AWPEL Peeling the onion top-down: Language policy in Serbia between power and myth

This paper considers the issue of language policy and planning in Serbia, as managed by the main competent institution, the Serbian Language Standardization Committee, a trans-state, national institution dealing with vital issues of language policy and planning. Specifically, assuming a Bourdieusian perspective, it investigates the ideology behind the Committee’s policies, grounded in a series of language myths, and the way these policies influence professionals and everyday language users. The effects of a rigid, strict educational system and a standard language culture by educators are shown in detail focusing on the Torlak dialect in Southern Serbia. The Serbian case reveals a constant promotion of censorship and a heightened understanding of the benefits of self-censorship in the language market. This can be seen in the pressure exerted on certain speakers and the threat their mother tongue represents for their status in the labor market.


Introduction
This paper considers the issue of language policy and planning in Serbia, as managed by the main competent institution, the Serbian Language Standardization Committee. 1 Specifically, it investigates, from Bourdieu's (1977Bourdieu's ( , 1991 perspective, the ideology behind its policies, grounded in a series of language myths, 2 and the way in which AWPEL Vol. 2.1 2018, 17-41 these policies influence professionals and everyday language users, enforcing speakers to employ certain strategies. First of all, there is the myth that languages are separate, discrete entities, and not social constructs invented by states and nations through the production of powerful discourses of certain language ideologies, as argued by Makoni and Pennycook (2007). Speakers in so-called standard language cultures (Milroy 2001), such as Serbian, consider it crucial that their language exists in a standardized form and their language attitudes are conditioned by powerful ideological positions (Milroy 2001, Milroy & Milroy 1999Milroy 2007), even though they "believe their attitudes to language to be common sense and assume that virtually everyone agrees with them", as Milroy (2007: 133) has put it. Thus, the existence of a shared belief system ensures that the particular set of language practices does not appear as if it were chosen by some specific groups to protect their interests or promote their ideologies. The top-down language policy in Serbia is based on the myth that only certain exclusive language varieties should be valued as acceptable, desirable or attractive, while the rest of them are unacceptable, undesirable, and unattractive, both in the public and private domains. The language authorities are convinced by common beliefs that the media are ruining the language, that speakers with certain geographical or socioeconomic backgrounds are verbally deprived, and that the grammar of certain varieties is bad, truncated or that it simply does not exist, as illustrated by Bauer and Trudgill (1998) for English.
Using a Bourdieusian theoretical frame, in which every language exchange is seen as expressing power relations, we will shed light on the issue of (re)production of the legitimate language (Serbian standard language) and its role in establishing a specific language variety hierarchy. We will analyze the dichotomy of legitimate standardized language vs. non-legitimate non-standardized varieties (dialects), which is an opposition that necessarily emerges from the laws of the language market (Bourdieu 1991;Blommaert 2005Blommaert , 2015a, and then illuminate the myth (or Bourdieusian 'illusion') about Serbian standard language, which is taken for granted by prescriptivist linguists, but mostly defines language policy in Serbia. As a matter of fact, we will demonstrate that Serbian language policy, through its legitimate representatives (Serbian Language Standardization Committee) is perpetuating and imposing these myths. The symbolic power and violence (Bourdieu 1991, Bourdieu andPasseront 1977), identified in imposing the legitimate language (qua standardized Serbian), will be observed from a top-down perspective. At the very top, it will be based on a discourse analytic approach of the language of the so-called codifiers (legitimate language authority) on a large set of data, consisting of official documents as enacted by the Serbian Language Standardization Committee, but also interviews, public lectures, media shows, and other sources. This will be combined with insights into the way the symbolic dominance of these authorities is reflected upon educators and on the mechanisms of enforcement and reproduction of legitimate language through the educational system. Special attention will be paid to testimonies on linguicism 3 which is openly displayed in Serbian classrooms, resulting in the fallacious conviction of Southern students as linguistically inferior -their mother tongue allegedly being a sign of rurality-and that they should abandon it as soon as possible for the benefit of a AWPEL Vol. 2.1 2018, 17-41 brighter future in the job market, where another variety enjoys the prestige of approval. Following this, we will pay close attention to codifiers' measures and initiatives oriented to implementors, qua everyday speakers, and explore the experiences they share when exposed to the symbolic violence of the standard language.
Since language is a channel for expressing power relations (Mey 1985;Fairclough 2001, Mayr 2008, our goal is to try to gain insight into the ideological background of the concept of standard language and the way in which different ideologies are reflected in everyday language practices and speakers' strategies, as well as insight into the mechanisms which enable that the present power relations are maintained and reproduced. The analysis of purist academic discourse and academic practices concentrated on the perpetuation of the myth of legitimacy and naturalness of the imposed language hierarchy, and of "(standardized) language decay" instantiate the Bourdieusian language market, where "prices" are formed for various language products. The prestige of the Serbian standard is imbued by a nationalistic ideology (cf. Wodak 1989), although Serbia is neither emblematic nor an exception to what is commonly observed in standard language cultures. In Section 2, we introduce the foundations of Bourdieu's theory, which is our basis for interpreting the data in Section 3.

Language and symbolic power
Following Saussure's (1916) canonical principle that language should be investigated in isolation from the social context, structural and formal linguistics see language as an abstract, closed, and self-sufficient structure, deprived of any extra-linguistic relations. Saussure's dichotomy between langue vs. parole suggests that all speakers are in principle replaceable, assuming commonality vis-à-vis use of language and everyday speech practices. This means that language itself, as an independent structure, is exempted from social, economic, and cultural context -as corroborated by Chomsky: AWPEL Vol. 2.1 2018, 17-41 interchangeable (Schiffman 1996). The viewpoint of "the impartial observer", as assumed by formal, descriptive, and structural linguistics, excludes the performative power of language, reducing it to a "dead letter" with little practical use save its potential for interpretation, much as one interprets a work of art (Bourdieu & Wacquant 1992: 141).
Therefore, both ordinary speakers and experts presuppose the validity of the myth of the common standard language, which posits a particular set of linguistic practices as a normative model of "correct" usage, even though such a thing as "linguistic communism" (Bourdieu 1991: 43) 6 or "linguistic democracy" (Mey 1985) is, demonstrably, not the case. On the other hand, this kind of approach to the study of language, "divorced from both the wider socio-historical and socio-political context, and issues of language use, power and inequality" (May 2011), suggests that all languages should be treated as being equal. Bourdieu (1991) sees language as a product of the relation between a "linguistic habitus" and a "linguistic market", as a medium of power through which individuals pursue their interests and display their practical competence. The more linguistic capital (capacity to produce expressions for a particular market) a speaker possesses, the more (s)he is able to exert symbolic power (Bourdieu 1977(Bourdieu , 1991Bourdieu & Boltanski 1977). As previous studies have shown, language usage depends on many factors, such as class, age, ethnicity, geographical region, etc. (Bernstein 1972;Labov 1972;Trudgill 1972;May 1985;Fairclough 2001;Holmes 2013;Blommaert & Rampton 2013). 7 Competence in the official standardized language can be considered a kind of linguistic capital which affords its holders symbolic power (Bourdieu 1991), so consequently all non-standard variety speakers possess less of this symbolic power.
