The Link Between Case and Event Structure in Russian

Kylie Richardson, Harvard University

In my paper, I explore how a syntactic analysis can account for the fact that some direct internal arguments in Russian occur with instrumental, genitive or dative case marking (a phenomenon that is referred to as "quirky" case marking in the syntactic literature on case), while all others occur with accusative case marking, as seen in examples (1) (dative case marking on the internal argument) and (2) (accusative case marking).

(1) Oni sluzhili emu.

They served him–DAT

(2) Ja kupila knigu.

I bought (a) book–ACC

I will explore the link between event structure and morphological case to provide a solution to this problem. (Note that event structure is often also referred to as semantic or inner aspect. In Vendlerian (1957) terms, event structure refers to whether a verb and its arguments is perceived as a state, activity, achievement or accomplishment.) I will also suggest that by incorporating event structure in syntactic phrase structure we can account for many other case marking patterns in Russian that otherwise seem anomalous.

If we analyse the event structure of the verbs that take quirky case marked arguments, it not only becomes clear that these verbs are almost exclusively non–eventive—they are either states or activities (versus eventive achievements and accomplishments), but also that their object arguments never play a role in determining their event structure syntactically. It is generally accepted that, unlike external arguments and indirect internal arguments, the direct internal argument can play a role in determining the event structure of a predicate syntactically. While a verb like dance, for instance, in its base meaning is an activity, once we add the definite noun phrase the jig, the verb has an ambiguous interpretation between an activity and an accomplishment, as seen in examples (3) and (4). Dowty (1979) formulated a test to determine whether a verb was an activity versus an accomplishment. If a verb can occur with the adverbial in x time, it is an accomplishment, since the action has an inherent endpoint. If, instead, the verb can occur with for x time, it is an activity, since it denotes an ongoing action. Dowty's test shows that example (3) is unambiguously an activity, while (4) has an ambiguous interpretation. Crucially, it is the direct object that affects the event structure of the verb syntactically.

(3) Alli danced for an hour/*in an hour. (activity)

(4) Alli danced the jig for an hour/in an hour. (activity/accomplishment)

Dowty's test can also be applied to Russian. Notice in Russian that the verb tancevat' 'dance' without its direct object present is ungrammatical with the time expression za chas 'in an hour', once we add the direct object dzhiga 'jig', however, as in English, the verb is aspectually ambiguous. This has nothing to do with the imperfective or perfective morphological aspect of the verb in Russian. Example (7) shows that the verb is ungrammatical with the perfective aspect, za chas, and no internal object (note that example (7) is only grammatical if an elided object is understood contextually).

(5) Ona tancevala chas/*za chas.

She danced for (an) hour/*in (an) hour.

(6) Ona tancevala dzhigu chas /ona stancevala dzhigu za chas.

She danced (the) jig for (an) hour/she danced (the) jig in (an) hour.

(7) *Ona stancevala za chas.

She danced–PERF in (an) hour.

The direct internal argument of a verb like tancevat' 'dance' therefore also plays a role in determining the event structure of the verb syntactically.

If, however, we apply Dowty's event structure test to verbs that take so–called quirky case marked objects, we see that they are ungrammatical with za chas 'in x time', regardless of whether their internal argument is interpreted as definite or not, as examples (8) and (9) show, with instrumental and dative case marked internal arguments, respectively.

(8) *Ja gordilas' studentami za pjat' minut.

*I (was) proud (of the) students–INSTR in five minutes.

(9) *Ja doverilas' studentam za chas.

*I trusted (the) students–DAT in an hour.

In my paper, I will show that genitive, dative or instrumental case marked direct objects never play a role in determining the event structure of a predicate in the syntax. Verbs that take accusative case marked arguments, however, fall into three types: (1) inherently eventive predicates; (2) predicates whose internal argument plays a role in determining the event structure of the predicate syntactically; and (3) pure non–eventive predicates like those that take quirky case marked arguments. I will posit a syntactic configuration for case marking that captures the role that an internal argument can play in the event structure of a predicate and the relationship between morphological case marking and this role.

Such a syntactic account of event structure also provides a solution to other curious case marking phenomena in Russian. Consider, for instance, the distribution of the instrumental case marking that occurs on some secondary predicates in Russian. This instrumental case marking is only grammatical with external argument antecedents (both nominative and dative "subjects") and some accusative case marked antecedents, as examples (10)–(14) show.

(10) Polina prishla na vecherinku p'janaja/p'janoj.

Polina–NOM arrived at (the) party drunk–NOM/drunk–INSTR

(11) Mne ne rabotaetsja p'janym/?p'janomu.

I–DAT NEG work drunk–INSTR/?drunk–DAT

'I don't feel like working'.

(12) Ja nashla egoi p'janymi/p'janogoi.

I found him–ACC drunk–INSTR/drunk–ACC

(13) Ja tolknula egoi p'janogoi/*p'janymi.

I pushed him–ACC drunk–ACC/*drunk–INSTR

(14) Ja pozvonila emui p'janomui/*p'janymi.

I phoned him–DAT drunk–DAT/*drunk–INSTR

Notice that an instrumental case marked secondary predicate is only possible with a nominative (example 10) or dative (example 11) external argument antecedent, or the accusative case marked antecedent in example (12). Despite the fact that in (13) the internal argument also has accusative case, a secondary predicate with instrumental case marking is ungrammatical, as it is with the dative antecedent in (14). I will show that this restriction on the so–called predicate instrumental is linked both to the structural location of an antecedent, and to the event structure of both the primary and secondary predicate. Consider the ungrammaticality of the predicate instrumental with an object antecedent in examples (13) and (14). Notice that the primary predicates in these examples are non–eventive, as our event structure test shows, i.e., neither verb can occur with 'za + time expression': *Ja tolknula ego za pjat' minut '*I pushed him in five minutes'/*Ja pozvonila emu za pjat' minut '*I called/phoned him in five minutes'. The primary predicate in example (12), however, is eventive, since it is grammatical with 'za+ time expression': Ja na&shachek;la ego za pjat' minut 'I found him in five minutes'. In my paper, I will claim that herein lies the solution to the distribution of the predicate instrumental in Russian, namely the instrumental case on a secondary predicate is also sensitive to the event structure of the primary predicate with which it occurs.

Thus, my paper will explore how a syntactic analysis of event structure can account for otherwise anomalous case marking patterns in Russian. Most of the generalizations I will present in my paper are based on research recently conducted in Russia. My findings are the result of the grammaticality judgments of over thirty native speakers I interviewed there.