Annotation of the results obtained in 2020
Providing structural support to minority languages is one of the important foci of the social policy of the Russian Federation. The number of speakers of all languages of Russia other than Russian is constantly diminishing to a varying degree. State authorities, non-governmental and non-commercial organizations, language activists, people working at various cultural and educational establishments are coordinating efforts to counteract the status quo. However, practical efforts on language revitalization should be backed by theoretical research, and here is why. It is often the case that middle and younger generations of minority language speakers acquire it only partially: limited vocabulary, most basic grammatical constructions. If a person wants to improve their language competency at a later stage (as a grown-up), learning new words turns out to be fairly easy (especially in agglutinative languages where forms are derived in a regular way). On the other hand, syntax poses a problem. If a construction specific to a given language has not been acquired early on, this can turn out to be a problem: not only language textbooks, but also existing theoretical grammars of the minority languages of Russia often lack the description of syntactic phenomena which make the given language unique. Our project aims at filling the gap. Over the past three years, we studied the syntax of non-finite verbal forms: deverbal adjectives (or participles), converbs, deverbal nouns (or nominalizations). Such forms usually combine the syntactic properties of a verb (for example, adverbs as modifiers as “quickly” in “a man quickly scanning the book”) and of some other part of speech (for example, of adjectives in the case of participles; just like adjectives, participles can modify nouns and in some languages, e.g. Russian, Andi or Avar, agree with the noun in number or gender). This is a general scheme; at the same time, languages of the world show quite an extreme diversity with respect to which exact properties of the verb and non-verbal parts of speech a non-finite form can obtain and how these properties can or cannot be combined with each other. To describe and explain these kinds of phenomena we used various linguistic theories, both functional and formal. Let us provide examples for each case in turn. The first example concerns the study of participle and converb forms in Andi (a Nakh-Daghestanian language, spoken primarily in Daghestan) which was conducted based on the collection of texts – the translation of the Gospel of Luke and a collection of fairy tales. All the uses of the forms in question in the texts were selected and analyzed. Andi employs five frequently used participial forms and four converbal forms. Based on the way these forms are derived morphologically, they can be divided into two groups: “dedicated” participles and converbs (similar to their counterparts in other languages) and forms which are derived from the finite form of the verb with the help of a special affix, an attributivizer or an adverbalizer. In other words, adding such an affix to a past tense verb turns the verb into a participle. The aim of the study was to understand the functions and distribution of these many forms in use. It was shown that they reveal subtle semantic differences with respect to tense and aspect. The second example is a study devoted to another Nakh-Daghestanian language – Avar – which focused on participles. Participial forms in Avar can function 1) as modifiers to nouns; 2) as components of a periphrastic progressive form consisting of a participle and an auxiliary verb (similarly to the English “is running”); 3) in certain contexts (e.g. in questions) they must replace finite verbal forms. In all these uses participial forms demonstrate distinct clusters of grammatical properties. For example, in the first and third types of use participles in Avar can bear a negation suffix, while in the second use they cannot. On the other hand, the first and third types contrast with the second one in the interpretation of tense. The ensuing question is: are these combinations of syntactic properties arbitrary or are they rooted in some more general structural properties of a sentence? We argue that these properties can be predicted by the formal analysis of Avar participial forms based on the idea of a complex syntactic head (see for similar approach Shimada 2007; Zwart 2009; Bruening 2019). Our analysis is elaborated further to predict unusual case marking in the clauses with the periphrastic progressive form. In such clauses a noun denoting the agent is marked not with Ergative as is usually the case for transitive verbs, but with Absolutive, the same case that is used for marking the object. Therefore, prima facie different phenomena (the absence of a negation suffix, the interpretation of tense, the choice of argument case) get an explanation based on the same syntactic properties of a clause. In 2020/2021, the participants of the project organized a workshop Multifunctionality and syncretism in non-finite forms as part of the 53rd Annual Meeting of the Societas Linguistica Europaea (SLE, 26 August – 1 September 2020). This workshop was aimed at fostering scientific discussion, at the international level, of the intermediate empirical and theoretical results accumulated within the project with fellow colleagues working on similar phenomena in the world’s languages and integrating the data of minority languages of Russia into the broader linguistic discussion. Based on the results of the workshop a special issue of the general linguistics journal Folia Linguistica on the topic of Multifunctionality and syncretism in non-finite forms is being prepared, it will be co-edited by the members of the project team.

