Contractions, constructions and constructional change

Abstract

In this paper, I argue that construction grammarians may have to consider integrating modal contractions into the English modal system as distinct constructions rather than variants of their uncontracted forms. Based on data from COHA, it can be shown that the contractions investigated here have emancipated themselves from the full forms both in terms of relative usage frequency as well as function over the past two centuries, thus yielding a series of constructional changes. From a usage-based, constructionist perspective, these results contribute to modelling the modal network as possibly represented in the minds of speakers, but they also support the understanding that this network appears to be much more heterogeneous than perhaps desired (by some linguists).