DAIVA ALIŪKAITĖ

Kaunas Faculty of Humanities, Vilnius University

ERIKA MERKYTĖ

Kaunas Faculty of Humanities, Vilnius University

A SOCIOLINGUISTIC STUDY OF THE CORRECTNESS OF DIALECT CODES

INTRODUCTORY NOTES

The concept of linguistic correctness is primarily associated with the standard language. In linguistics, this view is determined by one of the relevant features of the standard language—codification (for more on relevant and non-relevant features of the standard language, see Pupkis 2005: 16–19). It should be noted that in the theory of the Lithuanian standard language, the criterion of correctness has not been fully clarified: some researchers understand this criterion as conformity to the laws of language or to codified rules (cf. Girdenis, Pupkis 2000 [1978]: 109; Kniūkšta 2001 [1979]: 65); others formulate the definition of correctness by specifying its autonomy. For example, Jonas Šukys distinguishes two main components of the criterion of correctness—purity and consistency (see Šukys 2003: 19). Such theoretical refinements are not particularly important for the question raised in this article, since, one way or another, definitions of correctness make it clear that the concept is relevant only when speaking about a codified code¹, because only codification allows for the possibility of error.

Dialects are spontaneous systems. They testify to the inexorable development of language, reveal features of usage, and so on.² Despite the fact that a dialect may absorb

¹ As Ronald Wardhaugh states, code, unlike dialect, style, standard language (=standard language), pidgin, or creole, is an emotionally neutral term (see Wardhaugh 2006: 88). That is precisely why the term code is often chosen in this article as well. For the sake of textual coherence or stylistic considerations, the terms dialect (dialectal speech) versus code are used interchangeably for the concept of the standard language.

² It should be said that, to some extent, the standard language could also be viewed in this way; however, when speaking about the standard language, the relationship between deliberate influence on language and spontaneous language development must not be forgotten (see Rita Miliūnaitė’s illustrative diagram depicting the standard language

D. ALIŪKAITĖ, E. MERKYTĖ. Sociolingvistinis tarminių kodų taisyklingumo tyrimas

259

Although these are systematic features of foreign languages, the latter do not make a dialect incorrect³. Thus, on the basis of the aforementioned definitions explaining correctness, it can be argued that this kind of assessment—the criterion of correctness—is fundamentally inapplicable to dialects. However, as the experience of researchers in other countries attests, studies of perceptual dialectology⁴ are primarily meant here, and assessing dialect codes according to the criterion of (in)correctness is common practice⁵.

In almost every study that follows the principles of perceptual dialectology, tasks assessing speech include questions about the correctness of the codes used by the participants—the members of linguistic communities. This method of assessing and ranking codes⁶ is well-founded. Several so-called language myths reveal people’s inclination to rank speech varieties according to various features. In the book Language Myths (Bauer, Trudgill 1998), linguists use scholarly arguments to refute false beliefs found in individual linguistic communities, or even universal beliefs shared by speakers of different languages, concerning the features and value of language varieties or languages⁷. However, it is undeniable that scholarly discourse, which emerges in response to what we might call the linguistics of language feeling, operates only in certain sectors of society (of course, primarily among philologists), while ordinary language users’ attitudes arise, spread, and change without heeding either the laws of grammar or linguistic reality.

funkcionavimo visuomenėje schemą – Miliūnaitė 2009: 17); vadinasi, negalima neįvertinti to, kad kai kuriais atvejais kalbos patirtis yra šiek tiek dirbtinai sukuriama, o apie tarmes to pasakyti negalima.

³ As R. Miliūnaitė observes, “it makes sense to speak of errors only in the use of the standard language, not of dialects or sociolects, although phenomena uncharacteristic of them may also occur in the latter” (Miliūnaitė 2008: 155).

⁴ For an account of the principles of perceptual dialectology (English perceptual dialectology), see Preston 1988: 373–395; 1989; 1999: xxiii–xl; also Aliūkaitė 2006a: 88–89; 2006b: 567–583; 2007: 52–61.

⁵ Tirti: amerikietiškasis anglų kalbos variantas (Preston 1989: 51–70); Vokietijos dialektai (Dailey-O’Cain 1999: 229–231); Prancūzijos dialektai (Kuiper 1999: 243–262); Turkijos dialektai (Demirci, Kleiner 1999: 263–282); prancūzų kalbos variantai Kvebeke (Kanadoje) (Evans 2002: 83–86); anglų kalbos variantai Kanadoje (McKinnie, Dailey-O’Cain 2002: 282–284) ir kt.

