2. Hawick and Southern Scots
a. Geography, history and population
b. Southern Scots and Braid Haaick
3. Methodology and attitudes
a. Attitudes and its measurement
4. Linguistic situation and language use in Hawick
a. Linguistic situation
i. Status of standard speech and local varieties in general
ii. Regional variation in the Borders
iii. Attitudes towards BBC London English and local speech
iv. Status of local speech versus Standard English
v. Language change in Hawick and the Borders
vi. Local identity
b. Language use and attitudes at Hawick High School
i. Language use in the school context
ii. Attitudes towards the vernacular at school
6. Appendix A
7. Appendix B
8. Appendix C
Firstly, I would like to thank pupils and teachers in Hawick alike for spending some of their time on my project. I also thank the dominie of Hawick High School, Mr. Neil Horne, who both welcomed and supported me in the friendliest way imaginable, Eileen and Ron Smith for their hospitality and kindness at their wonderful B&B, the chippies in Hawick and the canteen at HHS for providing me with good, substantial Scottish food during my three weeks’ stay, the staff of the Modern Languages Department for supplying me with a base for my study, the Dr. Sophie-Bernthsen Fund in Heidelberg for supporting my field work financially, and last but not least my mother and all my friends for supporting me.
Ich erkläre, dass ich die Arbeit selbständig und nur mit den angegebenen Hilfsmitteln angefertigt habe und dass alle Stellen, die dem Wortlaut oder dem Sinne nach anderen Werken entnommen sind, durch Angabe der Quellen als Entlehnungen kenntlich gemacht worden sind.
Heidelberg, im April 2003
During my stay as a Foreign Languages Assistant in the Scottish Borders in the exchange year 2000/2001, I used to work at three different comprehensive schools: Kelso High School, Jedburgh Grammar School and Hawick High School. Hawick, perhaps more than any other town in the South of Scotland, is an outstanding place to rouse the linguistic interest of a foreign learner of English. First of all, the rather peculiar and mystic local motto in the town crest - as can be seen in the preliminaries - and, secondly, the pronunciation of the place itself. According to Collins English Dictionary, the name Hawick is phonetically transcribed as /h]+wk/. However, as heard with my own ears, everybody in the Border region pronounces the name as /h]wk/.
Yet, apart from these features, one other local element was even more striking. While I had not too many problems with the pupils’ speech at Jedburgh and Kelso, the same thing could not be said of Hawick. In the first week, when asking the class to explain why they kept bullying a certain boy in class, a pupil responded: “because yin’s mingin!”. Obviously, I did not have the slightest clue and inquired what was meant by this remark. After some hesitation the children came up with the following explanation: “because this one is smelling!”, the Standard English equivalent. This was only one of many more similar incidents that made me curious. After consultations with the teachers in Hawick and in other towns of the Borders, it soon became clear that inhabitants from Hawick are very proud of their own speech; different, as they claim, from all the other Border towns. They insisted further that their vernacular is still one of now few remaining and very lively examples of the Southern Scots dialect. The following weeks and months in the Borders made me appreciate the local speech even more fully.
Back home, inspired by this ten months’ stay in Scotland, I soon considered the idea to write a thesis about the linguistic situation in Hawick, the attitudes attached with the dialect and how these attitudes are reflected at the local High School. Encouraged by Dr. Mohr’s seminar about Lowland Scots and Prof. Dr. Glauser, who helped me with the compilation of a questionnaire and surveyed the present study, it was decided that I should return to the Borders to administer a field study at Hawick High school, the results of which will be presented and discussed in the following study.
The present analysis will focus on two major aspects: the local vernacular as employed in Hawick and its usage in the school context. The emphasis is put on the present linguistic situation in terms of language usage and attitudes of both pupils and teachers. The second chapter will supply a brief social, linguistic and historical outline of Hawick and Southern Scots. Subsequently, in the third chapter the methodology and aspects of language attitudes and their measurement will be described. The fourth chapter, being the main part of the study, is subdivided into two sections. Firstly, the linguistic situation in Hawick in general will be illustrated, and extralinguistic variables are taken into account whenever they prove to be statistically significant. The final part of the analysis, chapter 4.2, will clarify language use and attitudes towards the vernacular in the school context. The conclusion in chapter five will summarize the most important results and might also serve as an impetus for further dialect and attitude studies. Appendix A contains the questions of the two questionnaires, Appendix B lists tables of additional figures and data not included in the study, and Appendix C contains a town map of Hawick and locality scores needed for social stratification.