Bourdieu describes the social processes related to language using the metaphor of the economic market and associating language practices with notions such as interest, price and capital. The language market, as any other market, is characterized by a certain mechanism for establishing price values. The linguistic market represents "a system of relations of force which impose themselves as a system of specific sanctions and specific censorship, and thereby help fashion linguistic production by determining the price of linguistic products" (Bourdieu & Wacquant 1992: 145). Morevoer, Grenfell notes that "[f]or a value to be set, a 'base rate' needs to be defined: this is the linguistic norm. Such a 'norm' can be particular to a specific field, or indeed a microcosm within it, for example, the language of science, or culture, or politics" (2011: 128).
Standardized language, as a product of its privileged creators, is a code regulating language practices according to a set of ground rules. It is important to highlight that standardized (official) languages are always deeply connected with the state and its institutions: The official language is bound up with the state, both in its genesis and in its social uses. It is in the process of state formation that the conditions are created for the constitution of a unified linguistic market, dominated by the official language. Obligatory on official occasions and in official places (schools, public administrations, political institutions, etc.), this state language becomes the theoretical norm against which all linguistic practices are objectively measured.
AWPEL Vol. 2.1 2018, 17-41 Ignorance is no excuse; this linguistic law has its body of jurists -the grammarians-and its agents of regulation and imposition -the teachers-who are empowered universally to subject the linguistic performance of speaking subjects to examination and to the legal sanction of academic qualification. (Bourdieu 1991: 44) Standardized language is accorded legitimation through the educational system, which is, in turn, tightly related to the job market. Bourdieu even considers the possibility that speakers are actually induced to collaborate in the "destruction of their own instruments of expression" (Bourdieu 1991), in order to obtain both a better position in the scholastic language-market and a better position in the job market.
Therefore, every standardized language can be seen as a form of capital enabling symbolic power and leading to a situation in which all non-standardized varieties are devalued in the local market; non-speakers of the standardized variety possess less symbolic power, and are thus handicapped, when not totally blocked, on their path to success in all fields.
Consequently, if we stop describing linguistic structures and focus on the conditions governing the use of language (Mey 1985), several questions arise: who produces that myth? Or, how is the common-standard-language illusion maintained? How does this power game work? 8 Who made up the game in the first place? Or: How are the rules applied? And (maybe the most interesting of them all): Does anybody get cheated at this game? If yes, who, how, and why? And who does the cheating, how, and why? (Mey 1985: 208) In order to answer these questions, we should consider the path and the visibility of power hidden both behind standard language production (Mey 1985) and behind its utilization in communication. Since "the decisions about which languages will be planned for what purposes ultimately reflect power relations among different groups and sociopolitical and economic interests" (Ricento 2006a: 5-6), we will first demonstrate how these power relations affect language planning and policy in Serbia (executed by the so-called language codifiers). Following the top-down direction, we will display how power is "steered", reflecting on the language educators and ordinary speakers qua implementors, as well as their everyday communication. We will focus on all three levels of the "pyramid" codifiers-educators-implementors, in order to capture the language myths that maintain and enforce these power relations.

Language policy in Serbia
Language policy in Serbia is "an excellent case of language management or language leadership model of language policy" (Filipović & Vučo 2012), that includes language leaders and managers (individuals and/or institutions), as well as "the explicit and observable effort by someone or some group that has or claims authority over the participants in the domain to modify their practices or beliefs" (Spolsky 2009: 5). 9 8 More recent work by a variety of scholars has of course tackled these issues, especially in the framework of Critical Discourse Analysis (Fairclough 2011, Wodak 1989). 9 Greenberg (2010) identifies three different groups of Serbian linguists: i) advocates of the status quo (most SLSC members), ii) advocates of rediscovering the Serbian of nineteenth-century Serbian language reformers Vuk Karadžić and Đura Daničić (Radoje Simić and Miloš Kovačević), and iii) extreme Serbian nationalists seeking an "Orthodox Serbian" language and orthography (Radmilo Marojević).
AWPEL Vol. 2.1 2018, 17-41 In this paper we analyze the activities of the first group, representing official language policy and planning in Serbia. Filipović and Vučo (2012) demonstrate that the official Serbian language policy is rooted in certain nationalistic ideologies, which gained prominence after the disintegration of Yugoslavia, 10 when the language formerly known as Serbo-Croatian split up into many standardized languages bearing national names (Bugarski 2012). 11 These authors highlight issues regarding the Serbian standard language 12 which are significant for our discussion. First, they emphasize that, like the majority of European standard languages, the Serbian standard language has been created following the nationalist model of standardization (Anderson 1983;Wright 2007) and that it is viewed as a symbol of national statehood, i.e. as the most important cultural and historical heritage (in footnote 16 we present the statement of the current Serbian president Aleksandar Vučić regarding the historical relation between language and nationhood). They also identify the top-down language policy as a model for promoting the direct outcomes of the ideology of the standard. Furthermore, Filipović and Vučo note that language cultures, such as the Serbian one, "are resistant to any type of change in standard language structures and insensitive to social, cultural and political shifts in cultural models of given societies" (Filipović & Vučo 2012: 290).
Considering the above, in the following subsection we will analyze some official statements, decisions, recommendations, measures, and initiatives enacted by the Serbian Language Standardization Committee and its representatives. This will corroborate the nationalistic discourse of Serbian language policy authorities and reveal how this discourse creates a specific, ideologically based linguistic culture.

Codifiers
We have already introduced the notions of language codifiers (kodifikatori), educators (edukatori) and implementors (realizatori 13 ), as initially defined in Normativna gramatika srpskog jezika ('Normative Grammar of the Serbian language', 14 Piper & Klajn 2013). The term codifiers refers to "linguistic institutions and well-experienced experts in the field of normative linguistics" (Piper & Klajn 2013: 7), whose direct role is determining the language norm. The educators "as intermediaries between those who make decisions and those who implement them" (ibid.), have a task to disseminate and implement codifiers' solutions. Finally, implementors are those members of the language community who are supposed to apply the prescribed language norm as consistent as possible. The three terms, according to Piper and Klajn,10 It is important to notice that the Serbian Language Standardization Committee was established in the middle of the nineties, i.e. in the post-Yugoslav era. 11 Bugarski (2012) gives a very informative overview of the history of language policy in the South Slavic terrain from the beginning of 19 th century. 12 Serbian is standardized on the basis of the so-called novoštokavian (i.e. neostokavian) variety, which consists of two vernacular segments: the ekavian variety, spoken mostly in Serbia, and the ijekavian variety, used in parts of Serbia, Bosnia and Herzegovina, Croatia and Montenegro. 13 The authors of the paper translated the Serbian terms kodifikatori, edukatori and realizatori as codifiers, educators and implementors (we thank an anonymous reviewer for suggesting the translation for the last term, implementors). These notions, as far as the authors are aware, were introduced for the first time by Piper and Klajn (2013). 14 The grammar was published 2013 by the initiative of the Serbian Language Standardization Committee.
