The newly born definite article spreads along two paths – a theoretical discussion and a case study in Old Scandinavian

Abstract

This article contains an investigation of the spread of the definite article in Old Scandinavian on its way towards obligatorification from the very scant use of the newly born article. On an overall level, it turns out that the process is significantly earlier in West Norse than in East Norse. The empirical analysis is preceded by a theoretical discussion of which referential categories there is reason to distinguish, alongside the (direct and indirect) anaphora, where the development is generally assumed to begin, as definite articles normally arise from demonstratives in anaphoric use. When it comes to the non-anaphoric NPs, it is argued, with inspiration from Löbner 2011, that there is reason to pay regard to whether or not the head noun expresses inherent uniqueness and/or relationality. This view is confirmed by the empirical analysis, which shows that scribes can be significantly more as well as significantly less inclined to use the definite article in NPs with such features than in NPs without. Moreover, this result suggests that definiteness marking does not progress strictly linearly along one single path, contrary to what various models in the literature presume (including the scale of uniqueness in Löbner 2011). Instead, a two-path model is proposed.