2 Hawick and Southern Scots
2.1 Geography, history and population
The Scottish Borders Region, a political administration area founded in 1974 and thus replacing the old counties of Scotland, comprises the districts of Roxburgh, Berwick, Ettrick and Lauderdale, and Tweeddale. It covers an area of 1,800 square miles. The Region stretches from the English borderline north to the outskirts of Edinburgh and from the Tweedsmuir Hills in the west to the North Sea cost in the east (cf. Map 2.1.3).
As can be seen in Figure 2.1.1, in 1891 around 130,000 people lived in the Borders. Over the years, the population dropped to a mere 98,477 in 1971 (Baldwin 1995: 107). In the last twenty years, however, the population has increased slightly, and the Borders Council counted 106,311 inhabitants in 1996. Calculations on future population figures project a further rise in numbers, and efforts like the Borders College in Hawick or a dependence of Heriot-Watt University Edinburgh for Textile Studies in Galashiels surely help to restrict any further migration of young people to the big cities owing to the lack of prospects in the fields of employment or vocational training.
Figure 2.1.1: Population Change in the Borders Region
illustration not visible in this excerpt
Another striking characteristic of the populate in the Borders is the fact that it has significantly more people older than 45 years of age and fewer youths compared to the rest of Scotland, as illustrated in Figure 2.1.2 below.
Figure 2.1.2: Age Groups in the Borders
illustration not visible in this excerpt
This fact can be explained by mainly three reasons. Firstly, most pupils who have gained University entrance qualifications at school leave Hawick for the University cities in Scotland or in neighbouring England. This obviously accounts for the comparatively low percentage figures for the group of 16-24 year olds. Secondly, more and more retired English people move to the Borders Region to enjoy a relaxed country life in the quietness and natural beauty of the Borders. In 1986, for example, 30% of the populace of Town and Kirk Yetholm were born outside of Scotland (Baldwin 1995: 114). Finally, Edinburgh commuters have moved south to the Borders in order to avoid the high council taxes and prizes of housing in the capital city. Overall, these are three factors that might account for the “ups and downs” of population figures in the region over the past twenty years.
Although Hawick belongs to the Scottish Borders Region politically, this should not be confused with the linguistic boundaries. In fact, the dialect boundaries of Southern Scots are not identical with the political borders as drawn by the Scottish Executive. More, however, shall be said about dialect division and local speech in chapter 2.2.
Hawick is a municipal burgh, town and parish in the old county of Roxburghshire. According to a prominent study on Scottish place names, “the name is first recorded as Hawic between 1165 and 1169 in the Liber Sancte Marie de Melros, and probably goes back to an Old English haga wīc ‘hedge (or enclosure) farm’” (Nicolaisen 1976: 3). Nowadays, the town has not much in common with a farm anymore. On the contrary, Hawick has the largest population of all settlements in the newly formed Scottish Borders Region with a total number of approx. 15,700 inhabitants. The town is situated on the river Teviot, approximately 40 miles south-east of Edinburgh and 35 miles north of Carlisle. Hawick has gained a reputation for producing fine knitwear, hosiery, tweeds, woollen produces and thus its appearance as a town is still “uncompromisingly industrial” (Steel & Steel 1996: 111). While the main Border towns have abandoned most of their traditional industries, Hawick, unfortunately, “has relied overlong on its now depressed textile industry” (Baldwin 1995: 101), which resulted in a comparatively high rate of unemployment. Knitwear aside, Hawick, “Queen of a’ the Borders”, boasts two masculine traditions: rugby and the Common Riding.
Hawick Rugby Football Club (HRFC), founded in 1873, have won the last two Scottish Championships in 2000 and 2001 and a great majority of Teris (Singular: Teri or Terie, the spelling is not consistent) - people who are born and bred in Hawick, taking the name from a corruption of the town motto - show an enormous emotional and even patriotic interest in the local club’s campaign. It was also Hawick that has brought forward Bill McLaren, the famous rugby commentator for the BBC, who not only has “popularised the game worldwide” (Coutts 1995: 250) but has also proved that commentating on sports in the local vernacular can be as vividly and competently as in any other variety of English without lacking expressions of competence and ingenuity.
