CONTINGENCY, CONTIGUITY, AND CAPACITY: ON THE MEANING OF THE INSTRUMENTAL CASE MARKING IN COPULAR PREDICATIVE CONSTRUCTIONS IN RUSSIAN by VALERIIA TRETIAK A DISSERTATION Presented to the Department of Linguistics and the Graduate School of the University of Oregon in partial fulfillment of the requirements for the degree of Doctor of Philosophy September 2020 DISSERTATION APPROVAL PAGE Student: Valeriia Tretiak Title: Contingency, Contiguity, and Capacity: On the Meaning of the Instrumental Case Marking in Copular Predicative Constructions in Russian. This dissertation has been accepted and approved in partial fulfillment of the requirements for the Doctor of Philosophy degree in the Department of Linguistics by: Cynthia Vakareliyska Chairperson Volya Kapatsinski Core Member Eric Pederson Core Member Katya Hokanson Institutional Representative and Kate Mondloch Interim Vice Provost and Dean of the Graduate School Original approval signatures are on file with the University of Oregon Graduate School. Degree awarded September 2020. ii © 2020 Valeriia Tretiak This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs (United States) License. iii DISSERTATION ABSTRACT Valeriia Tretiak Doctor of Philosophy Department of Linguistics September 2020 Title: Contingency, Contiguity, and Capacity: On the Meaning of the Instrumental Case Marking in Copular Predicative Constructions in Russian. This study investigates the use of the Instrumental case marking in copular predicative constructions in Russian. The study endeavors to explain why the case marking whose prototypical meaning cross-linguistically is that of an instrument, occurs with predicative nominals (nouns and adjectives), what meaning it has in predicative constructions, and how this meaning resonates with the rest of the Instrumental meanings in the language. While cross-linguistically the Instrumental case marking is notoriously known for a wide array of meanings and functions, only in Slavic and Baltic languages it is used to mark predicative nominals. On a broader scale, I use the Russian Instrumental case marking as a case study to examine the internal organization of a complex grammatical category. The study uses the prototype model based on Wittgenstein’s (1953) family resemblance to establish semantic relatedness among the various meanings of the Instrumental case marking. The study also proposes a general meaning of the Instrumental case marking, which I define in cognitive terms as relations of contingency and contiguity. iv Using evidence from Early East Slavic manuscripts, the study demonstrates that the Instrumental case marking in predicative constructions has as its semantic source the Instrumental case marking in similative constructions. I propose that besides denoting the manner of motion, the referent of the Instrumental noun phrase in similative constructions also denotes a new capacity of the subject referent which emerges when the subject referent metaphorically adopts the most salient features associated with the referent of the Instrumental noun phrase, that is, its particular manner of motion. This emerging capacity is contiguous with and contingent on the specific mode of acting. In predicative constructions, the referent of the Instrumental noun phrase is a capacity, as opposed to an inherent or essential property, of the subject referent and is realized through acting/ performance. That acting/ performance is crucial in delimitating Nominative vs. Instrumental-marked properties in predicative constructions is supported by the semantic unacceptability of the Instrumental case marking in instances where the implied acting is negated in the conjunct clause. Capacity is a role which has its designated function and purpose. Function links the meaning “capacity” with the meaning “instrument”. Inasmuch as function is what delimitates instruments from other physical objects, function is what tells apart, respectively, capacities from properties in Instrumental vs. Nominative predicative constructions. That all the individual meanings of the Russian Instrumental case marking, including its meaning in predicative constructions, are interrelated and form a coherent grammatical category is further corroborated by the analysis of Instrumental constructions with predicative nouns and adjectives. v CURRICULUM VITAE NAME OF AUTHOR: Valeriia Tretiak GRADUATE AND UNDERGRADUATE SCHOOL ATTENDED: University of Oregon, Eugene Karelian State Pedagogical University, Petrozavodsk, Russia, 1999-2004 DEGREES AWARDED: Doctor of Philosophy, Linguistics, 2020, University of Oregon, Eugene Master of Arts, Slavic Linguistics, Russian, East European, and Eurasian Studies, 2013, University of Oregon, Eugene Bachelor of Arts, English and German Philology and Teaching Foreign Languages, 2004, Karelian State Pedagogical University, Petrozavodsk, Russia PROFESSIONAL EXPERIENCE: Instructor of record (Graduate teaching fellow), Department of Linguistics, University of Oregon, 2013–2019 Instructor of record (Graduate teaching fellow), Russian, Russian, East European, and Eurasian Studies, University of Oregon, 2011-2013 Senior instructor for first-year Russian, Center for Language Studies Immersion-Style Program, Beloit College, Wisconsin, June 12, 2014 – August 9, 2014 Teaching Assistant for second-year Russian, Beloit College, Wisconsin, 2009-2010 (under Fulbright FLTA scholarship) vi FELLOWSHIPS, AWARDS, AND HONORS: John L. and Naomi Luvaas Graduate Fellowship, University of Oregon, 2015-2016 Graduate Teaching Fellowships, Department of Linguistics, University of Oregon, 2013 – December 2018 Graduate Teaching Fellowships, Russian, East European and Eurasian Studies Program, University of Oregon, 2011-2013 Fulbright Scholarship, Foreign Language Teaching Assistant, Beloit College, 2009-2010 Diploma with Honors, Karelian State Pedagogical University, 2004 WORKS IN PROGRESS: Valeriia Tretiak and Cynthia M. Vakareliyska, “Productivity and semantic distribution of English loanblend [N[N]] constructions in the East Slavic languages vs. Bulgarian” (lead author; expected submission to Contrastive Linguistics, Sofia in late July or early August 2020) Valeriia Tretiak, “Grammatical metonymy in case marking: A case study of the Accusative-Instrumental alternation in Russian” (expected submission to the Journal of Slavic Linguistics in July 2020) RESEARCH INTERESTS: Morphosyntactic representation of conceptual structures Historical and comparative evolution of morphological case paradigms Cognitive approaches to case and case acquisition by L1 speakers vii ACKNOWLEDGEMENTS This dissertation would have been impossible without many people and ideas that I have encountered along the course of my graduate journey. I wish to acknowledge the faculty of the Department of Linguistics at the University of Oregon for their pivotal role in my growth as a scholar. The impact of the generous financial support that I received as a graduate teaching fellow cannot be overestimated in my ability to reach this major academic milestone. I am deeply indebted to my advisor, Professor Cynthia Vakareliyska, for her support, guidance, and intellectual exchange. I wish to express my appreciation for her continuous encouragement, flexibility, and availability through each step of the research and writing process. I am extremely grateful to my committee members: Professor Volya Kapatsinski for providing insightful and sophisticated comments; Professor Eric Pederson for posing challenging, thought-provoking, and heart-wrenching questions, and Professor Katya Hokanson for making me think about the practical implications of this study to teaching Russian. I would like to thank Professor Scott DeLancey for the exciting lectures on Syntax and Semantics. Klamath examples were my favorite; sepk’eec’a! I also had great pleasure working with Professor Doris Payne, who introduced a whole new world of Field Linguistics to me; Ashe oleng! I must also thank Professor Lisa Redford for the Praa(c)tical knowledge of how to interpret frequency, formants, and amplitude into sounds, words, and even sentences. It never Hz to point out how much mystery and viii intrigue “the Fourier analysis window” creates in a conversation! Special thanks should go to Professor Spike Gildea for being understanding and considerate about my parenting needs. I will recall all of you with fundamental frequency! Thanks also goes to my fellow graduate colleagues Becky Paterson, Amos Teo, Manuel Otero, Sara Pacchiarotti, Jeff Kallay, Richard Griscom, Charlie Farrington, Marie-Caroline Pons, Zahid, Tiago, Misaki Kato, and many more. I wish to thank my family in Russia and here for their invaluable and unwavering support all along the way. My deepest gratitude goes to my husband Frank and son Bruno for letting me get off the grids of family life for several months to complete the manuscript. Without you all, I could not have done it. ix To the spatio-temporal anchors in my life – my family x TABLE OF CONTENTS Chapter Page I. INTRODUCTION .................................................................................................... 1 1.1. The Problem ................................................................................................... 1 1.1.2. The Proposal .......................................................................................... 4 1.2. Copular Predicative Constructions ................................................................. 4 1.2.1 The Variation in Case Marking of Predicative Nominals in Russian .... 5 1.3. Models for Studying the Russian Instrumental Case Marking ....................... 7 1.3.1. An Alternative to the Existing Models ................................................. 8 1.3.1.1. Conceptualization and Generation of Meaning ........................... 10 2. Organization ..................................................................................................... 11 II. CASE MARKING OF PREDICATIVE NOMINALS IN A CROSS-LINGUISTIC PERSPECTIVE ..................................................................... 14 2.1. Constructions with Predicative Nominals ..................................................... 14 2.2. Cross-linguistic Variation in Case Marking .................................................. 16 2.2.1 Two Major Case Marking Strategies ..................................................... 16 2.2.1.1. Identical Case Marking ................................................................. 17 2.2.1.2. Differential Case Marking ........................................................... 18 2.2.2. Identical and Differential Case Marking: Variation within a Language .......................................................................................... 20 2.3. Case Marking of Predicative Nominals in the Balto-Finnic Languages ....... 20 xi Chapter Page 2.4. Case Marking of Predicative Nominals in the Slavic Languages and Lithuanian ............................................................................................................. 22 2.4.1. Polish ..................................................................................................... 23 2.4.2. Lithuanian ............................................................................................. 24 2.5. Cross-linguistic Correspondences of the Russian Instrumental Case Marking in Predicative Constructions ................................................................................. 26 2.5.1. Telugu ................................................................................................... 26 2.5.2. French ................................................................................................... 28 2.6. Conclusion ..................................................................................................... 33 III. WHAT IS CASE? CASE AND CASE MARKING ............................................. 36 3.1. Traditional Approaches to Case and Case Marking ....................................... 37 3.1.1. Hjelmslev .............................................................................................. 37 3.1.2. Jakobson ............................................................................................... 38 3.1.3. Limitations of Hjelmslev’s and Jakobson’s Models ............................ 39 3.2. Formal Models ............................................................................................... 39 3.3. Semantic Roles .............................................................................................. 41 3.3.1. Principles Underlying Semantic Role Models ..................................... 41 3.3.1.1. Issues with Semantic Roles .......................................................... 41 3.4. The Instrumental Case Marking in Russian: Existing Models ...................... 43 3.4.1. Semantic Roles and Semantic Lists ...................................................... 43 xii Chapter Page 3.4.2. The Formalism of Švedova’s Academy Grammar ............................... 43 3.4.3. Wierzbicka: Metalanguage and Semantic Invariance .......................... 46 3.5. Case as a Cognitive Phenomenon. Case and Morphological Case Marking ........................................................................................................ 47 3.5.1. Cognitive Definition of Case ................................................................ 47 3.6. The Instrumental Case Marking vs. the Semantic Role of Instrument .......... 49 3.7. The Instrumental Case Marking in a Cross-linguistic Perspective ............... 52 3.7.1. The Prototypical Meaning of the Instrumental Case Marking Cross-linguistically ..................................................................................... 53 3.8. Conclusion ..................................................................................................... 55 IV. ALTERNATIVE MODEL FOR STUDYING THE MEANING OF THE INSTRUMENTAL CASE MARKING IN RUSSIAN ............................................... 57 4.1. Organization of Meaning ............................................................................... 57 4.1.1. Polysemy and Prototypes ..................................................................... 57 4.1.2. Cognitive and Semantic Relations ........................................................ 59 4.2. Application of the Proposed Model ............................................................... 61 4.2.1. Referents of Instr Nouns as Accessories to the Action Described by the Verb ..................................................................................... 62 4.2.1.2. Referents of Instr Nouns after Verbs of Controlled Motion ........ 65 4.2.2. Referents of Instr Nouns with Verbs of Government and Possession ................................................................................................. 67 xiii Chapter Page 4.2.2.1. Some Diachronic Issues .............................................................. 68 4.2.2.2. Conceptual Similarities in CSR ................................................... 68 4.2.3. Referents of Instr Nouns as Accidental and Involuntary Performers ................................................................................................. 69 4.2.4. Referents of Instr Nouns as Causes or Stimuli ..................................... 70 4.2.5. Referents of Instr Nouns as Spatio-temporal Continuity and Entirety ................................................................................................. 72 4.2.5.1. Continuous Pathways .................................................................. 74 4.2.5.1.1. Literal Pathways ................................................................. 74 4.2.5.1.2. Metaphorical Pathways ...................................................... 76 4.2.5.2. Referents of Instr Nouns as a Course of Time ............................ 79 4.2.5.2.1. Diachronic Evidence .......................................................... 82 4.2.6. Referents of Instr Nouns as Literal or Metaphorical Capacities .......... 83 4.2.6.1. Referents of Instr Nouns in Similative Constructions ................. 85 4.2.6.2. Referents of Instr Nouns with Verbs of Naming ......................... 87 4.2.6.3. Referents of Instr Nouns in Predicative Constructions ............... 89 4.2.6.3.1. Diachronic Evidence .......................................................... 91 4.2.6.3.2. Motivation for Change ....................................................... 92 4.2.6.3.3. Co-temporality vs. Temporariness ..................................... 97 4.2.6.3.3.1. Evidence from Polish and French ........................... 98 4.3. Conclusion ..................................................................................................... 102 xiv Chapter Page V. THE INSTRUMENTAL CASE MARKING WITH PREDICATIVE NOUNS ........................................................................................... 105 5.1. Variation in Case Marking in Contemporary Standard Russian ................... 105 5.2. Copula in the Present Tense .......................................................................... 107 5.2.1. No Copula, no Instrumental Case Marking? ........................................ 109 5.2.2. Javljat´sja: The Alternative to Zero Copula? ....................................... 112 5.3. A Brief Diachronic Sketch ............................................................................ 115 5.3.1. The First Attested Instances of the Instrumental Case Marking with Predicative Nouns ..................................................................................... 115 5.3.2. The Instrumental Case Marking with Predicative Nouns from the 15th Century Onward ..................................................................................... 116 5.3.2.1. The 15th – 17th Centuries ............................................................. 116 5.3.2.2. The 18th Century .......................................................................... 117 5.3.2.3. The 19th Century .......................................................................... 118 5.3.2.4. Intermediate Conclusion .............................................................. 119 5.4. Previous Research: The Instrumental Case Marking with Predicative Nouns in Contemporary Standard Russian ...................................................................... 120 5.4.1. Nichols (1981) ..................................................................................... 121 5.4.2. Timberlake (2004) ................................................................................ 122 5.4.3. Roy (2006, 2013) .................................................................................. 125 xv Chapter Page 5.4.4. Krasovitsky et al. (2008): A Quantitative Analysis .............................. 126 5.5. An Alternative Analysis ................................................................................. 128 5.5.1. Deverbal Nouns and Nouns Denoting Occupations ............................. 128 5.5.2. Nouns Denoting Nationalities and Personal Traits .............................. 131 5.5.2.1. Nationalities .................................................................................. 131 5.5.2.2. Personal Traits ............................................................................. 132 5.6. Conclusion ...................................................................................................... 136 VI. THE INSTRUMENTAL CASE MARKING WITH PREDICATIVE ADJECTIVES ................................................................................. 139 6.1. Short and Long Adjectives in Contemporary Standard Russian: A Grammatical Sketch ......................................................................................... 139 6.2. The Emergence and Development of Long Adjectives ................................. 141 6.2.1. Grammatical Functions of Short and Long Adjectives ........................ 143 6.2.1.1. Disappearance of Short Adjectives from Attributive Positions ................................................................................................. 143 6.3. Long Adjectives in Predicate Positions ......................................................... 144 6.4. Long and Short Predicative Adjectives in Contemporary Standard Russian ................................................................................................................. 146 6.5. Instr and Nom Predicative Adjectives in Contemporary Standard Russian: An Overview of Literature ................................................................................... 149 6.6. An Alternative Analysis ................................................................................ 153 xvi Chapter Page 6.6.1. Testing the “Subjective-Objective Characteristic” Hypothesis ............ 153 6.6.2. Testing the “Observability” Hypothesis ............................................... 156 6.6.3. Properties as Capacities ........................................................................ 