Relations Between Different Linguistic Variables example essay topic

What is sociolinguistics? - It is the study of language in its social context. It is a field of investigation which describes all areas of the study of the relationship between language and society. In the late 1960's and the early 1970's, a widespread interest in sociolinguistics developed. The study of language in relation to society has a long tradition, but a recognizable growth in sociolinguistics took place in the 60's and 70's. As most other fields of investigation as well, sociolinguistics is partly theoretical and partly empirical.

The development of quantitative studies of speech has coincided with that of sociolinguistics and, for many linguists whose main interest is the structure of language, this part of sociolinguistics apparently makes the most relevant contribution, providing new data which need to be reconciled with current linguistic theories. The work which is done quantitative studies is all based on the study of spoken rather than written language (though in some cases the speaker is reading from a written text, such as a list of words), and its aim has been to find out about everyday speech of ordinary people, in reaction to the high degree of idealization that is typical of transformational-generative grammar. The aim of this branch of sociolinguistics, like that of the 'dialect geography' branch of dialectology, is explicitly comparative - to compare texts with one another, rather than to make some kind of 'total' analysis of each text without reference to others. It is the purpose of studying texts - to test hypotheses about relations among linguistic and social variables. The fact that the investigator starts with a predetermined list of linguistic variables and their variants shows that he expects the variants in his list actually to occur in the sort of texts he has collected, and he also generally starts with a range of hypotheses about the social variables to which those in his list are related, such as region, social class, or sex. If each text contained instances of only one variant for each variable, then it could be located in the relevant multi-dimensional linguistic space without using quantitative methods.

Different variants of the same variable occur together in the same text, and texts can be arranged on a continuous scale according to how often the variants occur. The relations between different linguistic variables are also a matter of degree, some being more closely related than others; and the same is true of relations between linguistic and social variables. It is rare indeed to find any linguistic variable whose variations exactly match those of any other linguistic or social variable, though it is common to find variables which match each other sufficiently closely to convince one that there is some kind of causal connection between them. A 11 these facts call for a quantitative treatment of the "data, using appropriate statistical techniques. The person mainly responsible for the use of quantitative methods in the study of texts is the linguist, William Labov. In his studies of linguistic variation, William Labov paid a good deal of attention to methodology.

The questions he was concerned with were how to collect data, how to analyse it well, and how to interpret the results successfully. Following the "classical Labovian" approach to quantitative studies, we differentiate five different stages. In a sociolinguistic text study, we differentiate five different stages. Those are: A -selecting speakers, circumstances and linguistic variables; B - collecting the texts; C - identifying the linguistic variables and their variants in the texts; D - processing the figures E - interpreting the results. A - selecting speakers, circumstances and linguistic variables - The selection of speakers, circumstances and linguistic variables involves some extremely important decisions, which are to a certain extent dictated by hypotheses about the expected results. It is similarly important that all the speech should be collected under the same circumstances, so far as this is possible.

There is a major problem of definition here, both for social variables relating to speaker and circumstances, and for the linguistic variables themselves. How can we define 'manual worker'? How can we distinguish old from young? Even worse is the problem of defining the community to be studied, since 'speech communities' are not self-defining. The researcher has to provide solutions which are at least reasonably satisfactory, to avoid the real danger that his results will be valueless because of ambiguities in defining the variables. B - collecting the texts - The collection of texts necessitates finding appropriate speakers who are willing to participate.

This means finding people willing to be interviewed and recorded for about an hour in their homes, but many alternatives are described in the literature. C - identifying the linguistic variables and their variants in the texts - At this stage, one might expect the least difficulty, since we already know what the variants to be distinguished are, and all we need do is listen for them. However, there is a considerable degree of subjectivity in recognizing phonetic variants, and different researchers can produce different analyses of the same text, even when they are highly trained phoneticians. One may also need to record information about the linguistic environment in which each instance of a variable is used since this often influences the choice of one variant rather than another, but this is only possible if there is already a clear hypothesis as to which aspects of that environment are relevant.