At the end of the rugby season in April or May, another event takes over. The Common Riding at the beginning of June has a place in every Teri’s heart. Its historic roots date back to the year 1514: one year after the Battle of Flodden, young people of Hawick won a minor skirmish in the vicinity of Hawick against English troops and are said to have captured the enemy’s flag (Steel & Steel 1996: 116). A statue in town, erected in 1914 and now widely known as “The Horse”, commemorates this event. This “capture” is re-enacted every year with ridings along the commons, and to be chosen to lead the festivities as the Cornet, a representative of the young unmarried men of Hawick, is surely one of the greatest honours the town can bestow on a Hawick lad. The first verse of the song sung at the festival can be seen in the preliminaries. As Lady Judy steel puts it, “[t]he Common Riding there is one of the last bastions of male exclusivity in twentieth century Britain” (Steel 1995: 225). Whether this male supremacy plays a part in language usage and attitudes will be interesting to investigate in the study. The fervent support for these two activities and the emotional bonds that go along with it might be a major reason for the close-knit relations in town and for the local pride in tradition and - even more interesting to us - in the local speech, sometimes called the yow and mei dialect. Why this name is used by both laymen and scholars of English shall be explained in the next chapter.
Map 2.1.3: Scottish Borders Council (http://www.scotborders.org.uk 10.01.03)
illustration not visible in this excerpt
2.2 Southern Scots and Braid Haaick
As mentioned before, the rather curious town  motto shall be examined very briefly. Tyr-ibus ye Tyr ye Odin, a peculiar phrase translated by Murray as “Tyr keep us, both Tyr and Odin!” (1873: 18). In other words, the motto might be a direct address to the heathen gods of the North Anglian kingdom, which had its spheres of influence also in Southern Scotland in the 7th and 8th century (Murray 1873: 17f). Before analysing the connection between the North Anglian dialect and the local vernacular any further, let us first consider the dialect classification of Scots and some of its most striking features.
Linguistically, Scots can be divided into five districts on the basis of distinctive phonetic features: (1) Insular Scots, as spoken on Orkney and Shetland, (2) Ulster Scots, as spoken in parts of Northern Ireland, (3) Northern Scots, subdivided in South Northern Scots, Mid Northern Scots and North Northern Scots , (4) Mid Scots, which can be sub-divided in East Mid Scots, West Mid Scots and South Mid Scots and, most important for the present analysis, (5) Southern Scots, the vernacular of the old counties of Mid and East Dumfriesshire, Roxburghshire and Selkirkshire (Grant 1931: xxiv). Today, the political region of the Scottish Borders comprises more or less two dialect areas: East Mid Scots in Peeblesshire and Berwickshire and Southern Scots in the western counties of Roxburgshire and Selkirkshire. Map 2.2.1 illustrates the area in question. The map is an adaptation from the introduction of the Scottish National Dictionary. The line that divides Scotland in roughly two parts is called the “Highland line”. It separates the Scots speaking people from the English speaking inhabitants in the Highlands and has not changed significantly over the last three centuries (Johnston 1997: 433). Although the above-mentioned classification by Grant is the “most accepted” one (Johnston 1997: 437), minor refinements, especially for Southern Scots, might be useful.
Map 2.2.1: Dialect Boundaries in Scotland (Grant 1931: xxv)
Abbildung in dieser Leseprobe nicht enthalten
Johnston claims that “Southern dialects have been losing their traditional characteristics for some time” (1997: 433) and that “[t]wentieth-century immigration from the Central Belt has not helped the dialect’s survival at all” (1997: 445), a development that surely has affected Hawick as well. Even Murray, over a hundred years ago, cites informants from the towns of Jedburgh, Selkirk and Kelso who complain that the distinctive features, especially the pronunciation of yuw and mey in words like you and me, are not longer observed and that the greater mobility in those places have led to a more “pretentious pronunciation” (1873: 84). However, stressing the dialect dominance of Hawick, another informant claims that among all the towns there are no natives who speak Southern Scots as “emphatically as do the people in and around Hawick” (Murray 1873: 86). Despite all these shifts away from the old dialect forms, some of the most important grammatical, lexical and phonetic features of Southern Scots will be illustrated.
In grammar, Southern Scots still maintains the distinction between the present participle, with the ending – an (/-cn/), He’s chappan at the door, and the gerund, which ends in – een (/-wn/), He likes chappeen at the door (Grant 1931: xxx, see also Murray 1873: 81). Of equal importance is the preservation of inflectional distinctions as in meae and mair in the sentences meae bairns ‘more children’ and an mair tui gie them ‘and more to give them’, the former being a plural and the latter a singular form (Murray 1873: 81). Although Murray deals with several more grammatical features (1873: 150-230), most of them seem to have been lost, and Grant only mentions the distinction between the gerund and the present participle (1931: xxx).