160 6.6.3.1. Stimuli ......................................................................................... 160 6.6.3.2. Instruments and Conduits ............................................................ 163 6.6.3.3. Functions ..................................................................................... 168 7. Conclusion .............................................................................................................. 169 VII. SUMMARY AND CONCLUSIONS ................................................................. 171 APPENDICES ............................................................................................................. 179 A. LIST OF ABBREVIATIONS .......................................................................... 179 B. TRANSLITERATION TABLE ........................................................................ 180 C. OLD CHURCH SLAVONIC AND EARLY EAST SLAVIC SOURCES ...... 181 D. CONTEMPORARY STANDARD RUSSIAN SOURCES (from the National Corpora of the Russian Language) ........................................ 183 REFERENCES CITED ............................................................................................... 185 xvii LIST OF FIGURES Figure Page 1. Semantic functions of the Instr across languages ................................................. 54 2. Long and short adjectives in CSR ......................................................................... 140 xviii CHAPTER I INTRODUCTION 1.1 The problem This dissertation examines the Instrumental case marking (Instr) in copular predicative constructions in Russian and attempts to explain why specifically the Instr, as opposed to the other Russian case markings, is the other case marking strategy alongside the Nominative case marking (Nom), what meaning it evokes in these constructions, and how this meaning resonates with its other meanings. While the Russian Instr has never lacked scholarly attention, the issues concerning its meaning and use with predicative nouns and adjectives that this study investigates have not been fully addressed in the research to date. The Russian Instr, as well as the Instr cross-linguistically, is known for numerous meanings.1 I follow the general assumption of cognitive linguistics that the way we construe the world around us is reflected in the linguistic structures available in the language. However, the opposite is also true: the existing linguistic structures impose specific conceptualizations. As evidenced by Old Church Slavonic (OCS) and Early East Slavic (EES) manuscripts, by the time the Instr is first attested in Slavic with predicative nouns in the second half of the 13th century, it had already been used to express a wide range of meanings, almost all of 1 Note that I reserve the term ‘Instr(umental)’ to mean “related to the Instrumental case marking”. 1 which are attested in modern Slavic languages.2 While it is feasible that a single grammatical means occurs in two or more different constructions by chance, the cross- linguistic parallels that the Russian Instr has in predicative constructions point to its likely relatedness with other Instr constructions in the language. This potential semantic and cognitive connection seems even more probable in light of the findings made in the field of cognitive linguistics, more specifically, in the study of polysemous effects in morphology and syntax (e.g., Taylor 1989, Sweetser 1990 on modality, Janda 1990, Nikiforidou 1991 on the Genitive case marking (Gen) in Indo-European languages; Verhagen 1992 on passives, Dąbrowska 1997 on the Polish Dative case marking (Dat), etc.). The investigation of the Instr in predicative constructions thus starts on the premises that case marking is a polysemous category and that the meanings that it conveys are interrelated in a motivated fashion. The implications of this assumption are such that the Instr in predicative constructions is expected to reflect in one way or another the semantic and cognitive regularities that underlie its other meanings in individual Instr constructions. It is generally believed that the variation in the marking of predicative nominals, which in Russian manifests itself as the Nom-Instr variation, reflects the differences in perception of properties they express as inherent or temporary (Stassen 2001, Comrie 1997). When the predicative nominal is marked identically to the grammatical subject, 2 OCS is the first documented Slavic language and the first literary language of the Slavs. It is a South Slavic language, based predominantly on Macedonian dialects (South Slavs). OCS was in use in the 9th-11th centuries. EES is a term that refers to the dialect of East Slavs spoken during the 10th -15th century. The name reflects the fact that this dialect was shared by the speakers of modern Russians, Ukrainian, and Belarusian. 2 which corresponds to the Nom–Nom pattern in Russian, the property at issue is discerned as inherent or permanent. When the predicative nominal is marked differently from the grammatical subject, which corresponds to the Russian Nom–Instr variation, the property it denotes is perceived as temporary. However, because the grammatical means that capture the semantic distinction in question differ cross-linguistically, it is plausible to assume that the inherent/permanent–non-inherent/temporary distinction is only a rough approximation. For example, in Slavic and Baltic languages, the proposed semantic opposition is captured by the Nom-Instr variation. Historically, the Instr with predicative nominals is a late innovation for Balto-Slavic languages, as originally the Nom is attested in such constructions. The occurrence of the Instr with predicative nouns and adjectives poses challenges for at least two reasons. First, it is attested specifically in Slavic languages and Lithuanian. Second, while both cross-linguistically and in Slavic the Instr expresses a wide scope of meanings, none of them conveys ‘temporariness.’ Issues like the above warrant a separate investigation in individual languages to determine why a specific grammatical means occurs with predicative nouns and adjectives and whether the meaning it elicits in these constructions conforms to the typological findings. On a broader scale, using the Russian Instr as a case study, this dissertation examines the internal organization of meanings in a complex grammatical category. Additionally, because the Instr in predicative constructions has cross-linguistic parallels (functional similarity), the generalizations made for its semantic function in Russian could possibly deepen our understanding of the semantic and cognitive organization of predicative constructions in general. 3 1.1.2 The proposal I demonstrate in this study that the Instr in predicative constructions is related to other meanings of the Instr in Russian. More importantly, the predicative Instr is linked to the meaning of “instrument”, which is the most prototypical meaning of the Instr attested cross-linguistically, through the notions “function” and “purpose”.3 Using evidence from OCS and EES manuscripts, I establish that the Instr in predicative constructions has as its semantic source the Instr in similative constructions. I propose that like the similative Instr, the Instr with predicative nouns and adjectives implies that the property they express unfolds co-temporally with the subject referent’s mode of acting, which makes it a capacity. A capacity is a role which has its own purpose and function, and which is instantiated as such through performance. The referent of the Nom predicative nominal, on the other hand, denotes a property, which is an intrinsic quality or an essential characteristic of the subject referent and which, due to its nature, does not presuppose performance and hence is atemporal. The capacity is not bounded in time but is rather contiguous with and contingent on the mode of acting and therefore is co-temporal rather than temporary. 1.2 Copular predicative constructions Constructions with predicative nouns and adjectives are a type of copular predicative expressions. These constructions are called “copular” because in many languages the grammatical subject and the predicative nominal are separated by a 3 I will use “the predicative Instr” interchangeably with “the Instr in predicative constructions” 4 copulative verb. In some languages, however, including Russian, a copula is not always required, and the construction simply consists of two nominal elements. English equivalents of predicative constructions are given below:4 (1) a. Ms. Smith is a baker. b. Ms. Smith is nosy. 1.2.1 The variation in case marking of predicative nominals in Russian In Russian no copula is used in predicative constructions in the present tense.5 In the absence of the copula, Russian predicative nouns and adjectives are marked by the Nom. However, in environments where the copula byt´ ‘be’ is required, the predicative nominal is marked by either the Nom or the Instr: (2) a. Marija byl-a balerin-a. Mary-NOM was-FEM ballet.dancer-NOM.FEM.SG ‘Mary was a ballet dancer.’ 4 Another type of a predicative construction is the English sentence “Ms. Smith is from Mississippi.” Because such constructions in Russian do not exhibit any variation in case marking, they lie beyond the scope of this dissertation. 5 A zero copula in Russian is the marker for the present tense indicative mood of “be.” 5 b. Marija byl-a balerin-oj. Mary-NOM was-FEM ballet.dancer-INSTR.FEM.SG ‘Mary was a ballet dancer.’ (3) a. Marija byl-a grustn-aja. Mary-NOM was-FEM sad-NOM.FEM.SG ‘Mary was sad.’ b. Marija byl-a grustn-oj. Mary-NOM was-FEM sad-INSTR.FEM.SG ‘Mary was sad.’ As can be seen from the examples above, predicative nominals in Russian show the variation in case marking: identical case marking of both the grammatical subject and the predicative nominal (Nom-Nom) and non-identical case marking of the grammatical subject and the predicative nominal (Nom-Instr). The grammatical subject is always marked by the Nom, while the predicative nominal is marked by either the Nom or the Instr. The variants in (a) and (b) examples above are translated into English in the same way, but for native Russian speakers they are not identical in meaning (see, for example, Timberlake 1986, Philip 2001, Henterhölzl 2001, Richardson 2001). The Instr in copular predicative constructions has been puzzling researchers for many decades and has yielded volumes of scholarly works (Potebnja 1888, Jakobson 1936, 1958, Bernštejn 1958, 6 Mrázek 1964, Švedova 1980, Wierzbicka 1980, Nichols 1981, Timberlake 2004, among others). Traditionally, however, these studies examine the Instr in terms of its semantic opposition with the Nom and point out the semantic and grammatical contexts in which either the Nom or the Instr occurs. This dissertation endeavors instead to look at the Instr in predicative constructions from a different angle: that is, to investigate how the meaning of the Instr in predicative constructions relates to the other Instr meanings in individual Instr constructions. I take the position that speakers select the Instr or the Nom not because there is a choice between the two but because each of the two case markings is associated in their mind with a specific meaning. Because the differences in meaning between predicative nominals are captured in Russian by the variation in case marking, a part of this dissertation addresses some theoretical questions concerning the distinction between case and case marking. 1.3 Models for studying the Russian Instrumental case marking The Russian Instr is believed to express around twenty different meanings (Potebnja 1888, Bernštejn 1958, Mrázek 1964, Wierzbicka 1980, inter alia). The question that arises in this respect is how these twenty or so meanings are related to one another. In Jakobson’s famous 1936 model, the meanings of the Instr are organized in such a way that there exists one general or aggregate meaning (Gesamtbedeutung) and other, specific meanings, which Jakobson treats as contextual variants of the former. However, Jakobson’s model is concerned not necessarily with how the specific meanings of the Instr are connected to one another, but with how the aggregate meaning of the Instr makes it distinct from the other case markings in the Russian language. Despite the 7 insightfulness and economy of Jakobson’s approach, its definition of case as a bundle of abstract features makes the occurrence of the Instr unpredictable in a communicative situation. 1.3.1 An alternative to the existing models A different model can be adopted, however, for exploring more comprehensively the relations among the various meanings of a given case marking. According to this model, when a given case marking has multiple meanings, they are assumed to be organized as a network, or family, in which meaning A is similar in some respects to meaning B and meaning B shares some commonalities with meaning C, but in which meaning A may not necessarily share any traits with meaning C. This alternative model, which the dissertation proposes, has one significant advantage over Jakobson’s (1936) and similar models, in that it treats all the meanings of a given case marking as a priori interrelated in various fashions, rather than as situational. As such, this alternative model does not warrant on-the-fly explanations for instances where the specific meanings of the case marking in question do not fully line up with one another. Moreover, unlike Jakobson’s (1936) or Wierzbicka’s (1980) models, discussed in chapter III, which presuppose the existence of one basic, core meaning of the Instr, this dissertation proposes instead that some meanings are more prototypical than others, while at the same time both more prototypical and less prototypical meanings are related to each other on semantic and cognitive grounds. It is agreed in the literature (see, for example, Bernštejn 1958, Mrázek 1964, Wierzbicka’s 1980, Janda 1993, Pečenyj 2012) that the core meaning of the Russian Instr 8 manifests in situations in which a referent of the Instr noun is envisioned as an instrument, in the sense of the semantic role, as illustrated by example (4) below. (4) Ivan-Ø rubil drova topor-om. John-NOM chopped wood ax-INSTR.MASC.SG ‘John chopped the wood with an ax.’ Compare in this respect the following example in (5): (5) Ivan-Ø torguet apel´sin-ami. John-NOM trades oranges-INSTR.PL ‘John sells oranges.’ Should it be assumed that apel´sin-ami in (5) is a specific contextual realization of the core “instrument” meaning? How does the context in (5) help us to understand that the forms topor-om ‘(with) ax-INSTR’ and apel´sin-ami ‘oranges-INSTR’ are related to each other, rather than just happening to have phonologically homonymous suffixes? Note that if the verb in (5) is replaced with the synonymous verb prodavat´ ‘sell,’ the Instr is ungrammatical with the noun whose referent denotes an entity being sold: (6) Ivan-Ø prodaet apel´sin-y/* apel´sin-ami. John-NOM sells oranges-ACC.PL/*oranges-INSTR.PL ‘John sells oranges.’ 9 Contextually, examples in (5) and (6) are very similar: both describe John as a seller and oranges as the product that he sells. However, if the context in both examples is identical, why are there two different case markings? More importantly, how does the same context make the oranges in (5) more of an instrument than the oranges in (6)? A simple contrastive analysis like this reveals the limitations of Jakobson’s model and warrants the alternative approach that I advocate for in this dissertation. This model departs from the assumption that specific meanings of a given case marking are derived contextually, and instead is based on the proposition that grammatical representation, including case marking, is motivated semantically and cognitively. Although grammatical meaning itself certainly is abstract and cannot be categorized in terms of specific conceptual content, the choice of a specific grammatical structure in a given situation is motivated by the meaning associated with it in the mind of the speaker. 1.3.1.1 Conceptualization and generation of meaning Human language is fundamentally symbolic, and language capacity is not an autonomous module, but rather an integral and integrated part of human cognition. As such, language use reflects our way of categorizing the outside world. This can explain why in the similar truth conditions in (5) and (6), the noun denoting the product that is being sold occurs with different case markings in Russian. While prodavat´ ‘sell’ and torgovat´ ‘trade, sell’ are synonyms, the scenes they evoke are not the same. Thus, prodavat´ refers to the final stage in a business cycle, that is, the actual exchange of the product for money, whereas torgovat´ refers to the business cycle in its entirety and entails finding, ordering, and buying a product for the purpose of its subsequent 10 realization for profit. Note also that because torgovat´ represents a business cycle as such, it does not necessarily imply that the product is in fact sold. While selling for profit is the end goal of any commercial activity, whether the act of selling happens or not in the situation described by the verb depends on the scene that this verb evokes. Understanding the significance of conceptualization in the generation of meaning allows the researcher to better appreciate how linguistic representations which are similar or even identical, in terms of objective reality they portray, can yield fine-grained and yet profound distinctions in meaning. 2. Organization The dissertation is structured as follows. Chapter II presents an overview of the differences in the case marking of predicative nominals in a cross-linguistic perspective. Section 2.1 lays out some preliminaries concerning the structure and meaning of copular predicative constructions. Section 2.2 discusses two major case marking strategies attested in predicative constructions, which are identical and non-identical case marking of the predicative nominal and the grammatical subject. While identical case marking is the most common strategy attested cross-linguistically (Comrie 1997, Stassen 2001), some language groups, like Slavic, Baltic, and Finnic, show variation in the case marking of predicative nominals. In section 2.3 I talk about case marking of predicative nominals in Balto-Finnic languages, while section 2.4 describes the same phenomenon in Slavic languages and Lithuanian. Section 2.5 examines cross-linguistic correspondences of the Russian Instr in predicative constructions. Section 2.6 presents conclusions for the chapter. 11 Chapter III analyzes the distinction between grammatical case itself and morphological case marking. Section 3.1 discusses traditional approaches to case and provides an overview of Hjelmslev’s (1935) and Jakobson’s (1936, 1958) models and their limitations. Section 3.2 summarizes the findings of formal models of case such as Chomskian’s Government and Binding (1981) and the Minimalist Program (1993) and explains why these models cannot be applied successfully in the present study. In section 3.3 I analyze semantic role models. Section 3.4 discusses models for studying the Russian Instr. In section 3.5 case is defined as a cognitive phenomenon. Section 3.6 emphasizes the distinction between the Instr and the semantic role of Instrument. Section 3.7 talks about the meanings of the Instr in a cross-linguistic perspective. Section 3.7 presents conclusions. Chapter IV proposes an alternative model for studying the meaning of the Instr. Section 4.1 talks about organization of meaning: I discuss polysemy and prototypes as well as cognitive and semantic organization of the meanings of the Russian Instr. In section 4.2 I apply the proposed model and demonstrate that the meanings of the Russian Instr are interrelated and form a semantic network with radial structure. I adduce diachronic evidence to establish in this section that the Instr in predicative constructions is related semantically to the similative Instr. I talk about similar correspondences cross- linguistically. Section 4.3 provides conclusions for the chapter. Chapter V explores the meaning of the Instr with predicative nouns. Section 5.1 discusses the variation in case marking in Contemporary Standard Russian (CSR). Section 5.2 addresses the restrictions on the occurrence of the Instr with the copula byt´ ‘be’ as well as explains the occurrence of the Instr with the semi-copula javljat´ sja ‘be 12 someone/something, serve as.’ In section 5.3 I provide a brief diachronic sketch concerning the emergence of the Instr in Russian predicative constructions. Section 5.4 reviews the literature concerning the meaning and use of the Instr with predicative nouns in (CSR). In section 5.5 I examine the Instr with deverbal nouns, nouns denoting nationalities and personal traits. Section 5.6 provides conclusions. Chapter VI scrutinizes the Instr with predicative adjectives. Section 6.1 lays out a grammatical sketch of uses of short and long adjectives in CSR. Section 6.2 provides historical background, including the development and use of short and long adjectives. This section also discusses the role that long-form adjectives played in the emergence of the Instr with predicative adjectives. Section 6.3 focuses on the semantic and grammatical distinctions between long and short predicative adjectives in CSR. Section 6.4 reviews research to date concerning the distribution of the Instr with predicative adjectives. Section 6.5 tests some hypotheses proposed in the literature for the meaning of the Instr with predicative adjectives in CSR and presents an alternative analysis. Section 7 provides the conclusion. Chapter VII provides a summary of the dissertation and presents conclusions from the discussions in the previous chapters. The appendices towards the end of the dissertation contain a list of the abbreviations used in the dissertation, a transliteration table, a list of OCS and EES manuscripts, and a list of contemporary sources cited in the study. 13 CHAPTER II CASE MARKING OF PREDICATIVE NOMINALS IN A CROSS-LINGUISTIC PERSPECTIVE This chapter opens with a brief description of copular predicative constructions. I then discuss the most common case marking strategies attested cross-linguistically with predicative nominals. These are the identical and non-identical case marking of the predicative nominal and the grammatical subject. The differences in the case marking of predicative nominals are believed to reflect the differences in the perception of the properties that they express as inherent/permanent or non-inherent/temporary. 2.1 Constructions with predicative nominals Predicative constructions are a type of copular expressions.6 What makes predicative sentences differ from the rest of copular constructions is that, like non- 6 Usually, four types of copular sentences are defined: predicative, equative, specificational, and identificational (see, for example, Higgins 1979, Mikkelsen 2011, among others): (i) Mary is an actress. Predicative (ii) Mark Twain is Samuel Clemens. Equative (iii) The best actress is Mary. Specificational (iv) This actress is Mary. Identificational Equative expressions are defined in Higgins (1979) as denoting identity between the two noun phrases located on both sides of the copula, as in (2). However, equative constructions involving two proper nouns (“true equatives” in Heycock and Kroch (1999: 373)) are not very productive, since individuals in general rarely have two different names. Another type of an equative sentence is illustrated by the constructions in (v) and (vi) below: 14 copular sentences, they assign a property to the subject referent. Predicative expressions differ from non-copular sentences, in that the assigned property is expressed in them by a non-verbal lexeme. Predicative nominals can be expressed, by a noun, an adjective, or a prepositional phrase, as in these English examples below: (8) Mary is an actress. (9) Mary is beautiful. (10) Mary is from London. Unlike in English where predicative nominals are syntactically uniform, in some languages of the world, including Russian, predicative nominals display different case marking patterns. It is generally agreed that variation in case marking of predicative nominals captures differences in meaning. This chapter reports that cross-linguistically, the semantic distinction between differentially-marked predicative nouns and adjectives is defined in terms of “relative time stability” (Givón 1979). This means more specifically that the variation in case marking of predicative nominals reflects differences in the perception of properties they (v) Sylvia Obenauer is HER. Mikkelsen (2011:1805) (vi) She is Ms. Doherty. Specificational sentences are described in Higgins (1979) as specifying who or what a given referent is. In contrast to predicative sentences, specificational expressions do not predicate any feature of the pre- copular referent. Here the referent to the left of the copula functions as a variable, whereas the post-copular referent specifies a value for that variable (Akmajian 1979: 162–165). In particular, in construction in (iii) above the value is Mary and the variable is the best actress. Identificational expressions, such as (iv) above, teach names of people or places; here the demonstrative pronoun or the expletive it (Mikkelsen 2011: 1812–813) is a deictic reference and is not anaphoric (Higgins 1979: 237). 15 express as inherent, permanent, or essential vs. temporary, impermanent, or contingent, as illustrated by the following Telugu examples: (11) neenu president-gaa unnaanu. I-NOM president-gaa EX.PRES.1SG ‘I am (now) president.’ (Balusu 2016: 8) (12) neenu president-ni. I-NOM president-1SG ‘I am president.’ (Balusu 2016: 8) Thus, the morpheme -gaa attached to the predicative noun ‘president’ in (11) indicates that being president is a temporary capacity of the subject referent, while the absence of - gaa in (12) implies that it is a permanent role. 2.2 Cross-linguistic variation in case marking 2.2.1 Two major case-marking strategies While case marking of the predicative nominal identical to the grammatical subject is the most common pattern attested cross-linguistically (Comrie 1997, Stassen 2001), differential case marking of the predicative nominal and the grammatical subject is attested not only in Slavic and some Baltic languages, but also in Dravidian (e.g., Balusu 2016) and Daghestanian languages (e.g., Kalinina 1993), in the Carib languages 16 of Amazonia (e.g., Meira and Gildea 2009); in Chadic languages of West Africa and in some Ibero-Celtic languages, e.g., Northern Basque (Stassen 2001: 579).7 Cross-linguistically, most languages are reported to have either of these two case marking strategies (e.g., Comrie 1997, Stassen 2001, Bailyn 2001, Barker 2003). However, some languages are known for having both. 2.2.1.1 Identical case marking Where the predicative nominal occurs with the same case marking as the grammatical subject, the case marking is the Nom or a form functionally similar to the Nom (e.g., the Absolutive), as illustrated, respectively, by the Latin, Finnish, and Northern Basque examples below: (13) Latin Ver-ae amiciti-ae sempitern-ae sunt. true-NOM.PL friendships-NOM.PL eternal-NOM.PL are ‘True friendships are eternal.’ [Cicero, Laelius 9.32] (Comrie 1997: 39) 7 Stassen notes that differential marking of predicative nominals is a “fringe phenomenon” in the Indo- European mega-area (2001: 570). 17 (14) Finnish Kirja on valkoinen book.NOM is white.NOM ‘The book is white/weak.’ (Comrie 1997: 39) (15) Northern Basque Zakurr-a beltz-a da dog-ABS.SG black-ABS.SG is ‘The dog is black’ (Saltarelli 1988: 62) From a cognitive perspective, the identical case marking of the predicative nominal and the grammatical subject is a perceived as an identity relation between the two, which is why the property expressed by the predicative nominal is envisioned as pertaining inherently and permanently to the subject referent. 2.2.1.2 Differential case marking In addition to using identical case marking for subject and predicative nominal, a language may case-mark the two differently. Thus, in Slavic languages and Lithuanian the subject is marked by the Nom, while the predicative nominal is marked by the Instr. In Arabic the subject is marked by the Nom, while the predicative nominal is marked by the Acc. In Balto-Finnic, Daghestanian, Chechenian, and Dravidian languages, predicative nominals can be marked with a special adverbial case marking (Stassen 2001, 18 Comrie 1997, see also Kalinina 1993 for Avar, Daghestanian language; Balusu 2016 for Telugu, a Dravidian language). Consider the following examples: (16) Arabic wa yakuuna r-rasuul-u ‘alaykum šahiid-an. and may.be the-apostle-NOM on.you witness-ACC ‘And the Apostle be a witness in regard to you.’ (Stassen 2001: 40) (17) Avar (a Daghestanian language) Mun jaka bercina-go j-ugo. 2SG.FEM today beautiful-ADV II-be:PRES ‘You look (lit., you are) beautiful today’ (Kalinina 1993: 94) (18) Finnish Hän oli siellä opettaja-na 3SG was there teacher-ESS ‘He was (worked as) a teacher there.’ (Lehtinen 1963: 373) Differential case marking of the predicative nominal and the grammatical subject indicates a different relation between the subject referent and its property than the one captured by the identical case marking. This different relation can be explained on cognitive grounds as such that the property is detached from the subject referent and therefore is perceived as acquired and temporary rather than inherent and permanent. 19 2.2.2 Identical and differential case marking: Variation within a language In some languages predicative nominals are case-marked using both case marking strategies. Although the variation in the case marking of predicative nominals is attested in some languages outside Slavic and Baltic, Stassen (2001: 569) proposes that this is a Circum-Baltic phenomenon, since in all the language families in the Circum-Baltic area except Germanic, predicative nominals can be marked with either the Nom or some non- Nom case marking.8 2.3 Case marking of predicative nominals in the Balto-Finnic languages In Finnish, Estonian, and Votic predicative nominals are attested with either the Nom or the Essive case marking (Ess). The Nom with predicative adjectives in Finnish and Estonian indicates that a property is perceived as “time-stable,” whereas the Ess implies that the property is temporary, transient, or contingent (i.e., that it holds for a specific time/place only), as illustrated by the following Estonian and Finnish examples: (19) Estonian a. Ta oli noor. 3SG was young:NOM.SG ‘S/he was young’ (Diana Krull, p.c. in Stassen (2001: 570)) 8 The Circum-Baltic area is the area around the Baltic Sea. Linguistically, it is associated with three major Indo-European families, such as Baltic, Germanic, and Slavic (Dahl and Koptjevskaja-Tamm (2001)). 20 b. Ta oli seal noore-na. 3SG was there young-ESS ‘S/he was there (as/when) young’ (Diana Krull, p.c. in Stassen (2001: 570)) (20) Finnish a. Ole-n opettaja. be-1SG teacher.NOM ‘I am a teacher.’ (without time dimension) (Turunen 2011: 394) b. Ole-n opettaja-na. be-1SG teacher.NOM ‘I am a teacher (now).’ (Turunen 2011: 394) Lehiste (1972: 216) also notes that besides marking an inherent-temporary distinction, Estonian has an additional contrast, which is a new state or capacity, and which is marked by the Translative case marking (Transl), as illustrated by the following examples: (21) Estonian a. NN on meie saadik London-is. NN is our ambassador:NOM London-INESS ‘NN is our ambassador in London’ (Lehiste 1972: 216) 21 b. NN on meie saadiku-na London-is. NN is our ambassador-ESS London-INESS ‘NN is our ambassador in London’ (ibid.) c. NN on meie saadiku-ks London-is. NN is our ambassador-TRANSL London-INESS ‘NN is our ambassador in London’ (ibid.) Lehiste (1972: 216) points out that the Nom in (21a) implies that “ambassador in London” is a life-long occupation, whereas the Ess in (21b) indicates that “ambassador” is a temporary capacity of NN: that is, the subject referent is not necessarily the permanent or regular ambassador in London, or he may be in London occasionally in other capacities. The Transl in (21c) signals that “ambassador in London” is a new official capacity of NN. 2.4 Case marking of predicative nominals in the Slavic languages and Lithuanian In the Slavic languages and Lithuanian, which is a Baltic language, the perceived differences in meaning between predicative nominals are captured by the Nom-Instr variation. Brugmann’s (1911: 537) work was the first to propose that the Instr with predicative nominals was a common Balto-Slavic innovation. This hypothesis was, however, challenged by Fraenkel (1926), who argued that because the Instr with predicative nominals is not attested in OCS, which predated the modern Slavic languages, it most likely had developed in them separately. Moreover, Fraenkel (1928: 198) 22 suggested that in Baltic languages, the Instr with predicative nominals developed internally and under the influence of Polish.9 2.4.1 Polish In Polish only predicative nouns can be marked by the Instr, while predicative adjectives are invariantly marked by the Nom. Compare the following examples: (22) Polish a. Marek-Ø to ( jest)/ to był szef-Ø. Mark-NOM this (is)/ this was boss-NOM.MASC.SG ‘Mark is/was a boss.’ (Bondaruk 2014: 334) b. Marek-Ø jest/był szef-em. Mark-NOM is/ was boss-INSTR.MASC.SG Marek is/was a boss.’ (Bondaruk 2014: 334) 9 In fact, Fraenkel’s claim about Polish influence on Baltic languages is supported in this respect by the facts that in modern Latvian the Instr is not attested with predicative nominals and that Latvian lost the Instr altogether. As noted in Holvoet (2004: 76), the Instr has merged with the Acc in the singular and with the Dat in the plural. However, Endzelīns points out that the Instr is preserved in Latvian in several fossilized predicative expressions. e.g., puišu man bij būt ‘I had to be a farmhand,’ kalpu gāju ‘I worked as a farmhand,’ māsiņām saucamies ‘we are called sisters’ (1951: 585). 23 (23) Polish a. Ew-a jest ładn-a. Eva-NOM is pretty-NOM.FEM.SG ‘Eva is pretty.’ b. *Ew-a jest ładn-ą. Eva-NOM is pretty-INSTR.FEM.SG ‘Eva is pretty.’ The semantic distinctions between the Nom in (22a) and the Instr in (22b) is reported to conform to the cross-linguistic pattern. Thus, the identical Nom-Nom implies that the property expressed by the predicative noun is inherent or permanent, while the Nom-Instr indicates that the property in question is temporary or occasional (Freese 1959, Jakobson 1964, Bacz 1993). However, Bondaruk (2014) argues that this semantic distinction is not accurate for Polish and proposes instead that the Nom defines the property as a class membership, whereas the Instr characterizes the subject referent through the property in ways other than identification. 2.4.2 Lithuanian In modern Lithuanian, the differences between Nom and Instr predicative adjectives are attributed to the differences in the perception of properties as stative versus 24 dynamic (Stassen 2001: 573, Senn 1966: 430). Thus, after the stative copula būti ‘be’ predicative adjectives usually occur with the Nom, whereas after dynamic copulas with the meaning ‘become, turn into,’ predicative adjectives are marked by the Instr, as illustrated by the following examples: (24) Lithuanian a. Arklys yra geras. horse:NOM.SG is good:NOM.SG ‘The horse is good.’ (Senn 1974:18) b. Norėjo turtingu tapti. want:3PAST rich:INSTR.SG become:INF ‘He wanted to get rich.’ (Senn 1966: 429) (25) Lithuanian a. Jis buvo mokytoj-as. he-NOM was teacher-NOM.SG ‘He was a teacher’ (Senn 1974: 118) b. Jis buvo mokytoj-u. he-NOM was teacher-INSTR.SG ‘He was a teacher (worked as a teacher)’ (Senn 1974: 118) 25 However, Semėnienė (2004) notes that in modern Lithuanian, the differences in meaning expressed by the Nom and the Instr are not the same as the proposed permanent- temporary distinction and that both the Nom and the Instr can mark either a permanent or a temporary state. 2.5 Cross-linguistic correspondences of the Russian Instrumental case marking in predicative constructions As has been noted above, the Instr in Russian has cross-linguistic parallels in predicative constructions, corresponding to the Ess in Finnish (Matushansky 2012), the Ess and the Transl in Estonian (Stassen 2001), the adverb morpheme -gaa in Telugu (Balusu 2016), among others.10 Roy (2006, 2013) proposes that the Russian Instr not only corresponds to French and German bare predicative nouns (nouns without an indefinite article) but also to the Spanish locational copula estar ‘be’ and the Irish locational copula ta/tha ‘be.’ Because of time and space limitations, I will provide only several of these correspondences (for the parallels with Finnish and Estonian, see section 2.3 above). 2.5.1 Telugu Balusu (2016: 4) notes that the morpheme -gaa which occurs in predicative constructions in Telugu is attested also in contexts where Russian has the Instr (e.g., with semi-copulative verbs, with verbs of naming, with resultative verbs, etc.): 10 Note that the Ess and the Transl are not used in these languages to mark the role of instrument, and these case markings are not equivalent to the Russian Instr. 26 (26) Telugu a. kukka balam-gaa anipistaandi. dog strength-gaa seems.3FEM.SG ‘The dog seems strong.’ (Balusu 2016: 4) b. Russian Sobak-a kažetsja sil´n-oj. dog-NOM.FEM.SG seems strong-INSTR.FEM.SG ‘The dog seems strong.’ (27) Telugu a. meemu kukka-ni hero-gaa ennukunnaamu. we dog-ACC hero-gaa elected-1PL ‘We elected the dog the hero.’ (Balusu 2016: 4) Russian b. My vybrali sobak-u gero-em. we-NOM elected dog-ACC.FEM.SG hero-INSTR.MASC.SG ‘We elected the dog the hero.’ Balusu (2016) proposes that -gaa in Telugu is an eventive marker. In predicative constructions this morpheme indicates that the property expressed by a given predicative noun or adjective does not hold of the subject referent per se but rather of the subject 27 referent in relation to the event, which can endure for a long or short period of time. As will be demonstrated in chapter IV, the Russian Instr has a similar function in predicative constructions. 2.5.2 French Roy (2006: 47) points out that when French predicative nouns occur without an indefinite article, they behave like Instr predicative nouns in Russian.11 For example, only bare predicative nouns in French occur after the aspectually-marked copula, as illustrated by the following French and Russian sentences in (28) and (29): (28) Paul a été (*un) prince pendant 5 minutes, Paul had been (*a) prince for 5 minutes et il est redevenu (*un) baron immédiatement après. and he had become (*a) baron immediately after ‘Paul had been a prince for 5 minutes and he became a baron again immediately after.’ (Roy 2006: 47) Similarly, only Instr predicative nouns are grammatical in these contexts in Russian: 11 Note that Russian, as well as most of the Slavic languages, does not have articles. 28 (29) Pavel-Ø po-byl knjaz-em/ *knjaz´ v tečenie 5 minut Paul-NOM had.been prince-INSTR.MASC.SG/ *prince-NOM.MASC.SG for 5 minutes i potom snova stal baron-om/ *baron and then again became baron-INSTR.MASC.SG/ *prince-NOM.MASC.SG ‘Paul had been a prince for 5 minutes and then he became a baron again.’ Roy (2006: 47) proposes that because only bare predicative nouns in French and only Instr predicative nouns in Russian can occur after the copula marked for aspect, as illustrated by the examples above, the property they express is understood as bound in time, and therefore it characterizes the subject referent rather than identifies it. As I will propose in chapter four, however, the property expressed by the referent of the Instr NP is not bound in time, but is instead contiguous with it, and as such is co-temporal with the situation rather than being temporary. The difference between the two is that “bound in time” implies that the property holds for a certain discrete amount of time, while contiguity implies that the property holds for the entire duration of the situation described in the predicative clause. Moreover, Roy (2006, 2013) observes that only bare predicative nouns in French are compatible with locative and temporal modifiers, whereas predicative nouns with the indefinite article are not grammatical in such contexts. Compare the following examples: 29 (30) French a. Paul est médecin à Paris. Paul is doctor in Paris ‘Paul is a doctor in Paris.’ (Roy 2006: 53) b. *Paul est un médecin à Paris. Paul is a doctor in Paris ‘Paul is a doctor in Paris.’ (Roy 2006: 53) (31) French a. Max est étudiant le jour, et gardien de sécurité la nuit. Max is student the day and guard of security the night ‘Max is a student by day, and a security guard by night.’ (Roy 2013: 40) b. Max est un étudiant * le jour, et un gardien de sécurité * la nuit. Max is a student the day and a guard of security the night ‘Max is a student by day, and a security guard by night.’ (Roy 2013: 40) While in Russian the Nom is not ungrammatical with locative and temporal modifiers, it does sound marginal. Note also that Russian does not require a copula in the present tense, which bans the Instr from such constructions. However, the Instr is appropriate with the addition of the verb rabotat´ ‘work’: 30 (32) ?Maks-Ø oxrannik-Ø v Pariže/ dnem. Max-NOM security.guard-NOM.MASC.SG in Paris/ by.day ‘Max is a security guard in Paris/ by day.’ (33) Maks-Ø rabotaet oxrannik-om v Pariže/ dnem. Max-NOM works security.guard-INSTR.MASC.SG in Paris/ by.day ‘Max works as security guard in Paris/ student by day.’ Roy also argues that only bare predicative nouns in French presuppose participation in an activity, whereas predicative nouns with the indefinite article do not, which explains why example (34a) below is infelicitous, whereas (34b) is acceptable. Compare the following: (34) French a. ?Paul est médecin mais il ne pratique plus. Paul is doctor but he not practices anymore ‘Paul is a doctor, but he does not practice anymore.’ (Roy 2006: 31) b. Paul est un médecin mais il ne pratique plus. Paul is a doctor but he not practices anymore ‘Paul is a doctor, but he does not practice anymore.’ (Roy 2006: 31) Roy (2006: 31) proposes that (34a) is marginal because the second conjunct clause negates the activity that is implied by the bare predicative noun. Conversely, (34b) is 31 felicitous, because the second conjunct clause negates the activity that is not implied by the predicative noun with the article. The unacceptability of (34a) and acceptability of (34b) are explained on the premise that the property “doctor” is interpreted in (34b) as a “label.” In other words, the bare predicative noun (34b) implies that Paul is a doctor by training but may not be practicing medicine, whereas the predicative noun in (34a) indicates that the subject referent practices medicine. Roy maintains that properties expressed by bare predicative nouns are characterizing, because they pertain to a given individual, whereas predicative nouns with the indefinite article are defining because they indicate membership in a class of referents (2013: 35). While I do not completely share Roy’s intuition with respect to the function of the Instr with predicative nouns in Russian, her observations captured in (34a) and (34b) are compatible with the Russian data. Compare the following examples: (35) Russian a. ?Pavel-Ø rabotaet vrač-om, no on bol´še ne praktikuet. Paul-NOM works doctor-INSTR.MASC.SG but he any.more not practice ‘Paul works as a doctor but he does not practice any longer.’ b. Pavel-Ø vrač-Ø, no on bol´še ne praktikuet. Paul- NOM doctor-NOM.MASC.SG but he any.more not practice ‘Paul is as a doctor but he does not practice any longer.’ 32 I propose, however, that the marginality of (35a) arises on different grounds. The Instr marking on the referent implies that the property it expresses is a capacity. A capacity is always a role which has a purpose and a function, and which is realized as such through performance (in the sense of carrying out an action, duty, or task). The Nom referent of the predicative nominal, on the other hand, denotes a property, that is, an intrinsic quality or an essential characteristic which, due to its nature, does not presuppose performance. 2.6 Conclusion The cross-linguistic literature reviewed in this chapter reports that the variation in the case marking of predicative nominals is driven by the differences in the perception of the properties they express as inherent and permanent or temporary and less-time stable. The cross-linguistic pattern in case-marking languages is that identical case marking of the grammatical subject and the predicative nominal implies that a property is envisioned as inherent and permanent, while differential case marking indicates that the property is temporary and less-time stable. The phenomenon of the alternating case marking can be explained cognitively on the grounds that identical case marking of both the subject referent and the predicative nominal is discerned as a symmetrical and equative relation between the two, whereas differential case marking is perceived as an asymmetrical relation in which the subject referent is cognitively distanced from the property. It is not coincidental, then, that properties expressed by predicative nominals marked identically to the grammatical subject are envisioned as inherent and permanent, while properties expressed by the differently-marked predicative nominals are perceived as temporary or accidental. 33 While these cognitive premises can generally account for the Russian Instr, the data from Telugu (Balusu 2016) and French (Roy 2006, 2013), as well as the observation made for Lithuanian (Semėnienė 2004) support the proposal made in this study that the Instr with predicative nominals may have a much broader scope than simply marking a property as temporary. For example, the Instr in Slavic and Lithuanian, also occurs in a variety of other, non-predicative constructions, and among other functions, marks referents as instruments. However, in none of its uses in Slavic or Lithuanian does the Instr convey temporariness. Note also that in Telugu the morpheme -gaa that marks predicative nominals is also attested to form adverbs of manner and adjectives (Krishnamurti & Gwynn 1985:126). I take the position that grammatical categories, including case marking, are polysemous and the meanings they express are interrelated to each other in a motivated fashion. It is therefore legitimate to assume that the meaning of the Instr with predicative nominals should be related in one way or another to its semantic functions in other Instr constructions. Hence I argue that temporariness as the meaning of the Instr with predicative nominals is only a rough approximation of a more complicated semantic phenomenon. I establish in chapter IV and demonstrate in chapters V and VI that temporariness cannot account fully and accurately for the meaning of the Instr with predicative nominals in Russian. Since the differences in meaning between predicative nominals in Russian are reflected in the variation in case marking and because the Instr and the semantic role of Instrument are often treated as the same phenomenon, although they are not, marking predicative nominals by the Instr may seem quite odd. In the following chapter I draw a 34 distinction between case, case marking, and semantic roles, to provides insights on why the Instr is possible in predicative constructions. 35 CHAPTER III WHAT IS CASE? CASE AND CASE MARKING This chapter analyzes the distinction between grammatical case itself and morphological case marking. It also discusses traditional approaches to case and provides an overview of Hjelmslev’s (1935) and Jakobson’s (1936, 1958) models and their limitations. I further summarize the findings of formal models such as Chomskian’s Government and Binding (1981) and the Minimalist Program (1993) and explain why these models cannot be applied successfully in the present study. The chapter then proceeds to an overview of semantic role models, following which I provide a motivation for treating case as a cognitive phenomenon. Following the distinction made in this study between case and case marking, I demonstrate that the scope of the Russian Instr is much broader than the semantic role of Instrument. The phenomenon of case has been a subject of linguistic investigation for centuries, probably since the time of Pāṇini’s grammar of Sanskrit. Importantly, since linguists have yet to agree on how the notion of “case” resonates with grammatical relations, meaning, and morphological form, the term “case” has been used differently in different models. Thus, in the works of the Neogrammarians (e.g., Delbrück 1883), of Hjelmslev (1935), and of Jakobson (1936), the term “case” is meant to refer to an abstraction, while in formal models, “case” is used to mean a grammatical relation. 36 Because of space and time limitations, I will restrict the overview of the literature to the studies and approaches that have influenced the study of case the most. In the sections that follow, I will present traditional approaches to case, such as Hjelmslev’s and Jakobson’s models, and formal grammar. These sections are organized chronologically rather than hierarchically, and the order in which the models appear is not intended to underscore or diminish their significance in the linguistic scholarship. I then discuss the distinction between case and case marking, and advocate against approaches in which case is understood as a syntactic relation. 3.1 Traditional approaches to case and case marking Since the Renaissance, case marking systems in inflectional languages have been traditionally described as lists of individual meanings and functions. However, at the end of the 19th century the Neogrammarians proposed a new model, in which they introduced the notion of Grundbegriff, or a generalized, basic meaning (Delbrück 1883). The idea behind this model is that case is an abstract notion and that it is realized in individual languages as a set of morphological markings associated with a given case. Case as an abstraction has one generalized meaning. At the beginning of the 20th century, this idea was further advanced by Hjelmslev (1935) and Jakobson (1936, 1958). 3.1.1 Hjelmslev In Hjelmslev (1935), case is defined as an abstraction on the premises of which concrete uses can be inferred. While Hjelmslev does not make a distinction between case 37 and case marking, it is clear from his account that the concrete uses refer to case markings pertaining to a given case. The abstract meaning of case is understood in terms of the opposition between sets of case markings in the language. Because Hjelmslev was a localist, the meaning of a given case marking is described in his analysis in local terms. Thus, Hjelmslev (1935) juxtaposes case markings along two dimensions. The first dimension is direction, and it is realized as the opposition between rapprochement (bringing nearer) and éloignement (taking away) (1935:128). The second dimension is the degree of intimacy, which can be coherent or noncoherent (ibid.). Coherence involves contact or penetration, whereas incoherence involves proximity. For example, the Nom is defined as incoherent but neutral with respect to direction. 3.1.2 Jakobson Like Hjelmslev (1935), Jakobson’s (1936) model presupposes one invariant meaning of a given case marking and numerous individual meanings which are conditioned lexically and syntactically. Following the Neogrammarian tradition, the value of each case marking is envisioned as aggregate meaning, or Gesamtbedeutung. Since the aggregate meaning is an abstraction, it is not conditioned by the sentential environment and cannot be deduced from the individual sub-meanings (Sonderbedeutungen). Among the individual meanings, there exists one core meaning (Hauptbedeutung). Jakobson’s model resonates with Hjelmslev’s in that it envisions cases as correlative. This means more specifically that the value of a given case marking is elicited in its opposition to other case markings in the language. Applying his model to 38 Russian, Jakobson states that the Nom is always unmarked. The Acc is opposed to the Nom, in that it is subordinated to it. The Acc also implies direction or goal. The Instr and the Dat are marginal or peripheral (Randkasus) and as such they are opposed, respectively, to the Nom and the Acc, which are called central case markings (Vollkasus). Jakobson’s (1958) article organizes all case markings in Russian in a cube and describes them in three dimensions in terms of features, e.g., [± direction], [± marginal], and [± quantification]. These necessary features convey the extent to which the referent of a noun with a given case marking takes part in or is affected by the action described by the verb. 3.1.3 Limitations of Hjelmslev’s and Jakobson’s models One of the disadvantages of Hjelmslev’s and Jakobson’s models is that they do not explain how the aggregate meaning of a given case marking is related to its individual meanings. Like the Neogrammarians, Hjelmslev and Jakobson thus treat the individual meanings as randomly connected. Another limitation is Jakobson’s and Hjelmslev’s intuition that the meanings of a given case marking are organized in a hierarchical fashion, presupposing the dominance of a Hauptbedeutung rather than prototypicality of one meaning among others. 3.2 Formal models In formal models, case is envisioned as a purely syntactic phenomenon that has to do with a specific structural position that noun phrases (NPs) occupy in a sentence. Case Theory, as proposed in Government and Binding (1981) and the Minimalist Program 39 (1993), presupposes the existence of abstract Cases, which are universal, and which are responsible for all kinds of movement and transformations (e.g., passive from active). Importantly, Chomsky (1980, 1981) argues that abstract Case may or may not coincide with the actual morphological marking that is attested in some languages. In Case Theory, Case is assigned in the syntactic surface structure if certain structural conditions are met. Nominative Case is assigned in finite clauses to the NPs in the subject position by the head of the inflection phrase. Accusative case is assigned to NPs in the object slot by the governing verb. This type of assignment is referred to as structural Case. Additionally, Case Theory states the existence of non-structural Case, which can be either lexical or inherent. Lexical Case is idiosyncratic, in that it is lexically selected and licensed by certain lexical heads (specific verbs and prepositions). Conversely, inherent Case is more regular, and its assignment is accompanied by a theta- role assignment. In these formal models, the Russian Instr is lexical and both idiosyncratic and inherent. According to this view, the Instr with predicative nominals is as idiosyncratic as its occurrences in other constructions and therefore is unpredictable (but see Bailyn 2011, Babby 1987, 1991, Freidin and Sprouse 1991 who argue that the Russian Instrumental Case is a structural Case, like the Nominative Case). Conversely, I will demonstrate in chapters V and VI, that the Instr occurs with predicative nominals on the same semantic premises as the Instr in other Instr constructions. Moreover, because formal models refute the Neogrammarian and Structuralist assumption that case marking has meaning, they explain differences in the case-marking of predicative nominals as being determined by language-internal structural principles and therefore interpretive rather than motivated semantically and cognitively. 40 3.3 Semantic roles Semantic roles or “case frames” (Fillmore 1968), also known as thematic roles or theta roles, are lexico-semantic representations depending on their syntactic distribution. Influential papers by Gruber (1965), Fillmore (1968, 1977), and Jackendoff (1972, 1976), among others, made semantic roles a significant contribution into linguistic theory. Traditionally, semantic roles are organized as lists of discrete roles, such as Instrument, Theme, Experiencer, Agent, and Patient, among others, and are intended to record differences and similarities in meaning among various verbs in terms of possible participant roles. 3.3.1 Principles underlying semantic role models Dowty (1991) formulates a set of fundamental principles that underlie any theory of semantic roles: 1) every argument of every verb is assigned a thematic role; 2) every argument is assigned only one thematic role; 3) every argument of every verb is distinguished from the other arguments by the role it is assigned; 4) each semantic role is provided with an exhaustive definition that applies to all verbs and all situations; Importantly, the role definition is independent of the meaning of a given verb or other thematic role that this verb can assign. 3.3.1.1 Issues with semantic roles Despite these idealistic expectations set forth in Dowty (1991) and discussed in the paragraph above, semantic role models are not uniform, and there is no agreement 41 among the proponents of semantic role theories as to the number of roles needed for successful application across the board or the nature of the roles themselves. For example, Anderson (1971) argues for a set of just three semantic roles, Source, Location, and Goal, from which other non-local roles emerge. On the other end of the spectrum, one finds models like Pollard and Sag’s (1994), in which semantic roles are understood as specific to a given verb. Thus, the verb “love,” for instance, assigns two semantic roles, “lover” and “lovee” and these roles cannot be assigned by any other verb. What follows is that there are no generalizations as to the types of semantic roles, which makes lists of possible roles in a given language endless. A related controversy is the degree of granularity, or how finely semantic roles should be fragmented. For example, in Jackendoff’s (1983) model, the semantic role Agent is further divided into Agent and Actor. Van Valin (1990) distinguishes between Agents and Effectors; Cruse (1973) differentiates among volitive, effective, initiative, and agentive, and Lakoff (1977) goes even further by suggesting around 14 possible sub- roles of Agent. Additionally, the boundaries among the types of semantic roles turn out to be fuzzy rather than clear-cut. A well-known problem is how to differentiate between Instruments vs. Comitatives in sentences such as i) John cut the meat with a knife vs. ii) John burgled the house with an accomplice vs. iii) John won the appeal with a highly- paid lawyer. The issue here is that in English the argument with with can introduce a range of roles instead of a single role. 42 Moreover, even as a set, semantic roles appear to be disconnected from one another, which makes it more difficult to register any similarities among individual roles and thus commonalities among various verbs. 3.4 The Instrumental case marking in Russian: Existing models 3.4.1 Semantic roles and semantic lists The long-established tradition has been to describe the meaning of the Instr in terms of individual disconnected meanings (e.g., instrument, means, comparison, cause, among others), which are labeled in accordance with the lexical meaning of the referent of the Instr noun (Potebnja 1888, Bernštejn 1958, Mrázek 1964, among others; see also Pečenyj 2012 for an inventory of semantic roles conveyed by the Instr). 3.4.2 The formalism of Švedova’s Academy Grammar The Academy Grammar of the Russian language (AG) (1980) envisions case as a syntactic relation. The meaning of a given case marking depends on the syntactic function of the noun in the clause or phrase. Two types of subordinate syntactic relations play a role in shaping the meaning of case markings. In the first type, case marking does not depend on the syntactic position of the NP but is governed instead by the lexeme on which it syntactically depends (prislovnaja svjaz´): (49) Mal´čik-Ø gordit-sja sestr-oj. boy-NOM.MASC.SG proud.of-REFL sister-INSTR.FEM.SG ‘The boy is proud of (his) sister.’ 43 (50) Mal´čik-Ø dumaet nad knig-oj. boy-NOM.MASC.SG thinks over book-INSTR.FEM.SG ‘The boy is thinking about the book-Instr.’ In (49) above the Instr is assigned by the verb gordit´sja ‘be proud of,’ while in (50) the Instr is assigned by the preposition nad ‘over.’ In the second type of the subordinate relations, case marking is determined solely by the syntactic positions of the NP in a clause (neprislovnaja svjaz´): (51) Mal´čik-Ø čitaet. boy-NOM.MASC.SG reads ‘The boy reads.’ The AG further proposes that the meanings of case markings are derived from the syntactic functions of NPs. Three general meanings are postulated: complement (ob´´ektnoe), subject (sub´´ektnoe), and attribute (opredelitel´noe) (1980: 475). The central meanings of the Instr are defined in the AG as attribute and complement, as illustrated by the following examples: (52) Brat-Ø budet student-om. (attribute) brother-NOM.MASC.SG be-FUT student-INSTR.MASC.SG ‘Brother will be a student.’ 44 (53) My šli bereg-om. (attribute) we-NOM walked shore-INSTR.MASC.SG ‘We walked along the coast-line.’ (54) Mal´čik-Ø gordit-sja sestr-oj. (complement) boy-NOM.MASC.SG proud.of-REFL sister-INSTR.FEM.SG ‘The boy is proud of (his) sister.’ (55) My napisali pis´m-o karandaš-om. (complement) we-NOM wrote letter-NOM.NEUT.SG pencil-INSTR.MASC.SG ‘We wrote the letter with a pencil.’ The subject meaning is peripheral for the Instr and is restricted to passive constructions: (56) Dom-Ø stroit-sja raboč-imi. house-NOM.MASC.SG being.built-REFL workers-INSTR.PL ‘The house is being built by workers.’ While the AG submits that the Instr has several meanings, these meanings are understood as specific syntactic functions rather than cognitive or semantic relations that exist among the referents of the NPs, i.e., actual people and entities, in the situation 45 described by the verb, and therefore this approach is fundamentally incongruent with the model proposed in this dissertation. 3.4.3 Wierzbicka: Metalanguage and semantic invariance Wierzbicka (1980) builds her model around the meanings evoked by the Instr in various constructions. In particular, she distinguishes among18 such meanings and postulates that it is more appropriate to talk about 18 related constructions with the Instr rather than 18 meanings of the Instr per se. Like Potebnja (1888), Jakobson (1936), Janda (1993), and Mrázek 1964, Wierzbicka (1980: 147) attempts to formulate an invariant meaning of the Instr, which is defined in her metalanguage as “something which is acted on not in order for something to happen to it, but in order for something else to happen,” and which semantically corresponds to the role of Instrument. While Wierzbicka admits that the Instr is polysemous, she fails to articulate explicitly how the different meanings of the Instr in individual constructions are interrelated. Moreover, Wierzbicka’s model does not take into consideration the role of construal, which is especially significant in situations when two different case markings are possible, e.g., the Russian Nom-Instr alternation in predicative constructions or the Nom-Instr alternation with verbs of controlled motion (see Langacker 1987: 138–141, Taylor 2002: 11 for further detail on conceptualization). 