Language ideologies in the mirror of some Serbian students’ attitudes towards variation and change in British English

Popular beliefs and value judgements about language underlie attitudes that may reflect various language ideologies. In this light, the paper presents some results of two questionnaire-based surveys into students’ overt attitudes towards varieties of English. The first survey was carried out in 1997 amongst first-year students of English at the English Department of the Faculty of Philology, Belgrade University, while a replication survey was conducted in 2016. The primary aim of this comparative study is to assess whether any ideological and attitudinal shifts may have occurred in the intervening nineteen years. The main focus of this paper is on the results related to attitudes towards varieties of British English and towards language change. The comparison of the results has indicated that, although a lower percentage of the 2016 survey respondents expressed stereotypical views of some social and regional varieties of British English, their attitudes towards language change are very similar to those expressed by the participants in the 1997 survey.


Introduction
English is used by hundreds of millions of people across the world either as a mother tongue or as a second or foreign language. As a language of global communication, English is also the object of various perceptions and attitudes held by both native and non-native speakers. Socio-psychological images of English and its manifold varieties may thus reflect diverse social, cultural, historical and geopolitical dimensions.
Broadly defined, language attitude is "an affective, cognitive and behavioural index of evaluative reactions toward different language varieties and their speakers" (Ryan, Giles and Sebastian 1982, 7). Since the 1960s, language attitudes have been widely researched using a number of investigative techniques within the three following approaches: the content analysis approach (i.e. the analysis of societal treatment of language varieties), the direct approach (questionnaires and interviews) and the indirect approach within which the matchedguise technique (MGT) and the verbal-guise technique (VGT) are used for implicit attitude measurement (for an overview see Garrett 2010).
Numerous language attitude studies, employing the matched-guise or the verbal-guise techniques to measure non-linguists' evaluative responses to various speech styles (mostly accents), have shown that speakers of standard English are rated significantly higher on the evaluative dimension of social status (e.g. intelligence, competence, social class), but not always on the evaluative dimension of solidarity (e.g. pleasantness, social attractiveness, trustworthiness), while speakers of non-standard varieties are usually (but inconsistently) rated favourably on the evaluative dimension of solidarity (for overviews see Giles and Billings 2004;Marlow 2011, Fuertes et al. 2012).
Similar results have also been obtained in other empirical studies that have used questionnaires and interviews for direct elicitation of explicit attitudes towards varieties of English (see e.g. Coupland and Bishop 2007), as well as in some perceptual or folk dialectology studies that have employed mapping techniques, content-oriented (discourse) analysis of conversations about language-related issues, and various other tools (see e.g. Niedzielski and Preston 2000;Preston 2007;Montgomery 2012). Thus, as Preston (2016, 182) points out, "attitudes towards and folk beliefs about language are not isolated instances, but reflect patterned and structured ideologies within cultures and speech communities".

Language ideologies
The study of language ideologies also referred to as 'linguistic ideologies' and 'ideologies of language' is a separate field of linguistic-anthropological inquiry that emerged in the last few decades of the 20th century. It has since influenced many sociolinguistic studies (for a survey of such studies see McGroarty 2010).
It was Silverstein (1979, 193) who first defined the concept of "linguistic ideology" as "any sets of beliefs about language articulated by the users as a rationalization or justification of perceived language structure and use". He maintained that linguists could not properly explain linguistic changes without taking into consideration speakers' awareness and conceptions of their language. Since then, many other definitions and explanations of various language ideologies have been postulated (e.g. Milroy and Milroy 1985;Irvine 1989;Irvine and Gal 2000;Lippi-Green 1997;Woolard 1998;Wolfram 2001;Kroskrity 2016).
