Diachronic syntax

My research primarily focuses on the syntax of Old French and Old Occitan. These two languages share numerous similarities, making them the perfect testing ground for investigating microvariation both synchronically and diachronically.

I am particularly interested in the distribution and evolution of clitic placement, notably clitic climbing and the transition from enclisis to proclisis with infinitives, which both of these varieties underwent. My research demonstrates that in French, this transition occurred as infinitival morphology eroded. In an article published in the Journal of Historical Syntax, I have argued formally that the non-finite IP-domain was stronger, or richer, in earlier French, and that the aforementioned erosion led speakers to cease acquiring high V-movement (across the clitic) in this context. In languages where non-finite morphology remains intact (e.g., Italian, Spanish, and to a certain extent, Catalan), this change never occurred, and enclisis remains the standard order. Concerning Old Occitan, the empirical work I have been undertaking on clitic placement reveals additional constraints that are connected to prosody. Eventually, I am interested in investigating and defining the connections and differences between clitic placement and cliticisation.a

Additionally, I have studied infinitive fronting, a phenomenon present throughout all Old Romance varieties that subsequently disappeared from each. In an article published in Isogloss, I argue that this word order is not a case of Stylistic Fronting, as previously assumed in the literature. Given that infinitive fronting is only available in monoclausal structures where the matrix verb is a modal, I propose that this word order is, in fact, a Transparency Effect.

My focus also extends to other Transparency Effects and the structure of Restructuring clauses. I am particularly interested in examining the interaction of clitic climbing and auxiliary switch, which has been observed empirically but not formally addressed. Considering that this interaction weakens when the clitic is a locative, partitive, or impersonal, I am considering whether the availability of person-features on v*-heads can account for both climbing and HAVE-BE switch.