The Scope of Our Study
The point made by Wyld about a difference between two major kinds of London English in the Tudor and Stuart period is something that will be tested below. The linguistic features to be analysed consist of the 14 changes that we have discussed in the previous chapters. It is noteworthy that linguistic atlases of both historical and modern dialects suggest that, despite typical majority variants, alternative expressions 'for saying the same thing' commonly occur in different regions. This is also the case with our data. The features we study are usually not strictly confined to a given area at the time when their diffusion begins but may sporadically occur elsewhere, too. However, once it begins to gain momentum, the incoming feature typically diffuses faster in a particular region or locality, not everywhere at the same rate.
Some hypotheses may be proposed as to how a change in progress diffuses. The process can in principle follow the wave model, that is, changes spreading gradually outwards from a centre. This is the basis for linguistic similarities between neighbouring dialects. Dialects typically constitute continua, 'a series of systems where those nearest and most in contact show only slight differences' (Samuels 1972: 90). As suggested in the Gorlach quotation, above, the principle also applies 'vertically', to sociolects. But the wave model can only be expected to apply normally when the population is distributed equally.
There may be geographically isolated or otherwise self-sufficient areas that are bypassed by changes in progress. With supralocalizing features it appears that some of these basic assumptions of dialect diffusion may be specified further on the basis of modern dialect data: the new urban dialects discussed in section 8.1. all diffuse out of large urban centres.
One of the hypotheses advanced in this chapter is that London was instrumental in both promoting dialect mixtures and spreading linguistic innovations. The population of the metropolis grew rapidly, and social networks were looser and more uniplex in London than in small towns and in the countryside. They must have typically been based on a large number of continual contacts between nonadjacent dialects rather than on stable and continuous interaction between adjacent regional or social varieties (Samuels 1972: 92-93). The diffusion of changes is therefore expected to have been faster in London than elsewhere. In order to study this claim empirically, a point of comparison will be established between the capital city and one of the more peripheral dialect areas, East Anglia. The reason for selecting East Anglia rather than, say, the West Country is simply the amount of extant material. East Anglian letters are available throughout the period covered by us, while material from areas such as Somerset and Dorset is harder to come by, especially in an edited form.
The second part of our argument is related to linguistic innovations spreading from urban centres to the surrounding rural areas. In modern language communities these processes have also been found to follow a mechanism of diffusion that differs from the wave model, namely the gravity model. Trudgill (1986) has shown in detail how language changes may skip from one urban centre to another, bypassing the countryside in between. This model is based on the population sizes of the communities in interaction and the distances between centres (Chambers & Trudgill 1980: 189-204). Dialect skipping may be a powerful means of promoting and speeding up supralocalization: when the urban centres in a given area converge on the use of a particular form, it may diffuse to the countryside from several directions simultaneously.
In our case, migration in sufficient quantities can also constitute a form of dialect skipping. When a process of dialect diffusion following the classical wave pattern is under way percolating from the north to the south, it can be greatly accelerated by the immigration to the capital of a large number of northerners - or of people from the intervening regions already in possession of the feature. This is an aspect of dialect diffusion that can be tested by including a third point of comparison in addition to London and East Anglia. The North, consisting of the counties north of Lincolnshire, was selected because we wanted to explore the mechanisms by which supralocal features spread across the country.9
In their study of the diffusion of the third-person singular -s, Ogura and Wang (1996) propose a refinement to the traditional wave model by suggesting that the rate at which a change diffuses may vary in different localities. More precisely, they propose that those localities that are later in adopting the change in progress will, when the process advances, not only catch up with the ones that started earlier, but surpass them. One of the aspects of dialect diffusion that we shall pay particular attention to in our analysis is whether we can observe a snowball effect of this kind or whether the changes progress at a constant rate.
However, as shown in Chapter 7, a real success story in supralocal terms also had to involve vertical diffusion in Tudor and Stuart England. This social aspect of dialect contact will come into the picture when we compare the City of London and the Royal Court. While the two other regions, East Anglia and the North, are normally accounted for in dialect research - East Anglia including Norfolk and Suffolk - the capital is not included as a matter of course, or at least not divided into localities. But as we have seen, there is good reason to try to find out whether there were linguistic differences between Westminster and people associated with the Royal Court, and the City of London.10
In this chapter, we have included all people living in the City and in Southwark as Londoners. On the social scale they tend to occupy the ranks from the lower gentry downwards with a considerable merchant contingent, though the odd nobleman letter writer could be found living in the City. At Court, by contrast, there is a heavy concentration of the uppermost ranks from the royalty down to the upper gentry. Those who constituted the Royal Court in the Tudor and Stuart periods were mostly resident in Westminster and consisted of courtiers, members of the royal household, high-ranking government officials and diplomats directly reporting to the ruling monarch or the Lord Chancellor. In East Anglia and the North no similar social divide can be found. The high overall literacy rate of London discussed by social historians like Cressy (1980: 72-76) may be reflected in the number of lower- ranking letter writers in the City and Southwark; they included men like Philip Henslow and his theatrical circle, writing around 1600.
Unlike in Chapter 7, both men and women writers are included in our regional subcorpora. As nonmobile older rural males, the 'NORMs' of Chambers and Trudgill (1980: 33), have traditionally served as English regional dialect informants, including women in our data ought perhaps to be justified.11 There is some research suggesting that females typically migrated shorter distances than males in early modern England (Clark & Souden 1988: 19). In this respect women might be thought of as better regional dialect informants than men. But, for several reasons, this need not be the case.
Especially in the Tudor period, it was customary among the higher ranks to send both boys and girls away to complete their education in the households of family friends or relatives. Women also typically moved house when they got married. In the seventeenth century, the higher ranks became increasingly mobile and attended the London social season on a regular basis. All these facts suggest that those who were fully literate in the early modern period were not typical rural dialect informants: they were either socially privileged and travelled for business and pleasure, or their training and occupation, such as apprenticeship or the merchant's trade, were apt to promote their regional mobility. It is these kinds of people that one would expect to be exposed to supralocal tendencies in the language community, to be in a position to adopt ongoing changes, and by doing so to contribute to dialect levelling.
For all these reasons our classification is not aimed to make a basic distinction between those inhabitants of a locality who had migrated from elsewhere and those who were born there. Nor does it include those who had emigrated from their native area and permanently settled somewhere else. All that matters is where they had settled down to live and work on a permanent basis. It is nevertheless telling of the migration history of the period that three quarters of our Norfolk and Suffolk writers whose place of birth is known to us were native East Anglians. Also, as far as we can ascertain, more than 80 per cent of the Northerners in our data were born in the North. But the picture of London is very different: out of all the writers included in the CEEC who permanently lived and worked in London, and whose place of birth has been documented, about one fifth are recorded as having been born there. This is also the case with only one quarter of those who were attached to the Royal Court. These figures come close to the estimates of the proportion of native Londoners based on the sixteenth- and seventeenth-century records discussed above.