Timing Linguistic Changes
The accurate timing of linguistic changes has not been a central issue in historical linguistics. There is still a great deal of truth in Chen and Wang's argument (1975: 256) that 'one of the most neglected aspects of historical linguistics, which professes to be a study of language evolving across time, is the time element itself' (see also Raumolin-Brunberg 2003).
Timing linguistic changes is not necessarily such a straightforward matter as one would expect at first sight. It is clear that, if a change has reached Labov's fifth, completed stage, the new form has triumphed and language has changed. But it is not unusual that the old form lingers on in some linguistic environments, some dialects or a genre or two. For instance, the possessive determiner its3 has not ousted its postnominal variants of it and thereof from the language, and the third-person singular suffix -th is found in King James Bible, which is still in use today. Can we speak of language change, although these processes have not been completed? If we can, as we assume is the case with our examples, what then is the limit, how much linguistic and nonlinguistic variation is allowed? If we look at the question from the angle of the speech community, how large a percentage of speakers should have adopted the innovation or used it as their main variant before we can say that language change has taken place?
There is no one answer to these questions. In what follows we shall look at a number of processes of change as we see them in our data. In some cases we are tempted to speak about change when totally new elements are introduced, although it takes time until they become firmly established in the language. This is what happened, for instance, when its was introduced into the possessive paradigm around 1600 (Nevalainen & Raumolin-Brunberg 1994a). On the other hand, shifts in uses and frequencies of existing elements, such as the object form you replacing ye as subject, tend to make us feel that language change has taken place only when most people have changed their usage. One might even suggest that there are two points of change, namely when the variation leading to a shift begins and when the process comes to an end, or at least when the old form is understood to be archaic or only limited to a specific use or type of writing.
Timing linguistic changes cannot be discussed without taking into account their multidimensional character. Besides linguistic and social embedding, textual aspects also need attention. At the same time as changes spread among people and across linguistic environments, they also find their way from one register or genre to another.
As pointed out in Chapter 3, by choosing personal letters as the material for this study, we have aimed at capturing linguistic innovations as quickly as possible after their emergence in colloquial spoken language. However, there was probably some time lag before spoken-language innovations found their way into writing, at least to any larger extent.4 It seems plausible, however, that the CEEC dates change of this type earlier than more formal and literary genres would do. Some changes discussed in this book most likely in Labovian terms came from above and followed a written-language model. Their introduction into personal letters presumably occurred later than their innovation in formal genres. Systematic research on the time courses of linguistic changes in different types of writing is needed for the corroboration of these assumptions.