⁶ Not necessarily according to the criterion of correctness.

⁷ In the book Language Myths, Ray Harlow refutes the myth that some languages are not good enough; Edward Carney, that English pronunciation is catastrophic; Lars Gunnar Andersson, that some languages are more complex than others; and Howard Giles and Nancy Niedzielski, that Italian is beautiful, whereas German is ugly, and so on (see Bauer, Trudgill 1998: 9–14; 32–40; 50–57; 85–93).

260

KALBOS KULTŪRA | 82

It can be argued that the attitudes of members of linguistic communities are influenced by various stereotypes, and that school knowledge (more or less objective) may also have some bearing on them. It is plausible that the factor of linguistic security (or insecurity) also shapes differing views of different codes; this factor is undoubtedly connected with the social security and self-esteem of community members. At times, as researchers’ experience attests, attitude studies may reveal that a dialect inferiority complex⁸ is also present in the community, and this may be associated with both linguistic preferences and linguistic attitudes. Meilutė Ramonienė’s study of linguistic attitudes in Joniškėlis is worth mentioning. The researcher’s observations provide good grounds for believing that a dialect inferiority complex also exists among speakers of other dialects. Presenting the study data, the researcher notes that respondents in Joniškėlis regard the dialect as an incorrect linguistic variety as well (“at school you have to speak correctly, not in dialect”; “our dialect is especially incorrect” (for more, see Ramonienė 2006: 144)). Undoubtedly, linguistic identity, for example, affects not only linguistic behavior but also attitudes (cf. Aliūkaitė 2005: 101–108). In light of this, it becomes clear that the question of the correctness of a dialect code must at least be raised. Whether correctness can be measured is another matter.

The subject of this article is the correctness of dialect codes. It is important to emphasize that, on the basis of perception-study data, the aim is to describe the correctness of dialect codes not from the researcher’s perspective, but from that of the language user, that is, the perceiver (evaluator). To provide a scholarly basis for the concept of correctness in dialect codes, aspects for determining correctness must be identified. Although, as A. Girdenis and A. Pupkis observe, “the norm of each dialect taken separately is usually even stricter than that of the standard language” (Girdenis, Pupkis 2000 [1978]: 98), the relevant feature of dialect codes—their lack of codification—renders such considerations meaningless. One might argue that assessing the distribution of distinctive dialect features in a particular dialectal expression could have some validity, but this would not constitute requirements for the correctness of dialect speech. Thus, it must be acknowledged that scholarly evaluations of linguistic facts and the attitudes of ordinary members of a linguistic community are different things.

⁸ Dėl Takesi Sibatos dialekto nevisavertiškumo koncepcijos daugiau žr. Inoue (1999: 148).

D. ALIŪKAITĖ, E. MERKYTĖ. Sociolingvistinis tarminių kodų taisyklingumo tyrimas

261

ATTITUDES RESEARCH: (UN)CORRECTNESS OF DIALECTAL SPEECH

The concluding insights on the concept of correctness are formulated on the basis of a study of the perception and evaluation of dialectal speech conducted in October and December 2008 in Kaunas and Šakiai districts. The study—a corpus of stimulus texts and a questionnaire—was prepared in accordance with the theoretical requirements of perceptual dialectology.

It should be noted that perceptual dialectology studies are most often based on analysing attitudes as responses to a particular linguistic stimulus (cf. Shen, Baykal 2004). Undoubtedly, a user’s (evaluator’s) relationship with a particular variety is best revealed when they hear an example of that variety. To establish the attitudes of members of the linguistic community towards the correctness of dialect codes, a corpus of six stimulus texts was created. Five excerpts, each approximately one and a half minutes long, were selected for this corpus from LKTCH (2004) and Part Two of LT (2005): Text A⁹—from the Tirkšliai site of the Telšiai subdialect of Northern Samogitian (LKTCH 2004: 251); Text B—from the Varėna site of Southern Aukštaitian (LKTCH 2004: 91–92); Text D—from Eastern Aukštaitian Vilnius subdialect (LT 2005); Text E—from the Karklėnai site of the Kretinga subdialect of Northern Samogitian (LKTCH 2004: 266–267); Text F—from the Verkšnionys site of the Anykščiai subdialect of Eastern Aukštaitian (LKTCH 2004: 147).