AWPEL Vol. 2.1 2018, 17-41 denote "the participants with their specific communication roles" within the normative linguistics discourse. Since interpreting the listed roles properly is the key for maintaining the quality of the language norm (consisting of normative solutions, normative education and normative practices), "constant, full awareness of [the] complementary roles [of participants] is very important, as well." (ibid.).
The 3-level structure postulated by the two linguists perfectly illustrates the dominant course of language policy and planning in Serbia. On the one hand, it appears as if a certain non-political, non-ideological, pragmatic and even technicist paradigm (May 2006) is offered -each participant has his/her own place, function, and tasks within an already defined system. Moreover, this may give the impression that the paradigm is only descriptive, as if it describes a genuine, existing unified communication model. On the other hand, this concept is presented within a certain dominant ideology which advocates for the linkage between notions of nation and language, with common language being one of the properties perceived as essential for a nation. As a result, language is reputed to be a warrant of national unity and stability: The stability of the standard language largely depends on the stability of the language norms, just as the stability of the national culture depends on the stability of the standard language, the way that the entire social stability in one society depends on the stability of the national culture. (Piper & Klajn 2013: 9) A good cooperation between the scientific institutions and the state is a precondition for a successful language policy in each country, including ours. And -and this is very important-the concern about the status of Serbian language, the language planning tasks, language policy and language standardization, they all represent a continuous process. Language is the best guardian of the past of one nation, a treasury of its spiritual treasures, the most reliable witness of the material goods and the history of a nation; national language is the companion of the society in every current moment, as well. A well-arranged language, with well-designed language policy, is the precondition or a good and cultured communication in the country in all aspects of social life. Finally, language, with its identity, preserves the national identity and culture in the best manner.
[…] One can freely say that there is no successful national politics if language policy is not embedded in it, because that is the way to pursue a responsible policy in the preservation of the national integrity, protection of the cultural heritage and cultural identity. (Tanasić 2014: 48-49) Sreto Tanasić, a prominent member of the Serbian Language Standardization Committee, considers the disintegration of Yugoslavia a threat to the status of the Serbian language, although Serbian had already been enduring "unfavorable consequences" during the Yugoslav federation because of the "self-proclaimed right of other republics and nations, which used to claim right over the language [Serbian] by labeling it with some other, different names" (ibid.). 15 Nevertheless, Tanasić concludes that although "the state did not react timely" on these "creators of new languages", "today the question of Serbian as used among Serbs does not depend" on AWPEL Vol. 2.1 2018, 17-41 these "other nations", which are "actually using Serbian as their means of communication." 16 The logic of the proposed concept is based on the myth that there is a particular autochthonous, pristine language variety that ought to be nurtured and saved (e.g. in campaigns such as "Let's nurture the Serbian language" (see 3.3), because it represents the key element of national identity.
Every society tends to be organized in the best possible way, including the domain of language knowledge and usage […] one of the main features of civilizations, both ancient and contemporary ones, is their explicit orderliness in all important areas, including the domain of language. (Piper & Klajn 2013: 5-6) It is exactly this approach to language that Makoni and Pennycook (2007) address when stating that the notion of languages as separate, discrete entities is a social construct and that languages are invented; they are invented by states and nations through the production of powerful discourses, i.e. language ideologies. Such discourses or ideologies suggest that some languages or language varieties (dialects) are eligible and some are ineligible for the state, i.e. they are "positioned as relatively 'good, useful, valuable' or 'bad, useless, valueless' within the state system" (Ricento 2006b: 233).
In addition to this, the top-down hierarchy presented in Piper & Klajn (2013) clearly ignores actual linguistic practice and the communicative demands of the speakers. Such abstract construction suggests that speakers are in the service of language, and not vice versa.
A somewhat similar approach, containing the idea of hierarchical dominance between different varieties, is offered in Brborić's (2001) "three-floor building" metaphor. The former secretary of the Serbian Language Standardization Committee uses the "three-floor building" metaphor to describe language stratification into distinct non-standard, substandard, and standard varieties. Non-standard varieties are located on the first floor of the building, urban substandard varieties are on the second floor, and the highest floor is reserved for the most prestigious, most representative language variety, which is highly uniform, consciously nourished and standardized. Even though Brborić himself admits that his "three-floor building" has been criticized as promoting linguicide, with the standard language being "of all others idioms", he dismisses these remarks as part of the standard language mythology, 16 Similar rhetoric can be heard from Serbian statespersons and politicians, and not just language professionals. It is symptomatic that identical phrases and sentences are repeated in the media. For example, a report in the newspaper Danas quotes the current Serbian president Aleksandar Vučić, regarding the Declaration of Serbian identity, which should appear soon in Parliament both in Serbia and in Republika Srpska, reminiscing of the 19 th century language reformer Vuk Karadžić: "Without Vuk Karadžić, we would not exist today, we would not know who we are, nor where do we come from, we would not have our roots -said Vučić, adding that Serbia gained its name and identity in 1847, when Karadžić published a new translation of the New Testament. It is due to this, that we have become a nation, a Serbian nation -stated Vučić, reminding that all great reforms in history started with language. As he emphasized, this is the reason why we will take care, nurture and protect what has been bequeathed to us from Vuk Karadžić, our beautiful Serbian language, illuminating not just Serbia, but everywhere where its people reside. Vučić also pointed out that Vuk Karadžić has urged us not to forget that our future is language itself" (Danas, 09/17/2017). AWPEL Vol. 2.1 2018, 17-41 labeling the criticizing linguists "mythologists" and "turmoil-causers". It is a paradox that the abstract language building is actually mapping the reality -the standardized Serbian language is a prestigious variety and all other varieties are worthless in the public domain and usually associated with the notions of uneducated rurality, and economical and cultural backwardness or lower social classes (in section 3.3. we will illuminate the modus in which this hierarchy affects the ordinary speakers, i.e. the implementors).
As a matter of fact, from all of the varieties of Serbian language, only the socalled Torlak dialect (the non-standard variety located in the South-East of Serbia) is listed by UNESCO as an endangered language with vulnerable vitality. 17 The problem is that the exposed linguicism viewpoint manifests a clear tendency for maintaining the hierarchy and its hegemonistic ideology. In a context where linguistic unification is promoted, it is neglected that this kind of policy "would favour those who already possessed the official language as part of their linguistic competence" unlike those "who knew only a local dialect" and who "would become part of a political and linguistic unit in which their traditional competence was subordinate and devalued" (Thompson 1991: 6).
The preferred status quo is justified by the idea that Serbian linguistics should be on a mission of preserving the authentic, genuine Serbian language. And, again, we are witnessing a preposterous dialectic: the reality in which language stratification is associated with extra-linguistic context has been declared a myth, while the myth of one special, dominant, and legitimate language is promoted as a fact. Nevertheless, since these statements come from a legitimate authority, the statements themselves are taken as legitimate.
