatures as separate entities with unique historical and cultural trajectories. Following such developments in the former Yugoslavia, and inspired by the discussion among teachers and students at the University of Vienna, the Croatian Slavist Radoslav Katičić, in an article published in 1995, called for the division of Serbo-Croatian studies abroad into three separate study programmes: Croatian studies, Serbian studies (including the closely related yet independent Montenegrin studies) and Bosnian studies, proclaiming Croatian, Serbian and Bosnian as "three objects of knowledge" (Katičić 1995: 61). Katičić's appeal, however, was not met with acceptance either by universities and education authorities in Austria (cf. Neweklowsky 2004), or by the majority of other Slavonic departments outside of the former Yugoslavia (cf. Bugarski/Hawkesworth 2004: 235-282).8 7 The ethno-national and religious heterogeneity of Bosnia and Herzegovina, Croatia, Montenegro and Serbia, and the fact that the borders of the regional dialects do not correspond with ethnic or national delineation in the region, are two phenomena that often get overlooked in the literature on linguistic matters in this area. This is conducive to spreading a misleading and erroneous identification of "successor" languages with ethnic or national groups, rather than with the group of speakers of a certain variety who happen to share the same linguistic repertoire regardless of their ethnic allegiance. 8 It has to be added that the approaches presented in the volume Language in the Former Yugoslav Lands (Bugarski/Hawkesworth 2004) are not reflective of the stances taken by all foreign institutions. Namely, at a smaller number of universities Serbo-Croatian studies have been divided. According to the Croatian Ministry of Sciences, Education and Sport, there are three centres for Croatian studies, in Australia, Canada, and Scotland. Serbian and Croatian studies are separated at most universities in Poland, and at some universities in the Czech Republic, Slovakia, Hungary, Bulgaria and Italy.
The debate about the status of the Serbo-Croatian language and its varieties has recently shifted (again) towards a position which looks at the internal variation within Serbo-Croatian through the prism of linguistic pluricentricity (Greenberg 2008(Greenberg [2004 ;Gröschel 2009;Kordić 2010;Bugarski 2018). This framework endorses a communicative conception of language based on the historical and current usage of the common language. What is more, within this framework, the conventional sociolinguistic knowledge produced by linguists who supplied arguments for the treatment of national varieties of Serbo-Croatian as separate languages is critically examined. As recently (re-)applied9 to the Serbo-Croatian language, the pluricentric model therefore presents itself as a framework through which, in my view, the focus of linguists' activities can be diverted from the endless elaboration of differential varietal features, orthographic issues and ensuing local prescriptivisms, to an exploration of linguistic and sociolinguistic phenomena in the Serbo-Croatian speaking area across recently imposed linguistic, cultural, scholarly and institutional boundaries.
It has been argued that taking the model of pluricentric standardisation as a framework allows us to disregard the treatment of Serbo-Croatian as an exceptional or unique linguistic case (Kordić 2010). General issues such as the role of the nationstate in legitimising the national standard variety/language, the (in)equality in status of the national varieties of pluricentric languages, speakers' attitudes and the role of linguists in accepting/discarding the pluricentic model have already been addressed, widely discussed, and their significance recognised with regards to languages such as English, German, French, Spanish, Portuguese, Dutch, Swedish, Chinese, to name a few (cf. Clyne 1992; Soares da Silva 2014).
Looking at similar pluricentric scenarios, such as those of the German language and the English language to which Serbo-Croatian is most frequently compared, offers an opportunity to critically approach the current sociolinguistic situation in Croatia, Bosnia and Herzegovina, Serbia, and Montenegro, and to establish how the relationship between what is common and shared and what is distinctive and particular can be conceptualised. For instance, as argued by Schneider (2014) and Grzega (2000) in the English and German cases, from the strictly linguistic point of view we face similar difficulties in delineating national varieties of Serbo-Croatian in an absolute sense as lexical and grammatical differences depend on speakers' preferences, or on the spoken or written variety chosen for analysis. Furthermore, factors such as intra-varietal variation and varietal convergence, arising due to the mutual influence of varieties, make it harder yet to draw clear-cut lines between varieties of both English and German, as well as Serbo-Croatian. In the case of German, the focus on norms and the 9 The term has been used since the late 1960s to conceptualise the linguistic unity in this area. Stewart (1968) and Kloss (1967) mention Serbo-Croatian as an example of a pluricentric language, albeit a unique one, as it is used within one and the same country. Before Yugoslavia's break-up the term had been used by Yugoslav linguists such as Bugarski (1983), Radovanović (1986), Karadža-Garić (1990) and Brozović (1992). state as the framework that authenticates national varieties' standard language norms (Ammon 1989), means that regional variation and actual language use are ignored. As Auer (2014) demonstrated, this leads to differences between national standard varieties of German being presented as "natural" rather than created. The same argument could be applied to the normative approaches to the national varieties of Serbo-Croatian.
