Aurelija Čekuolytė

Research Institute of the Lithuanian Language

[email protected]

Abstract

The current sociolinguistic enterprise is preoccupied with the local meaning of the linguistic resources, however, the global meaning is equally important, because any linguistic resource becomes socially meaningful only when it is recognized as such by the others. Therefore, the main objectives of this article are (1) to advocate for the need to investigate not only the local meaning, discovered through the in-depth ethnographic fieldwork, but also the global meaning of the linguistic resources, (2) to demonstrate how by inclusion of other methodologies, in this case, the verbal guise technique, we can investigate the global meaning of the ethnographically derived data, and (3) to present results of the study of Vilnius adolescents’ perception of their peers’ linguistic identity which encompassed these two methodologies. During the course of the fieldwork in a school in Vilnius, five main social categories of Vilnius adolescents were distinguished: active schoolwise girls, cool girls, cool boys, streetwise girls, and streetwise boys. Different linguistic resources are incorporated in construction of different adolescents’ social categories. But are those linguistic differences local or could they be recognized as having this particular social meaning in other communities of practice? In order to answer this question, the verbal guise experiment was conducted in 3 other schools. Most of the adolescents’ identities were recognized by the adolescents in the verbal guise experiment. This implies that the linguistic variation, involved in the identity construction, has the same meaning in Vilnius dormitory neighborhoods.

1. Introduction and research objectives

Bernelis su 3 paloskėm^[1] (A youngster with three stripes), gerai besimokanti, mokytojų numylėtinė, tėvai ja didžiuojasi  (she is doing well in school, teacher’s favorite, her parents are proud of her), pasikėlus, atstumianti “ne savo lygio” bendraamžius (she’s arrogant, who rejects the peers who are not “on her level”) – these descriptions of adolescents have been provided not by their friends, teachers, parents, not even by the key source of information in a dormitory Vilnius neighborhood – a female neighbor from the third floor who watches what is going on in courtyard night and day. No, theses descriptions have been provided by their peers who have never met any of them. Even more, the adolescents arrived at such images of their peers just by listening to a few seconds of their speech. It is even more astonishing that adolescents’ perception of their peers corresponds to the identity which their peers are constructing in daily interactions. How is it all possible?

In the school year 2012–2013 (8 months in total), I carried out an ethnographic study in a secondary school in one of socially unmarked dormitory neighborhoods in Vilnius (Čekuolytė forthcoming). The main objective of the ethnography was to define Vilnius adolescents’ social categories and what resources adolescents employ in the construction of these categories. The ethnographic method enables the researcher to perform the in-depth analysis of the category construction which is based on the directly observed interactions and practices, not the presupposed ones. However, the scope of the ethnographic inquiry is limited – in my case, it was three 8^th grader classes, 90 pupils in total, in one secondary school. So with the help of ethnography we can directly examine the local construction of the social identities, however, we cannot be sure if the resources, associated with certain identities will be recognized as markers of those identities in other communities of practice. As I will point out later in this article, the recognition and perception part is essential in the identity construction. Therefore, the main objectives of this article is (1) to account for the need to investigate not only the local, but also the global meaning of the resources, (2) to demonstrate how by inclusion of other methodologies, in this case, the verbal guise technique, we can investigate the global meaning of the ethnographically derived data, and (3) to present results of the study which encompassed these two methodologies, though the main focus of this article will be on the verbal guise technique and the perception of adolescents’ linguistic identities.

2. From local to global

Instead of operating with the predetermined generalized social categories, such as gender, social class, ethnicity, in the current sociolinguistics which sometimes is referred to as the third wave of variation study (Eckert 2012) researchers tend to carry out sustained ethnographies in various communities of practice in order to understand how people through their engagement in different practices and by taking different stances, give social meaning to linguistic variables (Eckert 2000, Maegaard 2007, Podesva 2008^[2], Quist 2012, Rampton 2006, Zhang 2005, just to name a few studies). In other words, the current sociolinguistic enterprise is preoccupied with the local meaning-making: How different meanings, such as feminine, masculine, streetwise, gay, immigrant, become associated with particular variables. However, the process of the meaning-making is only partly dependent on the person who is performing it. The process of the meaning-making is not only performed, it is also perceived by the others. As Agha (2006: 234) states: “But even when one’s self-conception (or, rather, a given timebound version of it) becomes fixed or definite for a while, it is only relevant to social life insofar as it is perceivable by others”. So according to Agha, one’s social identity and the resources involved in the construction of that identity only become socially meaningful when they are recognized as such by the others. Three paloskės, a specific clothing type I described in more detail in the first note, is socially meaningful because it is perceived as an identity marker by the others, even the ones who do not perform this identity. Three paloskės is associated with a stereotype of street culture. There are many people who wear sportswear; it is by no means restricted to street culture, but only streetwise personas through their continuous and visible engagement in the street culture made that sportswear socially meaningful in the Lithuanian context. It suffices to say A guy with three paloskės, that the majority of Lithuanians would be able to draw an image of a guy who you do not wish to mess up with. (This just shows how much three paloskės are loaded with the stereotype.) What is especially interesting here that the meaning that was once created locally, is now recognized globally.

So far I have dealt with non-linguistic resources, involved in the construction of certain identities. But what about the linguistic variation? The ethnographic studies that I referred to so far, have proved that linguistic variation is also involved in identity making. However, not all of them dealt with the recognition and perception part of the identity construction. Marie Maegaard, referring to Linell (1998, 2001^[3]), argues for the need to investigate the global ‘meaning potentials’ of the ethnographically derived variables in order to understand “stereotypes and their connection to language variation” (Maegaard: 2010^[4]: 189). But can such a small resource as certain linguistic variables be recognized as identity markers globally?

The variation between one speaker and another, or between the same person’s speech in one situation as opposed to another, is often unnoticeable to a particular hearer. In order to become noticeable, a particular variant must be linked with an ideological scheme that can be used to evaluate it in contrast to another variant. <…> the scheme to which a hearer orients may be one that links variation with class, carefulness, correctness, place, or any other framework in terms of which people position one another socially, each associated with a set of stereotypical personas (Johnstone 2009: 160).