It is interesting that one finds similar policy models in the post-Yugoslav terrain, for instance, in Croatia (Kordić [2010], Bertoša & Skelin Horvat [2012]). Kordić (2010) identifies a certain matrix, in which all authors (language authorities, codifiers) use the same strategies, arguments, and stereotypes in order to legitimize the changes in the standardized language, or to motivate the speakers to use certain language phrases, by creating a specific moral panic. "Moral panic is created when, all of a sudden, too much attention in the public discourse is given to a certain phenomenon, when these problems are discussed in an obsessed, moralizing and alarming fashion, as if bespeaking of a coming major catastrophe" (Busch 2004: 82 as quoted in Kordić 2010). Analyzing the nationalistic discourse of the "language authorities" and the initiatives for "saving the language" (for instance, the campaign "Let's nurture the Serbian language"), oriented to the educators and implementors, one reveals the same mechanisms for triggering public moral panic concerning language and its purity, and all based on its deep connections to nationhood.
At this point, we could gain a clearer insight into the power mechanisms related to language: Symbolic power (...) is a power that can be exercised only if it is recognized, that is, misrecognized as arbitrary. This means that symbolic power does not reside in 'symbolic systems' in the form of an 'illocutionary force' but that it is defined in and through a given relation between those who exercise power and those who submit to it, i.e. in the very structure of the field in which belief is produced and reproduced. (Bourdieu 1991: 170) AWPEL Vol. 2.1 2018,  As observable, the shared belief (doxa) that some variety is a priori dominant and that such language hierarchy is natural, demanded, and necessary is the key for understanding the power relations in language. This belief enables the invisibility of power, but it enables its recognition, as well, thus symbolic power is an 'invisible' power which is 'misrecognized' as such and thereby 'recognized' as legitimate (Thompson 1991: 23). Therefore, the existence of a shared belief ensures that the particular set of language practices promoted as dominant does not appear as if it was chosen by some specific groups to protect their interests or promote their ideologies. For instance, the mentioned "three-floor building" is explained as some sort of solution for the potential "language misunderstandings and conflicts", useful in "helping the broader interested parties to easily digest the whole thing" (Brborić 2001: 150).
This mechanism is at play particularly in so-called standard language cultures. Speakers in standard language cultures believe it is crucial that their language exists in a standardized form (Milroy 2001;Milroy & Milroy 1999;Milroy 2007). Moreover, in these cultures speakers' language attitudes are conditioned by powerful ideological positions, even though they "believe their attitudes to language to be common sense and assume that virtually everyone agrees with them" (Milroy 2007: 133).
Therefore, even though speakers of one language community use language in different ways, they do have to be in agreement about evaluative norms: The speech community is not defined by any marked agreement in the use of language elements so much as by participation in a set of shared norms; these norms may be observed in overt types of evaluative behavior, and by the uniformity of abstract patterns of variation which are invariant in respect to particular levels of usage. (Labov 1972: 120-121) In other words, in standard language cultures, the dominant and legitimate language is taken for granted as a common-sense concept. The very notion of "linguistic correctness" (the belief that the standard variety is the only correct language version) is most often simply accepted by the majority of speakers in a speech communitythis is a proof of its recognition and legitimacy.
Considering the above, we can reach the conclusion that language standardization consists of the imposition of uniformity. Actually, standard language can be described "as the consequence of a need for uniformity that is felt by influential portions of society at a given time" (Milroy & Milroy 1999: 22). Beside the principle of uniformity, Milroy (2007) singles out the following important characteristics of language standardization: the notion of correctness, the importance of authority, the relevance of prestige, and the idea of legitimacy. All of these features were identified in the Serbian language policy. Always on the edge of initiating public moral panic, language authorities are in a desperate need of enemies, so that their "protective" role in society can become prominent. After the "dissolution" of the Yugoslav federation, left without visible enemies at the local language market, language authorities oriented their moral panic attack of linguicism towards language minorities in the South, who had already been "stamped" as economically and culturally deficient (see 3.2). The spontaneous process of language change, present also in the (standardized) northern varieties, is observed as another kind of "balkanization" of the language, and as if it were managed by some well-organized and ideologically oriented AWPEL Vol. 2.1 2018, 17-41 market force coming from the South. Such declarative concern about the standard language is both excessive and unfounded: even more so since the variety that is understood to be a threat to the Serbian language (Torlak) is a language listed by UNESCO as endangered.
Building upon the considerations outlined above, in the next section we will present the brief, twenty-year-long history of the Serbian Language Standardization Committee and show how language authorities gradually gained power. This is important for conceiving the broader power relations established between the SLSC and the other two groups present on the language market, namely, educators and implementors, to whom we will turn next.

Serbian Language Standardization Committee (1997-2017)
When dealing with the issue of authority and de jure and de facto power, the actual power held by the Serbian Language Standardization Committee (SLSC) in the first decade of its existence (1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007) was minimal and its role really boiled down to that of an advisory council. In 2002, in the midst of the activities of the Commission for drafting the constitutional charter of the State Union of Serbia and Montenegro, SLSC was timidly recommending its own services to the current highest political figures, "considering that the legal act should be exemplary both from a linguistic and a stylistic point of view" (SLSC Documents 2002: 139). In these notes, members of SLSC express the hope that "their initiative will be interpreted correctly and that their offers will be willingly accepted" (ibid.).
It the middle of this ten-year period, Egon Fekete, a member of SLSC, raised the question of SLSC's institutional purpose: "Is the Committee a group of linguists publishing books or a broader forum with visible and recognized influence in the 'language culture' domain?" (SLSC Documents 2002: 9). Caught in the myth that the media promote bad language (Bauer & Trudgill 1988), Fekete expresses fear of the massive influence of television -"the announcers produce grammatical errors and prosodic/stress mistakes; much frivolousness is at play in the language domain, chaos almost" (ibid.). He advocates a broad campaign for visibility in the system that, otherwise, excludes influence on the "public word", but he also thinks that prominence in the public discourse implies certain conditions. First of all, SLSC members should establish close relations with political authorities -"we must establish a direct relation with the authorities, (as) we cannot allow this to remain at the level of proclamation alone, but rather at the level of personal contacts" (ibid.). The following years saw remarkable activity. A dozen letters, signed by the SLSC members, were addressed to the most important current political figures (the prime minister, the president, the ministry of education, etc.), resulting in frequent meetings with political authorities.
The Committee members admit that they are trained to implement the tasks of language policy and planning, including language standardization itself. However, they point out that they should not be held responsible for the "communicative functioning of the language in the native countries"; they further express hope that the collaboration between the Committee and the distinguished political institutions and figures will be advanced in the future (SLSC Documents 2002: 21-28).
It is important to note that, initially, Committee members used to describe themselves as volunteers-"without remuneration, because we already carry out a lot AWPEL Vol. 2.1 2018, 17-41 of activities without asking for money" (SLSC Documents 2002: 139). Once preliminary steps were taken to establish close relations with the authorities, an elegant method for financial benefit was designed. For instance, book items from a list approved by the Committee (all of them authored by SLSC members) were offered with a special 35% discount if purchased by the Ministry of Education and Sport, the Ministry of Religion, the Ministry of Culture and Media, the Ministry of Science and Environmental Protection and the Ministry for Diaspora (SLSC Documents 2007: 49-53). During 2006, considerable financial support was requested and obtained from the same sources. This indicates that established power legitimacy, in the sense of Milroy (2007), is converted and literally cashed in under the flag of "the mission of language maintenance and revitalization". Ultimately, this makes Bourdieu's "market" metaphor applicable to the situation in a banal way.
