In defense of a pragmatic view of reanalysis

This paper will defend a view of reanalysis as a key empirical phenomenon in grammatical change (defined broadly as including meaning change), arguing that an interactionist point of view offers independent justification for this assumption. I propose that reanalysis is a mechanism of change that is implemented at the level of individual language users, and which is fundamentally driven by hearers and by pragmatics. Having proposed a Constructionalist definition of reanalysis, I go on to distinguish two different subtypes, according to whether or not the hearer’s mental grammar already includes an existing analysis of the construction that is reanalyzed. Subsequently, I discuss the roles played by context and by frequency in reanalysis, and I seek to place the phenomenon within an overall typology of forms of language change, partly with a focus on forms of change that do not constitute reanalysis, and what makes them different from the latter, and partly with a focus on the relationship between reanalysis and grammaticalization.