The pluricentricity of both English and German has been increasingly acknowledged and the teaching of the two languages as foreign languages reviewed in order to achieve a better and fairer representation of both the linguistic and cultural practices of all speakers of the two languages. It has been proven that pedagogical approaches that encompass the acceptance of the pluricentric nature of German and English, although admittedly challenging, do not require radical changes of curricula, as they do not contest the usage of one standard norm in teaching. Rather, they advocate a critical choice of the norm, which needs to acknowledge the existence of other related norms, considering the context of teaching as well as learners' preferences (cf. Matsuda 2012;Lowenberg 2002;Muhr 2000;Studer 2002;Boss 2005).
However, the recent critical turn in approaching the Serbo-Croatian language "split" and linguistic nationalism in the region has remained largely on the theoretical level with little or no impact on the teaching of the language. Institutional approaches to the Serbo-Croatian language issue in the context of teaching it to foreign learners, both in the region and outside of it, continue to be clearly disparate and many questions still remain unanswered. How do we name the language that we teach to foreign students? Do we treat national varieties as part of one linguistic system or do we present students with different varieties as different and (un)related standard languages? If so, do we present the national culture as created by and transmitted through the national language? How do we approach teaching the shared Yugoslav culture in the new political circumstances and how do we relate it to Yugoslav successor states' national(ised) cultures?
These insights into the changing and challenging attitudes towards the pluricentric model served as a theoretical framework for my empirical study into different institutional and individual approaches to the teaching of Serbo-Croatian.

Methodology
In order to gain as comprehensive a picture as possible of the field of Serbo-Croatian language teaching, a survey method was chosen as the most efficient way to gauge a general understanding of the views of teachers of Serbo-Croatian (BCMS) on what the language used to be before Yugoslavia's collapse and what it was at the time I conducted the survey in 2014. A questionnaire was used to ask teachers to evaluate the existing, established attitudes that have become accepted representations of linguis-tic reality, and as such have defined both language use and language teaching. The questionnaire also contained statements that reflect attitudes at variance with these established and widely accepted attitudes. This has enabled me not only to determine what the most widespread attitudes are and what principles of language ideology inform these attitudes, but also to trace and analyse changes that have happened since the dissolution of Yugoslavia, and to locate reasons and present justifications for these changes.
The distribution of the online survey places the Serbo-Croatian language issue in both a domestic (within the Serbo-Croatian speaking area) and an international setting. This allowed the additional aim of testing whether institutions abroad continue to treat and present Serbo-Croatian as one entity, as evidenced in the previous section, and whether the opposite stance taken by institutions in the region has changed at all, given the recent critical turn in conceptualising the common language.
Participants in the study were teachers of Serbo-Croatian (BCMS) as a foreign language at higher education institutions.10 In the first phase of my study I obtained information about as many higher education institutions worldwide as possible that offer Serbo-Croatian (BCMS) language teaching, mainly through consulting the webpages of universities where the language is taught. Overall, I contacted 94 higher education institutions, out of which 60 responded. The number of teachers who completed the questionnaire is 83, giving a 43 % response rate (see Table 1). 10 The only exception is a language school in Sarajevo which I contacted due to the fact that the University of Sarajevo did not offer BCS as a foreign language at the time I conducted the survey. I could not identify any such schools in Montenegro. There were also no guest lectors from Bosnia and Herzegovina or Montenegro sent to teach the language abroad at that time. 11 This is the number of guest lectors teaching outside the region who responded to the questionnaire.