In compiling the corpus of stimulus texts, perceptual dialectology requirements were not the only consideration. Drawing in part on the principle of the so-called matched-guise technique¹⁰ developed by Wallace Lambert and his colleagues (see Swann 2003: 149), a Standard Lithuanian text (C) was also inserted among the dialect texts. Since the Standard Lithuanian stimulus text C was taken from Lietuvių kalbos bendrinė tartis (2004: 65), a publication prepared by Antanas Pakerys and Aldonas Pupkis for teaching correct pronunciation, the participants were in part misled: a non-dialectal text was presented as a dialectal one for evaluation. On the other hand, however, the participants were given a benchmark for correctness (which they could use—or not!). But, bearing in mind that the narratives in all the texts also differed,

⁹ The texts selected for the study were coded. Researchers adopt different approaches: in some cases stimulus texts are coded, while in others participants know which regions the texts they are listening to come from. Whether a study is open or covert depends on the researchers’ specific aims: for example, naming a location in the tasks may give rise to social stereotypes—when evaluating stimulus texts from a particular location, evaluators may be biased in one way or another. It can be argued that knowing the location influences not only ratings of the text being heard but also ratings of the region’s inhabitants. The regional anonymity of the stimulus texts is thought to ensure the possibility of recording participants’ dialect competence.

¹⁰ English: matched-guise technique.

262

KALBOS KULTŪRA | 82

and informants¹¹, W. Lambert’s idea of the matched-guise technique served only as a methodological point of reference for the spoken-language study design.

After listening to the stimulus texts, participants were asked to evaluate them according to their difference from Standard Lithuanian, (in)correctness, and (un)attractiveness¹², and to describe their choice in a few words. This article discusses the correctness of dialectal speech; the other data are therefore included in the analysis only as supporting arguments or as background to the insights¹³.

When conducting a perceptual study, it is important to choose an appropriate way of presenting questionnaire tasks. Participants were offered a typical seven-point semantic differential rating scale¹⁴. They were to indicate their opinion by marking a number from 7 (does not differ from Standard Lithuanian; completely correct; very beautiful language) to 1 (has nothing in common with Standard Lithuanian; completely incorrect; very ugly language) (see Fig. 1).

FIG. 1. Semantic differential rating scale

¹¹ This does not meet the requirements of the matched-guise technique.

¹² These are typical criteria used in perceptual dialectology studies (see Preston 1999: xxxiv–xxxv).

¹³ In addition, participants were asked to write down and mark on a map of Lithuania (the map in the task showed municipal boundaries and centres) the places they thought the speakers came from. At the end of the questionnaire, respondents were asked to describe in their own words a person who speaks dialectally; however, the data from the mental mapping and conceptualisation of dialectal speech tasks will be discussed in another article.

¹⁴ Studies may also use other scales, for example, five-, six-, or nine-point scales (see Al-Hindawe 1996: [1–9]).

¹⁵ For more on the use of the semantic differential scale, see studies by linguists (Wray, Trott, Bloomer, Reay, Butler 2005: 174–176).

D. ALIŪKAITĖ, E. MERKYTĖ. Sociolingvistinis tarminių kodų taisyklingumo tyrimas

263

The evaluations show respondents’ reactions to the text they hear. Such a rating scale leaves no room for any speculation among the participants that, for example, the researcher would prefer the highest possible score. It should be added that the semantic differential technique proposed by Charles E. Osgood in 1957 (Al-Hindawe 1996: [1–9]) is one of the more common methods in studies of linguistic attitudes; however, representatives of perceptual dialectology often combine this principle with open-ended questions—and in this study, too, participants were asked to explain the ratings they selected.

For the text-stimulus task, the participant sample had to be constituted so that it would be possible to assess whether respondents’ linguistic environment influences their evaluation of dialect discourse, i.e. the text stimuli. Two quotas of approximately equal size were established:

1) The first comprised participants from the KauMi¹⁶ group (46 in total); these are residents of a major city whose linguistic environment is (presumably) closest to the standard language.

2) The second comprised LukGr participants (40 in total); these are provincial residents whose linguistic environment (presumably) has more pronounced features of the Western Aukštaitian Kaunas dialect.

The attitudes of young people aged 14–18 (i.e. students in grades 9–12) were examined. The participants are representatives of a single age group. They are the generation of the future—their attitudes toward the correctness of the dialect code (and, more generally, all the attitudes that emerged in the study) can be regarded as indicators of one generation’s linguistic and cultural awareness.

