Language Change

This sociolinguistic study offers a theoretical framework for understanding the diffusion of language change. Advanced statistical analysis methods are used in rigorously testing the supposed norm-enforcement effect of social networks within a rural community. In this analysis, social factors are considered to operate simultaneously and variably on the individual, and the individual is considered to respond variably. Revisions to the social network model are proposed, allowing the effects of various social factors operating on the process of resistance to language change to be evaluated simultaneously. Most linguists have maintained that change in self cannot be observed; all that you can possibly hope to observe are the consequences of change.

The Traditional View

The traditional view of language change is the only changes that are important in a language are those that can be demonstrated to have structural consequences. Consequently, over a period of time a distinction between two sounds may be lost in a language, as occurred historically in most varieties of English in the vowels of meet and meat or horse and hoarse. A distinction may be gained where there was none before, as in a house with an [s] but to house with a [z], or finally in thin and thing, the [n] and []. In each of these cases a single phonological unit became two: there was a structural split. So we can find instances of phonemic coalescence, situations in which a contrast existed at one time but later was lost, and instances of phonemic split, situations in which there was no contrast at one time but a contrast developed. Variation is either controlled by circumstances, e.g. allophonic (as when the p in pin is aspirated but the p in spin is not), or it is free, i.e. random. Internal change in a language is observed through its consequences. A second kind of change in a language is external in nature. This is change brought about through borrowing. Of these two kinds of change, internal and

external, linguists view the former as being far more important even though it is the latter that is inclined to come to public attention, as when efforts are made to purify languages. The traditional view of language change also favors a family tree account of change and of the relationships among languages. Linguists tend to reconstruct the histories of related languages or varieties in such a way that sharp differentiations are made between those languages or varieties, so that at one point in time one thing (that is a language itself, or a variety, or even a specific linguistic item) splits into two or more, or it lost. In what the writer will call the traditional view of language change, the only changes that are important in a language are those that can be demonstrated to have structural consequences. Consequently, over a period of time a distinction between two sounds may be lost in a language, as occurred historically in most varieties of English in the vowels of meet and meat or horse and hoarse. In most dialects these vowels have fallen together (or coalesced). So we can find instance of phonemic coalescence, situations in which a contrast over-reporting refers to informants claiming to use prestige variant more often than they are actually observe to use it; underreporting is, of course, the opposite. Some Changes in Progress Various linguists have observed and reported on what they consider to be changes in progress. In the table 8.4 existed at one time but later was lost, and instances of phonemic split, situations in which there was no contrast at one time but a contrast developed. Such change, of course, is not restricted to phonology. The morphology and syntax of a language change in the same way. It is possible, therefore, to write internal histories of languages showing the structural changes that have occur over periods of time through use of this principles of contrast vs lack of contrast.

A second kind of change in a language is external in nature. This change brought about through borrowing. The traditional view of language change also favors a family tree account of change and of the relationships among languages. A third view of change is that particular changes diffuse throughout a language, sometime in rather idiosyncratic ways. It is in the last two views of change, through use of the concepts of wave and diffusion, that they see the possibilities that the study of variation opens up to us for understanding the process of change. Various linguists have observed and reported on what they consider to be changes in progress. Phonetician, Gimson (1962, pp. 83-5) has observed that in Received Pronunciation (RP) the first part of the diphthong in a word like home is tending to become increasingly centralized and the whole diphthong itself monophthongized. This tendency is seen mostly in the pronunciation of the younger members of fairly exclusive upper-class and professional social group, but it can also be observed to be spreading into less exclusive varieties of RP, e.g. the more general variety favored by the BBC (see also Rosewarne, 1994, for the development now referred to as Estuary English). Bailey (1973, p. 19) has pointed out that in the western United States the distinction between the vowels in such pairs of words as naughty and notty, caught and cot, and Down and Don is disappearing. In each of the examples just cited the factor of age seems to be important: younger speakers can be observed to use the language differently from older speakers. One study which was able to make use of roughly comparable sets of data from two periods of time is Labovs study (1963) of certain sound changes in progress on Marthas Vineyard. The explanation that Labov offers is that the change is merely an exaggeration of an existing tendency to centralize the first part of the diphthong. At the time of the survey for the Linguistic Atlas, it appeared that this centralizing tendency was being eliminated. It was virtually extinct in (aw) and in only moderate use in (ay). What has happened apparently is that, instead of eliminating the tendency, residents have exaggerated it to show their solidarity and their difference from the summer population.

Labov (1981, p. 185) points out that, when found, such behavior is a characteristic of the second-highest status group in a society. It is found in that group when its members adopt a formal style, and it is also found when they selfreport their linguistic usage, and respond to subjective reaction tests that require them to evaluate their own and others linguistic usages.