I attempted to contact all the language teachers I could find through the sources mentioned above. This means that I did not contact all existing teachers of the language and my sampling method could therefore be classified as convenience sampling. However, I contacted the most well-known and the most prestigious institutions throughout the world that offer courses in Serbo-Croatian and the data obtained from the survey should represent a snapshot of views on developments in the field at the time I conducted the survey. The sample includes teachers who teach Serbo-Croatian (BCMS) abroad, and also those who teach in Bosnia and Herzegovina, Croatia and Serbia. In the analysis of the collected data, the 16 Croatian and Serbian lectors, sent abroad by the Croatian state and the Serbian state to promote their national language and culture, were initially treated in the same way as the other teachers employed by these institutions. It was assumed that the former have to adapt their teaching to the policies and the curricula of the institutions at which they are placed.
The questionnaire was designed to collect both quantitative and qualitative data. There were two versions, one in Serbo-Croatian (Serbian, Ekavian variety) in the Latin alphabet and one in English. The questionnaire was distributed to invited participants through personalised emails as an online survey, using the survey programme Opinio. In the email, I informed the participants about the topic, the outline and the aims of my research.
With regards to thematic scope, the questionnaire contains four main parts as shown in Table 2. Quantitative data were collected through 28 close-ended statements, and qualitative data through 11 open-ended questions (see Appendix). Responses to close-ended items were measured on a five-point Likert-type scale (5=strongly agree, 4=agree, 3=neither agree nor disagree, 2=disagree, 1=strongly disagree). The type and range of closed responses reflected the range of perceptions and views that teachers in the predominant sampling frame (namely, domestic/international) might hold. The open-ended questions sought information about participants' profiles as well as the historical and current institutional approaches to teaching Serbo-Croatian as a foreign language. Qualitative data were also collected through comments which participants were invited to make after sections two and three in the questionnaire. The majority of them did leave additional comments and the data obtained from these comments have also been used in the analysis. The qualitative data were coded using the participant codes generated by the survey programme Opinio, in order to protect anonymity.

Survey results
The survey primarily sought to collect teachers' attitudes towards both the status of Serbo-Croatian and approaches to teaching the language (sections 2 and 3 in Table 2), with the additional aim to measure the relative dependence of such attitudes on variables such as the teachers' location (see Table 3). Attitudes to the sociolinguistic situation in the Serbo-Croatian speaking area are discussed in section 5.1 and attitudes to the teaching of the language and their distribution according to location in sections 5.2 and 5.3. Given the small size of the sample and the distinctly unequal sizes of the two groups, the data presented are the result of descriptive statistical analysis, and only show measures of central tendencies such as the mean and standard deviation.  Table 4 shows the four major themes covered by the questions. The mean observed for item 8, "The Serbo-Croatian language does not exist" (n=7812, mean = 3.04, SD=1.51) shows that opinions are still divided about the language, whose period of existence is frequently related to that of the common state of Yugoslavia, even more than two decades after its break-up. It seems that currently there is no consensus on this issue among participants (see Figure 1). Apart from the historical and political factors that may have influenced participants' views on this issue (to which I will return), it is important to observe factors relating to terminological and conceptual uncertainty with regards to understanding concepts such as standard language, common language and language variety revealed in participants' comments on the question. Five participants chose not to respond to item 8 and some commented that their choice not to do so was linked to the fact that, in their view, the statement was imprecisely formulated. This group (as well as some participants who did express their view on the statement) pointed out that the term language was not defined precisely enough, and as a result they were unsure whether the term Serbo-Croatian was meant to refer to standard Serbo-Croatian or not. As one respondent stated:

Attitudes to the sociolinguistic situation in the Serbo-Croatian speaking area
(1) I fail to answer several questions because the term "language" is ambiguous and creates room for misunderstanding and manipulation. Maybe you should use the term "standard language" instead. Serbo-Croatian did not exist as a unified standard language, not even immediately after the Vienna agreement.