One additional criterion was also taken into account when establishing the participant quotas. It may be assumed that gender can also influence respondents’ attitudes toward dialect speech. As noted in works on perceptual dialectology, gender affects not only speech but also attitudes toward other variants (see Demirci 2002: 41–50). Thus, efforts were made to ensure that the samples of male and female respondents were similar in size: KauMi—41.3 percent female respondents and 58.7 percent male respondents; LukGr—50 percent female respondents and 50 percent male respondents.

In summary, it can be stated that the participant groups were constituted on the basis of three assumptions—that attitudes and dialect competence may be influenced by the participants’

¹⁶ The article uses the following abbreviations for participant groups: KauMi—students of Kaunas Milikoniai Secondary School (assistance provided by teachers A. Babajan and R. Sventkauskienė); LukGr—students of Lukšiai Vincas Grybas Secondary School (assistance provided by teacher A. Urbaitienė).

264

KALBOS KULTŪRA | 82

linguistic environment, the age of the participants, and the participants’ gender¹⁷. It should be added that the distinguishing value of age would become most apparent only in comparative studies, that is, with data from several generations; thus, in this case, one can only formulate observations about the tendencies of the generation in question. In general, given the small sample of participants, it must be noted that the study’s results should be regarded as premises for further research¹⁸.

Before analyzing the data from the study under discussion, it should be noted that the work of scholars who have investigated evaluations of correctness, empirically or at least theoretically, reveals a tendency to link this evaluation with other evaluations of the dialectal code. First, the correctness of a variety is associated with its sociocultural connotation (cf. Dailey–O’Cain 1999: 227–242; Benson 2003: 307–330). Researchers emphasize that evaluations of a code’s correctness are also influenced by the so-called linguistic homeland¹⁹ factor, that is, people tend to regard their own region’s or their own code as more correct (cf. Diercks 2002: 51–70)²⁰. In addition, language studies show that varieties more similar to the standard language are evaluated more favorably (Aliūkaitė 2007: 152).

The correctness ratings for the texts used as stimuli in the study were calculated by adding the participants’ ratings of the texts and dividing the resulting sum by the number of evaluators. To present the data obtained precisely, the overall correctness ratings for the KauMi and LukGr groups’ texts used as stimuli were calculated. To reveal the influence of gender on ratings of dialectal-code correctness, ratings were calculated separately for female and male respondents in each group (see Figures 2 and 3).

¹⁷ When drawing samples, these assumptions are also taken into account in other perceptual dialectology studies conducted in Lithuania (cf. Aliūkaitė 2007: 143–146).

¹⁸ On the other hand, perceptual dialectology studies are characterized by an inductive principle: after a study has been conducted, the characteristics of a part (i.e., the community members studied) are transferred to the whole (i.e., the entire community); thus, evaluations of varieties are generalized as tendencies reflecting the attitudes of the entire community (cf. Preston 1989: 25–49; Dailey–O’Cain 1999: 227–242; Kuiper 1999: 243–262; Demirci 2002: 41–50). However, it must be noted that researchers formulate generalizations as tendencies, rather than unequivocal conclusions, thereby retaining their provisional nature. The same approach is taken in this article.

¹⁹ English linguistic homeland.

²⁰ Of course, factors such as linguistic insecurity or the perceived inferiority of a dialect must not be overlooked.

D. ALIŪKAITĖ, E. MERKYTĖ. Sociolingvistinis tarminių kodų taisyklingumo tyrimas

265

FIG. 2. Part of the KauMi M67 questionnaire (correctness ratings)

266

KALBOS KULTŪRA | 82

Part III

3.1. Evaluate the correctness of text A:

completely correct

completely incorrect

7 6 5 4 3 2 1

Why are opportunities wasted

3.2. Evaluate the correctness of text B:

completely correct

completely incorrect

7 6 5 4 3 2 1

Why Daina bat… “says”

3.3. Evaluate the correctness of text C:

completely correct

completely incorrect

7 6 5 4 3 2 1

Why

3.4. Evaluate the correctness of text D:

completely correct

completely incorrect

7 6 5 4 3 2 1

Why very short, other...