D - processing the figures - The processing of the figures involves counting the number of identified occurences of each variant in each text, and comparing the figures for different texts. The obvious step is to reduce all the figures to percentages, since this makes comparison much easier. The next step is to discover which differences between texts are significant, i.e. which would form a reasonable basis for general ising to other texts of the same types. The investigator has to use statistical tests in order to decide how significant the figures are. It is also important to understand that statistical techniques allow one to calculate the likelihood of some pattern of results occurring without any causal connection between the figures concerned, but never provide proof either for or against the existence of causal connection. Even where a causal connection between two factors is reflected by the statistics, it does not follow that one factor is the cause of the other.

It is possible that they are both results of some other factor. E - interpreting the results - The interpretation of the results is in some ways the most difficult stage, since this is where the findings have to be fitted into a general theoretical framework dealing with the structure of language and its relations to society and individuals. Success at this stage depends not only on correct methodology at all the previous stages, but also on having an adequate general theoretical framework. These methods were first developed by William Labov in his innovatory urban study The social stratification of English in New York City (1966). Labov attempted to attain representativeness in the Lower East Side of New York City by taking his informants from a previously constructed random sample of the population. A sample frame is any list which enumerates the relevant population, simple examples being electoral registers and telephone directories; the main principle of random sampling is that anyone within the sample frame has an equal chance of being selected.

Although Labov was by no means the first urban dialectologist to be sensitive to the need to give a representative account of urban speech, his sampling methods are, however, important and distinctive in that they were part of a larger, , principled programme for the quantitative study of language variation, which itself was designed to address important theoretical problems in linguistics. Labov's ultimate description of results was based mainly on 88 speakers - just over one quarter of the original random sample. The data was collected in just a few hours and which is a classic example of the method of rapid anonymous observation. The chosen variable represents the presence or absence ( (r): [r] versus (r): ) of a consonantal constriction corresponding to the letter r in words like farm and fair, where the next sound is not a vowel in the same word (as in very). At that time, New Yorkers sometimes used one variant and sometimes the other, which was of particular interest because the choice seemed to represent a change currently taking place, as New Yorkers moved from the previous norm of consistent (r): (as in British RP) towards a new and relatively consistent (r): [r] (as in many other United States accents).

The study of linguistic changes currently taking place has been one of Labov's recurrent interests, ever since his Martha's Vineyard work). Labov predicted that the proportion of (r): would be highest in the speech of older people (since (r): [r] is an innovation), and of lower-status people (since the new standard, (r): [r], is the result of influence from the high-status community outside New York). He farther predicted that (r): would be most frequent when speakers were paying least attention to their speech, since they would then be worrying less about how their hearers were assessing their social status; and finally that the linguistic context of (r) would influence the variant used, (r): being favoured more by a following consonant than by a following word-boundary as could be predicted on general phonetic grounds from the widespread tendencies to simplify consonant clusters. The results of Labov's survey were that the amount of r increases by social class and by formality of style. However, there is one noticeable exception: Lower middle class speakers outperform the upper middle class speakers on word-lists and pairs.

Labov calls this a crossover in the graph an explains it as the phenomenon of hypercorrection. Lower middle class speakers realise how prestigious r-pronunciation is and therefore outperform the next highest social class, whenever they are able to monitor their speech, that is in word-lists and list of pairs. Another investigation is that by James and Lesley Milroy in Belfast, Northern Ireland. The methods used are quite different from those of the classical Labovian approach.

The main difference between the Milroy's work and that of Labov is that Lesley Milroy, who did most of the field-work, was accepted as a friend by the groups whose speech she studied, which made it unnecessary to use the formal interview, technique. This had the great attraction that it was possible to study genuinely casual speech, as used between friends, because the researcher's presence did not increase the formality of the situation. By becoming a friend of the people one is investigating one becomes part of a network of relations among them, and can use the structure of this network as social data to which speech may be related. In her study, Lesley Milroy concentrated on the speech of working-class people in Belfast, only.