As Blommaert (2015: 9) points out, Bourdieu's important social-theoretical legacy is the "ethnographic invitation, in which longitudinal and slow processes of social structuration can be read, followed and appraised, so to speak, through the lens of register development and change in actual moments of social interaction". In the following two subsections we will assume an ethnographic approach in order to reveal the power relations established between codifiers, an essential, formative force, and educators and implementors, considered by the codifiers as undifferentiated groups of individuals who should be "handled" and trained in order to become users and/or instructors of correct language.

Educators
As discussed in 3.1.1, standard language, i.e. the consciousness thereof, to be more exact, needs to be "maintained and protected through authority and doctrines of correctness" (Milroy 2007: 138). A precondition for success of these processes of symbolic power is the belief in the legitimacy of both power and those who hold it.
The educational system has a key role in the processes leading to the construction, legitimation, and imposition of an official language.
According to Bourdieu and Passeron (1970), the educational system imposes the legitimacy of the dominant culture on the members of dominated groups or classes as a fait accompli. Thus, the aim of teaching a mother tongue is: to make them [those dominated groups or classes] internalize, to a variable extent, disciplines and censorships which best serve the material and symbolic interests of the dominant groups or classes when they take the form of self-discipline and self-censorship. (Bourdieu & Passeron 1970: 40-41) This kind of symbolic violence, legitimate because it is implemented by recognized legitimate authorities, enables the reproduction of the structure of power relations within a social formation (Bourdieu & Passeron 1970). Since pedagogical activities promoting the language and culture of dominant groups (e.g. the majority ethnic group) are institutionalized, that means that the path of displaying power is institutionalized and legitimized. As May (2011) observes, legitimation (as formal recognition accorded to the language by the nation state) and institutionalization (as the process by which the language comes to be accepted) are two main aspects of establishing a common national language which enables both linguistic and cultural homogeneity. This way the national (official, standard) language "becomes an arm of AWPEL Vol. 2.1 2018, 17-41 the state" (May 2011: 154) and the educational system becomes the means by which the nationalistic principle "one state-one culture-one language" is achieved. Furthermore, the aim of the state is to integrate all speakers into a single speech community that is a product of political domination, "reproduced by institutions capable of imposing universal recognition of the dominant language" (Bourdieu 1991: 46).
In this respect, the system of education is involved in "expressing or challenging particular kinds of relationships, value positions and identities" (Maybin 2007: 157), i.e. the classroom is a site of inculcation where teachers show students what is valued in society. Therefore, when it comes to language, school is where students learn how to use language in ways that are recognized and ratified by institutions.
In the previous section, we have sketched the position taken by the so-called educators within the framework of normative linguistic discourse -they are tasked with disseminating and implementing codifiers' solutions (this means verifying and correcting language "errors/mistakes"), but they also have a mission to inform those they educate about the significance of the standard language norm (Piper & Klajn 2013: 9).
Filipović and Vučo (2012) demonstrate how such strict top-down language policy can influence language education policy. As already mentioned, in 1997, after the breakup of Yugoslavia, SLSC was established and "has engaged in publication of a number of documents regarding the political and historical relevance of Serbian within the regional (former Yugoslavian) context" (Filipović & Vučo 2012: 16). The focus of these publications is the prescribed Serbian standard (folk Eastern Herzegovian variety), i.e. the national variety which "has always been a direct reflection of a consensus reached in the second half of the 19 th century, when the nationalist model of language standardization was applied" and "which has never become a native tongue of the speakers in major Serbian cultural centers" (ibid.). In accordance with top-down language policy, language education policy has been carried out as top-down language management in which a small group of experts (appointed by institutions of the state or institutions of higher learning) decides what language variety should be taught in schools, which approach to teaching Serbian as L1 or L2 should be applied, and what is the appropriate selection of teaching materials and literary works that are included into the obligatory national curriculum for Serbian as L1 and L2. (ibid.) Filipović and Vučo highlight the discrepancy between official language policy (which promotes one idealized language variety, almost nonexistent in reality) and real-life language practices and communicational needs.
This strict, non-flexible approach to language present in the Serbian educational system has numerous consequences. As already explained, if one particular language variety is "correct", all other varieties will be evaluated in comparison with it. Since the standard language variety is labeled as prestigious, all others varieties, and of course their speakers, risk being stigmatized. Thus, standard language, in most socially stratified societies, is associated with the most cultivated groups or higher social classes, while non-standard varieties are "incorrect", ugly or even vulgar, usually linked with the notion of cultural backwardness, attributed to the uneducated lower classes. In this respect, language displays the differences existing within one society, but also shapes, cements, and reproduces those differences which "most frequently AWPEL Vol. 2.1 2018, 17-41 go hand in hand with differing degrees of access to material resources, to knowledge, to power" (Montgomery 1995: 64).
Therefore, the school system tends to reproduce structural disparity due to unequal knowledge of the legitimate language, but mostly it tends to reproduce uniform recognition of this language (Bourdieu 1991: 62). In addition, the standard language is presented as some kind of mystery or, as Milroy (2007: 135) has remarked, "the possession of only a few persons (usually not clearly specified) who have the authority to impose the rules of language on everyone else". They fix and codify legitimate usage through publications (dictionaries, grammar books, pronunciation guides, and usage manuals), while teachers, "who have the privileged access to these mysteries" (ibid.), impose and inculcate it through innumerable acts of correction. Thus, as Bourdieu (1991: 61) puts it, "the educational system tends to produce the need for its own services and its own products, i.e. the labor and instruments of correction".
