Macro-sociolinguistics is concerned with societies and their various uses of linguistic resources. Variationist sociolinguistics is concerned with speech communities (however defined) characterized through the use of linguistic resources by (socially defined) individuals. The ultimate aim of any sociolinguistic theory must be to integrate the two. This means that macro-perspectives and macro-results must be brought in line with the results of variationist analyses of differences between groups of ‘informants'. This has to happen both synchronically and diachronically. A useful strategy may be to start with diachrony. What has to be integrated, then, is history, more specifically, the study of those historical societal processes which may be expected to have linguistic consequences in order to try out hypotheses on their relationships with the patterns revealed by variationist analyses.

In this paper, I attempt to bridge the gap by integrating analyses of historical processes defining the nation state Denmark in parallel and contrast to other Western European societies with the results of an analysis of the use of three phonetic variables by two generations of informants. 44 Copenhagen informants from the generations of 25-40 year olds and 15-24 year olds respectively were recorded in sociolinguistic interviews first in the mid 80ies and then again in 2005-07. This panel study will be supplemented by interviews with 14-16 year olds in 2007.

Copenhagen is first analyzed from the point of view of demography, urbanization, cultural centrality, as a state of mind, and as the centre of linguistic influence. The historical analysis shows that contrary to other Nordic nations Denmark is a monocentric society. Tore Kristiansen's results from the nationwide language attitudes study at the LANCHART centre lend strong support to this view (Kristiansen 2007). This leads to the hypothesis that all changes spread from the centre or are introduced in opposition to the centre, i.e. Copenhagen. Preliminary results from a dialect levelling study by Inge Lise Pedersen and Signe Schøning (Pedersen and Schøning 2007) support this.

One recent historical process which may be supposed to have linguistic consequences, however, sets Denmark apart from a number of European societies: institutionalization of socialization has progressed maximally in Denmark. Already in 1997 89% of the 3-6 year olds were enrolled in some form of day care. What influence does this have on the linguistic development of the new generations and hence on the patterns of variation in Copenhagen? The results of the LANCHART Copenhagen study and the results of a study by Marie Maegaard (2007) will be interpreted from this point of view: Can we e.g. find evidence of more age grading and less evidence of social background influence? And how could regular age grading be a part of the modified Labov model of linguistic change, (Labov 1994, Meyerhoff 2006, G. Sankoff 2005