Much more important and significant for Southern Scots is the “peculiarly developed vowel system” (Grant 1931: xxix). Old English ā, which is sounded [e] in other Scots dialect, is diphthongised in words like rope /rwcp/ , loaf /lwcf/ or soap /swcp/. A similar development occurred with the OE ă in open position as in made /mwcd/ or spade /spwcd/. Both Grant and Murray note the usage of this diphthong (1931: xxix and 1873: 105,144). It has to be pointed out that the second element of the diphthong is often lost in rapid speech. OE ō is now realized as [e:] in final position or before /r/, /v/ and /z/ in words like shoe /e:/, yaiz (use) /je:z/ or mair /me:r/ (Grant 1931: xxix). Perhaps the most striking peculiarities are the following two diphthongs as are best illustrated in the famous example given by Murray. The phrase “You and I will go over the wall and pull a pea” can be used to testify whether or not somebody speaks Southern Scots. In this very dialect the phrase should sound somewhat like “Yuw an’ mey ‘ll gang owre the deyke an’ puw a pey”. Apart from lexis, deyke ‘wall’, this sentence explains why the dialect is often referred to as the Yow and Mei dialect (Murray 1873: 82). The OE ū in final position is diphthongised to [u] often in connection with a loss of the final consonant. Examples are pull /pu/, through /tru/ and the notorious you /ju/. All other Scots dialects have oo [u:] in final position (Grant 1931: xxix). The second peculiar element is based on a refinement of all OE vowels that have become ee [i]. In final position they are diphthongised in ey [ei or ci] as in mey /mei/ ‘me’, sey /sei/ ‘see’ or tey /tei/ ‘tea’. Another marked feature is the [æ] sound, heard in (a) words with the spelling e as bed, led, pen, (b) where a appears before ss, sh, r as wash, ash, fasten, harvest, (c) in words with OE e in open syllables with later shortening as feather, fret, tread, (d) in other words with shortened long vowels as bled, bless, met, and finally in words with e [ e] as in vessel, pet (Grant 1931: xxx). Some other dialect features that are mentioned in Murray had already been lost to a large degree by the 1930s when Grant started his survey.
Consonants are normally more resistant to changes. Again most features have been lost since Murray listed them, and even then only very few words or place-names bore the peculiar feature of a λ voiced front lateral in words like tailor or collier (1873: 124) or the voice front nasal in words like lunzie ‘loin’, cuinzie ‘coin’ or gaberlunzie ‘wallet’ (Grant 1931: xxxi, Ellis 1869: 298). However, one consonant feature seems to have survived as I noticed the following pronunciation myself: the use of ch [¡] with back vowels in cough /k]u¡/ or daughter /d]u¡tcr/, and the ch [ç] with the front vowels [i, w, æ, e, e] in licht, micht or fecht (Grant 1931: xxxi, Murray 1873: 82). Although much more could be said of the dialect, the interested reader is referred to Murray’s excellent dialect monograph for an in-depth analysis, Grant’s “Introduction” in the Scottish National Dictionary for a quick overview or Johnston’s article for an up-to-date survey.
Having dealt with some of the most striking features of Southern Scots, it is only sensible to say a few words about the vernacular of Hawick. Although people in the Border towns know that a dialect form like Southern Scots exists, their vernacular is “first and foremost local […] [and] felt to be ‘Hawick’ or ‘Selkirk’ or ‘Jedburgh’ more than it is perceived as Scots” (Johnston 1997: 444). In order to understand what is meant by Braid Haaick, one has to look at the local speech more closely.
Braid Haaick, according to Elliot Cowan Smith, is a “potent and prolific part of the Upper Teviotdale vernacular, which itself is one of the purest and richest branches of the vernacular of Scotland” (1927: 8). Even some gypsy loan-words, though mostly restricted to the Kelso and Yetholm area, have found their way into the local dialect as in words like radgy ‘lewd’, ’wanton’ or manishee ’woman’ (Watson 1919: 9).
Phonetically, Braid Haaick is largely consistent with Southern Scots, and one of its chief characteristics is the strong and frequent use of the diphthongs in yow and mei. An excellent overview of more subtle phonetic differences is given in great detail in Smith (1927: 8-11). The use of oo for both ‘us’ and ‘we’ was not only a feature dominant at the beginning of the century (Watson 1915: 8) but, as will be seen later on, still has a wide-spread use in the town today.
In grammar, Hawick retains some peculiarities as in the irregular plurals bease ‘beasts’, claes ‘clothes’, een ‘eyes’, een-breen ‘eye-brows’, the use of ee for the second person personal pronoun (Smith 1927: 11) and a “faithful preservation of tenses of the defective verb” (ibid. : 13) as, for example, in greet, graat, grutten ‘to cry’. There are of course many more interesting features but a list of further items would definitely go beyond the scope of the present study on language attitudes.