Following the theorising of the standard language by Trudgill (2000), I view the standard language, in both written and spoken form, as a specially selected, codified variety of a language expected to be used in certain domains (education, publishing, media, for instance). In reality, however, the distinction between standard and non-standard features (usually regionally and socially specific) is linguistically arbitrary, and the demarcation of features is fluid and can be subject to change. According to Trudgill, Standard English can therefore not be equated with the English language, which he describes as "consisting of an autonomous standardised variety together with all the non-standard varieties which are heteronomous with respect to it" (Trudgill 2000: 117). Based on this understanding of the standard variety, the phrasing "Serbo-Croatian standard language" was deliberately not used in the questionnaire. Comment (1) also points to another problem with defining and conceptualising the standardness of the Serbo-Croatian language. The respondent takes the extent to which a (standard) language is unified as a criterion crucial for its existence. This stance reflects a widely-held view that Serbo-Croatian was, or was meant to be, a unified (standard) language. The idea of one unified language was, indeed, mentioned in the Novi Sad agreement in 1954. It was (later) used to refer to attempts to achieve unified solutions in continued, but often conflicting, work on the standardisation of orthography and scientific terminology. However, the Novi Sad agreement also stipulated that the Ekavian and Ijekavian pronunciations, as well as both alphabets, were to continue to have equal standing, and that the new dictionary of contemporary Serbo-Croatian should be based on the entire range of the Serbo-Croatian lexicon. Hence, from a normative point of view, the Western variety, with its centre in Zagreb, and the Eastern variety, with its centre in Belgrade, continued to exist, not as parts of a unified superordinate entity, but as two, interacting standard varieties.
Given this situation, the direct question about the existence of the Serbo-Croatian language (item 8) did not yield definitive answers, whereas the responses pertaining to the emergence of the "successor languages" and to reasons that motivated the alleged "dissolution" of Serbo-Croatian demonstrates more polarised views. Figure 2 shows the overall percentage of those who believe that Serbo-Croatian does not exist for political reasons (item 11: n=79, mean=3.91, SD=1.34) is 72.2 % (45.6 % strongly agree and 26.6 % agree). The absence of a consensus in response to the statement that Serbo-Croatian does not exist, compared to a clear majority who consider that political reasons stood behind its dissolution (pointing in the direction of the implicit acceptance of the language's fragmentation), may have to do with the way the question was formulated.13 It seems that, irrespective of whether respondents approve of or agree with the statement relating to the existence of Serbo-Croatian, they have accepted the fact (or the widely-held general view) that this is the way of describing the cause of the language's fragmentation (i.e. political reasons) and its consequences (newly established "successor languages").
Political changes in Yugoslavia and its disintegration are also stated as the decisive factor behind many Slavonic departments' decision to rename their Serbo-Croatian study programmes following the break-up of Yugoslavia, in response to an open-ended question in the last part of the questionnaire, which asks respondents to provide the reasons for the name change of the subject/department/programme. Political reasons aside, the proponents of the idea of separate standard languages in the Serbo-Croatian speaking area have claimed that these languages have in their essence always been what they are today, implying that Serbo-Croatian, (mis)construed as a mixture of these standard norms, never existed. For instance, a statement published by the Croatian Academy of Sciences and Arts (Hrvatska akademija znanosti i umjetnosti, henceforth HAZU) maintains that "one could never ever have written or spoken the 'Serbo-Croatian' (Croato-Serbian) language, because it never has existed as a concrete language nor does it exist today. Rather, texts are realised either as Croatian or Serbian" (HAZU, 1996, emphasis added). And yet, the survey results show that the majority of respondents would not agree with HAZU's assertion: 51.3 % strongly disagree and 21.8 % disagree with the statement that the language has never existed (item 12: n=78, mean=2.0, SD=1.30).
Despite the lack of consensus on the current existence of Serbo-Croatian, which, as shown above, can be ascribed to the interpretation of the standard language model and to reasons external to linguistic judgment, not to mention the strong agreement that political reasons played a part in the language's fragmentation, Figure 3 shows that 39.7 % of respondents strongly agree and 28.2 % agree (overall 67.9 %) that Serbo-Croatian is a pluricentric language (item 15: n=78, mean=3.83, SD=1.26).
In addition, 60.5 % of respondents agree with the statement that "National variation within the Serbo-Croatian language can be compared to that of the German language (German German, Austrian German, Swiss German)" (item 16: n=81, mean=3.67, SD=1.33). One possible interpretation of the above disparate and, to a certain extent, illogical results (how can we claim that a language does not exist yet attribute a descriptive category to it, as if it did?) is that the disappearance of the language name from the public sphere, and especially from official use, has given criteria such as political reasons and the official status of the language a greater prevalence in the acceptance of the "non-existence" of the common language. On the other hand, in the field of linguistic enquiry (the expert sphere), where the existence of one single language system, based on one and the same Štokavian dialect, is admitted (although with several caveats), it is "allowed" to claim that the classification of Serbo-Croatian as a pluricentric language is appropriate, as illustrated by comment (2). The respondent presents language-external factors as more significant in determining the status of the common language, but maintains that it can still be described as a pluricentric language: (2) Statement no.11 ["Serbo-Croatian does not exist any longer for political reasons"] sums up my opinion of the (non) existence of Serbo-Croatian. I consider it the only correct statement […]. So, Serbo-Croatian nominally no longer exists, regardless of my opinion, simply because no language under that name exists anywhere. But BCMS are, whatever they are called, four variants of one polycentric language -which is no longer called Serbo-Croatian.
