Previous approaches to Russian case may be divided into four groups: 1) the atomistic tradition, which merely lists contextual meanings of the six cases in Russian; 2) the Jakobsonian tradition, which advocates the principle of invariance and operates with binary feature oppositions; 3) the GB tradition, which generally distinguishes two types of case: structural case and inherent case; and 4) the newer cognitivist tradition, which has totally abandoned the feature approach and instead operates with prototypes/core meanings and submeanings. The four theories are briefly examined and tested against various types of parameters and data. It is found that they do not meet realistic requirements for a theory of Russian case and that they are unable to handle the data adequately, instead confusing levels which should be separated and treating contextual meanings as case meanings. Specific requirements for a theory of Russian case are set up and against this background a new theory is constructed which is based on the assumption that an isomorphic relationship exists between the structure of the nominal system and the structure of the verbal system. It is argued that Russian case sensu stricto is the nominal equivalent to mood. The theory includes two different case systems: 1) the propositionally defined system, which involves deep syntax and is universal; here a distinction is made between casus exterior, i.e., cases which function as underlying subjects (nom, acc, and gen) and casus interior, i.e., cases which function as underlying determiners (dat, instr, and loc); and 2) the referentially defined system, which involves surface semantics and mood as well, and is the specific Russian contribution to case semantics. Here a distinction is made between direct cases (nom and acc) and oblique cases (voc, gen, dat, and instr)&emdash;the latter are further divided into outer cases (voc and gen) and inner cases (dat and instr). All previous theories have been concerned far more with the relationship of Russian case to the universal system, i.e., deep case, and far less with the specific Russian system, i.e., surface case. They have dealt with what could be called participant roles as opposed to case roles, and been unable to connect the pure case system and the prepositional system, where the distinction between contact cases (loc and acc) and non-contact cases (gen, dat, and instr) replaces the distinction between direct and oblique cases.

aisrael@american.edu

This article investigates the distribution of clitic clusters in Bulgarian, Czech, Serbian-Croatian, and Slovak. It argues that clitic placement depends on both syntactic and prosodic factors. The syntactic factors include whether the clitic cluster is in I0 (verbal clitics) or C0 (second-position clitics) and whether there is a constituent which is a prosodic word, e.g., a complementizer or a topic, before the cluster. If the cluster is syntactically clause-initial, Prosodic Inversion occurs to provide a host for the clitics, resulting in the clitic cluster appearing after the first prosodic word in the clause. Differences among the languages reflect differences in where the clitic cluster is located syntactically and lexical differences in the clitic inventories and their prosodic properties. This analysis is contrasted with proposals arguing for Long Head Movement of participles to C0, and additional data involving optional participle movement, negation, and li questions are examined.

This paper discusses the composition of the *u°-stem class in Common Slavic. It is shown that if the lists of *u°-stems proposed in various specialist studies are combined with those found in more general works, nearly 150 nouns may be reconstructed as original *u°-stems, with varying degrees of probability. Forms usually neglected in the discourse (e.g., *kru°tu°-) can be shown to be almost certain original *u°-stems. Based mainly on cognates from Lithuanian, a similar number of adjectives may also be reconstructed as *u°-stems, giving a possible total of nearly 300. It is therefore proposed that the *u°-stems in Common Slavic were not a marginal class, but a fairly numerous, productive one, which strengthens the hypothesis of an early *u°-stem influence within the Common Slavic declensional system as a whole.

This paper examines case marking in Russian negative intransitive constructions, focusing on the lexicon and aspect. It treats two lexical hierarchies that together form a cline expressing the relative frequency of genitive case marking with different combinations of nouns and verbs. It thus demonstrates the extent to which case marking is predictable from the lexical content of sentences. The paper considers the effects of aspectual form and function, showing that submeanings of different morphological forms pattern together. This finding supports Timberlake's (1982) claim that morphological aspect has a limited role in the grammar of case, and that a grammatical description needs to include mapping rules from individual aspectual functions to morphological case.