46 3.5 Case as a cognitive phenomenon. Case and morphological case marking 3.5.1 Cognitive definition of case In contrast to the approaches described above, this study envisions case as a cognitive phenomenon. Contrary to the frameworks that treat case as an abstraction or a syntactic relation, in this dissertation, following Vakareliyska (1994), case is defined as a perception by the speaker of physical or cognitive relationships that hold among the referents of the NPs in the situation described in the utterance, rather than necessarily in objective reality. In inflectional languages, these cognitive relations among referents of NPs are expressed morphologically via suffixes or grammatical endings that attach to NPs in a sentence, or case markings. Some languages, including English, do not have case marking, but this does not imply that they have no grammatical case; in such languages case exists but simply is not marked morphologically. It is demonstrated in Vakareliyska (1994) that English, for example, does not have a Dative case marking (Dat), but does have the Dative case. For example, the marker to in Mary gave the book to John is functionally similar to the Dat marker na in the corresponding Bulgarian sentence Marija dade knigata na Ivan. However, the scope of the English to is restricted to marking exclusively indirect objects. While marking indirect objects is one of the relations conveyed by the Dative case, the range of the relations expressed by the Bulgarian marker na is much wider, as illustrated by the sentence Pomagam na Marija ‘I am helping Mary.’ In many Indo-European languages, the verb “help” requires a Dat case marking, whereas English disallows the use of to in the equivalent construction *I am helping to Mary. The absence of a similar morpheme in 47 English does not mean that in the well-formed English sentence I am helping Mary the NP Mary is Accusative; on the contrary there is no indication as to what case is assigned by the verb “help.” Similarly, certain semantic classes of verbs in Russian require their noun complements to occur with the Instr, e.g., verbs of government and possession, such as vladet´ ‘possess’ and upravljat´ ‘manage,’ among some others, as illustrated by the examples given below: (36) Mam-a vladeet turfirm-oj v Moskve. mom-NOM owns travel.agency-INSTR.FEM.SG in Moscow ‘Mom owns a travel agency in Moscow.’ (37) Mam-a upravljaet biznes-om otc-a. Mom-NOM manages business-INSTR.MASC.SG father-GEN ‘Mom manages father’s business.’ The NPs in (36) and (37) are marked for the Instrumental case, whereas, in corresponding English constructions, both NPs are marked for a grammatical relation, i.e., direct object. As is the case with the English verb ‘help,’ the English verbs ‘own’ and ‘govern’ are not marked morphologically for any particular case. 48 3.6 The Instrumental case marking versus the semantic role of Instrument In many languages that inflect for case, the Instr occurs on nouns to denote that their referents, which may not be typical instruments in objective reality, are envisioned in a situation described by the verb as tools or instruments with which the action is executed. Moreover, the Instr is readily recognized as referring to the role of Instrument when it occurs on nouns where their referents are prototypical instruments or tools in everyday life. For many such referents, the instrumental sense is indeed natural. Lexemes denoting such instruments and tools as brooms, shovels, scissors, knives, hammers, wrenches, and the like are most often found with the Instr in environments where they help another referent to implement the action described by the verb. The fact that speakers tend to discern real-world objects in a particular way, and that certain natural classes of nouns frequently occur with particular case markings has given rise to analyses in which case is viewed as an instantiation of a deep-level configuration. Note, however, that this is not always the case. In languages that express case morphologically, the Instr can also occur on nouns whose referents do not have the features associated with either a prototypical actual instrument or a prototypical Instrument role, as in these examples from Russian: (38) Ja tri dnja revela belug-oj. I-NOM three days cried beluga-INSTR.FEM.SG ‘I was bellowing like a whale for three days.’ 49 (39) Maš-a zabolela angin-oj. Mary-NOM.FEM.SG fell.ill strep throat-INSTR.FEM.SG ‘Mary fell ill with strep throat.’ Note that while the referent of the NP beluga is a referent from objective reality, that is, one of the species of whale, beluga in (38) is understood figuratively as a specific manner or mode of crying rather than the instrument with which crying is carried out. Similarly, angina in (39) is perceived as the cause of the illness rather than the instrument facilitating it. Even in English, where with typically marks the semantic role of Instrument, an illness causing physical discomfort can only marginally be called an Instrument. Consider also in this regard the three examples below in which the referents of the Instr NPs “year”, “arch”, and “boss”, like the Russian beluga or angina, respectively, in (38) and (39) are not compatible with the semantic definition of the prototypical Instrument: (40) Adyghe s-šə jəʎes-je-č’̣e səš’ nah-jə nahə-ẑ. 1SG-brother year-IFX-INSTR I than than-old ‘My brother is one year older than me.’ (Serdobolskaja & Kuznecova 2007: 4) 50 (41) Adyghe ar arke-č’̣e je-čə̣-ʁ. DEM-ABS arch-ERG-INSTR 3SG-enter-PAST ‘He went out through the arch.’ (Serdobolskaja & Kuznecova 2007: 4) (42) Polish Marek-Ø jest szef-em. Marek-NOM is boss-INSTR.MASC.SG ‘Marek is a boss.’ (Bondaruk 2014: 334) Importantly, lexemes that are ontological instruments often appear in Russian with case marking other than the Instr: (43) Maš-a kupila Ivan-u nov-yj molotok-Ø. Mary-NOM bought John-DAT new-ACC.MASC.SG hammer-ACC.MASC.SG ‘Mary bought John a new hammer.’ (44) My ne smogli vzlomat´ seif bez otmyčk-i. we-NOM not could break safe without latchkey-GEN.FEM.SG ‘We could not unlock the door without the (a) latchkey.’ 51 (45) My smogli vzlomat´ seif bez otmyčk-i. we-NOM could break safe without latchkey- GEN.FEM.SG ‘We managed to unlock the door without the (a) latchkey.’ While the referent of the noun molotok in (43) is a tool, it is not marked by the Instr because in this situation molotok is not intended to be used as an instrument. In a similar fashion, although otmyčka ‘latchkey’ in (44) and (45) would be the most typical instrument to break a safe in objective reality, it is not envisioned as such in the situation described by the verb, because in (44) a latchkey is absent from the situation in the first place, as is implied by the Gen, and therefore cannot be used as an instrument. In the scenario described in (45), the Gen suggests that a latchkey was not available and something else was used to break the safe. 3.7 The Instrumental case marking in a cross-linguistic perspective As has been pointed out in the preceding section, the scope of relations expressed by the Russian Instr is not exhausted by the semantic role of Instrument. In fact, Narrog (2009: 597) notes that it is quite atypical for the Instr to express only one function or semantic role. For example, in Adyghe, a West Circassian language of the North Caucasian family, besides the prototypical instrument meaning, the Instr is reported to convey 15 other meanings: a referent of the Instr noun can be perceived as a unit of measure, a cause, a source of information, a manner, a route, a basis of comparison, a point of identification, etc. (Rogava & Keraševa (1966), Kumaxov (1971), Xalbad (1975), Zekox (2002). Consider the following Adyghe examples: 52 (46) Adyghe (cause/reason) s-jə-ʁweš’ə-nəʁe čẹ sə-qə- čẹha-ʁ ŝwa-dež’. 1SG-POSS-make.mistake-VN-INSTR 1SG-DIR-enter-PAST 2PL-to ‘I came to you because of a mistake.’ (Serdobolskaya 2011: 521) (47) Adyghe (point of identification) se we w-jə- ḳwa-č’̣e-č’̣e wə-qe-s-ṣ̂e-ž’ə-ʁ. I you 2SG-POSS-go-NML-INSTR 2SG.ABS-DIR-1SG.A-know-RE-PAST ‘I recognized you by your step.’ (Serdobolskaya 2011: 523) In Manipuri, a Tibeto-Burman language, the Instr indicates that the referent of the Instr noun is a means of transportation, as in the following construction: (48) əy-nə bəjar-də-gi bas-nə laki. I-Nom market-Loc-Gen bus-Instr came ‘I came from the market by bus.’ (Bhat & Ningomba 1997:105) 3.7.1 The prototypical meaning of the Instrumental case marking cross-linguistically While cross-linguistically, the prototypical meaning of the Instr is that of an instrument (see Palancar 2002: 32; Guillaume 1992 for Indo-European languages; see also Wierzbicka 1980, Janda 1993 for Russian; Janda 1993 for Czech; Bacz 1993 for Polish), some researchers argue that it is not the only prototypical function of the Instr. 53 For example, Lehmann and Shin (2005: 20) suggest that expressing a means of transportation is the other major function of the Instr. Stolz (1996, 2001) names instrument and comitative as the two prototypical senses of the Instr. Similarly, Bernštein (1958) and Mrázek (1964) propose instrument and comitative as the oldest meaning of the Instr in Slavic languages. The broad cross-linguistic data analyzed in Stolz (1996, 2001) demonstrates a rich repertoire of meanings evoked by the Instr. Using a sample of 200 languages, Stolz shows that if the Instr evokes the instrument sense, it also evokes other senses, such as companion (comitative), cause/reason, manner, material, location, and point in time, among others. Figure I below captures the polysemous nature of the Instr across languages: Figure 1. Semantic functions of the Instr across languages (Narrog 2009: 599) duration physical proximity route co-participant locative cause/ reason clausal coordination NP-coordination comitative instrumental agent possession ablative material passive agent temporal ‘from’ ergative 54 The arrows in the network show the direction of extension from one function to another. This semantic map depicts only a minimal number of interrelations among the various senses of the Instr and many more connections are likely possible among the existing ones (Narrog 2009: 599).12 3.8 Conclusion The analysis of the examples presented in this section demonstrates that the Instr in Russian and other languages is much broader in scope than the semantic role of Instrument. That the Instr can occur on nouns whose referents are not actual instruments or tools, and that nouns whose referents are ontological instruments can occur with case markings other than the Instr suggest that the Instr and the role of Instrument are not the same phenomenon. While Instrument is one of the possible relations conveyed by the Instr in Russian, Polish, or Adyghe, it does not exhaust the range of the Instr meanings in these languages. Note also that in its traditional definition, Instrument is a discrete and delimited, in that it refers to a specific syntactic position of a given participant in the situation described by the verb, whereas the Instr in Russian, Polish, or Adyghe refers to a cognitive scope of possible physical and/or psychological relations among the actual people or entities in the situation described by the verb. 12 The literature on grammaticalization in individual languages unanimously agrees that the directionality of functional extension goes from comitative to instrumental and not vice versa (Heine et al. 1991: 166; Luraghi 2001; Stolz 2001). Less agreement is seen in identifying the directionality between the instrument meaning and the agentive one. Palancar (2001) surveys a sample from 137 languages to demonstrate that the agentive meaning has evolved from the instrumental one, via causation (see also Narrog 2009). Luraghi (2001) suggests that the relationship between the meanings of an instrument and an agent is bidirectional. 55 Also, in a description of polysemous grammatical categories such as case marking, the traditionally used labels can be misleading. More specifically, the name “Instrumental” unavoidedly brings to mind a reference to the semantic role of Instrument. Such associations are quite predictable for several reasons. Names for grammatical categories are intended to be mnemonic devices, which means that besides naming in itself they are supposed to reflect some of the most prototypical meanings evoked by a given case marking. For the Instr, the most common meaning emerges in a situation where a referent, typically an entity, i.e., a thing/physical object from the outside world, is used by another referent as an implement to perform the action described by the verb. This prototypical role is attested in many languages of the world. The long-standing tradition of describing case as a grammatical phenomenon in terms of semantic roles, makes the label “Instrumental” easily linked to the role of Instrument. However, as has been emphasized earlier, this study does not treat the Instr as a grammatical relation or a participant role. Even more so, the cross-linguistic data demonstrate that the scope of the Instr goes far beyond a syntactic relation or a specific argument role. Finally, while grammatical relations and semantic roles are discrete notions, the Instr is a network of interrelated meanings. 56 CHAPTER IV AN ALTERNATIVE MODEL FOR STUDYING THE MEANING OF THE INSTRUMENTAL CASE MARKING IN RUSSIAN This chapter propose an alternative, a two-level model which presupposes a cognitive and a semantic organization of meanings. In this model all referents of the Instr NPs, irrespective of their semantic relation in the situation, are contiguous with and contingent on the action described by the verb and another referent. The specific fine- grained Instr meanings (such as instrument, means, or unit of measure, among others) are meanings borne out in individual constructions and depend on the meaning of the verb and the lexical semantics of noun referents. 4.1 Organization of meaning 4.1.1 Polysemy and prototypes The analysis of the Russian Instr undertaken in this study presupposes that a given linguistic unit can activate more than one meaning. It is widely accepted in cognitive models that polysemy is pervasive in all aspects of language use. Polysemy is the association of two or more related meanings expressed in a single linguistic form. For example, Brugman’s (1981) study of the English preposition over and later analyses by Brugman and Lakoff (1988), Dewell (1994), Kreitzer (1997), and Tyler & Evans (2003) have established that prepositions are highly polysemous. More specifically, they have demonstrated that in each individual instance, the English preposition over elicits a 57 different scene: in some instances, it refers to different types of motion (e.g., The bird flew over the hill vs. The boy walked over the hill), in others it denotes a static scene (e.g., The lamp hangs over the table), yet in other instances it triggers a non-spatial reading (e.g., You made over a hundred errors). The examination of the English over suggests that while its senses are different in each individual construction, they are still shared in one way or another by other constructions with over, and thus can be represented as a structured network of interrelated meanings. Within this semantic network, the meanings of over are believed to form a “radial” category in which some meanings are envisioned as central, or prototypical, while others are argued to have stemmed from the central meanings by various types of links, i.e., semantic extensions. Thus, the meanings of over that seem to be unrelated at first glance can in fact be connected to one another in a motivated way. While classical polysemy refers to lexemes, polysemous effects have been reported in phonology, morphology, and syntax (e.g., Taylor 1989, Sweetser 1990 on modality, Janda 1990, Nikiforidou 1991 on the Gen in Indo-European languages; Verhagen 1992 on passives, Dąbrowska 1997 on the Polish Dat, inter alia). If one embraces the assumption that grammatical categories are as polysemous as lexical units, the next step in the analysis is to determine the nature of the relations among the different meanings of each polysemous category.13 In general, two principles of organization are 13However, like lexemes, case markings may be homonymous, in that a single case marking may express meanings that are not related to one another. For example, Taylor (1995:103) reports that some speakers perceive two meanings to be related, while others do not. However, while homonymy arises accidentally, when two distinct lexemes become phonologically identical following sound changes in the language, polysemy manifests itself as similar patterns of meanings cross-linguistically, also in genetically unrelated languages. As has been demonstrated in section 3.6 in chapter three, the meanings evoked by the Russian Instr are also attested cross-linguistically and can be represented as a semantic network. 58 possible: the meanings may all share a core meaning, as is proposed, for example in Hjelmslev’s (1935) or Jakobson’s (1936, 1958), Wierzbicka’s (1980) models for the Russian Instr, or meanings may be related on a one-to-one basis, forming a network where A shares some traits with B and B with C, but not necessarily A with C. This network organization is known as Wittgenstein’s (1953) family resemblance. I maintain that different meanings of the Instr can be organized in a semantically motivated way as a network with a radial rather than a hierarchical structure. 4.1.2 Cognitive and semantic relations Following Wierzbicka (1980), this model also argues that the meanings of the Russian Instr are interrelated. However, unlike Wierzbicka (1980) who seeks to find a semantic invariant of the Russian Instr, I propose a cognitive invariant. I define the general meaning of the Instr in cognitive rather than semantic terms. The Instr signals how the referents, actual people and entities are related to each other and to the action in time and space, and in what way they affect each other. These relations are contiguity and contingency. Contiguity implies that two (or more) referents are spatially adjacent and as such, they are simultaneously involved in the action described by the verb; hence the referent of the Instr NP is always contiguous with the action. Contingency means that between two (or more) referents, the emergence of one referent depends on the intention of the other. In some instances, contiguity and contingency may entail a collateral relation, which is defined such as that between the 59 two (or more) contiguous referents, the contingent referent is perceived as accompanying or mediating the action described by the verb. Thus, a referent of the Instr NP, irrespective of its specific semantic relation in the situation, is envisioned as cognitively adjacent in space and time to other referents in the situation through its contiguity with and contingency on the action described by the verb. The specific fine-grained meanings (e.g., instrument, means, unit of measure, pathway, among others) are instantiated in individual constructions and hinge on the verb meaning and the lexical semantics of the noun referents. Compare the following examples in this respect: (57) Det-i bolejut gripp-om. children-NOM be.ill-PL flu-INSTR.MASC.SG ‘The children are ill with flu.’ (58) ??Det-i bolejut molotk-om. children-NOM be.ill-PL hammer-INSTR.MASC.SG ??‘The children are ill with hammer.’ The utterance in (58) is grammatical because the verb bolet´ ‘be sick with’ requires an Instr complement, but it is infelicitous because molotok ‘hammer’ is not a typical cause of illness. A hammer can cause an injury or death, but in that event the hammer must come into direct contact with the body, as in udarit´ molotk-om ‘hit (with) hammer- INSTR’ or ubit´ molotk-om ‘kill (with) hammer-INSTR.’ 60 The individual meanings of the Instr are related to each other through metaphorical and metonymical extensions, which occur when the referent of the Instr noun is devoid of its concrete literal meaning and is understood figuratively, as in the following examples: (59) zabit´ gvozd´-Ø molotk-om. pin-INF nail-ACC.MASC.SG hammer-INSTR.MASC.SG ‘pin a nail with a hammer.’ (60) priletet´ samolet-om. fly-INF plane-INSTR.MASC.SG ‘arrive by plane.’ While “plane” in (60) is only marginally an instrument, the referents of the Instr NPs in both (59) and (60) are related semantically. This is possible because the referents of “plane” and “hammer” help to mediate the action described by the verb, which is pinning and flying, respectively. Since a plane is not a prototypical instrument, it is perceived metaphorically as such because of its similarity in function with the prototypical instrument. 