In contrast with the prevalence of language-external criteria that determine attitudes about the status of Serbo-Croatian as well as about its description, attitudes to language-internal criteria show there is almost unanimous agreement that differences between the standards are minimal. Overall, 96.3 % of respondents considered Serbo-Croatian varieties to be mutually intelligible (item 18: n=81, mean=4.86, SD=0.44; see Figure 4). We also note a very high level of agreement (overall 83.5 %) with the claim that linguistic differences between the four standard languages are small (item 20: n=79, mean=4.32, SD=1.07). Given the above, it is surprising not only that most respondents agree (overall 54.4 %) that the four standards are on divergent paths (item 19: n=79, mean=3.44, SD=1.11) but also that there is no consensus on the matter of whether the four standards are separate languages (item 21: n=79, mean=2.72, SD=1.47). The belief that the four standards will inevitably go their separate ways, leading to the development of mutually unintelligible languages, was prevalent in the period immediately after Yugoslavia's dissolution, when the existing standard norms and standard expressions were re-standardised. Similar to the moral panic caused by the fear of linguistic assimilation, present in the times before the re-standardisation processes were initiated in the 1990s, the scale and effects of the norm changes were exaggerated, both by the experts and by the public who accepted or opposed these changes. The current exchanges across the borders of Croatia, Bosnia and Herzegovina, Montenegro and Serbia in the spheres of culture, education, economy and politics also work against the scale of changes that would be needed for these standard languages to become mutually unintelligible. The above results can only be interpreted in the light of the prevailing conceptual and discursive strategy according to which Serbo-Croatian no longer exists, as the name of the language is no longer in use in the official domain. Furthermore, the belief about the disparate developments of the "successor languages" can be linked to the spread of the metaphor of "disintegration" or "dissolution" of the common language and the idea that its varieties are undergoing change and are therefore distancing themselves from one other. In line with what has been said above, respondents' most common reasons for the view that Bosnian, Croatian, Montenegrin and Serbian are separate languages are their status as official languages and the fact that these languages serve as the markers of distinct national identities. Table 5 shows the two major themes in this part of the questionnaire.  (items 24, 26, 30, 31, 33, 34) 2. Varieties of Serbo-Croatian to be taught together (items 27, 28, 29, 32) Despite the fact that the vast majority of respondents agree that Bosnian, Croatian, Montenegrin and Serbian are mutually intelligible and that the linguistic differences between them are small, only 55.6 % of respondents state that they should not be taught separately (item 24: n=81, mean=2.58, SD=1.42).

Attitudes to teaching Serbo-Croatian as a foreign language
Consistent with the (historically) strong normative orientation in language policy and language pedagogy in Bosnia and Herzegovina, Croatia, Montenegro and Serbia, the most stated reason for teaching varieties of Serbo-Croatian as separate languages (item 25) is to avoid mixing of the standard norms. However, even when a strong preference for teaching one variety is stated, as in comment (3), the implicit knowledge of the common core that all four standards share is pointed out as a bonus: (3) Exclusively one variety can be taught, one does not always have to teach all varieties, it depends on the participants in the course -some just want to learn Croatian, and some want to learn only Serbian. However, by learning only one variety they automatically learn a lot that is common to all of them.