Some of the phrases and idioms listed by Smith in the 1920s can be heard at and around school even today. The following examples might illustrate their liveliness and vitality. Instead of ‘why’, people in Hawick us ‘how’ all the time, causing serious problems of understanding with foreigners. Confronted with the phrase “hes ony o ee ony on ee?” by a teacher in 2001 who, as was explained to me afterwards, inquired if any of his pupils had any coins on them, I was lost completely, and one can surely regard this as a conservation of dialect features. Furthermore, the above phrase is even included in Smith’s word list (1927: 23). The brief glance at the local vernacular shall be concluded with an anecdote in Watson to explain the importance of the different burghs when it comes down to identity and language:
[…] a man, who, on applying for a situation, said he could speak three different languages, and on being asked which these were, he replied, “Hawick, Galshiels, and Selkirk”. (Watson 1915: 8)
3 Methodology and attitudes
3.1 Attitudes and its measurement
Since the present study tries to shed some light on language attitudes in a Scottish town, an overview of the major concepts underlying the word “attitude” shall be given here. Being very prominent in sociolinguistics studies today, research on language attitudes, however, did not begin until the seventies of the last century (Deprez & Persoons 1987: 125). Although there are at least two important and fundamentally different theories, “practically everybody agrees that attitudes are learned from previous experience, and that they are not momentary but relatively ‘enduring’” (Agheysi & Fishman 1970: 139).
The mentalist theory defines attitudes as a ‘mental and neural state of readiness’, and as such they are not “directly observable but have to be inferred from the subject’s introspection” (Agheysi & Fishman 1970: 138). Plausible as this view is in theory, it often shows deficiencies when applied to empirical measurement. Since an attitude is rather an internal readiness than a measurable response, the scholar must “depend on the person’s reports of what their attitudes are, or infer attitudes directly from behaviour patterns” (Fasold 1984: 147).
The behaviourist view simply explains attitudes as “responses people make to social situations” (Fasold 1984: 147), i.e. attitudes are overt responses or behaviour to a certain stimulus (Agheysi & Fishman 1970: 138). This theory of course does not pose too many problems in practice as the researcher only has to observe, collect and analyse the explicitly overt data. Critics, however, claim that if one accepts an extreme behaviourist approach, the collected data are not sufficient to predict other behaviour or responses as attitudes are seen as dependent variables (Agheysi & Fishman 1970: 138; Fasold 1984: 148).
Another difference between the two theories consists in the underlying structure of attitudes. While behaviourists regard attitudes as single, uni-dimensional units, mentalists distinguish between cognitive (knowledge), affective (feeling) and conative (action) components (Agheeysi & Fishman 1970: 138; Fasold 1984: 148). However, the interrelation between the components is neither simple nor straightforward, especially “[t]he cognitive and affective components of attitude may not always be in harmony” (Baker 1992: 12). Nevertheless, the present study mainly adopts the mentalist approach, i.e. the distinction between three components, outlined, amongst others, by Lambert & Lambert in 1964 and modernised and refined by Baker in 1992.
Having discussed the theoretical background very briefly, it is of equal importance to talk about the “how” of data acquisition, i.e. the measurement of language attitudes. Generally speaking, one can distinguish between three techniques: (1) Analysis of societal treatment of language varieties, (2) indirect measures of language attitudes and (3) direct measures.
The first technique has been widely ignored in discussions of attitudinal research as it does not involve inquiries. It includes, for example, observational, participant-observation, ethnographic studies, demographic and census analyses, analyses of government policies, and analyses of broadcasting media, etc. (Giles et al. 1987: 1068f).
Any research designed so as to prevent the subjects from knowing that their language attitudes are under scrutiny is labelled an indirect study (Fasold 1984: 149). The matched-guise technique, developed by Wallace Lambert and colleagues in 1960, is the most widely used indirect method.
The direct method, i.e. interviews, questionnaires and observation, is applied in the present study and will thus be explained in more details. Questionnaires have undergone a high level of formal development mainly because of their extensive use in sociolinguistics over the years (Agheysi & Fishman 1970: 147). Typically, one distinguishes between open and closed question items. The major advantage of open questions, e.g. what makes you a Teri?, is to enable the subjects to express their opinions about a chosen focal object more freely. However, open questions may not always succeed as planned. Firstly, writing down lengthy answers to questions may make the subject refrain from answering as comprehensively as expected. Secondly, though the subjects might provide answers, the responses may fail to meet the target of the question and are therefore useless for the researcher. Finally, not to be underestimated, answers to open question items are not easily evaluated and scored (Agheysi & Fishman 1970: 148). To avoid this crucial problem, researchers prefer closed questions.