Being socially and culturally constructed, language ideologies are not uniform and fixed, but variable and often contested. They are also internally di-L ... 243 versified and thus "inherently plural" and "multiple" not only "because of the plurality of meaningful social divisions (class, gender, clan, elites, generations, etc.) within sociocultural groups" that may produce divergent perspectives "expressed as indices of group membership", but also because of "continuous influence of nonindigenous ideologies" (Field and Kroskrity 2009, 6). Thus, different language ideologies offer different social, political and moral values and beliefs about language, particular languages, language varieties, language use, language structure, etc. This means that there may be a number of different, and often conflicting, language ideologies within a community. More often than not, however, one language ideology may gain dominance over others. Consequently, this dominant language ideology, though it may be resisted, "will exert an influence on language attitudes and the way in which language structure and language use are thought of in the community" (Watts 2000, 33).
In standard language cultures (see Milroy 2001, 530), the standard language ideology, also called the ideology of the standard language, is dominant. Standard language ideologies may vary from speech community to speech community. Thus, as Lesley  observes, social differences based on class are more salient in the standard language ideology in Britain, whereas those based on race and ethnicity are more salient in the standard language ideology in the USA. However, almost all ideologies of the standard language share the belief that the standard variety is the most correct and the best. Hence, this variety is also often perceived as the real language (Milroy and Milroy 1985;Wiley and Lukes 1996). As Milroy (2007, 133) explains, speakers are not usually aware that they are influenced and conditioned by this ideology, which makes them "believe their attitudes to language to be common sense and assume that virtually everyone agrees with them".
The ideology of prescriptivism (see e.g. Watts 2000) and purist ideologies (see Milroy 2005) are closely linked to and go hand in hand with the standard language ideology because the standard variety is based on prescriptive norms of the written language and is seen as uniform, internally consistent and pure. Prescriptivism aids standard language maintenance. Crystal (2006, 67) defines prescriptivism as "the view that one variety of language has an inherently higher value than others, and that this ought to be imposed on the whole of the speech community". He adds that those who use this favoured variety in speech and writing "are said to be using language 'correctly'" (Crystal 2006, 67).
Knowledge of prescriptive norms of Standard English is spread through formal education and the media. The school is the main institution that teaches Standard English and promotes the standard language ideology. The forms of Standard English are also taught to non-native speakers around the world. Since Standard British and Standard American English are the two preferred models in S R. S 244 teaching English as a foreign language (TEFL), elements of standard language ideologies dominant in the UK and the USA may also be, to a lesser or greater extent, transmitted through English language teaching. The normative approach in TEFL renders non-standard varieties practically invisible (see Irvine and Gal 2000, 38). In addition, "histories and grammars of the language usually describe not 'standard English' but simply 'English', as if no other varieties of the language had the right to the name" (Cheshire and Milroy 2013, 14).
Both native and non-native speakers are also exposed to Standard English through the media. Broadcast media, in particular, are believed to "have an impact on metalinguistic awareness of linguistic varieties and variation, standard and non-standard, and the ideologies surrounding them" (Stuart-Smith 2011, 224).
As the study presented below will show, a socio-psychological image of English may reflect a complex interplay of multiple language ideologies.

The present study: Instruments and methods
This study presents a comparative analysis of the results obtained from two surveys into students' overt attitudes towards varieties of English. The instrument used in the study is a paper-based questionnaire designed in 1996. The questionnaire contains both open-ended and close-ended questions as well as a number of agree-or-disagree statements, most of which offer a dichotomous response option (Yes/No).
The first survey was carried out in 1997. The questionnaire was distributed amongst 152 first-year students of English (16 male, 136 female, aged 18-28, M = 19.35) at the English Department of the Faculty of Philology, Belgrade University. At the time this sample formed part of a larger sample (see Stojić 2005).
A replication of the original survey was conducted in 2016. The same questionnaire was administered to 120 first-year students of English (25 males, 95 females, aged 18-24 years, M = 19.27) at the English Department of the same Faculty.
In this paper, the focus is on the results related to respondents' attitudes towards regional, social and ethnic varieties of British English, as well as towards language change.
The main purpose of this comparative study is to assess which attitudes and stereotypes are shared by two generations of first-year students of English and whether any ideological and attitudinal shifts may have occurred in the intervening nineteen years. For this purpose the data were compared and analysed using descriptive statistics. The results are presented as percentages, and where appropriate, as frequencies.