3.5. Rate the correctness of text E:

completely correct

completely incorrect

7 6 5 4 3 2 1

Why

3.6. Rate the correctness of text F:

completely correct

completely incorrect

7 6 5 4 3 2 1

Why

FIG. 3. Part of the LukGr V32 questionnaire (correctness ratings)

D. ALIŪKAITĖ, E. MERKYTĖ. Sociolingvistinis tarminių kodų taisyklingumo tyrimas

267

In similar studies, correctness ratings are compared with ratings of other codes; therefore, the article presents the overall data for both groups on difference from the standard language and the aesthetic value of the text stimuli (see Figures 4 and 5).²¹

CORRECTNESS BK PA RAv RAa Ž1t Ž2k KauMi 6,65 4,28 3,17 2,93 2,93 2,52 BK PA RAa RAv Ž2k Ž1t LukGr 6,15 4,2 3,52 3,37 2,67 2,6DIFFERENCE from BK BK PA RAa RAv Ž1t Ž2k KauMi 6,63 4,69 3,32 3,21 2,95 2,56 BK PA RAa RAv Ž2k Ž1t LukGr 6,25 4,22 3,4 3,25 2,65 2,37ATTRACTIVENESS BK PA Ž1t RAv RAa Ž2k KauMi 6,3 4 3,73 3,47 3,43 3,36 BK PA RAa Ž2k RAv Ž1t LukGr 6,2 4,35 3,5 3,4 3,37 2,9

FIG. 4. Correctness, difference from BK, and attractiveness

CORRECTNESS Ž1t PA BK RAv Ž2k RAa KauMiM 2,84 4,21 6,63 3,1 2,57 2,73 KauMiV 3 4,33 6,66 3,22 2,48 3,07 LukGrM 2,5 4,45 6,65 3,25 2,85 3,95 LukGrV 2,7 3,95 5,65 3,2 2,5 3,1

FIG. 5. Correctness, difference from BK, and attractiveness

Correctness ratings of dialect codes allow certain observations to be made. The study revealed that young respondents tend to regard dialect speech as more incorrect than correct. Only the text stimulus B=PA, judging by the overall data for the KauMi and LukGr groups, was rated as more correct than incorrect (more than 3.5 points out of 7). However, even these ratings indicate a certain deficiency of the code according to this criterion.

²¹ Coded names of the text stimuli: A – Ž1t (Samogitian Telšiai dialect text), B – PA (Southern Aukštaitian text); C – BK (standard language text); D – RAv (Eastern Aukštaitian Vilnius dialect text); E – Ž2k (Samogitian Kretinga dialect text); F – RAa (Eastern Aukštaitian Anykščiai dialect text).

268 KALBOS KULTŪRA | 82

Both Eastern Aukštaitian texts and both Samogitian texts (see Figures 4 and 5) received stricter ratings. It is likely that these ratings are related to difference from the standard language. The correspondence between these two ratings is evident. In this connection, one may recall Renée van Bezooijen’s so-called intelligibility, recognition, and similarity hypotheses²². The scholar links the aesthetic value of a code to the variant’s perceptibility, comprehensibility, proximity to the evaluator’s own way of speaking, or similarity to the standard language (see Bezooijen 2002: 14–15). Although the author proposed these hypotheses to explain attractiveness ratings, it is likely that the factors she identifies are also important for the correctness value of a code. This observation is also supported by the participants’ comments.

Respondents in both groups were reluctant to explain their ratings, but recurring arguments reveal certain tendencies. The comments make clear that (lack of) comprehensibility is a factor in correctness; for example, comments such as²³ – A=Ž1t: those words are strange (LukGr M36, 1 point), many words were impossible to understand (KauMi M67, 2 points); D=RAv: people use their own words, which sound distinctive (LukGr M8, 4 points), unclear endings (KauMi V58, 3 points); E=Ž2k: I understood every third word (LukGr V23, 1 point), hard to understand, many incorrect words and usages (KauMi M61, 2 points); F=RAa: not very correct, so you have to listen carefully (LukGr M8, 3 points), many new words and so on (KauMi V73, 4 points).

The participants’ comments show that the notion of the standard language (whether accurate or inaccurate is a question for other studies) is an imperative measure at the cognitive level; for example, A=Ž1t: doesn’t sound like standard language, incorrect stress (LukGr M10, 1 point), does not conform to standard language (LukGr V34, 2 points). In addition, the comments show that participants tend to trust their own competence and knowledge of language culture, evaluating the texts by identifying errors; for example, A=Ž1t: there are pronunciation errors, incorrect words (LukGr V18, 4 points); B=PA: there are words that should not be used (LukGr M29, 5 points), there are barbarisms, “t” is changed to “c” (KauMi M62, 5 points); D=RAv: full of barbarisms (KauMi M76, 2 points), incorrect, because the stress is placed badly (LukGr V19, 1 point); E=Ž2k: there are incorrect adverbs (KauMi V52, 5 points), many stylistic errors (KauMi V74, 2 points); F=RAa: the stress is placed badly (LukGr V19, 1 point).