What is important is that it is the women of a particular social class who seem to be in the vanguard of change so far as r is concerned in New York City. In this case, the change seems to be motivated by a desire to be like those who have higher social prestige. Trudgill found that the distribution of the variants of the (ng) variable showed that there were very marked differences between the usage of working-class males and working-class females: males favored the [n] variant (i.e., pronunciations such as singin rather than singing) much more than did females. He found similar results with other variables, with woman showing much stronger preferences for standard forms than men. He suggests that women may be more status-conscious because they are less secure and have less well developed social networks than men. Another important factor in this differential usage is that working-class speech has connotations of masculinity and women often want to dissociate themselves from it for that reason, preferring types of speech which are regarded as more refined. Trudgill devoted a considerable part of his research effort to investigating working-class speech and what he calls the hidden values associated with non-standard speech which may be particularly important in explaining the sex differentiation of linguistic variables (p. 183). A further analysis showed that both middle-class and working-class speakers produced very much the same levels of under-and over-reporting, so the phenomenon appears to be sex-linked rather than social-class-linked. Trudgill emphasizes that, though it may be correct that in certain communities middle aged-class women and the young are in the forefront of change toward the standard norm, in Norwich, at least, there appears to be considerable number of young WC working-class men marching resolutely in the other direction (p. 194). Cheshires (1978) finding in Reading, England, that lower-class boys use more nonstandard syntax than lower-class girls further supports the thesis that change may be motivated by a desire for solidarity.

A Spanish community Holmquist (1985) described how women show preferences for a standard variety of a language rather than a nonstandard one and for marital partners who speak that standard variety. Woman may be more active participants than men in some changes, but the situation may be reversed in others. Although the young are usually in the vanguard of most changes, in some it is the not-so-young who lead.

The Mechanism of Change

Bright (1960) examined the Brahmin and non-Brahmin caste dialects. His examination revealed that the Brahmin dialect seems to have undergone unconscious change, that is, changes in phonology and morphology. He suggested that the upper-class appear to originate sound change at the phonetic level and that, in their chase to initiate such phonetic changes, the lower classes bring about change at the phonemic level, that is, changes which eventually have structural consequences for the language. This is indeed what happens when the social distribution of the variants is caste-based rather than class-based. Labov (1981) has pointed out how difficult it is to get the right kinds of data on which to base claims about linguistic change in progress and how easy it is to make either false claims or incorrect predictions, giving several instances of the latter from Switzerland, Paris, and Philadelphia (pp. 177-8). He stresses the importance of having good data on which to base claims. Since individual linguistic usage tends also to vary in fixed ways according to the age of the individual, such age-grading must also be taken into account because this process is an independent one. That is, the relationship between diachronic (historical) matters and synchronic (descriptive) one is a two-way relationship. That is what Labov calls a dynamic dimension to synchronic structure, so that the past helps to explain the present and the present helps to explain the past. Labov (1972b, pp. 178-80) proposes a rather detailed outline of what he considers to be the best the basic mechanism of sound change. The mechanism has thirteen stages, and Labov points out that the first eight deal with what he calls change from below, that is, change from below conscious awareness, whereas the last five deal with change from above, that is brought about consciously. (p. 204)
5

All the model is useful in that it tries to deal with certain complex issues in an elegant way: the issues of both conscious and sub-conscious induced linguistic change; the place that social class plays in such changes; and such concepts as indicators, markers, stereotype, and hypercorrection. Change from above is conscious change. This change is to involve a movement toward standard linguistic norms. It is actually not initiated within the highest social group in society. This group is a kind of reference group to groups lower down in the social scale. Change from below is unconscious and away from existing norms. Many observers believe that in our society such as ours women are in the vanguard of the first kind of change and men in the vanguard of the second, because members of the two sexes have different motives. In this view women are motivated to conform to, and cooperate with, those who are socially more powerful whereas men are more inclined to seek solidarity with peers. Labov viewed (1994, p. 23) that cities have always been at the center of linguistic innovation. Labov adds that his conclusions are valid only for the speech of non-blacks in Philadelphia. Blacks do not use this vowel system at all, preferring instead one usually referred to as the Black English System. He suggests that the future direction of change in the vowel system in Philadelphia will depend very much on social changes that are occurring in the city. Bailey and Maynor (1989) have also proposed that Black English and Standard English are diverging in the Brazon Valley in Texas with only Black speakers using constructions like he always be tryin to catch up and resisting the adoption of post-vocalic r in words like farm. Butters (1989) argue that there is no evidence to support the divergence claim. He points out that the though there may be divergent features there are also convergent ones. Wolfram (1990) also discusses the idea that these varieties of English, are diverging and concludes that the evidence is flimsy (p. 131) James and Lesley Milroy (1992) are two others linguists who are interested in how change begins. They says that, groups linked internally mainly by relatively weak ties are susceptible to innovation and add that innovation between groups are generally transmitted by means of weak rather than strong network ties. They