Results and discussion
It is worth noting from the outset that only 23.3 per cent of the 2016 survey respondents (n = 28) reported they had had been to an English-speaking country, most of whom had been to Britain i.e. England (n = 15), while others stated they had visited the US (n = 9), Canada (n = 2) and Australia (n = 2). Although the aforementioned proportion is higher compared to 16.4 per cent (n = 29) of the 1997 survey respondents who had been to England (n = 9), the US (n = 7), Canada (n = 7), and Australia (n = 2), the results show that the majority of students who took part in the two surveys lacked lived, experiential knowledge about any English-speaking country.
For the purpose of this paper, the responses to other questions and statements are thematically grouped into four sets. As the percentage of item non-response was very low, it will only be mentioned where appropriate. When analysing the responses to the statements, only the percentage of those who agreed with them will be presented and analysed.
The first set of four statements refers to the social standing and a utilitarian value of BBC English i.e. Standard English spoken with Received Pronunciation (RP). A high percentage of respondents from both surveys agreed with the statement that 'The British people who speak BBC English can get a better job' (70.8% in 2016 vs. 69.7% in 1997), whilst 60.8% of the respondents in 2016 and 52.6% in 1997 also agreed that 'Those who speak this variety are seen as having a higher social status'. Yet, only 16.7% of the 2016 survey respondents and 17.8% of those who took part in the 1997 survey expressed their agreement with the statement that 'The British people who speak BBC English are more highly respected by their friends'. The results of the two surveys differ in responses to the statement that 'All British citizens should speak BBC English', for only 10.8% of the 2016 survey respondents agreed with it compared to 33.6% of those who participated in the 1997 survey (see figure 1). The above-mentioned results reveal interplay of overt and covert prestige. "Standard words, pronunciations and grammatical forms have overt prestige in that they are publicly acknowledged as 'correct' and as bestowing high social status on their users" (Trudgill 2003, 30). In contrast, nonstandard forms enjoy a kind of hidden or covert prestige because "they have become associated with the home, the neighbourhood, and with family and friends" (Cheshire and Milroy 2013, 18). Some speakers feel that the use of regional dialect forms is like "wearing a badge" (Hudson 1980), i.e. showing where they are from and feeling pride in being loyal members of their local community. Thus, non-standard varieties are still maintained and this "vernacular maintenance also implies competing ideologies that are in opposition to the standard (or more generally, institutionalised) ideology" (Milroy 2000, 13-14).
In addition, the responses to the first and the last statement reflect the concept of marché linguistique or the linguistic market (place). Put simply, this term refers to the commonsense view that some people feel a greater need to standardise their speech than others because of the market pressures and the jobs they (want to) do, which means that not all the people are motivated to use standard forms (see Chambers 1995, 178-184).
The questionnaire also asked the respondents to express their agreement or disagreement with the statements referring to the social stereotypes related to class-, age-, gender-, and ethnically-based linguistic variation of British English.
Although the word 'snob' can have different meanings to different people (see e.g. Thackeray 2005, 3;Brooker 2007, 48-9), 41.7% of the 2016 respondents agreed with the statement that 'Upper class people speak as they do because they are snobs'. Just a slightly higher percentage (46.1%) of the 1997 survey participants did the same (see figure 2).
Upper-class speech is often denoted as being "posh" and "snobbish" (e.g. Crystal 2004;Cheshire, Kerswill and Williams 2005). Nowadays, speaking with an upper-class accent is considered a disadvantage in many social situations and "there are actually now speech coaches who will help 'posh' people lose their accents, whereas thirty years ago, everyone else was trying to acquire them" (Smyth 2013, 207).