Johnstone makes an important argument here: She presents the theoretical model for interpreting the linguistic variation from the hearer’s point of view. Linguistic variant is only noticeable, i.e. recognized, when a hearer is able to place it on the social landscape (ways of talking which are characteristic for a particular social category), geographical landscape (dialect) or any other ideological scheme as Johnstone calls it. Even more, “a form that is enregistered (i.e. linked to a specific ideological scheme A.Č.) is one that is linked with a way of speaking or “register” associated with a personal or social identity” (Johnstone 2009: 160). In other words, the linguistic variable becomes noticeable when it invokes some kind of a stereotype. Stereotype is a way of social categorization. Stereotyping involves attribution of certain features such as various character traits, interests and occupations to different types of people (Garret 2010: 32).

There are plenty of methodologies developed for studying linguistic stereotypes – interviews, surveys, also ethnography^[5]. Just to pinpoint the critique towards the ethnographic studies, which I referred to earlier, I need to state that it is not impossible to study stereotypes and perception of stereotypes ethnographically, it is merely rarely done. However, most probably the most widely applied method in studying linguistic stereotypes is the so-called matched-guise technique and its various modifications (Garrett 2010).

3. The verbal guise technique

The matched-guise technique^[6] is an indirect method to investigate linguistic stereotypes, or put in another way, the attitudes to language use. An indirect method implies that informants are not aware of the fact that they take part in a linguistic study. Indirect methods were constructed to tackle the problems that might arise in application of the direct methods in the language research such as interviews and surveys. In the case of a direct method, informants are openly asked questions about different languages and speakers who use them and this may result in social desirability bias, i.e. when informants instead of giving their genuine attitudes, provide attitudes which they believe to be ‘socially appropriate’ (Garret, Coupland, Williams 2003: 8, Garrett 2010: 44). Informants who hold negative attitudes to a particular group of people, for instance, the Black Americans, newly-arrived immigrants, can conceal such information from the researcher. In the case of an indirect method, i.e. when informants are not aware of the real aim of the study, it is believed it is possible to arrive at the attitudes which should represent informants’ privately held linguistic attitudes (Giles 1976: 294).

How can we study linguistic attitudes indirectly, how is the matched-guise technique designed? The matched-guise technique is a speaker evaluation experiment. It is assumed that ‘listener’s attitude toward members of a particular group should generalize to the language they use’ (Lambert et al. 1960: 44). In the matched-guise experiment, informants listen to a number, usually five, audiotaped recordings of speakers. After they had listened to a recording, informants are asked to evaluate the speaker – to fill in the attitude-rating scales, i.e. to evaluate how friendly, interesting, clever, etc. they thought the speaker to be. What the informants are not aware of is that one speaker appears twice hidden under two different ‘masks’ (hence the term the matched-guise) during the experiment session, i.e. there are only four actual speakers, not five, as it was told in the presentation of the experiment^[7]. The guises usually differ in just one feature, for instance, one is in standard variety, the other is in local variety (for an overview for such studies see Garret 2010), one is purely national, the other contains a few words in English (Kristiansen 2006, Čekuolytė 2010). It is only the evaluations of that ‘guised’ speaker, or rather the differences of evaluations of the guises, that researchers investigate in their reports, the rest are just filler voices.

The possibility to elicit privately held attitudes and a possibility to compare one’s results to the results of other studies, conducted both nationally and internationally (Garret, Coupland and Williams 2003: 57, Garret 2010: 57) made the matched-guise technique a leading method in language attitude research. However, as any other method, the matched-guise technique is not unproblematic and has its own limitations regarding the accent-authenticity, the mimicking-authenticity, and the style-authenticity (Garret, Coupland and Williams 2003: 57 – 61, Garret 2010: 57–59). In order to keep other features (such as intonation, speech rate) constant, the same speaker presents both guises, so that evaluation of the speaker, i.e. a certain variety, would be solely based on the linguistic features. However, certain intonations and speech rate may co-vary with certain linguistic varieties. So if these features are eliminated, the variety, presented in the guise, does not represent the one that could be heard in real life. In some cases, for instance, if the objective of the study is to analyse language attitudes to five or more varieties, it is hardly possible to find a speaker who would be able to provide authentic recordings of five or more different varieties. In the original design of the matched-guise experiment, speakers were asked to read a prepared written text. However, reading style is usually more formal and less spontaneous than conversation. Therefore it could be assumed that the same variety provided in a more formal and in a more casual style, would be evaluated differently (based on Garret, Coupland and Williams 2003: 57 – 61, Garret 2010: 57–59).

In order to tackle the before-mentioned limitations, the matched-guise technique has changed a lot since its first application in Montreal. Nowadays researchers tend to use spontaneous speech rather than the prepared texts in their guises in order to present authentic linguistic styles which their informants (listeners) hear every day. As the objectives of the linguistic attitude inquiry became more complex and detailed, researchers use different speakers for different guises. As pointed above, it would be a difficult task (even if possible at all) to find a speaker who could provide authentic recording of five or more different varieties. Therefore in the current language attitude research the method is usually called verbal guise (technique) and is applied to study attitudes not only to different languages, but also to different dialects and sociolects (for an overview of such studies see Garrett 2010), to different linguistic variants (Pharao et al. (2014) studied perception of [s+] in Copenhagen youth speech, Campbell-Kibler (2008) studied the complexity of perception of /ing/ vs /in/ in American speech) and even to linguistic landscapes (Čekuolytė 2008, see also Garret 2010). In the following, I provide the version of the verbal guise, used in my study: choice of speakers, informants, questionnaire design, and performance of the experiment.