 

Publications

---

Annotation of the results obtained in 2018
The project pursues four distinct but interrelated research directions: (i) general theoretical, (ii) Caucasian, (iii) Finno-Ugric, and (iv) Chukotko-Kamchatkan. Thematically speaking, each subproject focuses on the following non-finite constructions: relative clauses, nominalizations (deverbal nominals, masdars etc.), and adverbial clauses. In 2018/2019, the Caucasian subproject investigated participial relative clauses in Agul (its Khpyuk and Tpig dialects) and Udi, morphologically bound complementation in Abaza, nominalizations in Abaza and Avar, and adverbial clauses in Adyghe. In particular, a large-scale corpus study of over 1000 relative clauses in Udi and Agul yielded quantitative data regarding participle types, syntactically identifiable verb classes and relativization targets. For Agul, the results were subsequently compared for written and spoken texts with the help of dedicated corpora. The study revealed an overall identity of relativization patterns in both registers with respect to the frequency ratios of participial forms, verbs of distinct syntactic classes as well as relativization targets of different types. The two registers only differ in the general frequencies of relative clauses, which was significantly higher in the written corpus. Another significant difference concerns the frequency of Agent relativization, which amounted to 15% in the written corpus and 6% in the spoken one. Relativization of peripheral thematic roles (e.g. Addressee, Location etc.) were also shown to be frequent. For Adyghe, a set of field data on the semantics and pragmatics of an anteriority converb was analysed to reveal the regularities of interaction between aspectual and taxis semantics of the anteriority converb with the semantics of negation that is introduced by a dedicated marker on the converb. The study confirmed the hypothesis viewing negation as making a compositional contribution to the semantics of the converbial form by taking as its scope the interval until the bounding point introduced by the semantics of the converb. The main research question regarding event nominalizations in Avar was whether the nominalizing functional element n° can attach to various phrasal categories or whether its complement is invariably an extended projection of the light verb (vP). Having considered the patterns of agreement and case marking as well as the considerations of locality in the application of syntactic operations, the second hypothesis was chosen to undergird the formal analysis of argument-predicate agreement and adpositional agreement within the Avar verb phrase. The results were subsequently used to reinforce the case against formalizing anaphoric binding in terms of Chomsky’s (2000) operation Agree by postulating a featural deficiency inherent in reflexive pronouns. The Northwest Caucasian language Abaza provided a case study of morphologically bound complementation, resulting in a preliminary analysis in terms of paths of grammaticalization involving semantic bleaching. Another aspect of Abaza at the core of this project is the relation between the event nominalization form and the four distinct construction types in which it participates (no marking vs. possessive marking on the nominalization; finite clause-like marking of arguments; definiteness marking on the nominalization). The investigation revealed a correlation between the event nominalizations’ syntactic properties and their morphological properties: for instance, adverbial modifiers more readily attach to those nominalizations whose arguments carry the same marking as in finite clauses than nominalizations with zero or possessive marking. The Uralic subproject focused on participial and adverbial clauses in the Khanty, Hill Mari, and Moksha languages. Temporal adverbial clauses expressing anteriority and simultaneity in Khanty, for example, were shown to exist in two semantico-pragmatic kinds, which were analysed as involving “overt” and “covert” simultaneity. We discovered that overt simultaneity constructions are incompatible with contexts of indirect evidentiality of the main event or with indirect modal contexts, or with predicates of mental activity expressing unobserved activity. An analysis whereby a neutral version of the temporal construction involving the semantics of “transparency” for observation on the part of the speaker is distinguished from a marked construction involving an event that is inaccessible for observation can capture the observed facts. A novel discovery regarding the Hill Mari anteriority construction is its ability to express the semantics of simultaneity which can be traced back to the availability of an implicit event signalling the main event. The Moksha anteriority converb was revealed to display regular correspondences between the temporal semantics of the converb and the Aktionsart interpretation of the verbal stem in the embedded clause, the converb’s carrying a subject possessive agreement marker as well as the nature of the semantic relation between the main and embedded clauses. Regarding the participial relative clauses in Khanty, two participial forms — past and non-past — were identified. With respect to linear order, participial relative clauses precede the head noun but are obligatorily clause-final. Because establishing the hierarchical position of the subject inside a participial relative clause is key to understanding its syntactic structure, adverbs belonging to various classes — and thus attaching at different structural heights (Cinque 1999) — were used to probe the internal structure of the participial relatives. The Chukotko-Kamchatkan subproject involved collecting and analysing field data on externally and internally headed relative clauses in Chukchi. Two distinct relativization strategies were identified, which are based on transitivity. The internal syntax of A relativization differs substantially from the internal syntax of S and O relativization. Relativizing the A argument results in an internally-headed relative clause, as evidenced by the subject’s linear position and case marking, an analysis that is further confirmed by the semantic interpretation of quantified nominals in the position of the head noun. Finally, the general theoretical line of inquiry involved an investigation into the interaction of main-clause grammatical categories with the grammatical categories of embedded clauses, and a study of finiteness and predicationality in exclamative constructions in Russian. On the basis of newly collected data from Russian and Abaza, a number of regularities were identified. In particular, several purely grammatical oppositions show that sentential arguments are sensitive to the syntactic status of subjects, direct objects and oblique objects: for instance, subjects allow initial placement more easily than oblique objects do. As regards the exclamative clauses, newly obtained quantitative data from the Russian National Corpus warrant the conclusion that exclamative constructions can be divided into two groups with distinct properties: exclamative constructions without a verb/predicate are likely to contain noun phrases, whereas exclamative constructions containing a verb/predicate typically do not. This conclusion is at odds with the results of a similar study reported in Siemund 2015, which can tentatively be interpreted as a locus of parametric variation in the syntax and semantics of exclamatives.