point out that change begins therefore in the middle of social class hierarchy is entirely consistent with Labovs finding that innovating groups are located centrally in the class structure, characterized by him as upper-working or lowermiddle class. Eckerts findings (1988, 1989, and 1991) what she calls jocks and burnouts. Life style also seems to be a factor of change. Jocks are either middleclass students or students with middle-class aspirations, and Burnouts are either working-class students or students who wish to identify themselves as such. Jocks tend to be college-bound and white-collar-oriented; burnouts will leave school for the blue-collar workplace. Jocks willingly participate in the activities of the schools; burnouts find activities outside school more attractive. Burnouts were much more active than Jocks in participating in the kinds of vowel shifting that are occurring in the northern cities of the United States.

What is particularly interesting about such shifts is that they are slow, unconscious, and systematic, i.e. they have a direction. Much Lobovs work is concerned with trying to understand these kinds of systematic changes and how individuals participate in them. The wave of theory of change and the theory of lexical diffusion are very much alike. Each attempts to explain how a linguistic change spreads through a language: the wave theory makes claims about how people are affected by change whereas lexical diffusion makes claims concerning how a particular change actually occurs.Labov insists that the best studies of change in progress look for different kinds of data sources, are very much concerned with assesing the accuracy of these sources, and are quite cautious in the claims they make. After conducting a number of investigations of sound changes in progress, Labov (1972b, pp. 178-80) proposes a rather detailed outline of what he considers to be the mechanism of sound change. The mechanism has thirteen stages, and Labov points out that the first eight deal with what he calls change from below, that is change from below, that is change from below conscious awareness, whereas the last five deal with change from about, that is, change brought about consciously. The thirteen stages are as follows.

1.

The sound changes usually originated with a restricted of the speech community, at a time when the separate identity of this group had been weakenead. The linguistic form which began to shift was often a marker of regional status with an irregular distribution within the community. At this stage, the form is an undefined linguistic variable.

2.

The changes began as generalizations of the linguistics form to all members of the subgroup; we may refer to this stage as change from below, that is, below the level of social awareness. The variable shows no pattern of stylistic variation in the speech of those who use it, affecting all item in a given word class. The linguistic variable is an indicator, defined as a function of group membership.

3.

Succeeding generations of speaker within the same subgroup, responding to the same social pressures, carried the linguistic variable further along the process of change, beyond the model set by their parents. We may refer to this stage as hypercorrection from below...

4.

To the extent that the values of the original subgroup were adopted by other groups in the speech community, the sound change with its associated value of group membership spread to these adopting groups..

5.

The limits of the spread of the sound change were limits of the speech community...

6.

Being expansion, the linguistic variable became one of the norms which defined the speech community, and all members of the speech community reacted in a uniform manner to its use (without necessarily being aware of it). The variable is now a marker, and begins to show stylistic variation.

7.

The movement of the linguistic variable within the linguistic system always led to readjustment in the distribution of other elements within phonological space.

8.

The structural readjustments led to further sound which were associated with the original change. However, other subgroups which entered the speech community in the interim adopted the older sound change as a part of the norms, and treated the newer sound change as stage 1. This recycling

stage appears to be primary source for the continual origination of new changes.. 9. If the group in which the change was not the highest-status group in the speech community, members of the highest-status group eventually stigmatized the changed from.. 10. This stigmatization initiated change from above, a sporadic and irregular correction of the changed form status group towards the models of the highest status-that is, the prestige model. This prestige model is now the pattern which speakers hear themselves using. 11. If the prestige model of the highest status group does not correspond to a form used by the other groups in some word class, the other groups will show a second type of hypercorrection, shifting their careful speech to a form further form the changed form than the target set by the prestige group. We may call this stage hypercorrection from above. 12. Under extreme stigmatization, a form may become the overt topic of social comment, and may eventually disappear. It is thus a stereotype.... 13. If the change originated in the high-status group of the community, it became a prestige model for all members of the speech community. The changed form was then adopted in more careful forms of speech by all other groups in proportion to their contact with users of the prestige model, and to a lesser extent, in casual speech. Whenever the change begins and whatever its causes, it is not an instantaneous event for the language as a whole. It has to establish itself. A number of linguists have proposed a theory of change called lexical diffusion. According to this theory, a sound change spreads gradually through the words in which the change applies. For example, a change in vowel quality is not instantaneous, affecting at some specific point in time all words in which that vowel occurs, as if you went to bed one night with vowel quality A in those words and got up next morning with vowel quality B. Instead, only some words that have the vowel will be affected initially, then others, then still others, and so on until the change is complete.

10