Three specific working-class areas were selected, between which there were important differences still. Two were unambiguously Protestant and one Catholic, and in one of the Vritestant areas the traditional local industry, the ship-yard, was still employing local men, whereas the traditional employer of men in the other Protestant area and the Catholic area was -the linen industry, which has declined, leaving men either unemployed or travelling outside the area to work. As a result of her efforts, Lesley Milroy became accepted as a friend who could 'drop in' at certain houses at any time, to sit in the kitchen listening or taking part in the conversation for as long as she wanted and even to use her tape-recorder, after explaining that she was interested in Belfast speech. Under such circumstances it seems unlikely that her presence, or even that of the tape-recorder, affected the way in which people spoke. The most obvious source of influence on linguistic variables is the speaker himself, i.e. the kind of person he is and the experiences he has had. An individual's use of a linguistic variable depends on the degree to which he is influenced by one or more norms in his society.

The Milroys have specifically investigated this aspect of variation. They selected their speakers through personal introduction's within a network of contacts, and they were able to spend a great deal of time in the households concerned, getting to know the structure of their social relations. The three communities studied were all typical poor working-class areas, and many of the families involved were typically working-class in being part of a 'closed network', a network of people who have more contacts with other members of the same network than with people outside it. This affects the kinds of relations they have, for, in a traditional working-class area, ties of friendship, work, neighbourhood and kinship will all reinforce one another. One effect of belonging to such a closed network is that people are very closely constrained by its behavioural norms and there is consequently little variation between members in their behaviour (or at least in the norms which they accept). Conversely, people who do not belong to a closed network, or who belong to a network united by fewer types o bond, might be expected to show a relatively low degree of conformity to the speech norms of any closed network.

Some of the people the Milroys recorded were from extremely closed networks, but others had looser relations to the community. Each speaker was therefore scored for the 'strength' of the network connecting him or her to the other members - a so-called 'network strength score' (NSS), which was calculated by taking account of five factors, for example, whether or not the person concerned has substantial ties of kinship in the neighbourhood, and whether he worked at the same place as at least two other people from the area. Five of the eight linguistic variables studied showed an overall correlation with NSS, i.e. were influenced by NSS in all subsections of the communities studied - whereas the other three were influenced by network strength in some subsections, though not in all. It is also possible to use the NSS to connect scores on some linguistic variables with known facts about social structure. For instance, there are clear differences between males and females for most of the variables in Belfast, and equally there are differences in NSS, where men score higher than women. Since the sex differences on linguistic variables show that men use more of the core variants than women, sex differences on the linguistic variables can be explained as an automatic consequence of differences on the network strength variable, and consequently we need no longer postulate sex as an independent social factor influencing this linguistic variable.

The theory of networks provides an easy answer: assuming that men go out to work more than women do, and that they work with men from their own neighbourhood, men form more work bonds than women, but have roughly the same number of other bonds. Overall, therefore, their networks have more bonds and their NSS will thus be higher. The differences in speech can therefore be explained, more or less directly, with reference to differences in employment patterns. Linguists have tended to select relatively focussed communities for their studies, and have consequently constructed theories of language which have relatively little room for variability. Even in the small, closely knit communities studied by the Milroys, there was a considerable amount of variation in detail, so we may expect relatively gross variation in more diffuse communities. But the results of the Milroys remain: the stronger the social network, the greater the use of certain linguistic features of the vernacular.

Milroy's hypothesis: a closeknit network has the capacity to function as a norm enforcement mechanism, could be confirmed. They also stated beforehand that a closeknit network structure appears to be very common in low status communities. Low-status varieties enable those who use them to show their solidarity with one another and achieve some kind of group identity. Sociolinguistics studies have only a short history in the canon of linguistic sciences and there remain some open questions that are worth to be asked. i.e. what social forces kept these norms alive and do they influence the middle or higher class? Can the same phenomenons of the Milroy's research be seen in the middle class too?

We may look forward to understanding these processes better after a few more decades of sociolinguistic research.