 

Publications

---

Annotation of the results obtained in 2019
In 2019/2020 the participants of the project continued compiling and systematizing empirical data on the grammar of non-finite predication in the languages of Russia (Abaza, Avar, Udi, Amguema Chukchi, Hill Mari, Moksha, Komi-Zyrian, and Russian), but also turned to the theoretical analysis of the data, linking individual tracks into joint subprojects and converging research results. From the theoretical point of view, non-finite verbal forms are interesting as they preserve some properties of a verb but also acquire some nominal properties. One of the tell-tale phenomena in this field is argument case marking. The question is if the arguments get the same case marking with a non-finite verbal form as they do with a finite one; what it depends on and what is the mechanism of case assignment. Research on case assignment was one of the directions pursued in the investigations of 2019/2020. Based on the theory of eventive nominalizations developed for Avar, project participants analysed the syntactic structure of Russian eventive nominalizations and case assignment in them in view of two competing theories of case assignment. Generalizations made on the basis of Russian data were then tested on the data from Meadow Mari, a language with differential object marking, which presents a challenge for both of these approaches. Also a tentative minimalist model for the syntax of “low” and “high” nominalizations in Avar was proposed. Another issue currently at the heart of a lively debate in the theoretical literature concerns the control of null subjects in non-finite clauses. The project’s contribution to this debate involved an integrated experimental and corpus study of control patterns in Russian converbial clauses with the focus on a shift in grammatical norm. The experiments have revealed that non-nominative control with a possessive prepositional controller in psych and mental contexts are rated lower than grammatical sentences but higher than ungrammatical ones. The corpus study has revealed, in turn, that the frequency of non-nominative control patterns is significantly higher in contemporary written texts from the mid-20th century onwards as opposed to older written texts from between the early 18th to mid-20th centuries. We have also identified a novel factor influencing acceptability, namely the linear position of the converbial clause: non-nominative control in preposed converbial clauses is consistently rated as more acceptable than in their postposed counterparts. Project participants compiled an inventory of interdependencies between the properties of matrix and dependent predicates in complex clauses. Mood, polarity and modality of the matrix verb can play a role in the choice of the verbal form in a dependent clause. Besides, in certain cases in Russian a phenomenon of morphosyntactic doubling is observed between the matrix and dependent predicates which cannot be attributed to the classic terms of government and agreement. The study of nominalization patterns in Abaza has revealed an especially rich system of nominalization comprising four distinct types and catalogued those verb types which are capable of embedding a nominalization. Furthermore, a contrastive study of Abaza and Russian verbs embedding a nominalization has identified certain similarities as well as established clear points of divergence. Moreover, a formal analysis of morphologically bound complementation in Abaza within the framework of Lexical Functional Grammar was developed. An investigation into relativized participial clauses in Udi showed that data from spoken, written original and written translated corpora demonstrate similar tendencies. The observed differences cannot be linked to the nature of texts in a corpus (spoken vs. written, translated vs. original). It appears that they can only be attributed to the idiolectal preferences of the authors. Research results of the year 2019/2020 will be published in the leading international peer reviewed journals, indexed in citation databases Scopus and Web of Science Core Collection.

 

Publications

---