Abstract
The goal of this talk was to open the "Converbs from a cross-linguistic perspective" workshop by presenting the viewpoint of the Prague Descriptive Linguistics group which organised this event. I pointed to the observation that while the notion of converbs is well established in the descriptive grammatical traditions of many Eurasian languages, it is often absent in other traditions, despite the fact that many converbial constructions usually have functional counterparts in languages of the world, even though they might show different morphosyntactic marking.
It is therefore worthwhile considering a concept which could grasp such constructions cross-linguistically without being discharged by formal morphosyntactic discrepancies. This lead our group to adopting the notion of constructional auxiliation (CAUX) proposed by Florian Siegl, in order to account for the phenomenon of a lexical verb acquiring a grammatical function once used in construction with another verb.
In the talk I tried to show how such approach is related to the existing descriptive notions applying to multiverb constructions, mainly that of serial verb constructions and auxiliary verb constructions. As regards the characterisation of such multiverbal constructions in terms of the type of the dependency between the verbs, it turns out that the notion of co-subordination (and pseudocoordination) is a useful device.
I illustrated that by examples taken both from colleagues documenting Asian languages and from my own Arabic data (while also including data from Czech, taken from Škodová (2009)). Finally, I presented the main research questions which our group has set for the ensuing research on the topic.
They are the following: 1.) Which verbs tend to become constructional auxiliaries and why? Which properties of their lexical semantics contribute to their constructional meaning? 2.) What are the functions typically expressed by CAUX and what is the mapping between them and the lexical semantics of the verbs? 3.) What are the dependency relations present in CAUX and what are their consequences for the typologisation of CAUX (and possibly differentiation from SVC)? What are the morphosyntactic and combinatorial constraints pertaining to different CAUX?