Inherent Case Features and Case Assignment in Double Object Constructions in Kazym Khanty

This dissertation contributes to the research on case by addressing questions about default case-marking and the mechanism of case assignment. This study is based on novel fieldwork data from the Kazymdialect of Northern Khanty (Ob-Ugric, Uralic) and makes three central claims. First, it makes an empirical and theoretical generalization that there exist two default cases in the language. I argue that the default case in argument positions is distinct from the default case in adjunct positions, e.g. on hanging topics (cf. Schütze 2001). Apart from the unmarked default, discussed in multiple previous studies (e.g. Schütze 2001; Legate 2008; Caha 2024 etc.), there exists a marked default case. Nominals in argument positions can bear the most marked case in the absence of syntactic case licensing, i.e. as default marking. To derive marked default, I argue that case features are inherently overspecified on arguments, which leads to an exceptional full spell-out in the absence of syntactic case checking. In particular, I suggest that arguments have an inherent [uCase]-feature, which is overspecified, i.e. valued for all case values available in the language. If a noun is not assigned case by a syntactic mechanism, all inherent case values are spelled out, which gives rise to oblique case morphology being used as the default. The second claim of this study addresses the mechanism of case assignment. As a baseline, I adopt a model where case is assigned under Agree with a valued Case feature on a functional projection. Consequently, case assignment happens under Agree between two valued features, which is considered impossible in all current theories of Agree (cf. both Preminger 2014 and Deal 2022). This problem is resolved by proposing a modified definition of Agree. I follow the theory that divides Agree into two parts: Agree-Link, which creates a pointer from a functional head to a nearby DP, and Agree-Copy, which transfers features from the goal to the probe (e.g., Arregi & Nevins 2012; Bhatt & Walkow 2013; Marušič, Nevins, & Badecker 2015; Baker & Camargo Souza 2020; Lyskawa 2021). However, the original definitions of Agree-Link and Agree-Copy must be modified to account for Agree between two valued features. Under the modified definition, two features can undergo Agree-Link if at least one of the features is uninterpretable and/or unvalued. This allows me to account both for φ-agreement between a valued and an unvalued feature and for case assignment, seen as agreement between two valued features. Furthermore, Agree-Copy is replaced by a more general Agree-Match that modifies the value of the uninterpretable feature so that it matches the value of the interpretable feature. Agree-Match either copies the value to an unvalued feature or deletes the redundant, “unmatching” values of the valued uninterpretable feature. The third claimaddresses the nature of DP-restriction on unlicensed arguments. This definition of Agree provides a motivation for the DP-restriction on secundative themes. An explanation is proposed as to why DPs must be case-licensed in Khanty, while structurally reduced nominals can survive unlicensed (cf. “derivational time-bombs” in Kalin 2018). I propose that D° in Khanty has an unvalued [uCase]-feature, which can enter into an Agree-Link relation with the valued [uCase]-feature on K°. Crucially, both features are uninterpretable, and Agree-Match cannot prefer impoverishment over copying the feature values. Such derivation crashes at PF. However, if a second Agree-Link with an interpretable Case feature takes place, Agree-Match can modify both uninterpretable features on K° and on D°, and the case is successfully assigned. The proposed approach is able to account for the inherent case overspecification on arguments, to model case assignment as Agree between two valued features, and to describe a mechanism behind the DP-restriction on unlicensed arguments.:Abstract Acknowledgments Table of contents List of tables and figures List of symbols and abbreviations 1 Introduction 1.1 General introduction 1.2 Design of the study 1.3 Overview of the dissertation 2 Background 2.1 Khanty language, dialectal classification and sociolinguistic situation 2.2 Morphology 2.2.1 Nominal morphology 2.2.2 Verbal morphology 2.3 Syntax 2.3.1 Passive 2.3.2 Object agreement 2.3.3 pro-drop 2.3.4 Word order 2.3.5 Position of wh-elements 2.4 Alignment alternation in Kazym Khanty 2.4.1 Information structure 2.4.2 Prominence hierarchies 2.4.3 Conclusion 3 Indirective alignment 3.1 Base-generation structure in double object clauses 3.1.1 Short scrambling in Kazym Khanty 3.1.2 Causatives 3.1.3 Condition C 3.1.4 Weak crossover and Variable binding 3.1.5 Floating quantifiers 3.1.6 Further tests 3.1.7 Summary 3.2 Indirective alignment 3.2.1 Dative as a structural case 3.2.2 Indirective alignment: derivation 3.2.3 High Applicative as the source of Dative 3.2.4 Summary 3.3 Causatives and HighApplP 3.3.1 Typology of causatives 3.3.2 Kazym Khanty causatives are VP-selecting 3.3.3 Indirective alignment in Khanty causatives 4 Secundative alignment 4.1 Secundative alignment derivation 4.2 Size of secundative theme 4.2.1 Structure of the Kazym Khanty nominals 4.2.2 Secundative theme is smaller than DP 4.2.3 Nominal licensing and DP-restriction 4.3 Alternative analyses of oblique themes 5 Marked default 5.1 Case feature hierarchy 5.2 Case assignment 5.2.1 Current theories of case assignment 5.2.2 KP and inherent case overspecification 5.2.3 Agree-Link and Case checking 5.2.4 Secundative alignment in Kazym Khanty: unifying oblique default and DP-restriction 5.3 Unmarked and marked default case 5.4 Surgut Khanty: Oblique marking of the theme 5.5 Complications for the Dependent Case theory 6 Typological parallels 6.1 Kalaallisut (Inuit) 6.1.1 Indirective alignment and High Applicative Phrase 6.1.2 Antipassive vs. secundative alignment: MOD/INSTR as default and the DP-restriction 6.1.3 Marked default case under noun incorporation 6.1.4 Intermediate summary 6.2 Eastern Mansi: DOM and case-licensing 6.2.1 Secundative alignment and DOM 6.2.2 Eastern Mansi 6.2.3 Testing existing analyses of DOM 6.2.4 Analysis of Eastern Mansi DOM 6.2.5 Intermediate summary 6.3 “Applicative construction” (Baker 1988) 6.4 Summary 7 Conclusion Appendix A Case system in dialects of Northern Khanty Appendix B Quantifier scope in double object constructions References

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.