The consequence of such standard language ideology is spreading the belief that individuals from certain speech communities are not capable of performing certain tasks and professions, on the basis of their alleged low standard language competencies. Such dogma is particularly widespread among educators, i.e. among the groups tasked to disseminate this language ideology through the school system. For instance, Petrović (2015) gives the example of a Belgrade university professor who supposedly advised a student that "given the fact that she is from Leskovac" (a municipality in Southern Serbia, located in the Torlak dialect zone), "she should not study the Serbian language" (Petrović 2015: 15). Moreover, Petrović quotes how the same professor declared that all people from the town of Leskovac are not real Serbs (given the immense language differences between the non-standard Torlak variety and standard Serbian). Petrović illustrates the doxa present among educators with an example of another university professor, in whose opinion Serbian language professors who are native speakers of non-standard old štokavian dialects almost never use the standard idiom in the classroom, "because they cannot learn it, and therefore they are bad teachers" (Petrović 2015: 16). In Piper & Klajn (2013) parents are also listed as educators. This implies that: firstly, parents are aware of the importance of competence in the standard language variety (as they adopt the standard language ideology); and secondly, that they already possess the required language competencies and can transmit them to their children. This type of "eyes closed" viewpoint (or, to quote Mey (1985: 26), "sociolinguistic blindness"), which puts aside diversity and differences among members of a language community (cf. Blommaert & Rampton 2013), in fact, creates perfect conditions for reproduction of these differences, i.e. existing power relations and, conditionally speaking, conflicts among different social groups. In such circumstances, where the educational system imposes the legitimacy of the dominant groups, for those children speaking a standard variety "the home-school link is close", while for the disadvantaged children "there often exists a marked home-school discontinuity, arising from that contact which is itself central to the very idea of disadvantage" (Edwards 2009: 126). This means that the linguistic capital of some children would be less valued right from the start of their education and that it can affect their performance and achievements (their linguistic competence even might be seen as linguistic deficit, or as a "restricted language code", as Bernstein (1972) AWPEL Vol. 2.1 2018, 17-41 argues). Moreover, since the standard language variety is prestigious, some parents even choose schools for their children in accordance with this criterion. For instance, Petrović (2015) quotes a conversation among a group of Belgrade parents on the advantages and disadvantages of private and state kindergartens in Serbia. As a disadvantage of choosing a state kindergarten over a private one, a certain parent singled out the non-standard "Southern" dialect, allegedly used by teachers in state kindergartens. In this context, the "Southern" dialect is a symbol of unacceptable deviation from the norm, associated with kitsch and trash culture, as well as uneducated rurality, thus the parent is ready to invest more money in order for children to learn normal language. In this regard, a mother tongue is literally commodified (Illich 1981).
As a matter of fact, the particular set of shared beliefs is actually a reflection of the discourse created and launched from above. Therefore, we can find similar messages delivered by so-called language leaders, i.e. undisputed language authorities. For instance, Pavle Ivić, 18 a renowned Serbian dialectologist and phonologist, notes that acquisition of the standard language for Southern dialects speakers is "an embarrassing process" and "a task that, as a rule, can never be carried out completely" (Ivić 1998: 11). So, again, we are dealing with a rather paradoxical situation. On the one hand, the SLSC members recommend that all speakers within the state should be competent in the standard language and use it in the public domain, because "Under better social circumstances, the individuals characterized by the knowledge of and respect for language norms could acquire greater public reputation and benefits, as is indeed the case elsewhere, particularly in the developed countries" (Decision No. 1 of the Committee, 1998); whilst, on the other hand, there is a widespread opinion that speakers from certain regions are a priori deprived of the ability to acquire the standard language (especially, in the media (Petrović 2015)).
The educational system promotes these contradictory facts side by side. Moreover, language planning and policy in Serbia, including language education policy, are focused on the creation of intensified measures for "minimizing the deviations from the language norm"; they are focused on the actions that will indicate the language errors/mistakes, as well as on improving methods ways for their correction. These actions are grounded on the policy of exclusivity, insensitive to both the interlinguistic and extralinguistic issues, but still ideologically oriented policy, which aims "to unearth" all the proof of the existence of one pristine, idealized Serbian language and all the reasons for keeping and maintaining this particular language version. In other words, language policy and planning are rather interested in "linguistics products" (grammar books, dictionaries, teacher manuals, seminars and trainings for teachers and lectors etc.), i.e. interested in initiatives which will intensify the existing pressure on the speakers (who are already convinced they are not competent in their mother tongue) much more than in adapting the prescriptive decisions and recommendations in accordance with speakers' demands and/or convenience. AWPEL Vol. 2.1 2018, 17-41 The findings of the ethnographic research of the Serbian Torlak dialect stigmatization in the linguistic market conducted by Stefanović and Stanković (forthcoming) show that both university professors and students, native speakers of the dialect, believe that their chances on the labor market outside their hometown or outside the region are lower because of their native non-standard dialect. A subject in the research, who happens to be a student of journalism, underlines that, for her profession, language background is almost essential: "If I want to work in some of the Serbian major media corporations, such as RTS [Serbian Public Broadcasting Service], Pink or so, using the correct language is very important, and this could stop me, because people from the North [area where vernacular close to the standardized variety is spoken] would surely have an advantage, and I probably wouldn't get the job." In her opinion, if she had the opportunity to be more surrounded with the standardized variety, she would have probably "start to talk correctly" after some period of time. Unlike the other participants, she is convinced that everyone can learn the standard language form, and that it only takes time and great effort. Therefore, she thinks that people from the North definitely have a language advantage at the job market, as this "gift" is delivered to them "by birth". As she explained, she volunteered on a local TV station as a newscaster, and found herself having doubts about the correct way to pronounce every third word in the text of the news. So, she was constantly talking on the phone to her sister (who possesses a master of philology degree), just to make sure that she would not embarrass herself and make an error, based on the dialect or some hypercorrection. Nevertheless, the same speaker thinks that if she would try to use the standardized language in everyday life, she would appear "sophisticated".
Unlike this participant, the majority of the subjects in the survey believe that they cannot learn the standard variety (particularly, its prosodic/stress system), that they will always be labeled as incompetent in the desired language competencies, and therefore, that their other professional competencies and qualifications will always be less valued. Thus, one cannot overlook the dialectical relation between the school system and the labor market. 19 To sum up, the educational system is essential for establishing consciousness about the standard language, the correctness doctrine and the language authorities. School is the place where the correct and pure way of using the language is learnt, where one should try to master the langue ideal that is recognized as correct and pure. It is in school where one learns about the sociolinguistic status of one's native variety, spoken at home, and its value and acceptability in the classroom. This opens a door for a massive abuse of the non-native speakers, reflecting on their potential academic or public career. It is regularly followed by contradictory assessments that, on one side, everyone should learn the standard variety, but, on the other side, that some speakers are deprived of gaining this skill, by birth. In some cases it leads to public discrimination by prominent figures, such as university professors, but this is, AWPEL Vol. 2.1 2018, 17-41 unfortunately, generally approved by the educational system. We saw that, as a consequence, most speakers from the Torlak zone feel unsecure, and not trained enough to use the standardized language.
In the following subsection, we will take a closer look at the relationship established between the codifiers and the implementors -first, to learn what initiatives and public campaigns are conducted by the language authorities to make an impact on the everyday speakers, and second, to present what kind of strategies are implemented by the common speakers of the Torlak dialect, as an individual type of reaction to the stigma they are exposed to.

