A non-standard form of the definite article is used by upwards of a million speakers every day and yet it has received little attention in terms of a phonological analysis. The most recent discussions have been a historical and contemporary survey, spectrographic investigations and sociolinguistic surveys. It is clearly a complex issue and has relevance to a number of theoretical questions, some of which are rarely addressed. We need to know the range of phonetic forms; the phonological distribution of the forms; the historical development and whether this helps with a synchronic analysis; and the number and nature of the lexical storage forms. The distribution varies between localities and between speakers; many speakers also use standard English the. The historical development is not at all clear, but does throw some light on what the synchronic grammars should be like. There appears to be a break between those grammars in which the phonetic forms of the definite article are predictable in terms of the phonological environment and those later grammars that have lexically specific forms. The changes that have taken place cannot be explained in terms of simple system-internal mechanisms and some knowledge of the sociohistorical background to the emergence of the new grammars is necessary. This article looks at historical sources from the mid seventeenth century onwards to try to determine the changes that have taken place in the northern dialects concerned, and proposes population movements and expansion as a determining factor in the changes that are in evidence. One possible scenario in terms of new-dialect formation during the nineteenth century is suggested. For the modern variants polysystemic, declarative grammars are proposed, with the possibility of grammar-switching under certain circumstances, such as repair or interaction with outsiders. © 2010 Cambridge University Press.