At the other end of the social scale, some working class people in Britain probably still remember being told by their teachers that their speech was "wrong", "bad", "careless", "sloppy", "slovenly", "vulgar" or "gibberish" (Trudgill 1975, 63). Even nowadays, working class children and young people may also experience linguistic prejudice and hear value judgements about their use of nonstandard English in school (see Williams 2007). Yet, the comparison between the survey results shows a significant difference in the responses to the statement that 'Working class speech is usually careless speech', as 38.3% of the 2016 respondents agreed with it compared to 72.4% in the 1997 survey (see figure 2).
This difference between the results calls for further empirical research in order to establish whether there are any real shifts in attitudes towards working L ... 247 class speech among the student population, or the responses might have been affected by social desirability bias (i.e. some respondents' tendency to misreport their opinions and attitudes by trying to make them more socially desirable). Different results are also revealed in responses to the statement that 'British people of Asian origin generally speak broken English' to which 44.2% of the 2016 survey respondents agreed compared to 63.8% respondents in the 1997 survey (see figure 2).
The adjective 'broken', apart from commonly referring to non-native speech, can also be used to describe 'anything from speech with frequent pauses to incomprehensible or very low-proficiency speech' (Lindemann 2005, 189). In a recent study, Asian-accented English has been evaluated by British speakers as one of the least prestigious as well as one of the least socially attractive varieties (see Coupland and Bishop 2007, 80).
Foreign-accent discrimination is embedded in what Shuck (2006) calls the ideology of nativeness. This ideology divides the linguistic world into native and non-native speakers of a language and exaggerates and rationalizes the problem of incomprehensibility of non-native speech (see also Gluszek and Hansen 2013). The ideology of nativeness is closely linked to the ideology of monolingualism which "sees language diversity as largely a consequence of immigration. In other words, "language diversity is viewed as imported" (Wiley and Lukes 1996, 519).
Although young themselves, the majority of respondents in both surveys (56.7% in 2016 and 52% in 1997) expressed a rather conservative attitude by agreeing with the statement that 'Generally young people do not speak as well as the older generation' (see figure 2).
Some quantitative studies (see e.g. Trudgill 1974;Cheshire 1987Cheshire , 2002Foulkes and Docherty 2014) confirm the view that young people, especially young adolescents, tend to conform to peer group norms by using non-standard forms of morphology and syntax, slang and non-standard pronunciation. Thus, the respondents who agreed with the aforementioned statement might have had in mind the use of non-standard English amongst young people, but their responses may also have been influenced by the normative approach in Serbian language teaching.
As for their perceptions of gender-based linguistic variation, 36.7% of the 2016 survey respondents expressed agreement with the statement that 'Generally women speak better English than men', as did 27.6% respondents in the 1997 survey (see figure 2). These findings show that the majority of both groups of respondents did not share a common belief that women speak better than men (see e.g. Romaine 1999, 173). Apart from this popular belief, some earlier quantitative sociolinguistic studies (see Trudgill 1974, Cheshire 1982, Milroy 1980 have shown that male speakers in the UK are more likely to use non-standard variants than female speakers. Conversely, female speakers "use linguistic S R. S 248 forms associated with the prestige standard more frequently than men" (Trudgill 1974). As gender perceptions have changed in recent times, young people may not necessarily see gender-based speech differences as important. In more recent studies on language and gender sociolinguists have abandoned research into speech differences between men and women and have adopted a social constructionist approach with the aim "to show how speakers use linguistic resources available to them to accomplish gender" (Coates 2007, 66).
The following set of statements refers to value judgements about speech and writing as well as about regional varieties of British English.
Since standardization is implemented through written forms of language, the standard written language is very highly valued and is often perceived as superior to spoken language. Nevertheless, a very high percentage of the respondents in both surveys (80.8% in 2016 vs. 75.7% in 1997) agreed with the statement that 'Spoken English is as good as written English' (see figure 3).
The egalitarian attitude towards spoken English expressed by the large majority of participants in both surveys may have been transferred from their own standard language culture (cf. Stojić 2005, 301-3). Every schoolchild in Serbia is familiar with the famous dictum 'write as you speak' proclaimed by Vuk Karadžić, the reformer of the Serbian writing system. Although the dictum actually refers to the phonological representation of sounds (one sound -one letter), it may have, to a certain extent, influenced the participants' responses.
