Regionality as an Independent Variable

Interlopers as Agents of Linguistic Change

§3 of "Dynamics of dialect convergence." Investigating Change and Variation through Dialect Contact,, ed. Lesley Milroy. Special issue of Sociolinguistics 6 (2002): 117-130.

For a fuller discussion, see "Region and language variation." English World-Wide 21 (2000): 1-31.

Dialectologists have always been aware that mobility is a potent force in leveling regional language variants, and for that reason, traditional dialect studies stipulated that their subjects be locals. The criterion of local nativity has deliberately been abandoned in my Dialect Topography project (Chambers 2000). As a survey of urban as well as rural areas, we seek a representative sample of the population, and that includes not only men and women of all classes and ages but also, obviously, residents of the survey area who are relative newcomers to it. Concomitantly, it was necessary to devise a metric for distinguishing indigenes, those subjects born and raised in the survey region, from interlopers, those who arrived there as adults, as well as the various degrees in between (subjects born outside but raised in the survey area, and so on). The metric is known as the Regionality Index (RI), and it is based on where the subject was born, where he or she was raised from 8 to 18, and where the parents were born. I have described the RI calculations in detail elsewhere (Chambers 2000: 10-13; Chambers and Heisler 1999: 40-46), and for my purposes here need only say that each subject receives an index score from 1 to 7, where RI 1 is a true indigene (as defined above) and RI 7 a true interloper, and the in-between scores indicate relative grades of nativeness. Conceptually, the easiest way to interpret the scores is in terms of major thresholds, as follows:

Our correlations of RI with regional variation are helping us come to grips with the influence (and lack of influence) of people who move into a region and bring with them lexical variants that differ from the ones in their new home area. I will discuss two case studies that indicate something of the range of possibilities.

We ask our respondents what they call the rubber-soled shoes worn when jogging. One of our survey regions, the Golden Horseshoe in southern Ontario, crosses the Canada-United States border, and among the answers to this question we find a clear border effect in the use of the majority responses, as shown in Table 2. The Canadian respondents prefer running shoes and its clipped form runners at 87 per cent, and the American respondents prefer sneakers at a whopping 99 per cent. However, Table 2 also shows that the Canadian response for sneakers is noteworthy at 13 per cent.

Table 2—Distribution of three lexical variants at the Niagara border.

When we look at the responses for Canadians only (excluding the Americans) for these variants, we find a significant VarbRul correlation with RI, shown in Figure 1. The responses for running shoes and for both running shoes and runners make a relatively flat trajectory across the graph from left to right, representing the usage of the people raised in the region (RI 1-5); the line in the graph then tails off gradually for people born and raised outside (RI 6 and the true interlopers, RI 7).

Figure 1—Use of three variants by Canadians in the Golden Horseshoe according to Regionality Index

These correlations provide empirical evidence for the source of the fairly sizable showing for sneakers in the region. The word is obviously imported by people moving into the region as adults. Many of those people are Americans, of course, immigrating across the international border in fairly large numbers as they have always done. Though their presence gives their word sneakers mild salience in the community, it does not seem be competing with running shoes as a potential change in progress. Instead, running shoes appears secure in the use of the people with deeper roots in the region. Under ordinary circumstances, we should expect the offspring of these interlopers to learn the indigenous variant rather than their parents’ interloper variant, and thus to perpetuate the communal norm in the next generation.

In the second case study, consider the results for a question that asks the respondents what they call a carbonated soft drink. Here, again, we find regional discrepancies, this time within three Canadian regions. Table 3 shows the responses for the two most common terms, soft drink and pop, in Montreal, the major city in the province of Quebec, in Quebec City, the capital of that province, and in the Golden Horseshoe in adjacent Ontario. (The tabulation ignores minor variants such as soda, sodapop, coke and cola, and hence the percentages are incomplete.) The Golden Horseshoe shows a clear preference for pop at 85 per cent, and Montreal shows a fairly large margin for soft drink at 62 per cent. We would expect Quebec City to show a pattern similar to Montreal, as the nearest city and one that shares much of its culture and history. Instead, Quebec City shows a strange bifurcation between soft drink and pop, suggesting an inherently unstable situation.