The great majority of teachers disagree that only textbooks published in the region should be used in teaching (item 33: n=82, 81.7 % overall disagree). Just over half agree that the languages should not be exclusively taught by native speakers (item 30: n=80, 57.6 % overall agree), with a greater number opposing the suggestion that Croatian should only be taught by a person who is Croatian (item 31: n=81, 75.4 % overall disagree). Most disagree that the ethnic composition of the diaspora in the country should be a decisive factor in deciding which variety should be favoured in language teaching (item 34: n=81, 70.4 % overall disagree), as comment (4) illustrates: (4) Regarding the diaspora, I noticed that at my university, although the largest diaspora is the Serbian diaspora, Serbs are not the most numerous group in our language courses. So, it would not be logical to teach the Serbian variety only. (906705) However, in many comments, where respondents explain how the teaching is organised and whether all varieties are taught, or only some, it emerges that the actual (ethnic) composition of classes and the predominance of heritage speakers in the classroom in some institutions abroad do dictate the approach to teaching the language. Hence, some institutions choose to organise teaching according to the political divisions that occurred in the 1990s in the territory of Yugoslavia, adopting the prevalent attitude that views language as a marker of ethnic identity: (5) By 2010, the language was taught under a different name. Since 1991, when the collapse of Yugoslavia began, the Department decided to separate Serbo-Croatian, and to start offering lectures in Serbian and Croatian. The Department decided to do that for political reasons and because of the pressures from the diaspora. (906705) On the other hand, a smaller number of respondents also comment on how they and their departments choose to resist this view: (6) Students with Croatian, Bosnian, Serbian background should be taught together. As they are linguists they should be well informed about the three or four varieties. (905675) Regardless of the above, the opinions on whether the varieties of Serbo-Croatian should be taught together are divided. Notably, the number of undecided responses is greater (item 27: 24.7 %) than among the responses to the question of whether they should be taught separately (item 24: 12.3 %). Also, the percentage of those who agree that they should be taught together is lower (item 27: 48.2 %) than the percentage of those who oppose the languages being taught separately (item 24: 55.6 %). This disparity may reflect the difficulty in putting simultaneous teaching of all four varieties into practice. Apart from the lack of (an agreed) theoretical framework of how to teach Serbo-Croatian in current circumstances, teachers face practical obstacles to teaching all standards together, such as the preferences of the potential heritage speakers in the classroom (as seen in the comments above), a general lack of teaching materials that offer content which covers all four standards and cultures in one teaching resource, and a lack of support due to the fact that Slavonic departments currently struggle with resources and the popularity of their programmes among students.
In the additional comments section, following the group of questions about teaching the language, the prevailing opinion is that students should be exposed to all four standards. However, differences in suggestions as to how to approach this way of teaching and also who can carry it out are noticeable. Some respondents suggest that beginners should not be exposed to all four standards "in order not to confuse students with phonetic and lexical variants" (906460) but should learn about the differences between them at the later stages in their studies. At the same time, some teachers are of the opinion that beginners should be made aware of the four standards and be exposed to them, and then choose one, which they will focus on later.
The lack of greater linguistic differences between varieties and the fact that students gain knowledge which enables them to communicate with native speakers across the region are the most frequently stated reasons for teaching the four varieties together, whereas the shared, common culture is the least stated reason for doing this, indicating acceptance of the strong link between national language and national culture (item 28). This does not come as a surprise, as teaching of the shared culture is currently not supported by the education systems in Bosnia and Herzegovina, Croatia, Serbia and Montenegro. After the collapse of Yugoslavia in the 1990s, philological studies in the successor states were delineated along national lines and students (the future teachers of the language) started to be educated in only one national language and literature, as indicated by comment (7): One should not fool oneself: a Serb from Serbia can with good will teach the Croatian language, at least to a certain level, but will not have the same competence as a lecturer from Croatia who has completed studies in Croatia (at least study of the Croatian language and literature). (917823)

Effect of the country where the language is taught