4. Ethnography and verbal guise combined

As I have already mentioned, the data, used in this article, was taken from the bigger ethnographic project, carried out in a school in Vilnius. During the course of the fieldwork, through continuous practices that pupils engage in and stances that they take, I was able to distinguish the following five main social categories of Vilnius adolescents: active schoolwise girls, cool girls, cool boys, streetwise girls, and streetwise boys. The labels of the categories have been given by me, based on the practices which are involved in the category construction. Pupils themselves either do not apply any labels when referring to a certain adolescent category or the category label is not well-established.

Active schoolwise girls perform identities which every single school strives to have. They engage both in classroom and extracurricular activities. Besides, due to their networking abilities, they are very popular which makes them very visible on the social landscape of the school.

Cool girls constitute a quite mixed group. Academically a few of them do so well (or a little poorer) in school as active schoolwise girls, but a few of them do very poorly in school. They do not perform in school plays. Besides, they secretly smoke cigarettes and drink alcohol. They are also popular girls.

Cool boys are popular boys who do academically relatively well in school. They show quite great interest in class activities, but at the same time they allow themselves sometimes to make loud remarks during a lesson and engage in other activities which might irritate a teacher, but as those activities don’t occur very often and are not very harsh, they are not perceived by the teachers and their classmates as troublemakers. Besides, they also play in class and school sport teams. A few of them, like cool girls, smoke and drink alcohol; however, they do not talk openly about it.

Streetwise boys and girls hold a clear anti-establishment attitude. They show very little interest in class activities. They do not have textbooks, pencils, do not do homework or they simply cut classes. They also smoke and drink alcohol and they practice it to such an extent that ‘the whole school knows about it’. Besides, quite a few of them have records in the police office (the most frequent cases are fights and drinking alcohol in public places).

However, pupils do not only differ from each other in engaging in different practices (or engaging to different extent to the same practice), they also differ linguistically. Active girls tend to monophthongize the diphthong /ie/ (/ie/ —-> /e/), especially in the discourse marker ‘tiesiog’ (simply). Some of their /r/’s are burred and some of their /n/ are distinctively guttural. Cool girls make great use of various discourse markers, for instance, ‘ta prasme’ (in that sense), ‘nežinau’ (I don’t know), ‘ten’ (like^[8]). They also tend to burr their /r/’s. Cool boys do not have distinctive linguistic features. Their speech is a mixture of the features, used by active girls and cool girls. Streetwise girls and boys tend to lengthen a short front vowel /i/ and a short back vowel /u/ in the stressed syllables. But are those linguistic differences local or could they be recognized as having this particular social meaning in other communities of practice?

With an inspiration in Maegaard’s study (2007, 2010) I decided to perform a verbal guise experiment in the nearby dormitory neighborhoods. Maegaard herself conducted her experiment in the same school where she carried out her ethnographic fieldwork (the verbal guise took place two years after the completion of the fieldwork in school) and in another school which is located in a very different Copenhagen neighborhood. Her aim was to test if the social meanings of certain linguistic variables have a global recognition: Will they be identified equally in economically richer and poorer neighborhoods of Copenhagen? Instead of carrying out the experiment in different parts of Vilnius, I, however, decided to focus on the dormitory neighborhoods. Therefore I performed an extensive verbal guise experiment in the dormitory neighborhoods which are very similar to the neighborhood where the ethnographic research was carried out. My assumption was that as the sociodemographic characteristics of these neighborhoods are similar, there should be the same (or very similar) adolescents’ social categories and the construction of them should involve similar resources. So the results of my study indicate not the global meaning of the linguistic variation, i.e. the meaning, which should be characteristic to all Vilnius adolescents’ speech, but the extended local meaning, i.e. the meaning which should be characteristic to adolescents’ speech in the dormitory neighborhoods of Vilnius.

5. Design of the verbal guise in my study

5.1. Stimuli (guises)

Speakers for the verbal guise experiment were chosen from the most distinguished adolescents’ social categories: active girls, cool girls, cool boys, streetwise girls and streetwise boys. They are the most visible categories on the social landscape of the school where the ethnographic study was carried out and have the clearly defined social characteristics. So their speech, if recognized, could be easily tied to a certain social category (or stereotype). Four speakers from the social category active girls, two speakers from the cool girls, two speakers from the cool boys, two speakers from the streetwise girls, and four speakers from the social category streetwise boys were included in the experiment.

Stimuli were prepared from the individual interviews with the pupils. Two researchers, including myself^[9], listened to the excerpts of the interviews where the pupils were talking about their friends and leisure activities and took notes what was specific about each pupil’s speech. The third researcher listened and commented on the prepared stimuli. On the one hand, an auditory analysis of the interviews is not very reliable, but on the other hand, it will be informants’ ears which will listen and evaluate the speakers. So in this case, the auditory analysis suited the aims of the study.

In order to make the content of the stimuli more or less similar, parts of the interviews, where pupils were talking about their friends and time off school, were used for the stimuli. However, it was sometimes hard to find a place in the interview where pupils don’t mention information which could easily give away their identity. Therefore I had to take different pieces from different parts of the interview and assemble them as if they were a natural spontaneous talk. Unfortunately, a few times ‘cut-and-paste’ resulted in a tiny pause, luckily, most of the pauses occurred in such point of the talk where they would have occurred naturally. Such pauses are marked grey in the transcriptions of the stimuli (.). Stimuli vary from 8 to 22 seconds.

In the table below all stimuli are presented in the order they were played in the experiment. The category of the pupil is noted in brackets. All names are changed. “I” and “U” marks lengthening of respectively short front vowel /i/ and short back vowel /u/. Monophthongization of the diphthong /ie/ (/ie/ —-> /e/) is marked as “IE”. “N” and “R” marks respectively guttural /n/ and burred /r/. Discourse markers are in bold.