The dynamics in Quebec City come clear in the correlations with RI, shown in Figure 2. The native respondents (RI 1-3 on the left) show considerable consistency in favouring soft drink, a result that falls into almost perfect accord with the Montreal results. However, there is a sharp reversal in favour of pop for the newcomers (RI 5-7), with the median group (RI 4) evincing a classic transition pattern.

The RI again allows us to explicate this odd result by revealing it as a disparity between the indigenes and the interlopers. This time the interloper variant has intruded deeper into the population with clear roots in the region, and it has also made a proportionately higher showing at all RI levels. These differences between the two cases indicate very different long-term consequences, as we shall see.

Linguistic variation of all kinds requires that competing variants be available in the community. As an independent variable, regionality, as in Figures 1 and 2, provides insights into one of the sources of the competing variant by registering the occurrence of interloper words, and it does so dynamically by tracking their distribution in the community. Neighbouring regions are an obvious source for new variants, and plotting their incursions as a function of regionality makes concrete what was previously known only impressionistically. In that way, it should bring us closer to an explicit understanding of how linguistic changes take place under the pressures of converging dialects.

It is possible to imagine an interloper word having fairly high frequency in the life-span of the interlopers themselves and then disappearing, or nearly disappearing, in the speech of their children. That pattern presumably has repeated itself countless times in the history of every speech community. If the interlopers fail to adopt the local words and persist in using the ones they brought with them from outside, either through obstinacy (perhaps out of lingering sentiment for the old home term) or oversight (perhaps out of ignorance of an infrequently heard local term), then their children normally will be the ones to adopt the local words, under pressure of peers and their own natal roots, filtering out, so to speak, their parental term in what is perhaps the homeliest application of the Ethan Experience.

But our detailed results show that it is not just the frequency of the interloper word that poses the threat to an indigenous term. The real threat comes, surely, when the interloper word insinuates its way into the middle and lower reaches of the RI, that is, when the word occurs in the speech not only of the people least indigenous in the community but also of people who have been there a while–people raised in the region, though they were born outside it (that is, RI 4-5). That is the pattern for pop in Quebec City. It is the preferred word not only for the rank outsiders (RI 7 63 per cent) but for everyone RI 5 and higher (about 65 per cent). It is also ominously common (45 per cent) for RI 4, that is, people with quite clear local roots as having been raised locally though born outside; in their speech, pop competes with soft drink on almost equal terms.

By contrast, in the Golden Horseshoe, sneakers makes a noticeable, though decidedly minor, showing in the speech of rank outsiders (RI 7 28 per cent, RI 6 17 per cent) but in the speech of all other groups it remains a very minor variant, almost nonexistent. It seems safe to predict that it will not persist in the speech of the interlopers’ offspring, who will rank, if they are born and raised in their parents’ adopted home, as RI 2 or RI 3. Presumably the real-time pattern for lexical replacement, or at least for one common type, is for an interloper variant to move across the RI scale, so to speak, starting in the speech of the interlopers and then gaining currency in the middle groups until it finally makes inroads in the speech of the indigenes.

These stages fit comfortably into a commonsense view of change as a gradual social process. In order for individuals to adapt their speech to the local norms, they must be able to identify what those local norms are. If an interloper form retains its currency in the speech not only of interlopers but also among people of longer standing in the community, then there are at least two social forces working in favour of the replacement of the indigenous term. In the first place, interlopers arriving fresh in the community may fail to recognize that the indigenous local variant is different from the variant they grew up with since their word does get used in the community, or they may recognize it but feel less pressure to change their usage when their own usage is tolerated in the community, albeit as a minor variant. In the second place, and more crucially, if the offspring of the interlopers, the next generation of indigenes, fail to recognize which variant is the local one because two variants appear to be competing on roughly equal terms, then the decline of the indigenous term inevitably gathers momentum.

Convergence and dynamics

Regionality provides an independent variable for charting with some precision one of the sociolinguistic effects of mobility. Exploiting regionality in sociolinguistic studies requires only registering it in subject profiles along with age, sex and the other attributes, and quantifying it for explicit application as an independent variable (in this case by the Regionality Index). In virtually all sociolinguistic studies, certainly all that take place in developed communities, subjects are likely to differ with respect to their natal roots. That is simply a fact of contemporary life. Grading the subjects with respect to their roots provides a measure for evaluating their linguistic and social similarities and dissimilarities.