The results obtained after comparing the views of the teachers in the region with the attitudes of their colleagues who teach the language abroad do not show clear differences in either the description of the sociolinguistic situation in Bosnia and Herzegovina, Croatia, Montenegro and Serbia or in the approaches to teaching the language. It transpires from participants' comments, as pointed out in the previous section, that the problems faced by universities abroad have to do with the ethnic composition of classes and the present lack of a framework to teach the national varieties of Serbo-Croatian simultaneously. In line with this, questions over where the differences in attitudes between the groups are more noticeable point to the predominant acceptance of the language "dissolution" model by the institutions abroad (items 10 and 19). This is shown by the results of the statistical analysis in Table 6. We see that the perception of national varieties as separate languages, which are increasingly diverging from each other, is more favoured by the teachers in institutions abroad (outside the region: n=69, mean=3.22/SD=1.50, compared with in the region: n=7, mean=1.86/ SD=1.21, where 5 = strongly agree, and 1 = strongly disagree). The main reasons teachers stated for seeing "successor languages" as separate languages are the different cultural traditions that these languages represent and the fact that languages serve as symbols of distinctive national identities. The opinions on teaching the languages separately (item 24) are more neutral in both samples and lean more towards disagreement with the statement (see Table 7). Where teachers think that the languages should be taught separately, they select avoiding mixing the standard norms as the main reason. In addition, fewer teachers from abroad think that differences between the standard norms are small, although there is not a great difference in opinions between the two groups (item 20, see Table 6). With regard to the initial assumption that institutions abroad predominantly still treat and present Serbo-Croatian as one entity in opposition to the stance taken by the institutions in the region, these results do not show the anticipated polarised views. This may partly be due to the size of the domestic sample (with 73 teachers from institutions abroad and only 10 from the region) and partly due to the composition of the samples. I therefore undertook an additional analysis of the data by comparing the attitudes of the 16 Croatian and Serbian guest lectors, who are sent abroad by the Croatian and Serbian states, with those of their colleagues abroad. The attitudes of these two groups are markedly different. Guest lectors predominantly agree that Serbo-Croatian no longer exists (item 8: n=15,mean=3.67,SD=1.42) whereas their colleagues have an undecided opinion on this issue (item 8: n=55,mean=2.91,SD=1.42). The latter strongly disagree with the statement that the Serbo-Croatian language never existed (item 12: n=54, mean=1.91, SD=1.25) while guest lectors seem to be less convinced (item 12: n=16,mean=2.38,SD=1.41 . These results show that the attitudes of guest lectors are more aligned with the language ideology that promotes separate status of national varieties. This finding prompts further investigation into the interaction between institutional frameworks that adopt different views of Serbo-Croatian and the teaching practices of guest lectors.

Implications and conclusions
The responses to the section of the questionnaire which explored attitudes to the language question among teachers in Serbo-Croatian studies at universities in the region and abroad confirmed the earlier insights about an undecided general view on the status of the common language. Although an evaluation based on the language-internal criteria of differences between varieties on all levels of language structure shows nearly complete agreement among respondents that Serbo-Croatian is one linguistic system, the language-external criteria (such as the changed political alignment in the region, change in language naming and in speakers' attitudes) take precedence over them. Hence, in line with the predominant discourse strategy that framed changes in language policy using the metaphors employed to describe changes in the geopolitical sphere in the 1990s, this study finds that a substantial number of participants in the survey (51.4 % overall) believe that the common language "disintegrated" into "successor languages" and that these languages are on divergent paths.
There is a strong link between attitudes that view a common language as not "unified" and not "concrete" and similar misconceptions circulated by linguists who propounded linguistic nationalism both before and after Yugoslavia's break-up (Gröschel 2009: 74). A similar type of correspondence can be observed with regards to the view that the "standardness" of the national varieties of Serbo-Croatian is the basis for their treatment as separate languages on one hand, and the strict normative linguistic culture established long before Yugoslavia's collapse on the other.
Given that most respondents agree that varieties of Serbo-Croatian are mutually intelligible, it is quite surprising that only 48.2 % of respondents agree that they should be taught together, with 24.7 % undecided on this matter (item 27: n=81, mean=3.37, SD=1.42). So, the survey neither shows strong agreement on this stance, nor does it present a unanimous view on how to approach teaching all four varieties simultaneously. Problems with issues such as the stage of learning at which students should be exposed to more than one variety and ways to approach variation in contexts where heritage speakers are part of the student cohort seem to be shared by a number of teachers.
Contrary to the initial expectation, the survey revealed that views on Serbo-Croatian as "one or more than one language" seem not to be so sharply polarised according to the criterion "in the region/outside the region". Additional exploration is therefore required of the challenge that teachers in the region face in the social and educational context where the varieties are supposed to be presented and taught as separate languages, but where students' exposure to other varieties through education, family ties or personal experience brings another view of the linguistic composition of the region. This study has detected the need to provide practitioners in the field with the theoretical and methodological apparatus which would enable them to evaluate all aspects of the pluricentric model, as well as the (beneficial) implications of its acceptance.