Table 1. Stimuli

Number	Name (category)	Stimulus	Translation
1	Pijus (cool boy)	kartais susitinku ten su kitais klasiokais bet va jie trejetas mano pagrindinis yra (.) ką veikiam? susitinkam (.) ten pagalvojam nu nueinam pas vieną pas kitą ten tennn dažnai va važinėjam į kokius akropolius ten	sometimes I meet like with other classmates but yea they this three is my core one (.) what do we do? we meet up (.) like we think well we go to one friend to another like liiike often so we go to akropolises like
2	Daiva (streetwise girl)	dažniausiai tai būna kaip susitInkam ir tada jau galvojam ką daryt ar einam kur nors pasivaikščioti ar tiesiog vienoj vietoj (.) ar einam kur nors užsiiminėti taip (.) tai būna taip kad susitInkam ir tada viskas labai spontaniškai	normally it is like how we meet up and then we think what to do or we go somewhere for a walk or we hang out simply in one place (.) or we go somewhere to occupy ourselves (.) it happens that we meet up and then everything happens very spontaneously
3	Daina (cool girl)	nežinau labai gerai vienas kitą pažinom ten buvom geriausi vos ne draugai (.) šiaip labai gerai ta prasme visada sutariam (.) ten šiaip nežinau nu į kavinę einam po parduotuves vaikštom jau į kokį prekybos centrą nuvažiuojam nu ten nežinau (.) filmus žiūrim	I don‘t know we knew each other very well like we were almost best friends (masculine) (.) anyhow we in that sense always get together very well (.) like anyhow I don‘t know well we go to a cafe, we go shopping we go to some shopping center well like I don’t know (.) we watch films
4	Rokas (streetwise boy)	tai per pažįstamus ten susipažinom (.) birželio septintą (.) mes geri draugai mes ta prasme taip daug bendraujam labai [(m) (.)] ten taip būna susitinkam ten kokį (.) savaitgalį	so we got acquainted like through friends (.) on the seventh of June (.) we are good friends we in that sense communicate so a lot [(m) (.)] like it happens we meet up like some (.) weekend
5	Rūta (active girl)	mes nuo galima sakyti nuo nuo smėlio dėžės kartu (.) e tokia šalia manęs netgi kaimynė (.) ne iš mano mokyklos (.) tai su ja vat kai grįžtu [(m) (.)] namo susitinkam arba jinai kartais pas mane užeina nes jinai metais jaunesnė yra↑ (.) ir dažnai labai susitinkam beveik kiekvieną dieną pasiliekam po pamokų↑ (.) tiesiog va pabūnam pasėdim	we you can say since since sandbox times are together (.) eh she’s such even my neighbor she lives nearby (.) not from my school (.) so yea when I come back [(m) (.)] home I meet with her or she sometimes comes to me because she’s one year younger↑ (.) and we very often meet up almost every day we spend time after school ↑ (.) simply yea we hang out together
6	Egidijus (streetwise boy)	aš kartu tiesiog su jais būnu gerai sutariu (.) ne kaip kiti ten dar kažką būnam susitinkam (.) sssėdim taip būnam tiesiog lauke valandą su puse tai (.) tris su puse valandos (.) man patinka	I am simply together with them I get along well with them (.) not like the others or something we hang out together we meet up (.) we sssit so we stay outside for one hour and a half so (.) three and a half hour (.) I like that
7	Renata (cool girl)	gal mes turim ta prasme daugiausia bendrų p tų pomėgių (.) ir šiaip tarkim mes buvom daržely vienoj grupėj ir mes jau iš anksčiau pažįstamos (.) bet kažkaip nebendravom nebendravom (.) paskui čianai mokykloj kokioj septintoj klasės pradžioj ar šeštos pabaigoj [(breathes in) (m)] vėl tenai pabendravom ir tarkim savaitgaliais ten (.) nakvojam vienas pas kitas	maybe we have in that sense the most alike those h hobbies (.) and so let’s say we were in the same group in the kindergarten and we knew each other from before (.) but somehow we didn’t talk didn’t talk (.) later here in school something like at the beginning of the seventh grade or the end of the sixth grade [(breathes in) (m)] we like talked again and let’s say in the weekends like (.) we sleep at each other
8	Urtė (active girl)	klasiokes pažįstu nuo pirmos klasės tai kaip ir aštuoni metai tai čia yra (.) gana ilgas laiko tarpas ir tiesiog tu sssusipažįsti šu su žmonėm (.) taip žinai kad tau jeigu reikės tave visada palaikys [(m) (.)] ir tai yra tikri draugai ir tiesiog aš nenoriu tikrai nenoriu daryti kitaip (.) ir (.) tiesiog aš su tokiais bendrauju	I know the classmates since the first grade so it’s like eight years so this is (.) a rather long period of time and you simply get fffamiliar wiz with the people (.) yes you know that if you need they gonna support you [(m) (.)] and so this real friends (masculine) and I simply don’t I really don’t want to do otherwise (.) and (.) I simply communicate with such (friends)
9	Arnas (streetwise boy)	su jais geriausiai susibendravau (.) nuo penktos klasės (.) manim pasitiki ir aš jais pasitikiu linksmiau man su jais nu nieko sėdim kalbam (.) nu kaip ir viskas čia	I got along the best with them (.) since the fifth grade (.) they trust me and I trust them it’s more fun for me with them well nothing we hang out we talk (.) well like that’s it
10	Kamilė (active girl)	nuo pat mažumės visi draugai tai dar vis draugaujam tos darželio laikų (.) tada mokykliniai draugai kiemo draugai (.) mes su jais susitinkam nes kažkaip mes turim dažnai visąlaik apie ką pakalbėti mum visiem yra linksma nes mūsų charakteriai visų yra beveik vienodi	(I have) all friends (masculine) since infancy and we are still friends since the kindergarten times (.) then school friends courtyard friends (.) we meet up with them because somehow we have often always something to talk about we all have fun together because characters of all of us are almost the same
11	Albertas (streetwise boy)	susipažinom kieme (.) nu susitikom visi ten (.) šiaip vaikščiojom kieme nuėjom pas kitus draugus ten (.) pakalbėjom pabuvom (.) nu ir taip buvom iki vakaro maždaug (.) nieko ypatingo neveikėm	we got to know each other on the courtyard (.) well we met up all there (.) you know we walked in the courtyard we went to other friends like (.) we talked we hang out (.) and well yea we hung out until like evening (.) we didn’t  do anything in particular
12	Tadas (cool boy)	laisvalaikiu būnu su draugaisss varau į prekybos centrą (.) stengiuosi (.) pabūt linksmai (.) geriausi draugai (.) su jais aš neturiu varžytis pavzydžiui aš su jais laisvai laisvai galiu kalbėti (.) tiesiog yra geriausi draugai kurie niekada neišduos patars visada	in my leisure time I hang out with friendsss I go to a shopping center (.) I try (.) to spend time merrily (.) best friends (.) with them I don’t have to feel cheap for example I can easily easily talk with them (.) Simply best friends are those who will never betray, always will give advice
13	Eglė (active girl)	nu mes nuo pirmos klasės kai bendraujam (.) e bet labiau susibendravom maždaug nuo trečios ketvirtos (.) irrrr mes abi dažnai būnam kartu nu mum tiesiog pan panašūs panašūs pomėgiaiii (.) ir dabar labai gerai sutariam ir tiesiog	well since the first grade we are friends (.) e but we became better friend since about the third grade (.) aaand we both hang out often together well we simply have sim similar hobbieees (.) and now we get along very well and simply
14	Samanta (streetwise girl)	mes su draugais muzikos dažnai klausom ir dar žaidimus žaidžiam (.) draugų būry (.) viskas labai linksma buvo (.) draugai man padėdavo ir niekad nepalikdavo bėdoje	we often listen to the music with friends and we also play games (.) in the circle of the friends (.) everything was so fun (.) friends used to help and always stuck by me