Implementors
In the circumstances described on the previous pages, where language has its value and price, created in relation to the market, and the market itself is established from above (from the authorities), the linguistic exchanges might be compared with the economic ones. In the Bourdieusian sense, every linguistic exchange or interaction expresses relations of power, no matter how innocent it seems (again, Bourdieu would say "there are no longer any innocent words"). The structure of some specific field has its own rules, i.e. the field itself "governs expression by governing both access to expression and the form of expression" (Bourdieu 1991: 138), and, as we already demonstrated in subsection 3.2., the key processes present are censorship and selfcensorship. Speakers are first taught and trained within the family, and then through the educational system and media what is "good" or "correct" language usage, i.e. what is legitimate language usage. Through these usage models and sanctions, speakers learn the value of linguistic products in different markets (e.g. school as a market). Moreover, Bourdieu emphasizes that "the sense of the value of one's own linguistic products is a fundamental dimension of the sense of knowing the place which one occupies in the social space" (Bourdieu 1991: 82), because this "sense of place" governs the degree of constraints within one field. The social position of the speaker governs his/her access to the language of the institution and the symbolic profit which he/she ought to obtain determines the manner he/she uses language or which language code she will use in a certain linguistic market, i.e. context. Therefore, language usage includes two aspects: the social one, i.e. the ability to identify what context or situation demands which language code (to know what is socially acceptable), and the linguistic aspect, the language knowledge (e.g. grammar), i.e. the capacity to produce language for a particular situation or market. The combination of these two elements -which, as we have already explained, both depend on numerous factors, such as socioeconomic status, cultural capital, geographical position, ethnicity, level of education etc.-determines the successes or profit on the linguistic market.
Considering the above, in this subsection we will see how certain concrete measures created by the language authorities affect the linguistic market, therefore the implementors' behavior in the market. Specifically, we will focus on those measures and initiatives launched from above in order to intensify the existing pressures and constraints for the speakers within certain fields (university, school, TV and media…), but special attention will be given to those language authorities' initiatives or messages that even tend to regulate the private domain of speakers' linguistic exchanges, outside the public domain. Finally, we will present certain AWPEL Vol. 2.1 2018, 17-41 language practices or, more precisely, three dominant language strategies that nonstandard variety speakers develop in order to gain symbolic profit in the language market.
It is interesting to follow how slowly, but gradually the pressure of the language authorities in Serbia proliferated to the public and the private domain. The constraints in both areas have been intensified for the last few years. Namely, in the second decade of its existence (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017), SLSC has implemented activities more oriented to the general public discourse. As has already been presented in subsection 3. 1.2., in 2002 Fekete raised the question about the position of SLSC as a broad forum with visible and recognized influence in the 'language culture' domain, encouraging all activities towards establishing tight relations with the political authorities (SLSC documents 2002: 9). Once this was accomplished, SLSC could concentrate on the "public word" for gaining immediate influence on the language employed in the public discourse. This is to be achieved by massive campaigns, such as "Let's nurture the Serbian language" (launched in 2015).
The campaign was launched by the Belgrade Secretariat for Culture, Faculty of Philology and Belgrade City Library, supported by the Serbian Public Broadcasting Service (RTS). It aims to inform "school, the non-school education domain, and media as well, that taking care of the language is a serious and responsible task" (Head of the Department of Serbian Language and Literature at the Faculty of Philology in Belgrade -official statement from November 29 th 2015, Serbian Public Broadcasting Service 20 ). Besides, the reason for launching the campaign -in which "a large number of wellknown people from public and cultural life promote the correct answers to some common language puzzles" (the official statement of the Minister of Culture and Information, 24.2.2017. [our emphasis])-is the myth that Serbian language is endangered.
In this particular project, a number of popular public figures are used as an impetus to obtain public attention. Via a series of posters and TV clips, the broad audience is informed with a list of everyday language "mistakes" or "misuses". The general nature of the synopsis in this material is supposed to be attractive, entertaining and didactic, at the same time. For triggering humorous effects, there is an intentional turnover in assigning the roles in the scenario. Namely, a street cleaner; an old, rustic saleswoman; and a famous artist with Albanian and Roma ancestry -all of who are "supposed" to use non-standard language varieties-are the one to correct the mistakes in the speech of the everyday character played by the famous actor Svetislav Bule Goncić. Therefore, the source of humor is letting down the stereotyped expectations about certain speakers, their geo-socio-economic background, and their deficit language capital at the language market. Although one could argue the opposite, this type of texts can be perceived as discriminatory, illustrating latent linguicism, as they are based on the shared belief that some variety is a priori dominant and that such language hierarchy is natural, demanded and necessary.
It is rather complex to determine and evaluate the effects of the conducted advertising activities, especially on the "Serbian language maintenance and revitalization mission of SLSC", but what seems more interesting is the modus in which AWPEL Vol. 2.1 2018, 17-41 the reputation of mass media public figures is employed in order to confirm the doxa in the given field, and to corroborate the established evaluative standpoint on language. The point of absurdity was reached once an international movie star such as Monica Bellucci, a non-speaker of the language, has appeared on an advertising poster, holding the official Orthography instruction book.
The conception of the analyzed campaign conceals an encroachment of the language right to use any desired variety in the intimate/personal space domain. Namely, Brborić is declaratively advocating for the freedom and right of the citizens to switch to the language variety that suits their communication needs.
Given the fact that within a certain area there is more than just official and public life domain […] -including the semiofficial, semi-public, as well as the non-public, private, personal, family, and various non-formal-group life domain, i.e. urban, semi-urban and rural life domain, the citizens have the freedom, and right, as well as the need to use not just standard language, but also other language varieties, maintaining them and expanding them, "changing" and "developing" them-"from here to eternity." (Brborić 2002: 87) Nevertheless, more than half of the scenarios in the "Let's nurture the Serbian language" campaign are based on a dialogue between a "bad" and a "good" speaker carried out via SMS, mostly between friends and partners, which indicates that even the private life domain is not isolated or protected from the impact of the well-known viewpoint of the doxa. Even while expressing an apology or presenting a plan for lunch in an intimate or confident manner, your co-locutor has the right (obligation, even) to correct your speech, for your own sake (actually, your market capital sake), for the sake of maintaining the language and culture you share, all the way to the stability of the entire society and state.
Language differences play an important role in revealing information about an individual. Particularly in urban societies, as Gumperz (1982) observes, these assumptions about values of different languages (language varieties or speech styles) can be crucial for the quality of an individual's life -whether it comes to formal (e.g. employment interviews, labor negotiation, taking exams at university etc.) or informal situations, an individual presents himself/herself by and through language. In this regard, language can be understood as a recognition code (for instance, a symbol of one) or a password that can provide the access to some groups.
The findings of the qualitative research (Stefanović & Stanković forthcoming) conducted among Torlak dialect speakers also reveals that the sense of value of one's own linguistic products -Bourdieu's "sense of place"-has a great influence on peoples' everyday social interactions. The so-called "Southern" dialects are stigmatized as "ugly" and "incorrect" due to discourse and the strict policy of the language authorities, and this image is perpetuated through the media, TV shows, movies, popular music etc. (Petrović 2015). Also, these varieties, associated with the lower economical and cultural development of the South Serbia region, have become a symbol of backwardness, and therefore the speakers of these dialects are labeled in the same way (Petrović 2015;Paunović 2009). Aware of such status of their dialect, the Southern speakers develop different types of strategies (in accordance with their linguistic capital -capacity to produce expressions for a particular market)-always taking into account several facts: who is their co-locutor (whether it is a member of their language community or not), where is the communication act performed (in their AWPEL Vol. 2.1 2018, 17-41 hometown or not), what is the nature of the communicative act (formal/informal). Considering the listed factors, speakers decide when to use the standard variety, and when it is more profitable to use the local dialect. These strategy types are also conditioned by the degree of speaker's psychosocial conformity or their willingness to be conformed to the socially acceptable variety.