It is true that the inclusion of variation as a topic and object of study complicates teaching, but since this practice exists in the context of other pluricentric languages, where questions about the modes of its inclusion in curricula have already been asked and solutions to the challenges suggested, it is reasonable to expect that this model could be used as a viable one for teaching Serbo-Croatian as a foreign language. With regards to teaching English and German as foreign languages, the recommendation that learners need to be exposed to different national varieties remains undisputed (Davies 2010;van Kerckvoorde 2012;Matsuda 2012;Marlina 2014). There are different opinions on when, to what extent and how to implement this, especially in a climate where constraints of time and resources have a bigger say in teaching than pedagogical recommendations. There is a broad agreement that, in order to increase communicative effectiveness, the emphasis in presenting variation should be put on developing receptive skills rather than productive skills, and also earlier in the learning process rather than later. Additional learning goals with regards to the inclusion of variation in teaching a foreign language are the incorporation of socio-political questions which underpin evaluative perceptions of variation, as well as building a perception of tolerance to linguistic variation, as encountered both on national and on regional levels (Matsuda/Friedrich 2011).
In the Serbo-Croatian case, the main advantage of the pluricentric model is that it incorporates both the common core -the raw material in a language -and the cultural and linguistic specificities of a national variety of a language, while not neglecting identity needs of the speakers of a national variety of a language (Bugarski 2018). As such, the acceptance of the pluricentric model currently encourages both the comparative approach to variation on a national level, and dialogue among linguists across the region. Teaching based on such an approach could better prepare students for communication with speakers across the Serbo-Croatian speaking area, and provide them with a broader insight into the rich and diverse linguistic and cultural history of the Balkans.
12: The Serbo-Croatian language has never existed. 13: Croatian and Serbian existed as distinct languages before the joint standardisation in the 19th and 20th century. 14: Serbian and Croatian were only artificially combined into the language of Serbo-Croatian. 15: The Serbo-Croatian language is a pluricentric language, i.e. a language with several national standard varieties. 16: National variation within the Serbo-Croatian language can be compared to that of the German language (German German, Austrian German, Swiss German). 17: The case of Serbo-Croatian is unique and therefore incomparable to other pluricentric languages. 18: Bosnian, Croatian, Montenegrin and Serbian are mutually intelligible. 19: The current four standard languages are on a divergent path. 20: Linguistic differences between the four standard languages are small. 21: Bosnian, Croatian, Montenegrin and Serbian are separate languages. 22: Bosnian, Croatian, Montenegrin and Serbian are separate languages because: (you can choose more than one option) They represent different historical and cultural traditions. They are official languages of their respective countries. The speakers of these languages think so. They serve as symbols of distinctive national identities. 23: Any additional comments: Part 3 24: Bosnian, Croatian, Montenegrin and Serbian should be taught separately. 25: Bosnian, Croatian, Montenegrin and Serbian should be taught separately because: (you can choose more than one option) They represent different cultures. There are linguistic reasons for such a decision. In order to avoid mixing of the standard norms. 26: Students who want to learn Croatian should not bother to learn the Cyrillic alphabet. 27: Bosnian, Croatian, Montenegrin and Serbian should be taught together. 28: Bosnian, Croatian, Montenegrin and Serbian should be taught together because: (you can choose more than one option) They share a common culture. The linguistic differences between them are small. There are not enough students to justify the creation of separate subjects. The students gain knowledge which enables them to communicate with native speakers across the region. 29: Bosnian, Croatian, Montenegrin and Serbian should be taught together but maintaining awareness of differences between them.
30: Bosnian, Croatian, Montenegrin and Serbian should be taught by native speakers only. 31: Croatian can only be taught by a person who is Croatian. 32: Teachers, who are native speakers of one of the successor languages, should be educated to teach the other three languages too. 33: Only textbooks published in Croatia, Bosnia, Serbia and Montenegro should be used in teaching. 34: If the predominant diaspora community in the country is for example Bosnian, then the language that the universities in the country offer should be Bosnian. 35: Any additional comments: Part 4 36: Since when has this language been taught in your department/school? 37: Under what name is the language currently taught in your department? 38: Has that name always been in use or has the language taught at your department/school been called by a different name? If the name changed, when did the change occur? 39: Could you explain the reason for the change of name? 40: If the name has not changed has a change ever been considered? Why? 41: Are teachers on your programme expected to teach only one variety (possibly their native variety) or do they teach all of them?