5.2. Questionnaire

In the matched-guise experiment, informants are usually asked to fill in the attitude-rating scales – semantic differential scales, i.e. how friendly, educated, energetic, they found the speaker to be. The speech evaluation instrument, based on Zahn and Hopper’s studies (Zahn and Hopper 1985) is comprised of three dimensions: superiority (traits such as educated / uneducated), attractiveness (traits such as friendly / unfriendly) and dynamism (traits such as confident / hesitant). Even though semantic differential scales are very easy to handle statistically, they put constraints on informants’ evaluations: Informants are forced to use the prepared scales. It raises a few methodological questions: (1) Would informants use the same adjectives if they were asked to evaluate speakers freely, (2) How important are the dimensions presented on the scales for the informants, for instance, the speaker might be evaluated very positively on a superiority dimension, but this dimension might be irrelevant for an informant. To avoid these limitations, the main question of the questionnaire was an open-ended one: Briefly describe the speaker. However, the main question was supplemented by three closed ranking questions: Do you think that this adolescent is popular / unpopular; Do you think that this adolescent has addiction; and Do you think that this adolescent plays in a school sport team / performs in school plays? The main purpose for the inclusion of these closed questions was to help informants to understand the task better, i.e. that they were asked to place the speakers in the adolescents’ or school’s social order (see also 5.4. Performance of the experiment).

5.3. Informants

Informants of the verbal guise experiment were 8^th grade pupils in the same dormitory neighborhood where the ethnographic fieldwork was conducted and in two other dormitory neighborhoods which share similar characteristics with that neighborhood. However, I conducted the experiment not in the same school where the ethnographic fieldwork was carried out. The experiment was conducted a few months after the completion of the fieldwork and I was afraid of that current 8^th graders,  who were 7^th graders by the time I worked in school, might recognize a few of the pupils’ voices which would result in bias, so I had to use the other school in that neighborhood.

5.4. Performance of the experiment

The pupils were told that they were going to listen to 14 adolescents, the same age as them, whom they had to describe. The pupils were also asked if possible to indicate the type of the adolescent, what kind of image of him or her they create in their minds. I deliberately avoided the terms category and identity because it might sound too scientific to adolescents. I have also made it clear that they should not provide the retelling of the story. So as much as possible the emphasis was put on the depiction of the speaker.

At first, adolescents listened to all 14 recordings at once, so they could get the impression of speakers and the task itself. During the second listening I paused after each recording. During the pause, the informants were asked to fill in the questions allocated to that speaker.

6. Results of the verbal guise experiment

In total, 274 pupils participated in the experiment. 3 pupils did not fill in questionnaires properly, so the analysis, presented in this article, is based on 271 questionnaires.

It turned out that to carry out the research where adolescents had to evaluate their peers was a quite risky enterprise. Quite a few informants provided negative and even derogatory evaluations of the speakers. Researcher group (Garret, Coupland, Williams 2003) who conducted a similar verbal guise experiment in Wales, was also confronted with negativity in adolescents’ responses. Adolescence is a period when individuals are ‘exploring a range of available identities, this may require an equivalent range of differentiating evaluative descriptors. And positioning themselves in this relation to this range of identities is likely to mean rejecting more than they find acceptable, and so lead to more negative than favorable reactions’ (Garret, Coupland, Williams 2003: 180). From the analysis of the open-ended data were excluded the questionnaires where at least 12 of 14 speakers were evaluated extremely negative or where the same trait has been applied to all 14 speakers, for instance, noob^[10] – this description (in English) was given to speakers from number 1 to number to number 8, the rest of the questions were left blank.

Firstly, I present the analysis of the open-ended data, or keyword comments (Garret, Coupland, Williams 2003), because it helps to understand the responses of the closed questions better.

6.1. Keyword comments of the verbal guise experiment

The answers to open-ended questions are very rich and diverse. In principle, informants can write whatever they feel like writing. In the analysis phase, the researcher has to bring structure to that diversity, i.e. to code the data without losing its richness. In my analysis, each response was divided into different parts based on the semantic content. Below I present a few examples of the coding.

Tai žalingų įpročių turintis paauglys – keistuolis (This is an adolescent who has addiction – weirdo). an adolescent who has addiction – coded under the label ‘addiction’; weirdo – coded under the label ‘weird’.

