LANGUAGE VARIATION AND CHANGE
This chapter has two main focuses.
1. First, it examines a key feature of the variationist school of linguistics-its interest in language change.
2. Second, it presents new methodologies and analyses which have modified and extended the approaches developed by the Labovian School.
TWO MODELS OF LANGUAGE CHANGE
The main of this section is on the way in which the sound changes spread through a speech community. And the two models of language change are:
1. Variationist Approaches to Change
Weinrich, Labov, and Herzog (1968) showed that tracking down changes required close attention to the language system as well as the social system. All preceded by variation. But it not means that all variation not always leads to change. Ex: –in, -ing. Labov (1989) notes that further –ing is stable in grammatical conditioning in three different continents that is USA, Australia and England. The variant –in also occurs in the following environments in decreasing order of frequency:
· Progressive and participles
Ex: She is playing
Ex: a flying fish
Ex: walking is good for health
Linguists have a greeter interest in variable undergoing change, as in the case of centralizing of diphthongs in Martha’s Vineyard and in the spread of postvocalic (r) in New York City. In principle, with minor changes, the model could be used to describe long-term grammatical changes in language. Sociolinguists find it necessary to distinguish between two kinds of sound change, changes “from above” and changes “from below”.
· Change from above
It involves by dominant social class. These are often consciously modeled on sounds used by other speech communities that have high prestige.
Ex: the using postvocalic (r) in New York City, and it is women involved as a head of men in their scores of the variants of change.
· Change from below
It represents the phonetic processes that are based on articulatory process that make pronunciation easier. An example is deleting one consonant before another.
Ex: the pronunciation (t) in half pas’ five, trus’ me, etc. In this case, the man leading the change.
2. Lexical Diffusion
Lexical Diffusion is the name of theory which is proposed that sound change occur word by word. The first hypothesis in Lexical Theory is that a sound change does not occur in all words or environments simultaneously. And the second hypothesis concerns the rate of sound change is affected in a language.
Real and Apparent Time
A community is divided into age groups which are studied intensively for a short period to examine whether any differences occur. Where older age groups show low of a variant while younger groups show increasingly greater use, we can assume that there is a change going on in ‘real time’.
A Real- time Verification of an Apparent-time Study
Joy Fowler (1986) and Jeff Macdonald (1984) separately repeated the departmental store study, two decades after Labov, in an attempt to ascertain the extent of change in [r] usage in New York City. Change in real time has taken place. In particular, the following is observable in the follow-up study:
· The percentage of speaker who used ‘all [r]’ (i.e [r] four times) increased in all three stores in real time.
· The greatest increase for ‘all [r]’ is shown at Saks, the highest- status store (though the overall increase is not very great).
· For ‘some[r] (i.e. use of postvocalic [r] in one, two or three of the four possible instances), all stores show a small increase.
· May’s shows the greatest increase for use of ‘some[r]’, though again this increase is not very great.
Labov’s (19994: 91) summary of this study is as follows:
The precise replication of the Department Store Study shows that the sociolinguistic structure of the speech community is perhaps even mare stable than anticipated. Two possibilities exist as to why change has not occurred in the intervening two decades to the extent predicted. The first possibility is that Labov’s ‘hypercorrection’ hypothesis has not held up. The second, and more likely, possibility is that change in (r) usage is still at an early stage.
VERNACULAR MAINTENANCE AND CHANGE
Sociolinguistic research has shown that the concept of social network is important for the understanding of such strategies of vernacular maintenance. Social network were first used in a sociolinguistic study of Belfast which will be discussed in some detail in the following section. Following Labov’s study of vernacular use by black adolescents in Harlem (1972b), the Milroys formed the hypothesis that the use of vernacular forms is associated positively with the speaker’s degree of integration into the community’s social network. The concept of social network has been used successfully in anthropological research and refers to the informal and formal social relationships that individuals maintain with one another. Two criteria are particularly important for the description of networks: density and multiplexity. Network density refers to the number of connections or links in a network. Miltiplexity refers to the content of the network links. Dense and multiplex network are typically found in rural villages and urban working-class areas. Anthropological research since 1950s has shown that dense and multiplex networks often act as norm-enforcement mechanism, imposing all kinds of behavioral norms (dress, conduct, language use) on their members. An important concern of the research was how to gain access to the vernacular in its most natural form. The situation was complicated by the political and social conditions in Belfast, where outsiders to the community were usually viewed with suspicion.
Milroy carried out a quantitative analysis of the data collected for forty-six speakers. The analysis showed that the use of several phonological variables was clearly stratified according to gender in the three working-class areas. To measure the integration of each informant into the social network of the community, Milroy (1080: 139-43) developed a six-point scale from 0-5 to indicate the degree of density and multiplexity of an individual’s network. According to Milroy’s analysis, male network scores notably higher than female scores, which implies that men gad more and stronger ties to the local community than women. In Ballymacarret, men tended to be strongly integrated into the local community network. Most women on the other hand, worked outside the area and their integration into the local network was therefore rather weak. Not only were the network patterns for men and women sharply different in Ballymacarrett, men’s ad women’s activities were also separated and gender roles were clearly defined. The link between integration into the network, gender and vernacular use also existed in the Clonard. On some linguistic variables, however, the situation was reversed for the younger generation, that is, young women showed greater use of certain vernacular variants than men.
