The EPP, Expletives, and Ukrainian -no/-to

Given the role that EPP-feature checking has continued to play in generative (minimalist) syntax, it is reasonable to ask how this checking requirement manifests itself in Slavic, particularly in the case of Slavic impersonal predicates where a thematic external argument is not available to perform this function. The EPP-feature of T(ense), or T’s “D-feature,” requires that the highest Spec of the verb’s extended projection be filled. Such a discussion necessarily involves a reconsideration of the role that expletives play (expletives have been argued to be the Spell-Out of T’s D-feature (Jonas 1996). It will be recalled from recent work in checking theory, that pure expletive-Merge is “less costly” from an economy perspective than Move. The main question this paper addresses is whether this less costly operation is available in Slavic impersonal predicates.

The Ukrainian -no/-to construction is chosen as an appropriate testing ground for the following reasons: (i) this is an impersonal passive construction that allows for comparison with Germanic impersonal passives, which rely on an expletive-associate chain for EPP-checking; (ii) it has been claimed that Ukrainian has an overt expletive (vono) Franks 1995), which has been argued to occupy the subject position in the -no/-to construction (Billings 1993); (iii) Ukrainian -no/-to is necessarily transitive and, thus, by hypothesis, the EPP-feature of T may be satisfied by a non-nominative internal argument (the fact that a non-nominative argument can check the EPP is predicted by the minimalist claim that subject properties can be “de-coupled” and checked independently); and (iv) an examination of EPP-checking in Ukrainian -no/-to allows for close comparison with Polish -no/-to, a construction that (I will argue) minimally differs from its Ukrainian counterpart by maintaining a thematic subject position and checking its EPP-feature with an (external) pro-arb argument.

An example of Ukrainian -no/-to is given in (1). Note that raising of the internal argument to check the EPP is not a case of topicalization. The EPP-position is viewed as “divorced” from discourse interpretation, following recent work by Babyonyshev (1996). Thus, the preverbal internal argument is optionally non-D(iscourse)-linked and felicitously responds to an out-of-the-blue “what-happened”-type question (e.g., Ukr.: Ščo stalosja?).

(1) Mašynu (acc) bulo zipsovano ([-agr]) v avariji.
    A car was damaged in an accident.

The argumentation presented relies largely on empirical facts, many of them from traditional descriptive sources (e.g., Shevelov 1969) where the possibility of pleonastic subjects in Ukrainian is discussed. The goal of this kind of research is to explore the descriptive generalizations that are made possible by an examination of problems that are theoretical in nature.

The central claim of this paper is that the expletive checking solution of pure-Merge is not available in Ukrainian -no/-to. This will account for the correct discourse-neutral word order found in (1). It will be argued that Ukrainian vono is not a D-feature-bearing expletive and, thus, is not involved in EPP-checking. Note that the verb-initial structure in (2), reported as awkward by informants, is not improved in (3) with vono (the judgments are for neutral discourse):

(2) ??Bulo zipsovano mašynu v arariji.
(3) ??Vono bulo zipsovano mašynu v arariji.

Next it will be shown that an analysis in which a null expletive is posited to satisfy the EPP forces the undesirable stipulation of a special, anomalous type of expletive-associate relationship (one that lacks the well-known “definiteness effect”) just for this Ukrainian construction. To summarize, based on word-order, the distribution of vono, and crosslinguistic expletive-chain facts, use of an EPP-satisfying expletive in Ukrainian -no/-to is shown to be simply an analytical tool of convenience motivated exclusively by theory-internal considerations.

Finally, the fact that Polish -no/-to does not require raising of an internal argument for EPP-checking provides an important insight into the underlying syntax of this Polish construction, why it differs so dramatically from Ukrainian -no/-to in terms of its distribution, and why it is generally believed to have an active interpretation. The discourse-neutral Polish -no/-to response to “what happened?” (Co się stalo?) given in (4) thus contrasts sharply with Ukrainian (1):

(4) (pro-arb) znaleziono ([-arg]) niemowlę (acc) na śmietniku.
    They found a newborn at the dump-site.


Babyonyshev, Maria. (1996) Structural Connections in Syntax and Processing: Studies in Russian and Japanese. Ph.D. dissertation. Cambridge, MA: MIT.

Billings, Loren. (1993) “A Note on Expletives in Ukrainian -no/-to Passives.” Slavic Syntax Newsletter 3(1): 1–8.

Franks, Steven. (1995) Parameters of Slavic Morphosyntax. New York: Oxford University Press.

Jonas, Dianne. (1996) “Clause Structure, Expletives, and Verb Movement.” In Minimalist Ideas: Syntactic Studies in the Minimalist Framework, W. Abraham et al., ed. Amsterdam: John Benjamins. 167–88.

Shevelov, George. (1969) “The Vicissitudes of a Syntactic Construction in Eastern Slavic: Impersonal Sentences in -no, -to with the Acc sg of Object.” Scando-Slavica 15: 171–86.