Draugiška, maloni, turi tik kelis ištikimus draugus. (Friendly, nice, has only a few loyal friends). Friendly – coded as ‘friendly’; nice – coded as ‘nice’, has only a few loyal friends – coded as the retelling of the content of the stimulus.

pasikėlus, atstumianti “ne savo lygio” bendraamžius. (she’s arrogant, who rejects the peers who are not “on her level”). The whole description was coded under the label ‘arrogant’.

rajonskas (slang term to depict streetwise boys and men of the dormitory neighborhoods) was coded under the label ‘street culture’.

Only those labels (personality traits and categories) which have been mentioned at least by 10 informants, were included in the analysis. However, if a certain label has been mainly applied to a certain speaker, I included it in the analysis, even though it was mentioned in fewer than 10 responses. Table 2 presents results of the keyword comments.

Table 2. Keyword comments about the individual speakers

Speaker	Keyword comment (frequency)
Kamilė (active)	Friendly (41), social, likes to socialize (33), fun, cheerful (17), has many friends (14), ordinary (13), boring, not interesting (12), nice, good (12), weird (10), kid, childish (8)
Urtė (active)	Friendly (34), social, likes to socialize (21), shy, modest, quiet (19), good, nice (18), ordinary (16), loyal, supporting, trustworthy (13), boring, not interesting (11), “scholar” (10), clever (9)
Rūta (active)	Friendly (40), nice, sincere, loyal (20), ordinary (17), social, likes to socialize (17), shy, modest (16), reserved (11), good girl (kid) (11), “scholar” (10)
Eglė (active)	Friendly (31), ordinary (22), social, likes to socialize (15), shy, modest, quiet (14), boring, not interesting (13), good, nice girl (10), “scholar” (9)
Tadas (cool)	Fun, cheerful (31), friendly (28), social, likes to socialize (17), ordinary (14), popular (14), “nolifer” (13), nice, sincere (13), addiction (9)
Pijus (cool)	Ordinary (53), shy, modest, quiet (28), social, likes to socialize (19), friendly (14), “swag” (5)
Daina (cool)	Friendly (25), social, likes to socialize (25), fun, cheerful (24), chatty, talkative (20), ordinary (16), active, energetic (14), popular (13), arrogant (13)
Renata (cool)	Friendly (30), social, likes to socialize (29), ordinary (23), shy, silent, quiet (14), arrogant, unfriendly (13), boring, uninteresting (12), fun, cheerful (12), popular (11)
Daiva (streetwise)	Friendly (32), ordinary (31), social, likes to socialize (27), good nice girl (15), modest, shy, quiet (13), has many friends (10), “scholar” (9)
Samanta (streetwise)	Friendly (19), boring, uninteresting (16), ordinary (14), shy, silent, quiet (13), weird (12), addiction (8), accent (1)
Arnas (streetwise)	Addiction (23), friendly (20), fun, cheerful (20), street culture (13), accent (3)
Rokas (streetwise)	Shy, modest, quiet (32), addiction (24), street culture (22), boring, uninteresting (18), reserved (16), social, likes to socialize (14), ordinary (11), incorrect Lithuanian (1)
Egidijus (streetwise)	Addiction (18), boring, uninteresting (16), quiet, silent, shy (15), weird (14), street culture (13), ordinary (12), accent (9)
Albertas (streetwise)	Addiction (28), boring, uninteresting (25), street culture (16), negative comments (13), shy, modest, quiet (13), ordinary (12), accent (3)

Strikingly, identities of 13 out of 14 speakers were recognized with a quite great accuracy which entails that the resources which are involved in construction of these identities have the extended local meaning. Instead of going into a detail analysis of the perception of each speaker, I would rather put emphasis on the difference of the perception of different groups.

The active girls – Urtė, Rūta and Eglė – were recognized as active schoolwise girls – they attracted the most evaluations of the ‘scholar’ type among all speakers, besides the streetwise girl Daiva. The active girl Kamilė who is the leader in the active girl circle, was not categorized as ‘scholar’ or the one who studies very well as the other three girls in her circle. However, she was depicted as ‘childish and kid’. This depiction could be caused by her rather child-like voice. But still she was categorized as a nice girl.

The cool girls Renata and Daina were the ones who received the most responses of the ‘arrogant’ type. Those two girls perform cool personas in school which naturally entails a little bit of arrogance. They were also the only girls characterized as popular, which is not surprising because in school pupils with a touch of arrogance (pasikėlus, atstumianti “ne savo lygio” bendraamžius (she’s arrogant, who rejects the peers who are not “on her level”)) are usually considered as popular.

The streetwise girl Samanta was evaluated the most negatively among all the girls. Also, of all the girls, she received the most responses of the ‘addiction’ type which might indicate that she was perceived as a streetwise persona to some extent. However, compared to the perception of the streetwise boys, Samanta’s link to the street culture is not very strong but this could be caused by a general societal bias that street culture is dominated by men. One informant also paid attention to Samanta’s speech. She noted that the speaker lengthened the /i/ in the word linksma (fun) and the speaker could be Polish: lynksma^[11] viskas buvo. lenkų nereikia. (everything was funny. we do not need Poles).

Pijus and Tadas, who construct the cool boys identity in their school, were evaluated quite differently. This is not surprising because their stimuli were also different: Pijus made great use of various discourse markers which were absent in Tadas’ stimulus. Tadas’ stimulus contained two instances of monophthongization of the diphthong /ie/. Pijus was identified as swag and Tadas attracted quite a few responses of the ‘nolifer’ type. Swag refers to a modern urban contemporary youth identity – the latest fashion clothes, going to trendy cafes and in general spending lots of time in the center of the city. Nolifer is an adolescent who spends most of his or her free time playing computer games. Pijus can be said to perform a swag identity, especially compared to the other boys in this school. Tadas, however, does not construct a nolifer identity. Of course, he plays computer games as any other adolescent nowadays but he is not labelled as such by his friends. Nolifer identity and cool identity are somewhat different identities: Playing computer games all day long would not be treated as cool even by nowadays adolescents. However, Tadas was ranked as the most popular speaker on the popularity scale. So it seems that Tadas was perceived as constructing the cool version of the nolifer identity. But what is very important in the perception of Pijus and Tadas identities is that they were perceived as creators of the modern identities – swag and nolifer – the ones which emerged relatively recently whereas the streetwise category which is being constructed by the others boys of the study, is a well-established adolescent social category.