According to Agheysi and Fishman, closed questions consist of three components: the focal object, dimension of valuation and a set of rating terms (1970: 148). Often rating terms range from simple “yes” or “no” answers to more sophisticated ones, i.e. the choice from a five-point or seven-point scale. Multiple choice questions are also part of close question items. Unlike the open question items, closed ones single out the respondents’ failing to focus on the expected dimension since they have to choose from a given set of categories. With regard to evaluation, closed questions can be scored much more easily than the open ones. Finally, more than one dimension can be covered by these questions. Unfortunately, one great disadvantage is not to be overlooked. If the questions only require simple “yes” or “no” answers or ticking boxes, subjects may tend to answer automatically; on the other hand, if the items are too complex, respondents may find their task too demanding and stop answering (Agheysi & Fishman 1970: 148f ; Fasold 1984: 152).
An interview is very similar to an open question questionnaire. A fieldworker asks attitude questions and tape-records the subject’s oral responses. The advantages of this method are obvious: the researcher can guide through the interview and thus reduce the chance of boredom or irritation; furthermore, the subjects may virtually feel free to answer as lengthily as they want. However, the major disadvantage is that this method is extraordinarily time-consuming and expensive (Fasold 1984: 152).
Since “language attitudes have tended to make more use of questionnaires than other methods” (Agheysi & Fishman 1970: 151), the present study is based on this method, and the next sub-chapter clarifies what kind of questionnaires were used.
Two questionnaires containing both open- and closed-question items form the basis of the present study. One questionnaire was designed for teachers, comprising 46 questions, and the other one for pupils, displaying 40 items. Most of the questions, however, are identical for the two groups. The two questionnaires are listed in Appendix A, subdivided into four sections. Yet, it should be noted that the subjects did not know about the subdivision and its purpose.
The first section of the two questionnaires mainly covers the subjects’ social background, i.e. age, gender, origin, place of birth, place of residence, partner’s and parents’ occupation and origin, education, newspapers and TV, teachers training, etc. (T1-13, P1-P14). The second part is composed of questions dealing with the linguistic situation in Hawick. For a more detailed analysis this section is further divided into attitudes towards the local vernacular and other dialects (cf. 4.1.1), regionalisation within the Borders (cf. 4.1.2), a comparison of attitudes towards Hawick speech and BBC London English (cf. 4.1.3), questions on the status of local speech and standard depending on situation (cf. 4.1.4), language change in Hawick and the Borders (cf. 4.1.5), and local identity (cf. 4.1.6). The third part further includes questions on the actual language use in Hawick depending on different contact situations. The fourth and final part focuses on the language use and attitudes towards the vernacular in the school context. Obviously, it is here that the questions are different to a large degree for the two groups (cf. chapter 4.2).
By taking together closed- as well as open-question items, subjects were expected to express themselves more freely in the open-question items. However, judging from some written remarks made on the questionnaire, e.g. your questionnaire is rather long!! Lost interest before the end, or unanswered open questions by some of the younger pupils, it became clear afterwards that a shorter closed-question item questionnaire might have provided a more satisfactory result and easier analysis. Problems of this kind will be addressed in the subsequent chapters. Yet, on the whole, the questions were answered seriously and even enthusiastically by most participants from the two groups.
In order to concentrate on the most striking findings without losing coherence and control, several questions were not analysed as they either went beyond the scope of the present study or did not contribute any relevant information. Since an evaluation of questions T26/P26: Do you sometimes use local speech? If yes, how would you describe your own use of the local speech?, T27/P27: What do you speak when you don’t speak local speech? If you don’t speak local speech, what would you say people from Hawick speak?, T28/P28: Which from the following statements is true for you?, and T29/P29: What language do you use when you speak to…? (cf. Appendix A) would have gone beyond the scope of the study, it was decided to concentrate on the language use and attitudes in the school context and exclude these four questions about general language use from the analysis.
Furthermore, questions about additional social data such as T12: What local papers do you read, T13: Which non-local papers do you read, P13: After school do you want to leave Hawick? What kind of exams are you planning to take?, P14: What are your favourite British TV-programmes?, were also not part of the analysis. Question T18/P19: Do you think men usually talk with a broader accent than women? was also not included in the evaluation since this item could not be grouped into any subcategory and did not produce any valid information.