Strict evaluation may also be associated with a desire to distance oneself from such a code. These observations would be better supported by larger samples. Thus, it is possible to conclude only

²² Angl. the intelligibility driven hypothesis; the familiarity driven hypothesis; the similarity driven hypothesis.

²³ Language not edited.

D. ALIŪKAITĖ, E. MERKYTĖ. Sociolingvistinis terminų kodų taisyklingumo tyrimas

269

the assumption that the so-called linguistic-homeland factor is an important determinant of the participants’ attitudes. This is shown by the participants’ comments on their C=BK ratings (high ones!), for example: correct, because I live (KauMi M83, 7 points), I make no mistakes, because I speak that way too (KauMi V64, 7 points)²⁴.

The C=BK text-stimulus ratings reveal that the young respondents more or less recognized the standard-language text²⁵. Their comments on the ratings support this claim, for example: everything is as it should be (KauMi M62, 7 points); the stress pattern matches that of standard language (KauMi M59, 6 points); the entire speech was impeccably correct (KauMi M67, 7 points); correct speech (LukGr M6, 7 points); the speech is very correct (LukGr M11, 7 points), and so on.

Although researchers note that code attractiveness is perhaps the most independent variable and should not be linked to other features (cf. Kuiper 1999: 243–262), the data obtained nevertheless suggest a tendency to associate correctness (and difference from standard language) with attractiveness (see Fig. 4). Only in the case of the E=Ž2k text-stimulus ratings could one claim that the correctness rating does not become an absolute prerequisite for judging the variety unattractive (or at least less attractive than attractive). The nearly 1-point difference in E=Ž2k text-stimulus ratings according to the criteria of correctness and attractiveness provides grounds for the further research hypothesis that correctness and attractiveness are unrelated variables.

It can be argued that the participants’ gender is not a significant determinant of correctness ratings (see Fig. 5). The research results cast doubt on the notion that female respondents pay closer attention to speech. There is also no reason to suppose that the younger-generation participants—the respondents—revealed greater covert prestige for the dialect codes (their correctness ratings are not higher than those of the female respondents). As the data show, differences in ratings by place of residence (i.e., the linguistic-environment factor) are also not significant; thus, it can be argued that linguistic attitudes were influenced more by the respondents’ shared age.

CONCLUSIONS AND ASSUMPTIONS

To summarize the research data, it can be said that young respondents primarily associate linguistic correctness with standard language. One question, however, would arise—

²⁴ Cf. D=RAv: similar to ours, only funny (KauMi V84, 6 points).

²⁵ Although only one respondent named it precisely—standard language (KauMi V49, 7 points).

270

KALBOS KULTŪRA | 82

— whether this is related to actual competence or merely to prescriptive school knowledge that the standard language is a correct system and serves (or can serve) as the point of reference for evaluating other forms of expression. The requirements imposed on the standard language tend to be applied to dialect codes; therefore, young respondents rate dialect codes as more incorrect than correct:

1) the rating of a dialect code’s correctness is determined by its similarity to the standard language – the more similar a dialect code is to the standard language, the more correct it is;

2) correctness is considered a condition for successful communication – an incomprehensible text is more or less incorrect.

The differing assessments of the correctness of dialect codes revealed by ratings of the stimulus texts suggest a mental map of dialects among young respondents²⁶, rather than a homogeneous image of dialect speech. The more divergent ratings for Aukštaitian and Samogitian codes provide a basis for further research into whether Aukštaitian dialect codes are considered more correct than Samogitian codes.

The ratings of the dialect stimulus texts by young respondents and the arguments they provide based on numerical values suggest that, for members of a linguistic community, correctness is not merely a principle at the level of speech; it is also a principle at the level of cognitive attitudes. The strict evaluation of dialect codes reveals the public prestige of the standard language.

LITERATŪRA

A1-Hindawe J. 1996: Considerations when Constructing a Semantic Differential Scale. – La Trobe Papers in Linguistics. Volume 9, [1–9].

Aliūkaitė D. 2005: Lingvistinės tapatybės veiksniai: tarminio kalbėjimo pasirinkimas. – Valoda dažādu kultūru kontekstā. Zinātnisko rakstu krājums 15, Daugavpils: Saule, 101–108.

Aliūkaitė D. 2006a: Perceptyvinės dialektologijos tyrimai: dialekto vaizdinys. – Valoda dažādu kultūru kontekstā. Zinātnisko rakstu krājums 16, Daugavpils: Saule, 87–95.