All streetwise boys – Rokas, Arnas, Egidijus and Albertas – were perceived as creators of the street culture identity with a great accuracy which means lengthening of short /i/ and /u/ is associated with street culture in the dormitory neighborhoods of Vilnius. Labels of the ‘addiction’ type were the most frequent in the pupils’ answers about these boys. These boys also received the most concrete descriptions of all the speakers, which is not surprising. Being a well-established social category, streetwiseness has quite a few globally recognizable and acknowledged features. Informants tended to mention in their answers the concrete social category, such as forsas, marozas, rajonskas (slang terms for streetwise boys and men of the dormitory neighborhoods), chuliganas (hooligan), or one of the most iconic features of this social category – their look: Nešioja Adidas, matosi iš balso (He wears Adidas. I can see it from his voice), Bernelis su 3 paloskėm (A youngster with three stripes), treninginis (slang term for boys and men who usually wear sport clothes). A few informants of the verbal guise study also paid attention to the streetwise adolescents’ speech. I will get back to that in the final section of the article.

The only speaker whose identity was not recognized was Daiva. She was perceived as performing a completely different identity than she actually does. Daiva is a streetwise girl, she smokes, drinks alcohol and cuts classes. However, in keyword responses she appeared to be a nice active girl, the one who is doing very well academically in school – ‘a scholar’. The incongruity of the persona, which Daiva performs through her daily practices, and perception of that persona could be caused by the linguistic variation, presented in the stimulus. She lengthened the least (just twice) of all the streetwise adolescents. Besides, her stimulus also contained the monophthongization of the diphthong /ie/ in the discourse marker ‘tIEsiog’ (simply) which is characteristic to the active girls’ speech. The perception of Daiva reveals that in order to be perceived as constructing a specific identity one has to continuously make use of the resources, associated to that identity.

6.2. Semantic differential scales

Informants were also asked to evaluate speakers’ popularity, tendency to addiction, and engagement in school activities. Statistical differences between individual speakers are either very small or insignificant, therefore in this section of the article, I will discuss speakers in groups according to their social category rather than individually (see Table 3).

Table 3. Ranking of adolescents’ categories across the differential scales^[12]

Popular – unpopular	Cool boys	>	Cool girls	>	Streetwise girls	>	Active girls	>	Streetwise boys	n=269
	2,25	/	2,53	***	3,07	***	3,56	/	3,59
Addiction – no addiction	Streetwise boys	>	Cool boys	>	Streetwise girls	>	Cool girls	>	Active girls	n=271
	1,7	***	2,53	***	3,33	/	3,49	***	3,95
Active – passive	Active girls	>	Cool girls	>	Cool boys	>	Steetwise girls	>	Streetwise boys	n=269
	2,49	/	2,79	/	2,83	*	3,26	/	3,63

Cool boys and cool girls were evaluated as popular. The difference between these two groups is statistically insignificant.  Active girls were perceived as unpopular, even though in their school they are very popular. However, it is not surprising that the active girls were perceived as unpopular as the keyword responses reveal that they identity construction is tied to school. Stereotypically, pupils with great engagement in school activities cannot be popular. It is also not surprising that the streetwise boys were deemed to be unpopular. Although they are very visible on the social landscape of the school, their popularity is of specific character. During the ethnographic interviews, when I asked pupils to tell me what’s involved in being popular in school, quite a few asked me to clarify the question: ‘Popular? In which way do you mean?’ Obviously, there two ways of being visible in school – being popular and being notorious. And streetwise boys are notorious.

There is no surprise that the streetwise boys were categorized as the most addicted to smoking and drinking among all the speakers. The next category which is prone to smoking and drinking, according to the survey results, is cool boys, not streetwise girls, what would be expected. There are two possible explanations for such perception. Firstly, Daiva was not perceived as a streetwise girl. Secondly, it could be possible that informants were influenced by a stereotype that smoking and drinking is a masculine practice.

Rankings of adolescents’ categories across the differential scale of engagement in school are statistically not very significant. The results of the verbal guise experiment divide speakers into two groups: pupils who hold pro-school attitudes or, at least, quite minor anti-establishment attitudes (active girls, cool girls, and cool boys) and pupils who hold very strong anti-establishment attitudes and whose engagement in school is very passive – streetwise adolescents. This division reflects the social order of the school where the ethnographic research took place.

7. Final remarks and perspectives for future studies

Most of the adolescents’ linguistic identities which they construct in employing different linguistic resources, to large extent were recognized by the adolescents in the verbal guise experiment. This implies that the social meaning of the linguistic variation, revealed through the sustained ethnographic research, is not locally bound. It is also recognized in other Vilnius dormitory neighborhoods, which in turn might suggest that the meaning of the active schoolwise, the cool, and the streetwise is being created using the same linguistic resources throughout the dormitory neighborhoods of Vilnius. The study also revealed that in order to be perceived as constructing a specific identity, the speaker has continuously to employ a necessary linguistic variation in the identity work.

However, the most interesting result of the verbal guise experiment was most probably Vilnius adolescents’ perception of the lengthening of the short vowels /i/ and /u/ in stressed syllables. Vilnius adolescents link lengthening of the short vowels to the ideological scheme of ‘street culture’. What is remarkable is that only stimuli of the streetwise adolescents attracted linguistic comments which might imply that lengthening is a distinctive feature among Vilnius adolescents. Informants, who paid attention to speech, would write down the words where the lengthening occurred, for instance, lynksma (fun), susytikom (we met), and would note that the speaker speaks with an accent (su akentu) or is Russian / Polish^[13]. However, a few informants provided more elaborative answers: Turi tokį akcentuką forsų (He has such a little accent of forsai) and Ne lietuvis, su akcentu arba specialiai padaro tokį balsa (Non-Lithuanian, with an accent or (he) deliberately makes such a voice). These explicit comments about the speech reveal the complexity of the social meaning of lengthening among Vilnius adolescents: Lengthening is associated with two social categories: Lithuanian Russians and street culture^[14]. But are these categories interrelated?

