Linguistics in Göttingen - A platform for empirical and theoretical linguistics

Hagen Pitsch (University of Göttingen):On Bulgarian DA

The talk is concerned with the Bulgarian verb-related particle DA. Despite a plethora of literature, there is still no widely accepted view on the syntax and semantics of this item. Existing analyses locate DA either in I or C, while its semantic contribution is mostly described in terms of modality. In the present talk, I will argue for the distinction of two variants of DA, both of which can, however, be subsumed under the broader notion of tense (finiteness) and thus be located in I:

#1: ?infinitival? DA specified [-tense]

#2: ?finite? DA specified [+tense]

The two variants differ wrt nominative case licensing and the availability of tense forms other than present tense. While #1 excludes both the nominative and other tenss, #2 co-occurs with the nominative and allows for a variety of tenses. I analyze structures involving #1 as non-finite IPs (reminiscent of infinitival VPs in other languages), whereas structures involving #2 are analyzed as finite CPs. Nonetheless, both variants can be modelled to share the same, partially uderspecified lexical entry.

The focus of the talk will be on the semantics of DA. As the claim that #2 introduces modality can be shown to be too strong, the question arises how to capture the fact that this variant generally comes along with some sort of ?modal? interpretation (irrealis, unveridicality, future orientedness, etc.). A plausible solution is to treat #2 as a subjunctive marker. The subjunctive might either be seen as a ?marked? mood (probably adding a presupposition saying that the world that the uttered proposition is part of does not conform to the actual world of the modal subject; cf. Zimmermann 2015). But such an analysis would still ?burden? #2 with rather ?strong? semantics. An alternative is Schlenker?s (2005) claim that certain moods are semantically vacuous, but are used in case competing forms would give rise to a presupposition failure. Along these lines, #2 is chosen in case the (semantically strong) indicative would yield such a failure. A promising way to explain the competition of #2 with the infinitive (i.e. #1) and the imperative, is Farkas? (1992) proposal that the subjunctive is used in environments where the afore-mentioned morphological forms are not available (presumably for strictly grammatical reasons).

The talk is meant as a basis for a fruitful discussion and evaluation of the indicated (and possibly more) options.


Farkas, D. (1992): On Obviation. In: I. Sag & A. Szabolcsi (eds.), Lexical Matters. Stanford: CSLI, 85-109.

Schlenker, P. (2005): The Lazy Frenchman?s Approach to the Subjunctive. Speculations on Reference to Worlds and Semantics Defaults in the Analysis of Mood. In: T. Geerts, I. van Ginneken & H. Jacobs (eds.), Romance Languages and Linguistic Theory 2003: Selected papers from ?Going Romance? 2003, Nijmegen, 20-22 November. Amsterdam: John Benjamins, 269-309.

Zimmermann, I. (2015): The Russian Subjunctive. In: G. Zybatow et al. (eds.), Slavic Grammar from a Formal Perspective. The 10th Anniversary FDSL Conference, Leipzig 2013. Frankfurt am Main: Peter Lang, 579-594.