Brian D. Joseph
I discuss here various ways in which one might devise a counting heuristic for grammaticalization with an eye to testing the quantificational claims that have been made against specific implementations of such a heuristic. More specifically, I address the question of grammaticalization as a phenomenon of individuals versus a phenomenon of speech communities versus a phenomenon of languages. Similarly, I hope to show, once the individual versus group issue is dealt with, that by adopting Haspelmath’s (2004) definition of grammaticalization as the tightening of internal dependencies, and thus a weakening of boundaries, between elements, we are in a better position to undertake a census since linguists have developed a reasonable idea of the sort of grammatical boundaries that need to be posited (word boundaries, clitic boundaries, morpheme boundaries, phoneme-to-phoneme transitions, etc.). Further, this view generalizes to offer a solution to the problematic notion of gradience in grammaticalization - cf. Kuryłowicz’s famous definition of grammaticalization as taking in movement from “less” to “more” grammatical - since linguists have long posited a hierarchy of boundary strength that can be appealed to.