To exemplify this complexity let’s return to Egidijus, stimulus no. 6, whose speech attracted the most comments. The third researcher, who only listened to the prepared stimuli, judged Egidijus as Lithuanian Russian because of his lengthening and his pronunciation of the word ‘sėdim’ (we sit). Egidijus is not Russian, he is Lithuanian and his competence in Russian is very limited. However, Egidijus, as well as Arnas and Albertas, spends a lot of his leisure time together with the core streetwise youth of the neighborhood who happen to be Lithuanian Russians (this information was gathered through the individual interviews and self-recordings). So could it be that in order to claim his membership to the streetwise community of the neighborhood he imitates the Russian accent? Is forsai accent actually based on the Russian accent? At this point, these are only hypothetical questions which call for the analysis of the development of the street culture in Vilnius and the in-depth ethnographic study of the street culture, its language and distribution of its resources, i.e. we need to return to the local meaning making of the resources.

Acknowledgments

The research which is article is based on, is part of a bigger project “Vilnius Speaking II: city and urban linguistic diversity”, funded by a grant from the Research Council of Lithuania (No. VAT–61/2012). I would like to thank my colleagues – project coordinator Loreta Vaicekauskienė and project participant Ramunė Čičirkaitė for their help in preparing the stimuli. I am grateful to Kristijonas Vaicekauskas who did the statistical analysis. I would also like to thank the anonymous reviewers for the immensely helpful comments. And last but not least, I am enormously indebted to my informants of both the ethnographic and verbal guise studies. I hope you had as much fun as I did!

References

Agha, A. 2007. Language and Social Relations. Cambridge: Cambridge University Press.

Campbell-Kibler K. 2008. I’ll be the judge of that: Diversity in social perceptions of (ING). Language in Society 37. 637–659.

Čekuolytė, A. (forthcoming). Vilnius Adolescents’ Social Categories and Language: An Outsider’s Look Inside.

Čekuolytė, A. 2010. Jaunimo požiūris į angliškus žodžius suaugusiųjų kalboje. Kalbos kultūra 83. 322–341.

Čekuolytė, A. 2008. Danskernes og litauernes holdninger til det engelske sprog på forretningsvinduer i København og i Vilnius. En masketest. Unpublished BA thesis. Vilnius.

Čičirkaitė R. 2014. Vilnietiško balsių ilginimo socialinė reikšmė. Moksleivių pasąmoningųjų nuostatų tyrimas. Taikomoji kalbotyra 6. http://taikomojikalbotyra.lt/lt/2014/12/vilnietisko-balsiu-ilginimo-socialine-reiksme-moksleiviu-pasamoninguju-nuostatu-tyrimas

Eckert, P. 2012. Three waves of variation study: The emergence of meaning in the study of sociolinguistic variation. The Annual Review of Anthropology 41. 87–100.

Eckert, P. 2000. Linquistic Variation as Social Practice. The Linguistic Construction of Identity in Belten High. Oxford: Blackwell Publishers.

Garret, P. 2010. Attitudes to Language. Cambridge: Cambridge University Press.

Garret, P., N. Coupland and A. Williams. 2003. Investigating Language Attitudes. Social Meanings of Dialect, Ethnicity and Performance. Cardiff: University of Wales Press.

Giles, H. and R. Y. Bourhis. 1976. Methodological issues in dialect perception: Some social psychological perspectives. Anthropological Linguistics 18. 294-304.

Johnstone, B. 2009. Pittsburghese shirts: Commodification and the enregisterment of an urban dialect. American Speech 84(2). 157–175.

Kristiansen, T. 2006. Nordiske sprogholdninger. En masketest. Oslo: Novus forlag.

Kudirka, R. 2012. Lietuvių kalbos žargono ir nenorminės leksikos žodynas. Kaunas: Technologija.

Lambert, W. E. et al. 1960. Evaluational reactions to spoken languages. Journal of Abnormal and Social Psychology 60(1). 44–51.

Maegaard, M. 2010. Linguistic practice and stereotypes among Copenhagen adolescents. Multilingual Urban Scandinavia. New Linguistic Practices. P. Quist and B. A. Svendsen (eds.). Bristol: Multilingual Matters. 189–206.

Maegaard, M. 2007. Udtalevariation og forandring i københavnsk. En etnografisk undersøgelse af sprogbrug, sociale kategorier og social praksis blandt unge på en københavnsk folkeskole. Copenhagen: C. A. Reitzels Forlag.

Pharao, N., M. Maegaard, J. S. Møller and T. Kristiansen. 2014. Indexical meanings of [s+] among Copenhagen youth: Social perception of a phonetic variant in different prosodic contexts. Language in Society 43. 1–31.

Podesva, R.J. 2008. Three sources of stylistic meaning. Proceedings from SALSA IV: Texas Linguistic Forum 51. K. Shaw, S. Wagner and E. Yasui (eds.) 134–143.

Quist, P. 2012. Stilistisk praksis. Unge og sprog i den senmoderne storby. Copenhagen: Museum Tusculanums Forlag.

Rampton, B. 2006. Language in Late Modernity. Interaction in an Urban School. Cambridge: Cambridge University Press.

Zahn, C. and R. Hopper. 1985. Measuring language attitudes: The speech evaluation instrument. Journal of Language and Social Psychology 4. 113–123.

Zhang, Q. 2005. A Chinese yuppie in Beijing: Phonological variation and the construction of a new professional identity. Language in Society 34(3). 43–466.

globali reikšmė | kalbėtojo vertinimo eksperimentas | sociolingvistika | trumpųjų balsių ilginimas | vietinė reikšmė | Vilniaus paauglių kalba | Vilniaus paauglių socialinės kategorijos