English stative possession has long been the site of variation: have (historical) vs. have (innovative) vs. got (recent). British varieties have been shifting toward have got (Tagliamonte 2003), but a resurgence of have is evidenced in Canadian English (CE) (Tagliamonte et al. 2010). Variation is also more constrained in CE, limited to present tense affirmative declaratives; elsewhere, only have occurs. Moreoever, the status of have marks a critical ‘transatlantic divide’ (Denison 1998), requiring do-support and resisting contraction in North American varieties.
These distintions are well established for contemporary use but their emergence in CE has never been considered from a diachronic perspective. The existence of large text collections in one of Canada’s youngest cities enables precisely this kind of longitudinal analysis, tracking the trajectory of change within the stative possessive system. Drawing on the full history of the local newspaper, this paper presents continuous data from over 140 years of evidence on stative possessive (1858-2000).
In contrast with current use, the early data reveal a system virtually invariant with respect to possessive verb (1858-1918: 98% have, N =2851). These data thus appear paradigmatically ‘Canadian’ in historical perspective (Dollinger 2008; Tagliamonte et al. 2010). However, given the ongoing retrenchment of have in CE, is seems clear that it participated to some extent in the change toward have got. The causation of the ‘reversal’ to have remains unclear, but do-support contexts suggest a possible explanation. Both historically and contemporaneously, do occurs strictly with have. Have got is most robustly attested in these data in traditional interrogatives, suggesting that the North American resurgence of possessive have was driven, at least in part, by the concomitant participation of stative have in a separate change: do periphrasis. These data thus have the potential to demonstrate the ramifications of one change on the developmental trajectory of another.