Grammaticalization and the puzzle of the longue durée

This is the handout for a plenary talk at the conference "Grammaticalization and Diachronic Construction Grammar (Université Montpellier Paul Valéry, 2026 Sept 30 – Oct 2)"

Abstract: Grammaticalization was originally associated with long-term typological changes (from isolating to agglutinating and fusional), and the unidirectionality of change is one of the more robust findings (e.g. Börjars & Vincent 2011). Still nowadays, grammaticalization is associated with typologically oriented linguists (see Heine et al. 2026), whereas corpus-based diachronic research tends to prefer ideas of Diachronic Construction Grammar (e.g. Traugott & Trousdale 2026). But how do the changes over the longue durée arise from the kinds of short-term changes that we can observe in our historical records? Instead of proposing a new answer to this big question, I will try to make progress by (i) advocating more precise definitions of key terms (such as clitic, affix, lexeme), (ii) calling for realism in understanding the mechanisms of change, (iii) situating phonetic reduction in producer-comprehender interaction, and (iv) redirecting the focus of attention to the core of grammaticalization: changes from free forms to grammatical markers.