Journal of Uralic Linguistics - Volume 2, Issue 2, 2023

The paper discusses the syntax of postpositional constructions in Meadow Mari (Uralic, Morkinsko-Sernur dialect; head-final, SOV). Building upon existing approaches to postpositions in the world’s languages, I propose that PPs in Mari may have one of two underlying structures depending on the nature of the dependent but not on the nature of the adposition. PPs with a pronominal dependent involve possession between the Ground and a location nominal. In PPs with a non-pronominal dependent the Ground is merged directly into the complement position of a P head. I further expand the dataset and show that the two configurations successfully capture the distribution of reflexive pronouns in PPs. The proposed analysis accounts for all the relevant data: examples with independent and affixal postpositions and referential and pronominal dependents.

This paper is dedicated to the debitive construction in Moksha Mordvin, which exhibits patterns unusual for a dative-infinitive construction (DIC). The non-finite verb can agree with its internal argument but not with the external one. The internal argument can bear both the definite nominative case, which is otherwise restricted to subjects of finite clauses, and the definite genitive case, which can mark both direct objects and definite possessors. Personal agreement is mandatory in the former case but prohibited in the latter. I propose that the Moksha debitive construction includes a null modal verb that assigns dative to the external argument of the infinitive. The case marking on the internal argument is determined by its structural position, in accordance with Dependent Case Theory. The case of Moksha demonstrates that the dative case in DICs can be inherent (contra what has been argued on the basis of Russian DICs) and that an argument for biclausality can be constructed based on case and agreement patterns.

This paper presents a descriptive account of reported speech, understood in terms of Spronck & Nikitina (2019), in Beserman (Uralic > Permic). Two phenomena are described in particular. First, it is demonstrated that, although Beserman generally prefers the direct speech strategy, the interpretation of indexicals in certain contexts may be affected by pragmatic factors. Second, Beserman allows for agreement mismatch between the subject and the predicate of the reported clause, known as “monstrous agreement”. This phenomenon has not been attested in Permic languages before. Apart from that, extended uses of two Beserman reported speech markers are described, whereby they convey pragmatic meanings that resemble English kind of or like. This semantic development reverses a well-known grammaticalization path that turns such pragmatic elements into reported speech markers. The study is based on both corpus and elicited data.

Juxtaposition as a machinery of building noun phrases is well-known to be widespread in Uralic languages: the modifier is left-adjoined to the head and does not bear any morphological marker of syntactic dependency. This strategy is used to attach adjectives, cardinals and modifying nominals to nouns (like in Beserman ǯ́aǯ́eg siĺ ‘goose meat’), and to build constructions with measure nouns (like Beserman odig kə̑də̑ gibi ‘one basket of mushrooms’, lit. one basket mushroom). However, there remains a question whether all these constructions share the same syntactic structure. We consider pseudopartitive constructions with measure nouns in comparison with NPs containing non-marked modifying nominals and NPs with adjectives, showing that they do not share the same syntactic structure. Constructions with unmarked modifying nominals show properties of compounding, while pseudopartitives arguably have a measure phrase (a cardinal with a measure noun) and a noun phrase (the quantified nominal), which is the head of the pseudopartitive construction. The syntactic properties we analyze include constituent properties, branching, pronominalization, fragment questions, linear ordering restrictions, particular properties of head ellipsis, and information structure effects.