The Belfast study is an important contribution to the understanding of language change and language maintenance in a community. Language use, according to the Milroys, is influenced by both status and solidarity. Use of the standard language is associated with high social status, while the use of the vernacular indicates solidarity with local people, customs, and norms. Vernacular use is typical in dense and multiplex network structures, which can be found in rural areas and the old urban working- class districts, where solidarity with the group encourages and demands the use of local, vernacular forms.
Dialect Loss and Maintenance in a Divided City: The Berlin Vernacular
The findings of the Berlin Urban Vernacular (BUV) project, which was carried out in the early 1980s under the leadership of Norbert Dittmar and Peter Schlobinski, give further support to Milioy’s hypothesis of the strong relationship between vernacular maintenance and integration in a local community.
Germany was divided into two political divisions after World War II, they are: the Federal Republic of Germany (FRG) in the west and the German Democratic Republic (GDR) in the east.
BUV (Berlinerisch) had received relatively little interest from traditional dialectologists, who saw it as a corrupted city slang, highly influenced by the German standard variety. Low German dialect influence is still clearly visible in BUV. An important example is the retention of the Low German voiceless stops /p/, /t/, /k/ as opposed to the High German fricatives.
Three districts of Berlin, they are: the traditional working-class districts of Wedding (West Berlin), Prenzlauer Berg (East Berlin), and Zehlendorf (West Berlin), a typical middle-class area.
The most important finding of the BUV project was that the pattern of linguistic variation reflected the political division of Berlin. While BUV variants had been maintained to a larger extent in the East Berlin district of Prenzlauer Berg, a clear loss of typical BUV variants was observed in the affluent middle-class district of Zehlendorf. Speakers in the West Berlin working-class area of Wedding were situated between the two extreme.
Furthermore, research on speakers’ attitude showed that BUV was clearly stigmatized only in West Berlin, where it was seen as vulgar, working-class and an indicator of lack of education. The adjectives employed for the characterization of BUV in west Berlin were, for example, ordinär (common), vulgär (vulgar), schnoddrig (brash) and falsche Grammatik (bad grammar), tierischer Slang (beastly slang), putzfrauensprache (charwoman’s language). Standard German was generally seen as the legitimate prestige variety. In the East Berlin district of Prenzlauer Berg, on the other hand, BUV was not only commonly used but also perceived as highly prestigious. The stigmatized variety in East Berlin was Saxon, a German dialect spoken in a region south of Berlin.
The reunification of German led to wide-ranging social changes in the East. BUV use in East Berlin has declined considerably in the younger generation. However some speakers refuse to assimilate and proudly insist on using Berlinerisch, even when talking to West Berliners.
NEW APPROACHES TO VARIATION AND CHANGE: THE NEED FOR INTEGRATION
Pattern of Variation and Change in Sydney: A Case Study
To investigate the sociolinguistic patterns of variation and change in Australian English, the Sidney social dialect survey was carried out in the early 1980s under the leadership of Barbara Horvath. The study, which is in many ways indebted to Labov’s work, contains important methodological innovations regarding the definition of what constitutes a speech community and the statistical procedures used for the analysis.
Horvath, therefore, decided to include in her survey not only speakers from different social backgrounds, but also speakers of different ethnic origin: Anglos (Australians of English speaking origin), Italians and Greeks. Including recent migrants into the study signaled that both native and non-native speakers of English were explicitly seen as being part of the Sidney speech community. This approach differs considerably from Labov’s original definition of a speech community, which excluded non-native speakers.
A second important innovation of the study concerns the analytical methodology. Horvath used a multivariate statistical procedure called principal components analysis, which allowed her to consider in her final analysis more than twenty different linguistic variables simultaneously.
Traditionally, linguists have distinguished three sociolinguistic categories of Australian English: ‘Cultivated’, ‘General’, ‘Broad’ Australian English. The inclusion of non-native speakers in the Sydney dialect survey, however, made it necessary to add two new categories: ‘Accented’ and ‘Ethnic Broad’ Australian English. Accented refers to variants which are the direct result of transfer from the native language of the speaker. Ethnic Broad consists of variants which have become ethnic markers of the English of immigrants and are frequently passed on intergenerationally. Five vowel variables (all diphthongs) were investigated in the study: (iy), (ey), (ow), (ay), and (aw).
There are two deep divisions of Sydney speech community: a core speech community and a peripheral speech community. The periphery was characterized by the dominant use of vowel variants in the categories ‘Accented’ and ‘Ethnic Broad’. All speakers in this peripheral group were adults of either Greek or Italian origin. Most of the members in this group were born outside Australia and had acquired English as a second language around the age 20.