Aliūkaitė D. 2006b: Tarminio kalbėjimo kultūrinis aspektas. – Acta Baltico-Slavica 30, 567–583.

Aliūkaitė D. 2007: Tarminis kalbėjimas ir bendrinė kalba: objektyviųjų ir subjektyviųjų skirtumų kultūrinė vertė. [Rankraštis: daktaro disertacija], Vilniaus universitetas.

Aliūkaitė D. 2009: Perceptyviniai dialektų tyrimai: mentalinis žemėlapis. – Respectus Philologicus 15 (20), 164–179.

²⁶ Tokias įžvalgas galima daryti ir iš kito tyrimo duomenų (žr. Aliūkaitė 2009: 164–191).

D. ALIŪKAITĖ, E. MERKYTĖ. Sociolingvistinis tarminių kodų taisyklingumo tyrimas

271

Bauer L., Trudgill P. 1998: Language Myths, New York: Penguin Books.

Benson E. J. 2003: Folk Linguistic Perceptions and the Mapping of Dialect Boundaries. – American Speech 78 (3), 307–330.

Dailey-O’Cain J. 1999: The Perception of Post-Unification German Regional Speech. – Handbook of Perceptual Dialectology. Volume 1. Ed. by D. R. Preston, Amsterdam, Philadelphia: John Benjamins Publishing Company, 227–242.

Demirci M. 2002: Gender Differences in the Perception of Turkish Regional Dialects. – Handbook of Perceptual Dialectology. Volume 2. Ed. by D. Long, D. R. Preston, Amsterdam, Philadelphia: John Benjamins Publishing Company, 41–50.

Demirci M., Kleiner B. 1999: The Perception of Turkish Dialects. – Handbook of Perceptual Dialectology. Volume 1. [etc.], 263–282.

Diercks W. 2002: Mental Maps. Linguistic-Geographic Concepts. – Handbook of Perceptual Dialectology. Volume 2. [etc.], 51–70.

Evans B. 2002: Attitudes of Montreal Students towards Varieties of French. – Handbook of Perceptual Dialectology. Volume 2. [etc.], 71–93.

Girdenis A., Pupkis A. 2000 [1978]: Bendrinės kalbos norminimo ir kodifikavimo principai. – A. Girdenis. Kalbotyros darbai 1, Vilnius: Mokslo ir enciklopedijų leidybos institutas, 97–110.

Inoue F. 1999: Classification of Dialects by Image: English and Japanese. – Handbook of Perceptual Dialectology. Volume 1. [etc.], 147–160.

Kniūkšta P. 2001 [1979]: Lingvistinis ir socialinis kodifikacijos santykis su norma. – P. Kniūkšta. Kalbos vartosena ir tvarkyba, Vilnius: Lietuvių kalbos institutas, 62–71.

Kuiper L. 1999: Variation and the Norm. Parisian Perceptions of Regional French. – Handbook of Perceptual Dialectology. Volume 1. [etc.], 243–262.

LKTCH 2004: Lietuvių kalbos tarmių chrestomatija. Sudarė R. Bacevičiūtė, A. Ivanauskienė, A. Leskauskaitė, E. Trumpa, Vilnius: Lietuvių kalbos institutas.

LT 2005: Lietuvių tarmės. II dalis. Sudarė R. Bacevičiūtė, K. Juškevičius, A. Leskauskaitė, K. Morkūnas, D. Mikulėnienė, E. Ožeraitis, L. Telksnys, D. Vaišnienė, A. Vidugiris ir kt., Vilnius: Lietuvių kalbos institutas, Matematikos ir informatikos institutas.

McKinnie M., Dailey-O’Cain J. 2002: A Perceptual Dialectology of Anglophone Canada from Perspective of Young Albertans and Ontarians. – Handbook of Perceptual Dialectology. Volume 2. [etc.], 277–294.

Miliūnaitė R. 2008: Dėl kalbos kultūros mokymo. – Kalbos kultūra 81, 151–169.

Miliūnaitė R. 2009: Dabartinės lietuvių kalbos vartosenos variantai, Vilnius: Lietuvių kalbos institutas.

272

KALBOS KULTŪRA | 82

P a k e r y s A., P u p k i s A. 2004: Lietuvių kalbos bendrinė tartis. Kompaktinės plokštelės tekstai, Vilnius: Gimtasis žodis.

P r e s t o n D. R. 1988: Methods in the Study of Dialect Perceptions. – Methods in Dialectology. Ed. by A. R. Thomas, Clevedon: Multilingual Matters, 373–395.

