Unaccusatives as Small Clauses

This paper argues against the commonly held assumption that subjects of unaccusative predicates enter the syntactic derivation in the canonical direct object position, as shown in (1a). Rather, I argue that subjects of unaccusatives are merged as the subject of their own Small Clause as shown in (1b), uniting them with copular constructions in Russian and other languages.

(1) a. Standard                 b. Small Clause
      Unaccusative                 Unaccusative
         VP                            VP
        /  \                          /  \
       V    DP                      V     SC(=PredP)
                                         /  \
                                       DP   (pred)

Moro (1997) proposes a similar syntactic analysis for unaccusatives based on evidence from Italian existential sentences involving the verb esserci ‘to be’. Like unaccusatives in Italian, ci-sentences select essere as their auxiliary in the past tense, show participle agreement with the subject, and allow for ne-cliticization, i.e., they pass all the standard diagnostics posited for Italian unaccusatives. Therefore, a unified syntactic account of these predicates appears to be a logical conclusion. Moro suggests, following Hale and Keyser (1991), that there are no true theta-roles. Instead, relations previously called “thematic” are determined by structural relations among categories and their projections. The role of theme, the role assigned to the subject of unaccusatives, is associated with the incorporation of a lower predicate into a higher V position, yielding the semantics commonly associated with unaccusative predicates, e.g., change of state, change of location, change of condition, etc. This is the analysis I adopt for unaccusatives in Russian.

While Unaccusativity has been discussed in great detail for the Romance languages, Slavic unaccusatives raise many new questions. This paper focuses on one such question: can the syntax of copular constructions and unaccusatives in Russian be accounted for with the single underlying structure shown in (1b)? I argue that this is indeed the case. However, the diagnostics posited for unaccusatives in Romance cannot be applied to Russian, where the situation is less than clear. One syntactic fact which Russian existential sentences with the verb byt′ share with unaccusatives is the licensing of Genitive of Negation on their sole syntactic argument, as shown in (2a–b).

(2) a. V  kvartire  ne bylo nikakix devušek. (byt′)
       in apartment no was  no      girls-GEN
       There were no girls in the apartment.

    b. Nikakix otvetov     ne  prišlo. (unaccusative) 
       no      answers-GEN not came
       No answers came.

However, Genitive of Negation is a poor diagnostic for true unaccusativity in Russian since it is dependent on factors such as definiteness, animacy, etc. Therefore, another diagnostic must be examined. Babyonyshev (1996) discusses the possibility of “conjunction agreement” for unaccusatives in Russian as in (3a). In Russian, a verb may show singular agreement with the first conjunct of a conjoined nominative subject if the conjoined subject is postverbal and the sole argument of an unaccusative verb. I argue that the small clause structure proposed in (1b) will still allow us to capture these facts in Russian, with the added advantage of allowing us to collapse the syntax of two constructions (copular sentences and unaccusatives) into one structure.

3) a. Unaccusative
      Na stole stojali/stojala/*stojal pepel′nica  i   pustoj stakan.
      On table stood-PL/FEM SG/MASC SG ashtray-FEM and empty  glass-MASC
      On the table stood an ashtray and an empty glass.

    b. Transitive
       Stixi pišut/*pišet Svetlov i Romanov.
       poems write-PL/*SG   Svetlov i Romanov
       Svetlov and Romanov write poems.

If time permits, I will discuss the implications of such an analysis for cases of “di-unaccusatives” in Russian as well, i.e., object experiencer psych verbs, which canonically select an experiencer and a theme. If we follow Moro’s proposal concerning the structure underlying all instances of “themes” in the syntax (à la Hale and Keyser), we force ourselves to adopt a very specific analysis of these constructions, i.e., the structure in (1b) above.

Babyonyshev, M. 1996. Structural Connections in Syntax and Processing: Studies in Russian and Japanese. Ph.D. dissertation, MIT.

Hale, K. and Keyser, S.J. 1991. On the Syntax of Argument Structure. Cambridge, MA: The Lexicon Project Monograph Series, Center for Cognitive Science, MIT.

Moro, A. 1997. The Raising of Predicates. Cambridge: Cambridge University Press.