Non-native speakers who are taught (and are mainly exposed to) Standard English certainly have more difficulty understanding speakers of some regional dialects of British English than native speakers. Thus, a high proportion of respondents in both surveys (83.3% in 2016 vs. 74.3% in 1997) expressed their agreement with the statement that 'In Britain there are some dialects of English which are mutually unintelligible' (see figure 3). In line with the results of many previous studies that explored evaluations of standard and non-standard varieties along the aesthetic or pleasantness dimension (see e.g. Edwards 1982, Trudgill and Giles 1983, Coupland and Bishop 2007, almost half (49.2%) of the 2016 survey respondents agreed with the statement that 'Regional accents are usually uglier than BBC English' as did an even higher percentage (59.9%) of those who participated in the 1997 survey (see figure 3).
In the eyes of many ordinary people Standard English stands in contrast to all other dialects as the only 'correct' variety and the 'real' language. Although linguists explain that regional dialects are rule governed, there is a long tradition of holding and expressing prejudicial attitudes towards dialect grammar, which is often perceived as a deviation from standard speech. The majority of the 2016 survey respondents (60.8%) also expressed such prejudicial attitudes by agreeing with the statement that 'Regional dialects (e.g. Cockney) are not grammatically correct', as did a much higher percentage (86.8%) of the 1997 survey respondents (see figure 3).
However, only 20.8% of the 2016 survey respondents agreed with the statement that 'Most people in Britain speak bad English', compared to 44.1% of the participants in the 1997 survey. Correspondingly, a high percentage of respondents (79.2% in 2016 vs. 68.4% in 1997) also agreed that 'British society should be more tolerant towards those who do not speak correct English' (see figure 3). The standard language ideology, the ideology of prescriptivism and purist ideologies underlie the long-standing tradition of complaints about declining standards of English. Speakers openly complain that English is in the process of decay and that language standards are slipping. While letters to the newspapers are still popular among the so-called custodians of the standard, it is the Internet that has become fertile ground for numerous blogs and forums where speakers openly express their prescriptive attitudes to and comments about 'incorrect' language usage or any other language issue (see e.g. Beal 2010; Heyd 2014). As Beal (2009, 47) points out "prescriptivism is not merely alive and well, but resurgent, in the twenty-first century". In this context, the following set of questions and statements refers to issues and perceptions of language change, pollution and decay as well as to language maintenance.
Answering the multiple response question 'Who pollutes good English?' a majority of the 2016 survey respondents (57.5%; n = 69) selected uneducated people as those who polluted it, while 23.3% (n = 28) chose foreigners, 21.7% (n = 26) immigrants, 10% (n = 12) journalists, 6.7% (n = 8) politicians and 2.5% (n = 3) bureaucrats. Only two respondents did not answer this question at all, while three stated that "no one pollutes English". Nine respondents added that, in their opinion, good English was polluted by: teenagers (n=2), young people and children, the working class, the Americans, teachers of English, the media, the Internet and rappers.
In accordance with the answers to the aforementioned question, a very high percentage of respondents from both surveys (71.7% in 2016 vs. 68.4% in 1997) agreed with the statement that 'English is in the process of decay'. Almost half (46.7%) of the 2016 survey participants also agreed that 'Standard English should not be allowed to change', as did 40.8% of the 1997 survey respondents (see figure 5). As can be seen, a slightly higher percentage of the 2016 survey respondents expressed prescriptive and purist attitudes towards language change and decay.