P r e s t o n D. R. 1989: Perceptual Dialectology: Nonlinguists’ Views of Areal Linguistics, Dordrecht: Foris.

P r e s t o n D. R. 1999: Introduction. – Handbook of Perceptual Dialectology. Volume 1. [etc.], xxiii–xl.

P u p k i s A. 2005: Kalbos kultūros studijos, Vilnius: Gimtasis žodis.

R a m o n i e n ė M. 2006: Nuostatos dėl bendrinės kalbos ir tarmės: Joniškėlio atvejis. – Kalbos kultūra 79, 137–147.

Š e n M., B a y k a l N., 2004: Evaluational Reactions of Standard Accented Turkish Speakers towards Accented Speech and Speakers. – TRANS. Internet-Zeitschrift für Kulturwissenschaften, No. 15/2003. Prieiga per internetą: http://www.inst.at/trans/15Nr/06_1/sen_baykal15.htm

S w a n n J. 2003: Language Choice and Code-switching. – R. Mesthrie, J. Swann, A. Deumert, L. W. Leap. Introducing Sociolinguistics, Edinburgh: Edinburgh University Press, 148–183.

Š u k y s J. 2003: Kalbos kultūra visiems, Kaunas: Šviesa.

V a n B e z o o i j e n R. 2002: Aesthetic Evaluation of Dutch. – Handbook of Perceptual Dialectology. Volume 2. [etc.], 13–30.

W a r d h a u g h R. 2006: An Introduction to Sociolinguistics, Oxford: Blackwell.

W r a y A., T r o t t K., B l o o m e r A., R e a y S., B u t l e r Ch. 2005: Projects in Linguistics. A Practical Guide to Researching Language, London: Replika Press Pvt. Ltd.

Gauta 2009 11 10

A SOCIOLINGUISTIC INVESTIGATION OF GRAMMATICAL WELL-FORMEDNESS OF DIALECTAL CODES

S u m m a r y

The concept of grammatical well-formedness is concerned with standard language. Dialects are spontaneous systems, they show the development of a language and usage peculiarities. Any dialect might adopt elements from foreign systems; however, they do not become ungrammatical. It can thus be claimed that the cri-

D. ALIŪKAITĖ, E. MERKYTĖ. Sociolingvistinis tarminių kodų taisyklingumo tyrimas

273

terion of grammatical well-formedness is in principle unsuitable for dialects. However, the experience of researchers from other countries has shown that evaluating dialectal codes in terms of grammatical well-formedness in the framework of perceptive dialectology is usual practice.

The present investigation focuses on the grammatical well-formedness of dialectal codes. Considering the data of perceptive research, the paper aims at defining the well-formedness of dialectal codes from the user’s/receiver’s (evaluator’s) point of view rather than the researcher’s. The concluding remarks on the well-formedness of dialectal codes are based on a research conducted in October-December of 2008 in the districts of Kaunas and Šakiai and focusing on the perception and evaluation of dialectal speech by secondary schoolstudents in their 9-12 year at school (aged 14-18). The research materials covered a corpus of 6 text-stimuli and a questionnaire; both were prepared in accordance with the requirements of perceptive dialectology.

To conclude, young respondents adhere to a certain prototype of a well-formed expression. For them, well-formedness is primarily concerned with standard language. However, in this context a natural question could arise: is well-formedness related with the school students’ actual experience or confined to the knowledge received at school that language is a well-formed system and that it serves (or can serve) as a reference to evaluate another expression? Since dialects are usually discussed in the framework of general requirements applicable to standard language, they are usually treated as ungrammatical rather than grammatical.

A major factor determining the well-formedness of a dialectal code is its affinity to standard language: the more similar the dialectal code is to standard language, the more grammatical it is. Young respondents consider the well-formedness as a major factor of communicative success: if the text is difficult to understand, it is also more or less ungrammatical.

Differing evaluation of the dialectal codes of Highlanders (aukštaičių) and Lowlanders (žemaičių) provides basis for further research that dialectal codes of Highlanders are treated as more grammatical than Lowlanders.

DAIVA ALIŪKAITĖ

Vilniaus universiteto Kauno humanitarinis fakultetas

Muitinės g. 8, LT-44280 Kaunas

[email protected]

ERIKA MERKYTĖ

Vilniaus universiteto Kauno humanitarinis fakultetas

Muitinės g. 8, LT-44280 Kaunas

[email protected]

274

KALBOS KULTŪRA | 82