The fieldwork in Scotland involved a three-week stay in Hawick with regular visits to Hawick High School on school days. According to the headmaster, Mr. Neil Horne, the school employs 73 teachers who educate and assess 1,000 pupils. Translated into the terminology of statistics, HHS has a teacher population of 73 and a pupil population of 1,000. These figures make HHS the largest school in the Border region. In general, pupils from ten different primary schools continue their secondary education at HHS. The schools in question are Burnfoot primary school, Drumlanrig, Wilton, Trinity, Stirches, Denholm, Newcastleton, Teviothead, Roberton and Hobkirk. The first five schools are all situated in Hawick, whereas the latter five can be found in villages in the vicinity. It was of course impossible to question the whole population in the short three-week period, therefore a random sample of 20 teachers and 29 pupils was taken. The questionnaires were filled in anonymously and each subject was later codified according to age and gender. It was decided to keep the letter key as simple as possible: the numerical digit just indicates a random listing; the first letter shows whether the subject is a (p)upil or a (t)eacher and the second one represents the respondent’s gender (cf. table 3.3.1). The results were then projected to the population as a whole. However, in order to obtain valid results, “the sample should be a microcosm of the whole population” (Fasold 1984: 86). As the sample contains both females and males, Teris and non-Teris in a fairly balanced number, the sample might well be regarded an adequate representation of the population.
Of the subjects, twenty-seven are female and twenty-two male. Thirty-three of all respondents are Teris and sixteen are not. In the present study, Teri will be defined as having grown up and spent the childhood in Hawick. Although some informants have spent a great part of their adult life in Hawick, they are not classified as Teris as it is the childhood, largely influenced by the parents and the peer groups, that has a huge influence on the subjects’ dialect and speech (Sandred 1983: 33).
Table 3.3.1 The Hawick sample
illustration not visible in this excerpt
On arrival, Mr. Horne welcomed the field project heartily. Together with him the modus operandi was discussed. It was agreed upon contacting the teachers personally and explaining the study to them. They were then asked to take the questionnaire along and fill it in whenever they got time. Although it might had been better to be with the teachers when they were answering the questionnaire, the tight schedule at school made this virtually impossible. Having been returned, the questionnaires were then checked to avoid obscurity. Apart from one teacher who retired one year ago because of health problems, 3tm, all teachers are still active at HHS. As the one retired teacher still has a close connection to the school and takes part in school festivities, he was counted among the active teachers.
Since the teachers did not seem to favour the idea of me choosing the pupils from individual classes, another modus had to be chosen. Pupils from school year 4, taking their Standard Grade exam at the end of the school year, were selected from several classes to assemble in a class room to fill in the questionnaires, both supervised by a teacher and me. Before they started to work on it, they were given a brief explanation of the questionnaire and its purpose. The same procedure was then repeated a week later with pupils from school year 5, taking their Higher exam at the end of term. It took the pupils approximately one school period, i.e. 40 minutes, to fill in the questions. Initially, the questionnaire was to be filled in by younger pupils as well but teachers’ consideration of the questionnaire as being to lengthy and complex led to a restriction to fourth and fifth year pupils. On the whole, the pupils showed great interest in the project, and the majority was very conscientious although for a few fourth year pupils the questions seemed to have been overly demanding.
In the present study, the data will be analysed according to the following social variables: Teri vs. non-Teri (origin), male vs. female (gender), teachers vs. pupils (age) and finally social class membership.
As regards age, the distribution has already been mentioned. There are twenty-nine pupils and twenty teachers. The youngest informant was fourteen and the oldest one sixty-two. The subjects could therefore be allocated to two age groups: Group I consisting of 14-16 years old pupils ((N)umber = 29) and Group II consisting of 28-62 years old teachers (N = 20). A further distribution of the teacher sample into age groups was not feasible as it would have required a much larger sample to be statistically relevant.
However, there are some problems linked to the extralinguistic marker social class. Although Hawick is definitely a settlement in a predominantly rural area, it was decided to apply the same social class divisions as in previous urban studies (cf. Trudgill 1972, Labov 1966). The reasons for this are the industrial past of the town and its being the largest settlement in the Borders. Moreover, there is probably no other town in the Borders where membership of a social class depends on locality, i.e. the part of Hawick one lives in.
An index was calculated based on two factors: parents’ occupation and locality. Following Macaulay’s concept of occupation (1997: 86), four groupings were then distinguished: Class I = professional workers, employers, managers; Class II = non-manual workers; Class III = foremen, skilled manual workers and Class IV = personal service, semi-skilled, agricultural. Scores were allocated to the classes with Class I scoring four points and Class IV valued one point. Locality in Hawick and its environment was divided into mainly three areas: significant for upper or upper middle class (3 points), not significant for any specific class (2 points) and significant for lower middle and lower class (1 point). The division was undertaken on the basis of Ron Smith’s rating as established via an e-mail enquiry. Appendix C shows the localities in town and the points attributed to them.