Since there is a widespread belief that standard language has to be "maintained and protected through authority and doctrines of correctness" (Milroy 2007, 138), one of the questions about the language maintenance was: 'Who should take care of the English language?'. As this was a multiple response question with 'all that apply' option, a great majority of respondents in both surveys (75% in 2016 vs. 80.9% in 1997) stated that taking care of English should be a team work of all the listed groups of people. In addition, the 2016 survey respondents also selected the following groups: linguists 15.8% (n = 19), teachers of English 14.2% (n = 17), lexicographers 2.5% (n = 3), journalists 2.5% (n=3) and government 0.8% (n = 1). The 1997 survey participants also chose: linguists 9.9% (n = 15), teachers of English 4.6% (n = 7), lexicographers 2% (n = 3) and government 1.3% (n = 2). No one selected journalists, while one respondent added that "everybody should take care of English".
In some countries, the principal standard language authority is a national academy, such as the Académie française or the Real Academia Española. Although there were some proposals for a language academy in the past, Britain has never had such an institution (see e.g. Watts 2011;Stojić 2005). Interestingly enough, more than a third of respondents from both surveys (35% in 2016 vs. 36.2% in 1997) expressed their agreement with the statement that 'Britain needs an academy for language to keep English from changing' (see figure 5).
It is quite odd that non-native speakers express concerns about the change, decay and pollution of the English language. A plausible explanation may be that the respondents transferred worrisome attitudes towards their own mother tongue to English.

Conclusion
The comparison of the findings yielded by the two questionnaire surveys reveal some differences as well as some similarities between the attitudes towards variation and change of British English expressed by the two generations of the first-year students of English. The findings indicate that, compared to the results of the 1997 survey, some positive ideological and attitudinal shifts may have occurred in the interim years, for a lower percentage of respondents in the 2016 survey expressed stereotypical perceptions of and attitudes towards some varieties of English and their speakers. Yet, a varying but significant percentage of them also expressed social and linguistic prejudice in their attitudinal responses. Therefore, the results also suggest that some deeply-embedded consensual attitudes are resistant to change.
It is worth noting that a large majority of respondents in both samples expressed conservative attitudes towards language pollution and decay, and almost half of the respondents in both groups expressed a view that English should not be allowed to change.
Overall, the attitudes expressed by the respondents in both surveys may be viewed as influenced by multiple language ideologies, particularly by the two standard language ideologies -one imported and associated with Standard English and the other associated with Standard Serbian. Although these two ideologies have unfolded from different histories and standardisation processes, they overlap in many aspects and are backed up by the ideologies of prescriptivism and purism. As mentioned above, the Standard Serbian ideology may have influenced students' positive attitudes expressed towards spoken English. This ideology may also have had some bearing on their rather conservative attitudes expressed towards young people's speech as well as towards language change; pollution and decay (see British students' responses to the same questions and statements in Stojić 2005).
Value judgements on language are here to stay, for they "form part of every competent speaker's linguistic repertoire" (Cameron, 1995: x). Yet, some persistent misperceptions, consensual stereotypes and prejudicial attitudes should be adequately addressed in language teaching in both primary and secondary schools. Primary and secondary school students should be acquainted with basic sociolinguistic concepts. The inevitability of linguistic variation and change, for instance, should be explained to them and recognised as axiomatic.
The students of English, who took part in the 2016 survey, should ideally reconcile the tension between prescriptivism in foreign language teaching and descriptivism in linguistic training. In their study of language variation and change, some of them may have to question their existing values, beliefs, preconceived perceptions and attitudes to language. L ... 253 Many of these students will become teachers of English and will further disseminate language attitudes towards English and, hopefully, also transmit new standard language ideologies that might be expected to emerge in the UK. Garrett et al. (2011, 61) call these new ideologies Positively Normative Language Ideologies (PNLIs) "where 'standard' ceases to entail 'correct and cultured usage' or 'maintaining standards'", but "comes to entail 'ordinary acceptable usage' (as in the phrase 'standard practice'), or 'usage agreed as fit for purpose in this particular discursive niche'".
Finally, the findings of this comparative study indicate the need for optimal and balanced instructional practices in English language teaching, which would enable students to master Standard English for the sake of intelligibility and academic and career success on the one hand, and to recognise the importance of linguistic diversity on the other. Such practices would heighten students' sociolinguist awareness and enhance linguistic tolerance.