This paper examines constructions consisting of a Dative NP + a predicate word ending in -o + an infinitive, as in (1). I will demonstrate that, despite their identical components, they have distinct syntactic structures. Their disparity results from the varying lexical properties of o-predicates, which I divide into three general lexico-semantic groups: i) PHYS/PSYCH: words expressing physical or psychological state; ii) WTH/ENV: words expressing the weather or one's surrounding environment; iii) EVAL/CHAR: words which make an evaluation or characterization:

(1) a. PHYS/PSYCH
       Mne     xolodno    exat&soft;.
       me(DAT) cold[-AGR] go(INF)
       'I am too cold to go.'
       [Pe&shachek;kovskij 1938:326-7]

    b. WTH/ENV
       Nam     vetreno     idti      po    beregu.
       us(DAT) windy[-AGR] walk(INF) along seashore
       'It is too windy for us to walk along the seashore.'
       [Galkina-Fedoruk 1958:290]

    c. EVAL/CHAR
       Gosudarstvu     nevygodno       pokupat&soft; zerno
       government(DAT) senseless[-AGR] buy(INF) grain(ACC)
       'It is unprofitable for the government to buy grain.'
       [Voinova 1965:154]

Both Experiencers of o-predicates and subjects of infinitives occur in the Dative Case in Russian. When both predicate types appear in the same sentence, as in (1), only one NP may occur. Some strategies independently motivated in syntactic theory which can account for this phenomenon are listed in (2):

(2)

a. Binding: The DAT Experiencer of the o-predicate binds (controls) the PRO subject of the embedded infinitive clause;

b. Raising: The DAT subject of the embedded infinitive clause raises to the matrix clause, leaving a trace;

c. Extraposition: The infinitive and its DAT subject together constitute the subject of the o-predicate. The infinitive is then extraposed later in the derivation.

In this paper I show which strategy from (2) corresponds to each sentence in (1), based on the lexical semantics of each type of o-predicate. First, of the three types only PHYS/PSYCH selects an Experiencer argument. I conclude then that sentences like (1a) must be an instance of Binding since this strategy is the only one involving an Experiencer argument:

(3) Mne_i xolodno [PRO_i exat&soft;].

The structure in (3), however, will not accommodate the other types. WTH/ENV words can cooccur with a DAT NP only when they are also accompanied by an infinitive as well:

(4) Nam vetreno *(idti po beregu). It is too windy for us (to walk along the seashore).

I argue that this is because WTH/ENV-types do not license Experiencers. The DAT NP that occurs in sentences like (1b), then, must be solely the subject of the infinitive clause. Therefore, the NP apprearing sentence-initially is an instance of Raising:

(5) Nam_i vetreno [t_i idti po beregu].

Finally, EVAL/CHAR words are similar to WTH/ENV words with respect to allowing a DAT NP only when a infinitive cooccurs:

(6) Gosudarstvu nevygodno *(pokupat' zerno). It is unprofitable for the government *(to buy grain).

I propose, however, that EVAL/CHAR words are distinct from the two previous types in that they select an external (theta-role and assign it to a grammatical subject, namely, the infinitive clause itself. This explains why the infinitive must be selected and overtly realized in the syntax. The infinitive clause, functioning as the grammatical subject, undergoes Extraposition:

(7) [_Subje]_k Gosudarstvu_i nevygodno [PRO_i pokupat&soft; zerno]_k

More proof of the subjecthood of the infinitive is lent by the fact that it can be replaced by a NP, with which the EVAL/CHAR word will agree in gender and number:

(8) Pokupka zerna gosudarstvu nevygodna. purchase_NOM.F.SG grain_GEN government_DAT unprofitable_F.SG The purchasing of grain is unprofitable for the government.

While the DAT NP is being selected also by the matrix o-predicate, it is not an Experiencer, but instead some sort of benefactive (theta-role or dative commodi. Evidence for this claim is the fact that with most EVAL/CHAR words the DAT NP can be replaced by the PP dlja + GEN.

References

Babby, Leonard H. (1996)

— (1997)

— (1998)

Bailyn, John F. (1991)

Billings, Loren F. (1998)

Bo&shachek;ković. &Zhachek;eljko (1996)

— (1997) The Syntax of Nonfinite Complementation. Cambridge, Mass. and London: MIT Press.

Chomsky, Noam (1995) The Minimalist Program. Cambridge, Mass.: MIT Press.

Galkina-Fedoruk, E. M. (1958) Bezli&chachek;nye predlo&zhachek;enija v russkom jazyke. Moskva: Izdatel&soft;stvo moskovskogo universiteta.

Grimshaw, Jane (1990) Argument Structure. Cambridge, Mass.: MIT Press.

King, Tracy Holloway (1995) Configuring Topic and Focus in Russian. Stanford: CSLI Publications.

Komar, Eric S. (1999)

Pe&shachek;kovskij, A. M. (1938) Russkij sintaksis v nau&chachek;nom osve&shachek;&chachek;enii, 3-e izd. Moskva: Gosudarstvennoe pedagogi&chachek;eskoe izdatel&soft;stvo.

Voinova, E. I. (1965)

— (1970)

Zaitseva, Valentina (1990)