Unfortunately, Weensland, Wilton and West End are very mixed areas and, according to my informant 3tm, contain pockets of all three social classes. As a compromise, they were attributed two points for locality. The villages around Hawick were given an average two points on the scale. The scores for locality and occupation were then added up to obtain a score ranging from a minimum of two to a maximum of seven points. Scores from 2-3 were labelled Group C, scores from 4-5 Group B, scores from 6-7 Group A. In order to avoid teachers being indifferently allocated to Group A, their scores were calculated differently. The teacher’s occupation score, i.e. 4, was added to the father’s occupation, then divided by two and finally added to the locality score. Subject 1tf, for example, achieves 4 points for her occupation plus 2 points for her father’s occupation. Divided by two we obtain 3 and added to the two points for locality we end up with 5 points. She is thus labelled Group B on the list. Pupils’ scores are more easily calculated. They simply consist of the sum of the score for the parent who is highest on the occupation scale plus the locality score. Pupil 1pf, e.g., achieves 4 points - two for her father’s profession and two for the locale. All together, Group A has 14 members, Group B comprises 25 informants, and Group C consists of 10 subjects.
As the collected data were obtained via different question types and scales, different evaluation methods had to be applied. The Statistics Package for the Social Sciences (SPSS), Version 11.0 for Windows, proved very helpful in the analysis as it computes all necessary calculation and offers a wide range of statistical tests.
The answers to the open question items are treated as nominal, i.e. discrete, data and can be analysed with the help of a χ²-test. Multiple answers are illustrated in cross tables. However, the responses had to be coded first, i.e. similar answers were grouped together.
Most answers to the closed question items consisted of either ordinal or interval data. Both these data are called continuous data and can be measured on an equal-interval basis (Hatch & Lazaraton 1991: 178-80). Parametric tests (t-test or ANOVA) can only be used, however, if the ordinal data are normally distributed, i.e. the mean and standard deviation are appropriate measures for central tendency (Hatch & Lazaraton 1991: 547) and if the intervals are equal. If these prerequisites were met, the ordinal data were analysed with a t-test for the variables gender, origin and age, and an AN alysis O f V ariance (ANOVA) for social class. Additional tests of strength of dependency have occasionally been used and are stated explicitly.
Non-parametric tests were applied to data were an equal interval or normal distribution could not be assumed. According to Hatch & Lazaraton, “when you are not sure [about normal distribution, etc.], it is best to select a nonparametric procedure” (1991: 270). In such cases, the parametric tests - t-test and ANOVA - were replaced by the equivalent non-parametric procedures of a Mann-Whitney U or Kruskal-Wallis tests.
On the whole, it has to be noted that the present sample may not be fully representative of the population and should not be over-generalized. In suchlike cases, the data can only be used for descriptive purposes (Hatch & Lazaraton 1991: 268).
 A brief explanation of origin and meaning will be given in chapter 2.
 Both tables are taken from the Scottish Borders Council Homepage. 10th January 2003. <http://www.scotborders.org.uk>.
 The data are taken from the 1991 Census as presented on the Scottish Borders Council Homepage. 10th January 2003. <http:// www.scotborders.org.uk>.
 The label Braid Hawick “is employed to designate the distinctive local speech of the town of Hawick” (Smith 1927: 8).
 For a detailed discussion of the three components cf. Deprez & Persoons (1987: 125-27).
 For a good introduction to the matched-guise technique cf. Fasold (1984: 152-58) or Agheysi & Fishman (1970: 145ff).
 In the following analysis the teachers’ questionnaire will be referred to as ‘T’, the pupils’ one as ‘P’ respectively. In discussion, a question will be preceded by a code, e.g. T6 = teachers’ question number six. Should a question be subdivided, small letters are used, as in P6a for example. If questions where asked with both groups, the code is given as P/T.
 The questionnaires are to a large degree based on questionnaires used in Birgit Splitt (1999). “Language Attitudes and Language Use in Orkney: An Empirical Investigation.” Unpublished Staatsexamensarbeit, Heidelberg.
 Other questions which are left out will be mentioned in the respective chapters.
 The parents’ and the partners’ origin was divided up into Hawick, Borders, Scotland, England and other. Additional information such as partner, year at school are listed but are not normally included in the evaluation of the questions.
 Unfortunately, two pupils did not state their date of birth. Since they both belong to school year 4, it is assumed that they are fifteen years old.
 This judgment was made by several teachers in Hawick, Kelso and Jedburgh.
 Groupings are based on the Registrar General’s classification of occupation and two other sociolinguistic studies (Romaine 1980: 175, Sandred 1983: 32).
 Ron Smith, subject 3tm of the sample, is a connoisseur of Hawick and its inhabitants.