The identified strategies among Torlak native speakers are labeled as mimicry, ignoring, and the eclectic strategy (Stefanović & Stanković, forthcoming). Mimicry implies rejection of regional features and the use of standard variety. This specific strategy can be perceived as an ethical act or simply as an attempt to profit on the linguistic market. Speakers who perceive this strategy as an ethical act state "I speak correctly, therefore, I preserve the purity and beauty of the language". Unlike these speakers, those using this strategy as a means to forming and maintaining a better public image, i.e. profiting on the linguistic market, state that "when I address the standard language speakers, I aim to be understood, not ridiculed". The ignoring strategy implies the retaining of the local language features. Within the group of speakers employing this strategy we differentiate those who cannot speak differently ("I don't know how to speak in any other way"), and those who do not want to speak differently ("in spite of every rule, I speak the way I like to"). Finally, the eclectic strategy is the most complex, as it implies the use a of whole spectrum of language variations, from standard to local ones, accompanied by developing a novel, supralocal and supra-regional variety. The speakers utilizing the eclectic strategy are fully aware of the different language varieties, their language features and markers, society prestige and communicative value. There are two ways of applying this strategy. In one approach, it implies an implicit transposition of the power pattern, relating specific varieties with the social identities they construct when used ("There is a wellknown, unwritten rule when to speak in which way"). In the other approach to the eclectic strategy, the speaker is consciously and intentionally disrespecting and breaking the established sociolinguistic formations, fields and doxas, by using a specific "misplaced" variety game ("Sometimes I purposely speak 'incorrect' in a formal situation").
Therefore, the market interest leads the game. Those who are familiar with the market, but also possess the capital, i.e. competence in the standard language, and know how and when to invest it, are the ones who will make a profit. One could anticipate that non-standard variety speakers will always try to change or correct their speech in interaction with those whose mother tongue is the variety which has been chosen as the basis for the standard language. But, what seems more interesting is to observe how the standard language ideology affects the communication among speakers from the same non-standard variety language community. Some of the participants, native Torlak dialect speakers, state that they are aware of the power of the standard language and know the standard language usage can bring some benefits. Therefore, they try to use it in formal situations or in the public space, not because they worry about language correctness (or because they think that Serbian language is endangered), but because they are aware of the symbolic power of that correctness. For instance, one subject states that he "always uses the standard variety in the post office". As a consequence, in his opinion, the clerks think he is a "polite" and "civilized" man and not "some peasant" like the rest of the clients, thus he can pay the bills skipping queues and waiting lines. "Maybe they think I am from Belgrade, AWPEL Vol. 2.1 2018, 17-41 I don't know, but this always works", states the subject. Another participant declares that she is aware of the fact that she could use "the two cases and one accent" 21 in court, in drugstore, bakery, like most of the people in her town, without any consequences ("nobody would tell me: Excuse me, lady, but please leave the bakery"), but the usage of the standard variety "can maybe provide me better treatment, maybe I can get a better service". There are even documented cases of relationship breakups initiated by variety differences, as the partner speaking the standard language idiom "was constantly being embarrassed by the partner's dialect speech" -this is a proof that language ideology can interfere even with the non-public, private, intimate domain. On the other hand, it is interesting that speakers that employ the eclectic strategy often decide to use the non-standard variety in similar environments, such as the market, because the common dialect can bring the speakers closer together, set an intimate tone, and thus the speaker can expect benefits in the market exchange relationship. Yet, in the very same settings the speaker can decide to switch to the standard variety in order to show symbolic power over the co-locutor, and express his/her hegemonic standpoint ("That is how I put the peasant in his place").
In this subsection, we identified the speakers, or implementors, utilizing the language ideology to gain some symbolic profit. Still, the described language strategies are actually only maintaining and confirming the present power relations. As Bourdieu explains: not only do the strategies of assimilation and dissimilation which underlie the changes in the different uses of language not affect the structure of the distribution of different uses of language, and consequently the system of the systems of distinctive deviations (expressive styles) in which those uses are manifested, but they tend to reproduce it (albeit in a superficially different form). (Bourdieu 1991: 64) As has been concluded in this final subsection, the two-faced relationship between the language authorities and everyday speakers can be interpreted as an interplay between an organized, ideologically based campaign for implementing the language standard, on one side, and various, formally and axiologically differentiated strategies that are self-conducted, and depend on a rich and dynamic set of factors and variables: who are the interlocutors, where is the communication act performed, in what circumstances, what is the topic etc. This way, we conclude investigating the cycle of power path that is initiated and enforced by the language codifiers, directed towards the educators and implementors, followed by an overview of the reverse "antistandard" speaker mechanism and the reactions of the latter two groups, as witnessed in their own words and experiences.

Conclusion
The paper examined the discrepancy between the minor executive power held by official Serbian language authorities and the considerable informal influence it has on Serbian society and public discourse, based on powerful mythologies regarding language and language policy. The myth of the language-culture-state chain leads to the conclusion that ignoring the decisions and recommendations of the Serbian AWPEL Vol. 2.1 2018, 17-41 Language Standardization Committee can result in social instability and anarchy, as "every society tends to be organized in the best possible way, including the domain of language knowledge and usage" and given the fact that "one of the main features of civilizations, both ancient and contemporary ones, is their explicit orderliness in all important areas, including the language domain" (Piper & Klajn 2013: 5-6). We demonstrated that, although "prescriptive tendencies of Serbian planners are most commonly explained as purely scientific decisions based on particular linguistic, structural features of Serbian" (Filipović 2012), the standard language culture in Serbia is often based on the myth that the language industry contributes to the gatekeeping function of social institutions. We saw that Serbian standard language culture is founded on a nationalist ideology, not unlike the dominant language ideologies in 19 th century Europe. A differentia specifica of Serbian language policy is its orientation towards the so-called balkanization of the language. On one hand, it is the "battle" against the tearing of Serbo-Croatian into many languages, which is observed as a threat for the status and the name of the language, the main symbol of nationhood. As a direct reaction to this threat, the Serbian Language Standardization Committee was established as a trans-state, national institution dealing with vital issues of language policy and planning. On the other hand, the "balkanization" of the language is presented as a novel threat to the standard language from within, i.e. from speakers of non-standard varieties, given their natural language deficit and/or mere indolence.
In the context of a rigid, strict educational system, standard language culture is inseminated by educators through constant promotion of censorship and a heightened understanding of the benefits of self-censorship in the language market. This can be seen in the pressure exerted on certain speakers and the threat to their status in the labor market. Finally, in order to reveal the point of view of the common speakers, we presented certain strategies they deploy in a situation when they are expected to use the standard. These strategies can be enriched by different axiological points of stance towards the standard language. Nevertheless, we witnessed the fact that they maintain and reproduce established power relations, sustaining nationalist ideology, which is advocated by the political elite and promoted by the codifiers, resulting in the hierarchical dominance of certain prestigious varieties over stigmatized ones.