On the other hand, ‘Accented’ and ‘Ethnic Broad’ variants were never use by members of the core speech community. An important non-linguistic characteristic of the core speech community was the disproportionare age distribution: ninety teenagers, but only forty adults were found in this group. While the teenagers came from all three ethic group, the adults belonged to the ethnic group ‘Anglo’.
The core speech community could be divided further into four sociolects. These four sociolects were characterized linguistically by quantitative but not categorical variation. In other word, speakers in the four groups used quantitatively varying mixes of Broad, General, and Cultivated variants, but there was no group which never used a certain type of variant exclusively.
The four sociolects were primarily correlated with the social variables age and gender, while the effect of class was less pronounced.
In 1965, Mitchell and Delbridge summarized the distribution of speakers across the three varieties of Australian English as follows: ‘Broad’=34%, ‘General’=55%, ‘Cultivated’=11%. Horvath modified the distributions for the early 1980s: ‘Broad’ (sociolect one) = 13%, ‘General’ (sociolects two and three) = 81% and ‘Cultivate’ (sociolect four) = 6%.
Furthermore, we have already mentioned the existence of generational differences: while adults were mainly found in sociolect four, teenagers dominated in sociolect two and three where the General variants of the vowel variables were used most frequently. He differences between the distribution figures and the different linguistic behavior of the two generation, can be interpreted as indicators of language change in progress. Speakers appear to have moved away from the two extremes of the Broad-General-Cultivated continuum and increasingly to use vowel variables which are characteristic of General Australian English.
VOWELS SHIFTS: TOWARDS A HOLISTIC APPROACH TO DIALECT AND CHANGE
Current attempts to study clusters of features such as Horvath’s Australian study are particularly promoting in the area of vowel shifts. Vowel shifts are changes that operate across a whole set of vowels: for e.g., they may all be raised (in contrast to older dialects or the prestige dialect) or lowered, fronted and so on. Such shifts are sometimes known as ‘chain shifts’ because each change is related of the previous one, like the links of a chain. The study of vowel shifts as pioneered by Andre Martinet.
Two vowel shifts in English dialect that are currently being studied by sociolinguists are the “Northern Cities Chain Shifts” (USA, including Detroit, Chicago, Cleveland, and Buffalo). This area includes western New England, New York State, the northern parts of countries in Pennsylvania, northern Ohio, Indiana and Illinois, Michigan, Wisconsin and a less, well defined area extending westwards. The shifts involved six vowel sounds( phonemes) which may be heard as members of another phoneme by listeners from another dialect area, with some resultant confusion of meanings: Ann as Ian, bit as bet, bet as bat or but, lunch as launch, talk as tuck, locks as lax.
a. Simplified sketch of the Northern Cities Chain Shifts( based on Labov 1991: 17)
b. Fuller Northern Cities Chain Shifts in Detroit (based on Labov 1991: 15 and Eckert 1991: 222)
The basic pattern of the shift in many cities is given in figure a. In Detroit a complete rotation occurs with the backing of [e] and [L] making a full chain shifts.
The southern Hemisphere shift is the term used to describe another chain shift which occurs in South Africa and New Zealand( and partially in Australia), whose impetus lies in the influence of working class British speech in these nineteenth century colonies. Roger Lass and Susan Wright (1986) analyzed the systematic shift which the short front vowel series of south African English has undergone compared the other varieties like RP. The two systems are geometrically related.
One of the key sets of variables is the vowel system which they typically use. L.W Lanham and Carol Macdonald (1979) did when they tried to characterize typical white speakers of the different sub varieties:
- Cultivated : Middle class speakers having associations with England.
- General : Middle class speakers.
- Broad : Mostly lower middle or upper working class; identifying with the outdoors and sport; and significant contacts with Africans’ speakers, and hence partially influenced by Africans’ norms.
CONCLUSION: THE LIMITS OF VARIATION THEORY
Labov’s model stressed changes in the linguistic stratification of particular variables. He posited two types of changes, one ‘from above’ and another ‘from below’ to explain the introduction of a new prestige variant in a community as opposed to the spread of a less prestigious vernacular variant
Chen’s model of lexical diffusion focused on the means by which a sound change spreads within the vocabulary of a particular speaker and/or a particular speech community.
Milroy’s model of social network seeks to understand maintenance and change as part of the same package, depending on the strengths of ties within a community and the nature of their contacts with outside groups. This approach was seen to apply in the dialect divide between East and West Berlin.
A final concern of this chapter was the increasing emphasis in sociolinguistics on studying sets of variables in relation to societal pattern. In relation to this approach, the methodological innovations made by Horvath in her study of Australian English were cited. The study of ongoing vowel shifts in English in different parts of the world also relates to the attempt to study sets of related variables.