4.2 Application of the proposed model The Instr is attested in a number of constructions in Russian. In some, the referent of the Instr noun is envisioned as an accessory for the action, in others as a pathway to a 61 destination, in yet others as a cause of emotional or physical discomfort, to name just a few. Contrary to the long-standing tradition of presenting individual Instr meanings atomistically, I seek to organize them in a more efficient and psychologically more plausible way. The referents of the Instr nouns in each grouping below may not share the same number or kind of features, but they are interrelated in one way or another. For example, while the referents of the Instr NPs “ax”, “chalk”, and “gun” in (61), (62), and (64), respectively, are perceived as hand-held by the human referent, the referent of the Instr NP “bus” in (63) is not. Moreover, it is implied by the Instr that the human referents in (63), unlike their counterparts in (61) and (63), are not performing the action described by the verb. However, because all four referents of the Instr NPs in these examples mediate and facilitate the action described by the verb, they can be grouped together. 4.2.1 Referents of Instr nouns as accessories to the action described by the verb As mentioned in the previous paragraph, the referents of the Instr nouns in the examples given below mediate and facilitate the action described by the verb: (61) Ivan-Ø rubi-l drov-a topor-om. John-NOM chopped-MASC.SG wood-ACC.PL ax-INSTR.MASC.SG ‘John chopped the wood with an ax.’ 62 (62) My narisoval-i kartin-u mel-om. we-NOM drew-PL picture-ACC.FEM.SG chalk-INSTR.MASC.SG ‘We drew the picture with chalk.’ (63) My vernulis´ domoj avtobus-om. we-NOM returned-PL.REFL home bus-INSTR.MASC.SG ‘We returned home (by) bus.’ (64) Ivan-Ø razmaxival pistolet-om u vsex na glazax. John-NOM brandished gun-INSTR.MASC.SG by everyone on eyes ‘John was brandishing the gun in front of everyone.’ Note that in addition to what has already been said about the referents of the Instr NPs in the examples above, an important feature that they share is purpose. Thus, in (61), (62), and (64), the human referent manipulates the referent of the Instr NP with a purpose. In (61) the purpose is to pin the nail; in (62) – to draw a picture; in (64) – to threaten. The purpose may not be as salient in (63) as it is in the other constructions from this grouping, but the fact that the motion described in (63) must be purposeful can be illustrated by the semantic unacceptability of (65) vis-à-vis (66) below: (65) ??My vernulis´ domoj sin-im avtobus-om. we-NOM returned-PL.REFL home blue-INSTR.MASC.SG bus-INSTR.MASC.SG ?‘We returned home by a blue bus.’ 63 (66) My vernulis´ domoj rejsov-ym avtobus-om. we-NOM returned-PL.REFL home shuttle-INSTR.MASC.SG bus-INSTR.MASC.SG ‘We returned home by shuttle bus.’ Sinij ‘blue’ in (65) only describes the color of the bus and does not say anything about its function. Conversely, rejsovyj ‘shuttle’ directly indicates the function of the bus, which is to bring passengers from point A to point B on a scheduled route. Since function presupposes a purpose, the construction in (66) is felicitous. The referents of the Instr NPs in this grouping are the most prototypical among all referents of the Instr NPs in all other Instr constructions. However, even here some referents are more prototypical than the others. For example, the referents of the Instr NPs in (61), (62), and (64) are more prototypical because they evoke very basic scenes from everyday human experience, where a human referent uses a physical object from the outside world as an implement in the action. The referents in these scenes are perceived to be related to each other in time and space through their concomitant involvement in the action described by the verb and their tandem work with the human referent. In these relations, the emergence of the referent of the Instr NP is within and at the human referent’s discretion, and as such, the Instr referent is envisioned as accompanying and facilitating the action performed by the human referent. Moreover, the referent of the Instr NP molotok ‘hammer’ in (61) is probably the most prototypical, since its perceived role of instrument fully overlaps with its most 64 prototypical ontological function.14 On the other hand, the referents of the Instr NPs in (63) and (66) are less prototypical because while it is implied by the Instr that the human referents and the referent of the Instr NP “bus” are spatially and temporally adjacent in the course of the motion described by the verb, they do not work in tandem, and the Nom human referents are envisioned as experiencing rather than facilitating the motion. Additionally, all referents of the Instr NPs in this grouping are collateral to the human referent and the action described by the verb because they mediate rather than instigate the motion. 4.2.1.2 Referents of Instr nouns after verbs of controlled motion A special place in this grouping is occupied by referents of the Instr NPs with verbs of controlled multidirectional motion. While razmaxivat´ ‘brandish’ in (64) requires an Instr complement, other verbs of controlled motion in Russian (e.g., vertet´ ‘twist, twiddle,’ vraščat´/ krutit´ ‘rotate, spin’ dvigat´ ‘move,’ dergat´ ‘pull, jerk’ kačat´ ‘rock, nod,’ ševelit´ ‘move,’ stir,’ trjasti ‘shake,’ voročat´ ‘roll over/ turn’) can take either an Instr or an Acc complement. When the complement is expressed by a noun denoting a body part, it invariantly occurs with the Instr. However, when the complement is expressed by a noun whose referent is not a body part, it is marked by either the Acc or the Instr, as in the following examples: 14 Empirical data from infant and children studies suggest that instrument is a robust cognitive category (see, for example, Tomasello 1987, Biro & Leslie 2007, Hofer, Hauf & Aschersleben 2005, Jovanovic et al. 2007, Träuble & Pauen 2007, Henrik & Csibra 2015, inter alia). 65 (67) Policejsk-ij trjas svo-ju furažk-u. policeman-NOM.MASC.SG shook his-ACC.FEM.SG cap-ACC.FEM.SG ‘The policeman shook his cap. (68) Policejsk-ij trjas svo-ej furažk-oj. policeman-NOM.MASC.SG shook his-INSTR.FEM.SG cap-INSTR.FEM.SG ‘The policeman waved his cap. The differences in the case marking in (67) and (68) activate different conceptualized scenes. Although in both scenes furažka ‘cap’ is assumed to be held in the hands, because the causative motion of shaking or waving is generally associated with hands rather than other part of the body, the Acc in (67) implies that the cap is conceptually distanced from the human referent, whereas the Instr in (68) indicates that the cap is contiguous with him. Moreover, because the Instr suggests that the cap is collateral to the motion described by the verb, it is envisioned that the cap in (68) facilitates the shaking. Conversely, the Acc in (67) implies that the cap is affected by the action. Thus, the motion executed with the help of the cap in (68) is interpreted metonymically as “drawing someone’s attention”, as opposed to the literal “causing the cap to move rapidly and jerkily in multiple directions” in (69). In a similar fashion, the referents of the Instr NPs in trjasti spravk-ami ‘wave documents in someone’s face,’ lit. ‘shake documents-INSTR’ or trjasti udostovereni-em ‘wave an ID in someone’s face,’ lit. ‘shake ID-INSTR’ are easily recognized as metonyms for ‘providing written evidence’ or ‘threatening.’ The specific interpretation depends on 66 the context. In a broader sense, such readings are possible because of the well-established cultural correspondences between the form and the meaning of co-speech gestures. A demonstrative, iterative back and forth motion with a hand-held accessory is commonly understood as “drawing someone’s attention”, or “providing written evidence”, or “threatening”. Similarly, the referents of the Instr NPs in dergat´ napil´nik-om, lit. ‘tug rasp- INSTR,’ dergat´ igolk-ami, lit. ‘tug/pull needles-INSTR,’ and dergat´ smyčk-om, lit. ‘tug/pull bow-INSTR’ are understood metonymically as filing, sewing, and playing a violin (or other stringed instrument). Conversely, their Acc counterparts are perceived literally as a caused motion performed on a physical object and therefore as tugging and pulling. 4.2.2 Referents of Instr nouns with verbs of government and possession In the following examples the referents of the Instr nouns are governed or owned by another referent either literally or metaphorically, as in (71): (69) Mam-a vladeet firm-oj v Moskve. mom-NOM owns company-INSTR.FEM.SG in Moscow ‘Mom owns a company in Moscow.’ (70) Otec-Ø rukovodit fabrik-oj v Moskve. father-NOM manages factory-INSTR.FEM.SG in Moscow ‘Father manages a factory in Moscow.’ 67 (71) Roditel-i ne vladejut inostrann-ymi jazyk-ami. parents-NOM.PL not own foreign-INSTR.MASC.PL languages-INSTR.MASC.PL ‘(My) Parents do not know/speak foreign languages.’ 4.2.2.1 Some diachronic issues It is noteworthy that diachronically verbs from this grouping used to behave differently. For example, verbs of government, as in (70), historically required an Acc complement. The Instr did not appear with these verbs until the 16th century, and then it took two more centuries for the Instr to oust the Acc from these constructions. Conversely, verbs of possession, as in (69) and (70), are attested with Instr complements already in the earliest EES manuscripts. The original differential case marking of complements after verbs of government and possession suggests that they were conceptually different from each other. These contrasts can be explained by the intuition that possession presupposes exploitation of the owned referent at one’s own discretion. Government, on the other hand, is not so much about the exploitation of one referent by another as it is about the leading and guiding of one referent by another by rule of authority. 4.2.2.2 Conceptual similarities in CSR The conceptual commonalities between verbs of government and possession in CSR are captured by their Instr complements. The Instr indicates spatial and temporal contiguity between the governor/possessor and the referent that is being 68 governed/possessed. Moreover, the contingency relation between the referents is understood in these constructions as mutually exclusive, since the absence of one referent excludes the occurrence of the other. Thus, both government and ownership are established only in relation to another referent. Referents of the Instr NPs with verbs of government share similarities with the most prototypical referents of the Instr NPs from the previous grouping in subsection 4.2.1. These similarities arise from the fact that in these constructions the Nom human referent and the referent of the Instr NP work in tandem. Thus, in Ivan rubil drova topor- om ‘John chopped wood (with) ax-INSTR,’ the action described by the verb is executed, to different extents, by both John and the ax. Similarly, because managing a factory in (70) implies managing the work of its employees, it is assumed that the manager and the employees work as a team. However, since the work of the employees is managed, they are only indirectly involved in the action described by the verb, and therefore facilitate rather than execute the managing. The referents of the Instr NPs with possession verbs are linked to the most prototypical referents of the Instr NPs via the notion of “control”. Thus, to the same extent that John exercises control over the ax in (61), the Nom referent mama ‘Mom’ in (69) has control over the company that she owns. 4.2.3 Referents of Instr nouns as accidental and involuntary performers In the following examples, while the referents of the Instr NPs carry out the action described by the verb, the Instr underscores their circumstantial role: 69 (72) Dom-Ø stroit-sja raboč-imi. house-NOM.MASC.SG build-PASS.SG workers-INSTR.PL ‘The house is being built by workers.’ (73) Kryš-u snesl-o vetr-om. roof-ACC.FEM.SG blew.off-NEUT.SG wind-INSTR.MASC.SG ‘The roof was blown off by the wind.’ The referents of the Instr NPs in the examples (72) and (73) are close to the prototype because their role in the situation is envisioned as accidental. Note that the role of the hammer in (61) is perceived to be accidental because it is contingent on the intention and volition of the human referent John to use the hammer as an instrument. The indirect and accidental involvement of the referents of the Instr NPs in the action described by the verb in (72) and (73) is underscored, respectively, by the passive morpheme -sja in (72) and the impersonal morpheme -o in (73). Note also that while the referents of the Instr NPs in (72) perform the action described by the verb, the Instr implies that they are not the instigators and that they work subject to someone else’s will. Similarly, the referent of the Instr NP “wind” in (73) is not the instigator but rather the cause of the event described in the utterance. 4.2.4 Referents of Instr nouns as causes or stimuli In the examples below the referent of the Instr NP is perceived as the cause of a particular state or a stimulus of a perceptive or emotional response in another referent: 70 (74) Det-i bolejut gripp-om. children-NOM be.ill -PL flu-INSTR.MASC.SG ‘The children are ill with flu.’ (75) Posle poxoda det-i paxnut dym-om. after camping children-NOM smell-PL smoke-INSTR.MASC.SG ‘After camping the children smell like smoke.’ (76) Mam-a uvlekaet-sja sadovodstv-om. Mom-NOM is.keen.on-REFL.SG gardening-INSTR.NEUT.SG ‘Mom is keen on gardening.’ In (75) and (76), dym ‘smoke’ and sadovodstvo ‘gardening’ are perceived as stimuli. The response to these stimuli is inexplicitly conveyed by the finite verbs paxnut ‘(they) smell’ and uvlekaetsja ‘is keen on’ and is understood, respectively, as ‘odor’ and ‘passion/interest.’ Response is contiguous with and contingent on the stimulus. In (74) the referent of the Instr NP gripp ‘flu’ is understood as causing physical suffering and discomfort in the human referents. The similarity among the conceptualized scenes in (74), (75) and (76) is that in all three portrayals, the referent of the Instr NP arouses a specific perceptive, emotional, or physical experience. The differences among these scenes emerge from the different perceptions of the stimulus, which depend on the kind of response it evokes. Thus, paxnut´ ‘smell’ in (75) and uvlekat´sja ‘be keen on’ in (76) 71 imply a neutral or positive response, whereas bolet´ ‘be ill (with)’ in (74) is associated with a negative physiological reaction. The referents of the Instr NPs in examples (74) through (76) are less prototypical than the referent of the Instr NP in Ivan rubil drova topor-om ‘John chopped wood (with) ax-INSTR’ in (61). This is because the situation portrayed in (61) involves at least three referents, two of which, John and the hammer work in tandem; also, the referent of the Instr NP ‘hammer’ is understood as facilitating the action carried out by John, and the action itself is geared towards a third referent. Conversely, the verb bolet´ ‘be ill with’ in (74) and the verb paxnut´ ‘smell’ in (75) describe states rather than actions and these states are contained within the referent that undergoes it. In a similar fashion, uvlekat´sja ‘be keen on’ which literally means ‘interest oneself (in),’ denotes the action that is directed toward the subject referent. At the same time, the referents of the Instr NPs in (74) through (76) share the sense of instrumentality with the most prototypical referents of the Instr NPs. More specifically, flu, smoke, and gardening are contiguous with and collateral to the human referent in such a way that they are not only envisioned, respectively, as the cause of the illness or the stimuli of the emotional response but also as metaphorical conduits of it. 4.2.5 Referents of Instr nouns as spatio-temporal continuity and entirety While in the examples (77) through (80) below the referents of the Instr NPs differ from each other in terms of the role that they are envisioned to have in the situation described by the verb, they can be grouped together on cognitive grounds, particularly through the notions of continuity and entirety. 72 Each of the referents of the Instr NPs in (77) and (78) is discerned as a continuous pathway that along its course coincides with the motion. The referent of the Instr NP in (79) is perceived as a non-discrete unit of measure, while the referent of the Instr NP in (80) is envisioned as a non-intervening course of time during which the action described by the verb unfolds. Consider the following examples: (77) Vs-ju dorog-u my šl-i les-om. all-ACC way-ACC we-NOM walked-PL woods-INSTR.MASC.SG ‘We walked through the woods the whole time.’ (78) U Ivana krov´ pošla nos-om.15 by John-GEN blood-NOM.FEM.SG went nose-INSTR.MASC.SG ‘John had blood gushing (through) the nose’ (79) Ivan-Ø pil vin-o stakan-ami. John-NOM drank wine-ACC.NEUT.SG glasses-INSTR.PL ‘John drank wine glassful after glassful.’ (80) Zim-oj my ezdil-i v Moskvu. winter-INSTR.FEM.SG we went-PL into Moscow ‘In winter we went to Moscow.’ 15 Note that (78) is a lexicalized expression as it can only appear with either nos ‘nose’ or gorlo ‘throat’. 73 Note also, that the referents of the Instr NPs in this grouping function as adverbials. More specifically, while morphologically they are still nouns (e.g., they fully retain their declension paradigm), semantically and syntactically they behave like adverbs, that is, they are optional, they serve to specify the circumstances of the verbal or sentential referent, and they are semantically restricted to specify palace, manner, and time. 4.2.5.1 Continuous pathways The Instr in (77) and (78) above implies that the referents of the Instr NPs les-om ‘(through) woods-INSTR’ and nos-om ‘(through) nose-INSTR’ are contiguous with the motion described by the verb and therefore are a pathway that along its course coincides with the motion. Note also that while the referent of the Instr NP in (77) is a literal pathway, the referent of the Instr NP in (78) is a metaphorical one. 4.2.5.1.1 Literal pathways Let me first address the example in (77). The closest equivalent to this Instr construction is the prepositional Dat construction po les-u ‘along/in the woods-DAT.’ However, the Dat construction does not convey precisely the meaning expressed in (77). This is because the referent of the Instr NP is associated with a purposeful motion, whereas the Dat referent is not. Recall that the notion “purpose” is evoked in the most prototypical scenes involving referents of the Instr NPs. More specifically, an entity 74 becomes an instrument or means only when the human referent has a purpose to use it as such. Compare the following examples in this respect: (81) ??My guljal-i les-om. we-NOM strolled woods-INSTR.MASC.SG ?? ‘We strolled through the woods.’ (82) My guljal-i po les-u. we-NOM strolled along woods-DAT.MASC.SG ‘We strolled in the woods.’ The example in (81) is infelicitous because “stroll” is typically associated with walking in a casual, leisurely way, while the Instr presupposes purposeful motion. The notion “purpose” emerges in (77) through a chain of metaphorical extensions. Thus, the Instr implies that the referent of the Instr NP les-om ‘(through) woods-INSTR’ coincides with the motion along its course, which makes the motion uninterrupted and progressive. Progression is typically associated with onward motion, as is achieving a goal. Significantly, Raxilina and Tribušinina (2011) observe that the Instr occurs only in instances where the function of an area as a pathway is established by the subject of the motion rather than being predetermined by the topographic features. This explains why in CSR, such expressions as ?plyt´ rek-oj ‘travel/sail (by) river-INSTR.FEM.SG’ or ?idti ulic-ej ‘walk (along/down) street-INSTR.FEM.SG’ are infelicitous. A river is a body of water intended for sailing/swimming, a street is a public road designed for walking, 75 whereas woods, as a natural area covered with trees is not topographically predetermined for walking or driving. That les ‘woods-Instr’ becomes a pathway is contingent on the intention of the subject referent to use it as such. In this sense, the intention is understood metaphorically as purpose. 4.2.5.1.2 Metaphorical pathways Let me now turn to the example in (78), which is reproduced here as (83). (83) U Ivana krov´ pošla nos-om. by John-GEN blood-NOM.FEM.SG went nose-INSTR.MASC.SG ‘John had blood gushing (through) the nose’ The Instr indicates that the blood in (83) undergoes a continuous motion. That the motion is necessarily continuous is supported by the unacceptability of the example in (84) below: (84) ??krov´ pošla kolen-om. blood-NOM.FEM.SG went knee-INSTR.NEUT.SG ??‘blood gushed through (the) knee.’ The marginality of the example above arises from the fact that anatomically a knee is a joint and its inside space is solid rather than hollow, which means that a knee cannot be a pathway for blood or any other liquid substance. However, because the Instr suggests 76 contiguity between the referent of the Instr NP and the motion, nos-om ‘through nose- INSTR’ in (83) is envisioned as a way that along its course cooccurs with the motion and serves as its path. The closest equivalent for the Instr construction nos-om ‘through (the) nose- INSTR’ in (83) is the prepositional Acc construction čerez nos-Ø ‘through (the) nose-ACC.’ However, while the preposition čerez ‘through, past, over’ also implies a motion through space and can be a rough substitution for Instr nos-om, in the infelicitous utterance ?krov´ pošla čerez nos-Ø ‘blood gushed through nose-ACC,’ the nose is perceived as a discrete space interval rather than a continuous pathway. Importantly, the referent of the Instr NP nos-om ‘through (the) nose-INSTR’ in (83) suggests an intensive, heavy bleeding, whereas čerez nos-Ø ‘through (the) nose-ACC’ does not have such associations. This is because the Instr indicates a complete contiguity not only between the referent of the Instr NP and the motion but also between the referent of the Instr NP and the subject referent. More specifically, the blood is envisioned as covering the pathway “nose” entirely. Consider the following examples in this regard:16 (85) Ivan-Ø pil vin-o stakan-ami. John-NOM drank wine-ACC.NEUT.SG glasses-INSTR.PL ‘John drank wine glassful after glassful.’ 16 Note that because I examine the examples in (85) and (86) here, I will not analyze them later in the section. 77 (86) Ivan-Ø el ikr-u ložk-ami. John-NOM ate caviar-ACC.FEM.SG spoons-INSTR.PL ‘John ate caviar by the spoon.’ (87) *Ivan-Ø pil vin-o dv-umja stakan-ami. John-NOM drank wine-ACC.NEUT.SG. two- INSTR glasses-INSTR.PL *‘John drank wine by two glasses.’ In the examples above, the referents of the Instr NPs in (85) and (86) are envisioned as units of measure. These units are discerned in their entirety, and both referents of the Instr NPs refer to capacity rather than to a discrete quantity: this is supported by the ungrammaticality of (87). Thus, pit´ vino stakan-ami lit., ‘drink wine glass-INSTR’ or est´ ikru ložk-ami lit., ‘eat caviar spoon-INSTR’ mean, respectively, ‘drink a lot of wine’ and ‘eat a lot of caviar.’ The addition of the numeral makes example in (87) ungrammatical. Note, however, that while the example in (88) below is acceptable, here the referent of the Instr NP is no longer perceived as a unit of measure: (88) Ivan-Ø el ikr-u dv-umja ložk-ami. John-NOM ate caviar-ACC.FEM.SG two-INSTR spoons-INSTR.PL ‘John ate caviar with two spoons.’ 78 In this way, the referent of the Instr NP in (88) becomes an instrument of eating. At the same time, the example in (87) is marginal both in Russian and English because of the inability of “two glasses” to be reanalyzed as an instrument. A glass is not a typical ontological instrument but rather a container, and because the semantic frame of the verb drink, contrary to eat, does not require an instrument of drinking, a glass cannot be perceived as such. Since the semantic extension “container –> instrument” is barred, the glass is understood as a unit of measure. However, the example is still unacceptable because the numeral “two” implies a specific discrete quantity, while a unit is traditionally understood as an entirety. 4.2.5.2 Referents of Instr nouns as a course of time In examples (88) through (91) below the referent of the Instr NP is discerned as a course of time during which the action described by the verb unfolds: (88) Utr-om deti smotrjat televizor-Ø. morning-INSTR.NEUT.SG children-NOM watch television-ACC.MASC.SG ‘In the morning the children watch television.’ Traditionally the Instr as in (88) above is referred in the literature as a “temporal” Instr. Note, however, that the temporal reference is contributed by the noun referent and not by the Instr per se. 79 One striking feature of (88) and similar examples below is that the referent of the Instr NP is exclusively expressed in them by nouns denoting either seasons or parts of the day. Compare the following examples: (89) Let-om deti ne xod-jat v školu. summer-INSTR.NEUT.SG children-NOM not go-3PL to school ‘In summer the children do not go to school.’ (90) *a. Avgust-om deti ne xod-jat v školu. august-INSTR.MASC.SG children-NOM not go-3PL to school ‘In August the children do not go to school.’ b. V avgust-e deti ne xod-jat v školu. august-LOC.MASC.SG children-NOM not go-3PL to school ‘In August the children do not go to school.’ (91) *?a. Polnoč´-ju my ne smotrim televizor-Ø. midnight-INSTR.FEM.SG we-NOM not watch television-ACC.MASC.SG ‘At midnight we do not watch television.’ b. V polnoč´-Ø my ne smotr-im televizor-Ø. midnight-ACC.FEM.SG we-NOM not watch television-ACC.MASC.SG ‘At midnight we do not watch television.’ 80 I propose that the unacceptability of (90a) and (91a) stems from the semantic incompatibility of the referents “August” and “midnight” with the meaning of the Instr. Recall that in instances where the referent of the Instr NP functions as an adverbial, it is entirely concomitant with the action described by the verb and portrays continuous, non- intervening motion. Recall also examples (85) and (86) above in which the referent of the Instr NP is discerned as a unit and cannot be modified by a numeral. While months are generally understood as units of time, these units are divisible and have clear-cut boundaries. More specifically, a month is a period of time that lasts as long as the motion of the moon, which roughly equals 29 days. Similarly, a week, while also a unit, is discrete and divisible. Nouns denoting parts of the day such as midnight and midday, which cannot occur with the Instr in Russian, are discrete because they refer to specific points in time. Conversely, seasons are non-discrete, in that they are understood as characteristic patterns of weather and daylight hours rather than particular months. Note also that the definition of a season is culture-specific. Thus, while the Georgian calendar is arranged on a four-season basis, some cultures are known to have more than four or only two seasons. Similarly, parts of the day such as day, morning, evening, and night have fuzzy boundaries because the exact time when they begin and end depends on the geographical location and varies through the year. In this way, referents denoting seasons and parts of the day that are not associated with specific points in time and are compatible with the notion of continuity and entirety expressed by the Instr. 81 4.2.5.2.1 Diachronic evidence That only referents denoting non-discrete units of time can occur with the Instr is also substantiated by the diachronic evidence. In OCS and EES manuscripts, the Instr with a temporal reference is attested in two different senses. In one, the referent of the Instr NP was viewed as a quantity of time necessary for the action described by the verb to unfold, as illustrated, respectively, by the following OCS and EES examples: (92) OCS i tr-ьmi dьn-ьmi sǫzdati jǫ. and tree-INSTR.PL days-INSTR.PL create it-ACC.FEM.SG ‘and build it in/for three days.’ [Matthew 26: 61] (93) Russian Church Slavonic (RChSlav)17 i dn-em odn-emь peremčali devjanosto verstъ. and day-INSTR.MASC.SG one-INSTR.MASC.SG raced ninety versts ‘and raced ninety versts in one day.’ [Life of Avvakum, 170] Constructions such as these above answered the question of how long it took for something to happen or someone to accomplish the action. However, this meaning of the Instr did not survive, going out of use in the 17th century (Bernštejn 1958: 359–360). In 17 Russian Church Slavonic is a later post-OCS Church Slavonic displaying East Slavic phonological dialect features. 82 CSR this Instr is replaced by the prepositional construction za + Acc, eg., za čas-Ø ‘in an hour-ACC.MAS.SG, za den´-Ø ‘in a day-ACC.MASC.SG,’ etc. In its other temporal meaning, which is the only meaning left in CSR, the OCS and EES a referent of the Instr NP was understood to be a course of time during which the action unfolds. The idea behind this Instr meaning is not how much time it takes for the action to develop or at what specific point in time the action happened but, instead that the action is envisioned as unfolding together with the course of time and occupying it entirely. I maintain that in this regard the referents of the Instr NPs in (88) and (89) above share the same cognitive premises as the referents of the Instr NPs denoting non- intervening motion (they were discussed in 4.2.5.1.2 above): continuity and entirety. This view finds support in Szucsich (2002: 9), where it is argued, albeit on different grounds, that temporal Instr expressions are unbounded with respect to time and do not delimit the temporal structure of the event. 4.2.6 Referents of Instr nouns as literal or metaphorical capacities Although traditionally, examples like (94) through (98) below are treated separately, I group them together because, as I will demonstrate in the following paragraphs, one of the features that they share is that the referent of the Instr NP in them is perceived as a capacity. In example (94) this capacity is metaphorical; in the rest of the examples it is literal. Consider the following: 83 (94) Maš-a tri dnja revela belug-oj. Mary-NOM three days cried beluga-INSTR.FEM.SG ‘Mary was bellowing for three days like a whale.’ (95) Maš-a byl-a medsestr-oj. Mary-NOM was-FEM.SG nurse-INSTR.FEM.SG ‘Mary was a nurse.’ (96) Maš-a stanet medsestr-oj v buduščem. Mary-NOM become-FUT nurse-INSTR.FEM.SG in future ‘Mary will become a nurse in the future.’ (97) Maš-u naznačil-i glavn-oj medsestr-oj. Mary-ACC appointed-3PL head INSTR.FEM.SG nurse-INSTR.FEM.SG ‘Mary was appointed a head nurse. (98) Maš-a prišl-a domoj grustn-oj. Mary-NOM came-FEM home sad-INSTR.FEM.SG ‘Mary came home sad.’ Another shared peculiarity is that in all the examples above except for (94), the Instr is a relatively recent innovation. In examples (95), (96), and (98), the Nom was the original case marking of the predicative nominal, but in CSR both the Nom and the Instr 84 are grammatical.18 With verbs of naming, as in (97), the original case marking of the complement was the Acc, but in contrast to (95) and (98), only the Instr is acceptable with these verbs in CSR. Moreover, where the variation in case marking exists in CSR, it is motivated by the differences in meaning between the Nom and the Instr. 4.2.6.1 Referents of Instr nouns in similative constructions Consider more examples with referents of the Instr NPs in similative expressions: (99) Poezd-Ø letit strel-oj. train-NOM.MASC.SG flies arrow-INSTR.FEM.SG ‘The train is flying like an arrow.’ (100) Ona smotrit na menja volk-om. she-NOM stares at me-ACC wolf-INSTR.MASC.SG ‘She stares at me like a wolf.’ Potebnja (1888: 500) and Xodova (1958: 183) suggest that the similative Instr is semantically related to the Instr in even older and now-obsolete constructions of 18 Example (98) is called a “depictive construction”. “Depictive” is the term first suggested in Halliday (1967: 63) for English constructions such as I like to drink my coffee cold, and which was made widely known by Jackendoff (1990). Although depictive constructions lie beyond the scope of this dissertation, in brief, the idea behind the Instr in these constructions is that the property expressed by the referent of the Instr NP is co-temporal with the event introduced by the lexical verb and because of this contiguity with the event, the referent of the Instr NP also refers to the manner of the motion (but see, for example, Philip 2001 and Richardson 2001 for different opinions). 85 transformation (tvoritel´nyj prevraščenija), which portrayed human referents’ transformations into various animals. In CSR, the referent of the similative Instr is understood metonymically as the manner of motion. Thus, revet´ belug-oj ‘bellow like a whale-INSTR’ in (94) means ‘cry profusely and loudly,’ letet´ strel-oj ‘fly like an arrow-INSTR’ in (99) means ‘moving very fast,’ smotret´ volk-om ‘stare at someone like a wolf-INSTR’ in (100) means ‘look angrily, scowl.’ I propose, however, that besides referring to the manner of motion, the referent of the Instr NP in these examples is also perceived as a capacity, which emerges when the subject referent metaphorically adopts some of the most salient features of the referent of the Instr NP and acts in the manner associated culturally with a given referent of the Instr NP. Thus, when a beluga whale cries, it cries deeply, loudly, and profusely; when an arrow flies, it moves very fast; when a wolf stares at someone, its glare looks angry and fierce. Note also another interesting feature about the referents of the Instr NPs in the examples above. Thus, while the referents of the Instr NPs belug-oj ‘like a whale-INSTR’ in (94), or strel-oj ‘like an arrow-INSTR’ in (99), or volk-om ‘like a wolf-INSTR’ in (100) are the properties of the corresponding events (crying, flying, and staring) because they refer to the manner of the motion, they are also the properties of the subject referents in relation to these events. This fact has two important implications. First, since the capacities belug-oj, strel-oj, and volk-om are instantiated through the subject referent’s action, the notion “acting/ performance” underlies the role of capacity. Second, capacity can be defined as a property which unfolds contiguously with the event. As I will 86 demonstrate in the following paragraphs, the occurrence of referents of the Instr NPs in predicative constructions and with verbs of naming is motivated by the same semantic principles. While the referents of the Instr nouns in similative constructions are less prototypical than instruments and means, they are related to these and other referents of the Instr NPs via the notion of contiguity with and contingency on the motion. 4.2.6.2 Referents of Instr nouns with verbs of naming As mentioned above, the Instr was a late innovation with naming verbs (e.g., OCS and EES naricati ‘name something/someone as,’ postaviti ‘appoint someone as’, sotvoriti ‘turn someone/something into,’ vzęti ‘take someone as,’). The pattern of case marking the complements of these verbs with the Acc was a calque from equivalent Ancient Greek constructions that dominated until the second half of the 13th century. Consider the following OCS example, in which the Acc with the second complement reflects the original Ancient Greek pattern: (101) Sъtvoriǫ va lovьc-a člov ěk-omъ. make-1SG.FUT you-ACC.DU catchers-ACC.DU men-DAT ‘I will make you both catchers of people (souls).’ (translation mine – vt) [Codex Zographensis, Matthew IV: 19] In the second half of the 13th century, verbs of naming began to appear in written East Slavic secular documents with the Instr in place of the second Acc. This change may 87 reflect the intention of chroniclers and scribes to eliminate semantic ambiguity which arose when verbs of taking, bringing, and having were used in place of a naming verb, as illustrated by the RChSlav example below: (102) Pojaša Mstislav-a knjaz-ja sobe. took.3PL Mstislav-ACC prince-ACC self.DAT [Ipat´evskaja Letopis´] One possible English translation of (102) is that ‘they took and brought prince Mstislav to their place,’ that is, at the time Mstislav was taken, he already was a prince. The other possibility, which is the intended meaning, is that ‘they took Mstislav and appointed him as their prince’ (with the reflexive pronoun sobe in this meaning being a Dat of possession). Such semantic ambiguity was less likely with verbs of naming because in addition to the referent who names or appoints and a referent who is named or appointed, these verbs require a third referent which is the kind of appointment or name per se. However, with verbs of taking, bringing, having, and the like, which were widely used in such constructions, the second Acc caused ambiguity. I argue that this is because the semantic frame of these verbs requires only two referents for a felicitous reading, i.e., a referent who takes/brings/has and a referent who is taken/brought/had. Because the Acc implies that the referent of the Acc noun is affected by the action described by the verb, ‘prince- ACC’ in (102) can be understood as affected by the action in the same way as the referent of the proper noun ‘Mstislav-ACC.’ Syntactically, the identical case marking on both complements suggests that knjaz-ja ‘prince-ACC’ is an attributive modifier of Mstislav-a 88 ‘Mstislav-ACC’. From a cognitive perspective, the referent of the Acc noun and its Acc modifier are perceived as one whole, which produces the reading ‘they took and brought prince Mstislav to themselves.’ The Instr, on the other hand, could be a good disambiguating tool for several reasons. First, the appointment is contingent on the incentive of the subject referent in (97) and (102) inasmuch as the role of instrument is contingent on the intention of the subject referent to use an artifact as such. Second, by the time verbs of naming began to occur with Instr complements, the Instr was already attested in adverbial expressions like the ones in (94), (99), and (100), as evidenced by RChSlav manuscripts (Potebnja 1888, Xodova 1958). In constructions with naming verbs, the role of the referent that designates the appointment is semantically congruent with the role of the referent of the Instr NP in similative constructions. Thus, to nominate or appoint means to propose for or establish the referent in an office or post, which is to assign a legal capacity. Moreover, just as the referents of the Instr NPs in similative constructions denote both a capacity and a mode of acting (manner), the naming or appointment, as a capacity, presupposes a particular mode of acting or performance, which is understood as fulfilling the duties associated with the naming or appointment. 4.2.6.3 Referents of the Instr nouns in predicative constructions Haspelmath & Buchholz (1998: 322) point out that cross-linguistically, expressions of comparison, similar to those discussed in subsection 4.2.6.1, are formally related to expressions of role, function, or a life stage. This is illustrated by the following Estonian examples with the Ess: 89 (103) rong möödus välgu-na. train.NOM go.along.PAST.3SG flash-ESS ‘The train passed like a flash.’ (Lutkat & Hasselblatt 1993: 193) (21/104) NN on meie saadiku-na London-is. NN is our ambassador-ESS London-INESS ‘NN is our ambassador in London.’ (Lehiste 1972: 216) Szucsich (2002: 15–16) notes for Slavic that only in those languages where the Instr is attested in adverbial expressions (Russian, Belarusian, Ukrainian, Polish, Czech, and Slovak), the Instr also occurs in predicative constructions, and that conversely, in Upper Sorbian, Burgenland Croatian, and Slovenian the Instr does not occur in any adverbial expression nor is it attested in predicative constructions. It has been propounded in the literature (Potebnja 1888, Xodova 1960, Mrázek 1964, Klein, Joseph & Fritz 2017) that in Russian the Instr in predicative constructions has originated from adverbial similative constructions. However, no semantic motivation has been set forth for this phenomenon. In the following paragraphs, I will show the semantic relatedness between referents of the Instr NPs in similative constructions and referents of the Instr NPs in predicative constructions. 90 4.2.6.3.1 Diachronic evidence That the Instr in CSR predicative constructions has as its source the similative Instr in adverbial expressions is corroborated by the diachronic data. The earliest attested instances of the Instr with predicative nouns display lack of number agreement between the Instr NP and the grammatical subject, which is a feature of adverbs, and which is illustrated by the following example in secular EES chancery language: (105) EES Ta dv-a byl-a posl-ъmь u riže. those two-NOM.DU were-DU ambassador-INSTR.MASC.SG in Riga ‘Those two were ambassadors in Riga.’ (translation mine – vt) [Smolensk Treaty, 1229, 26–27] Note that the singular predicative noun posl-ъmь ‘ambassador-INSTR.MASC.SG does not agree in number with the dual subject ta dva ‘those two’. This sort of number misagreement is disallowed in CSR:19 (106) CSR Te dvoe byl-i *posl-om/ posl-ami v Rige. those two were ambassador-INSTR.MASC.SG/ ambassador-INSTR.MASC.PL in Riga ‘Those two were ambassadors in Riga.’ 19 The dual number is an obsolete category in CSR. 91 The same structural peculiarity was shared by the earliest attested instances of Instr complements with naming verbs. Consider the following: (107) EES (chancery language) I postavilъ nasъ opekalьnik-omь muž-emъ i ljud-emъ […] and put us-ACC.PL defender-INSTR.MASC.SG men-DAT and people-DAT ‘And (he) appointed us as defenders (lit. a defender) to men and people […]’ (translation mine – vt) [Gramoty Velikogo Novgoroda i Pskova, 1389] Note the lack of number agreement in (106) between the singular Instr complement opekalьnik-omь ‘defender-INSTR.MASC.SG’ and the plural direct object nasъ ‘us-ACC.’ In CSR the number agreement is mandatory in such constructions. The unusual absence of number agreement in (105) and (107) suggests that the Instr NPs function in these constructions as adverbials. This can be explained by the implication that acting in a capacity presupposes a mode of performance (in this instance, as defender). 4.2.6.3.2 Motivation for change Constructions with verbs of naming in EES probably adopted the Instr earlier than predicative constructions, because the Instr could deem to be a disambiguating tool, as supported by the fact that in CSR only the Instr is grammatical with these verbs. Predicative constructions, on the other hand, were devoid of ambiguity, as evidenced by 92 the coexistence of both the Instr and the original Nom in these constructions in CSR. However, the Instr might have been used in predicative constructions as a stylistic device to differentiate between capacities (legal competencies or religious affiliations) and other properties of the subject referent, as illustrated by the following RChSlav example: (108) Bě bo u Jaropolk-a žen-a Grekin-i, was-SG PRTCL by Jaropolk-GEN.MASC.SG wife-NOM.SG Greek.woman-NOM.FEM.SG bjaše byla černic-eju. had been nun-INSTR.FEM.SG ‘Jaropolk’s wife was Greek, (she) had been a nun.’ (translation mine – vt) [Laurentian Chronicle, 23b: 15–16] In the example above, the predicative noun denoting the subject referent’s origin is marked with the Nom, while the noun denoting her station occurs with the Instr. This example is often cited in the literature as evidence that the Instr occurred in predicative constructions and still functions as such in CSR to indicate that the property expressed by the Instr predicative nominal is temporary, as opposed to inherent when the property is expressed by a Nom predicative noun (see Borkovskij & Kuznecov 1965, Mozer 1994, Mixajlov 2012 for the diachronic state; Švedova 1980, Nichols 1981, Strigin & Demjanow 2001, Timberlake 2004, Krasovitsky et al. 2008 for CSR). While this interpretation is possible in some instances, it does not apply across the board. More specifically, EES manuscripts have instances in which the Nom predicative noun denotes a temporary, transient property, as illustrated by the following example: 93 (109) Feodos-ij prestavisja, i bystь Stefanъ igumen-ъ, Theodosius-MASC.NOM passed.away and was Steven Father-Superior-NOM a po Stefanъ Nikon-ъ. and after Steven Nikon-NOM ‘Theodosius passed away, and Steven became Father-Superior, and Nikon became Father-Superior after Steven.’ (translation mine – vt) [Nicon Cronicle, 104] It can be inferred from the example above that the Nom property igumen-ъ ‘Father- Superior-NOM’ is temporary and not permanent because both Steven and Nikon served as Father-Superior at different times. Moreover, the interpretation by the researchers of example (108) such that it is the Instr that implies that Jaropolk’s wife was no longer a nun might have been heavily misled by the use of the pluperfect form bjaše byla ‘had been.’ Also, example (108) is taken out of the context. Thus, in the chronicle, there are two other clauses immediately following the clause containing the Instr predicative noun černic-eju ‘nun- INSTR.FEM.SG’: 94 (110) bě bo privelъ jǫ otec ego AUX PRTCL brought her-ACC father-NOM his-GEN.MASC i vda jǫ za Jaropolka. and gave her-ACC for Jaropolk ‘And his father brought her (the nun) and made her marry Jaropolk.’ (translation mine – vt) [Laurentian Chronicle, 23b: 15–18] Note that EES (and OCS) manuscripts lack punctuation marks and sometimes it is hard to determine whether or in what way two adjacent clauses are connected. If we analyze example (108) without its immediate context, it might be tempting to conclude that the Instr was used to underscore temporariness of the property “nun”. However, the analysis can yield a different reading if we take into consideration the immediate context provided in (110). Importantly, a proper interpretation of the pluperfect form bjaše byla ‘had been’ can also be crucial for the analysis. Thus, the imperfective auxiliary bjaše shows that the past action/state is coordinated with some other situation in the past; the participle byla indicates that the action/state had begun prior to another past situation but is still pertinent. Note that both clauses bjaše byla černic-eju ‘had been nun- NSTR.FEM.SG’ in (108) and its immediate context bě bo privelъ jǫ ‘AUX (had) brought her-ACC.FEM’ in (110) are in the pluperfect, which means that both situations happened before the nun became Jaropolk’s wife. Most importantly, it also means that when Jaropolk’s father brought the female referent she was a nun. The Instr can be accounted for in this case as signaling the capacity of the Greek woman. It is possible to hypothesize further that the distinction between the Nom and the Instr predicative nouns in (108) might be the 95 difference between properties and capacities, respectively, rather than between inherent vs. temporary properties. Note also that because a nun is bound by vows of chastity, Jaropolk’s wife would simply be defrocked by breaking the vow of not marrying, therefore there seems to have been little to no need to mark the property “nun” as temporary. However, there might be a need to indicate what the Greek woman was before she was forced to marry Jaropolk. Additionally, while capacities, such as professional, religious, occupational, and other affiliations are acquired rather than inherent, a capacity may hold over a lifetime. Similarly, one can be Greek or of any other origin by birth, which is an inalienable property, or by naturalization, in which case the property in question is acquired rather than inherent. My proposal that the Instr might have occurred to signal that a given property is a capacity rather than a property perceived as temporary can be further substantiated by the fact that while in CSR both the Instr and the Nom are grammatical with predicative nominals in copular constructions, except for the present tense with a zero copula, utterance like ?Ciceron-Ø byl Tull-iem ‘Cicero-NOM was Tully-INSTR’ is semantically unacceptable. This is because Cicero and Tully are the names of the same referent, while the Instr implies that Tully was a role of Cicero and therefore Cicero and Tully are perceived as the names of two different referents. Note that the Nom is a non-relational case marking (e.g., Jakobson 1936, Losev 1982), which means that the only possible relation it evokes is the relation of its referent to itself, which is an equative relation and therefore is understood as identity rather than a role. 96 A capacity is always a role which has its purpose and function, and which is realized as such through action (in the sense of carrying out an action, duty, or task). A property, on the other hand, is an intrinsic quality or an essential characteristic which, by its nature, does not presuppose action. Consider in this respect the following example: ?(35/111) Pavel-Ø rabotaet vrač-om, no on bol´še ne praktikuet. Paul-NOM works doctor-INSTR.MASC.SG but he any.more not practice ‘Paul works as a doctor but he does not practice any longer.’ The unacceptability of (111) stems from the fact that the Instr implies that vrač-om ‘as doctor-INSTR’ is Paul’s capacity, while the second clause negates any medical practice, which entails that ‘doctor’ is not realized as a capacity but rather holds as a property. Thus, the capacity vrač-om ‘as doctor-INSTR’ is realized only when Paul practices medicine, inasmuch as belug-oj ‘like beluga.whale-INSTR’ in (94) is realized only when the subject referent acts as such. 4.2.6.3.3 Co-temporality vs. temporariness As has been established in the preceding paragraphs, the source of the Instr in predicative constructions is the similative Instr, whose characteristic feature is that its referent denotes both a capacity and a mode of acting. Because a given capacity is contiguous with and contingent on a specific mode of acting, and because manner modifies the motion described by the verb, the capacity develops together with the motion and therefore is co-temporal with the event introduced by the verb rather than 97 temporary. Note that I use the term “event” in a general sense, meaning “situation described by the verb in which things happen” and that events can be either long-lasting or short-lived. In predicative constructions the referent of the Instr NP has the same semantic function as the referent of the Instr NP in similative constructions, that is, it denotes a capacity that develops concomitantly with the time introduced by the verb. In similative constructions it is a lexical verb, while in predicative constructions it is the he copula that provides a spatio-temporal anchor which enables the realization of a given property as a capacity. A property which is contiguous with and contingent on time and space is a capacity. Note that co-temporality and temporariness are not the same phenomenon. While temporariness is about the quantity of time for which the property in question lasts, co- temporality is about the quality or mode of time in which the property holds in relation to the event. 4.2.6.3.3.1 Evidence from Polish and French That the referent of the Instr NP in predicative constructions does not imply temporariness can be supported at least in part by evidence from Polish, in which, unlike in Russian, the Instr can occur in the present tense. Polish has two copulas: the verbal copula być ‘be’ and the pronominal copula to ‘this.’ In predicative constructions, the verbal copula requires an Instr complement, whereas the pronominal requires a Nom one, as is exemplified by the following constructions: 98 (112) a. Marek jest muzyk-iem. Mark-NOM is musician-INSTR.MASC.SG ‘Mark is a musician.’ (Bondaruk 2014: 342) b. Marek to muzyk-Ø. Mark-NOM this musician-NOM.MASC.SG ‘Mark is a musician.’ (ibid.) The Nom in (112b) indicates that “musician” is Mark’s property, i.e., that he is a musician by training, but he may not be practicing music, whereas the Instr implies that “musician” is Mark’s capacity and the construction can be translated roughly as ‘Mark works as a musician.’ Although a capacity is generally a non-inherent property, one can work in a given capacity for years or even for a lifetime. If temporariness indeed was the meaning of the Instr in predicative constructions, then the Polish Instr construction in (113a) below would be infelicitous: (113) a. Warszaw-a jest stolic-ą Polsk-i. Warsaw-NOM.FEM.SG is capital-INSTR.FEM.SG Poland-GEN.FEM.SG ‘Warsaw is the capital of Poland.’ b. Warszaw-a to stolic-a Polsk-i. Warsaw-NOM.FEM.SG this capital-NOM.FEM.SG Poland-GEN.FEM.SG ‘Warsaw is the capital of Poland.’ 99 Following the general view that the Instr in predicative constructions signals temporariness, we must assume that in example (113a) Warsaw is only temporarily the capital of Poland, while in (113b) the property “capital” is constant. However, according to Wikipedia, Warsaw has been the capital of Poland since 1596. I was informed by two native Polish speakers in personal correspondence that the choice between (113a) and (113b) depends on what the speaker intends to communicate about Warsaw. The Nom simply states the fact that Warsaw is the capital of Poland, while the Instr is used when something else is going to be said about Warsaw, e.g., that it is also an ancient, beautiful, vibrant city, etc. In other words, the Instr is chosen when the speaker wants to underscore the role of Warsaw in the country or in the life of its inhabitants. In fact, it has been argued for Polish (Bondaruk 2014: 346–347) that the distinction “inherent-temporary”, which is proposed cross-linguistically for differentially- marked predicative nominals in two otherwise similar predicative constructions, and which in formal models corresponds to individual- and stage-level predicates (Milsark 1974, Carlson 1977, 1980) cannot accurately capture the distinction between the Nom and the Instr in predicative constructions. Similarly, Roy (2013: 47) maintains that in French, where in the absence of the indefinite article predicative nouns are traditionally analyzed as stage-level predicates, the long-existing “inherent-temporary” opposition does not fully appreciate the differences in meaning between constructions with differentially-marked predicative nouns (see subsection 2.5.2 above). 100 The Polish example in (113a) is in line with the proposal put forward in this chapter. It shows, particularly, that like the Instr in similative expressions, the Instr in predicative constructions implies that the property expressed by the Instr NP is a capacity of the subject referent. The capacity is not bounded in time but is rather contiguous with and contingent on it and therefore persists for as long as the time indicated by the copula. Similarly, note these two examples: (34/114) French a. ?Paul est médecin mais il ne pratique plus. Paul is doctor but he not practices anymore ‘Paul is a doctor, but he does not practice anymore.’ (Roy 2006: 31) b. Paul est un médecin mais il ne pratique plus. Paul is a doctor but he not practices anymore ‘Paul is a doctor, but he does not practice anymore.’ (ibid.) Roy (2006: 31) argues that (114a) is marginal because the second conjunct clause negates the activity that is implied by the bare predicative noun, while (114b) is felicitous, because the second conjunct clause negates the activity that is not implied by the predicative noun preceded by the article (see discussion in subsection 2.5.2 above). The unacceptability of (114a) and acceptability of (114b) can be accounted for by the implication that in (114b) that Paul is a doctor by training, but that he may not be practicing medicine, whereas the predicative noun in (114a) indicates that the subject 101 referent practices medicine. In this sense, the semantic dichotomy “inherent/permanent- non-inherent/temporary” is a priori asymmetrical in Russian, Polish, and French, as well as in German and Dutch, since it assumes the distinction between properties of the subject referent, while, as evidenced by the Russian, Polish, and French examples discussed in this subsection, the distinction holds between properties and capacities. That is, properties are atemporal, whereas capacities are co-temporal because they hold of the subject referent in relation to time and space, and spatio-temporal situations can be long- lasting or short-lived. 4.3 Conclusion This chapter has presented an alternative analysis of the Instr in Russian, proposing that that the Instr per se implies a set of cognitive, rather than semantic, relations that are perceived to exist among the referents in a situation described by the verb. The Instr shows how these referents are related to each other and to the action described by the verb in time and space, and in what way they affect each other. These relations are contiguity and contingency. The individual meanings (e.g., instrument, cause, pathway, among others, are realized in individual Instr constructions from the interaction between the verb meaning and the lexico-semantic properties of the noun referents. I have demonstrated that the meanings of the Instr can be organized in a semantically motivated fashion as a family with a radial rather than a hierarchical structure. Such organization assumes a relative prototypicality of some rather than the dominance of one meaning. While, in principle, grammatical categories can be organized 102 in either way, polysemous categories tend to display a family resemblance structure rather than have one core meaning. Thus, as noted earlier, among the three meanings A, B, and C, A may be related to B and B may be related to C, but A and C may or may not be related, or they may be related differently to each other than A and B. Applied to Russian, the family resemblance structure can account, for example, for the occurrence of the Instr in constructions with verbs of government. While in these constructions the referents of the Instr NPs are not accessories that help to implement the action described by the verb, they are linked to the most prototypical referents via the notion of “tandem work”. The referent of the Instr NP and the subject referent in constructions with verbs of government work as a team inasmuch as the instrument and the subject referent work together. Similarly, although stimuli and causes are not prototypical instruments, they share the sense of instrumentality with the latter, in that besides evoking or causing, they also mediate a kind of emotional or perceptive response in the subject referent. Significantly, it has been demonstrated that the Instr in predicative constructions is not only related to the Instr in similative adverbial constructions but also to the most prototypical meaning of the Instr, that is to instrument. The meaning “capacity” is linked to the meaning “instrument” through notions “function” and “purpose”. Moreover, function delimitates Instr capacities from Nom properties in predicative constructions, inasmuch as function tells apart instruments from other physical objects. This chapter has also shown that the semantic relatedness between seemingly random meanings can be further substantiated by diachronic data. Using evidence from early Slavic manuscripts, I have defined that the Instr in predicative constructions has as 103 its semantic source the similative Instr, whose referent not only denotes a mode of acting but also designates a capacity of the subject referent which emerges through the acting. In the same way, the referent of the Instr NP in predicative constructions designates a property of the subject referent in relation to time and space, which is perceived as a capacity. A capacity is a role which has its purpose and function, and which is instantiated as such through performance. A property, on the other hand, is an intrinsic quality or an essential characteristic which, due to its nature, does not presuppose performance and hence is atemporal. The capacity is not bounded in time but is rather contiguous with and contingent on it and therefore persists for as long as the time indicated by the copula, which reflects such characteristic features of the Instr as a family as continuity and entirety. 104 CHAPTER V INSTRUMENTAL CASE MARKING WITH PREDICATIVE NOUNS In this chapter I examine the semantic function of the Instr in copular constructions with predicative nouns. In particular, the chapter addresses some controversial issues concerning the ungrammaticality of the Instr after a zero copula, as well as the semantic restrictions on the use of the semi-copulative verb javljat´sja ‘be someone/something, serve as someone/something’ as the alternative to the copula byt´ ‘be’. I also briefly attend to the question of why the Instr is much more frequent than the Nom after a copula that is marked for tense and mood. The chapter also sketches the emergence and evolution of the Instr with predicative nouns. Finally, the chapter challenges some observations about the occurrence of the Instr with particular semantic classes of nouns and proposes that the Instr in predicative constructions reflects the same semantic principles that motivate the Instr in other Instr constructions in the language. 5.1 Variation in case marking in CSR As has been mentioned earlier in this dissertation, predicative nouns in Russian can occur with either the Nom or the Instr if the copula is required, as in the following examples:20 20 Recall that as evidenced by early manuscripts, until the second 13th century, predicative nouns were exclusively marked by the Nom, as illustrated by these RChSlav examples below: 105 (115) Petr-Ø byl/ budet arxitektor-Ø. Peter-NOM.MASC.SG was-MASC.SG /will.be architect-NOM.MASC.SG ‘Peter was/will be an architect.’ (116) Petr-Ø byl/ budet arxitektor-om. Peter-NOM.MASC.SG was-MASC.SG/will.be architect-INSTR.MASC.SG ‘Peter was/will be an architect.’ Generally, the variation in case marking is more salient in the past tense, while in the future tense, imperative and conditional mood, the Nom with predicative nouns is less frequent than the Instr (Krasovitsky et al. 2008). In accordance with the proposal set forth in this study that the referent of the Instr NP in predicative constructions not only designates a capacity of the subject referent, but also denotes a mode of acting, the frequency of the Instr in imperative constructions, such as for example, Bud´ umnic-ej! ‘Be a good boy/girl-INSTR!’ can be accounted for, at least in part, by the fact that imperative constructions canonically express directive commands. (i) Běsta bo lov´c-a. were-DU for fishermen-NOM.DU ‘For (they) both were fishermen.’ [Ostromir Gospel, 1056-1057] (ii) Bě bo u Jaropolka žen-a Grekin-i. was for by Jaropolk wife-NOM.FEM.SG Greek.woman-NOM.FEM.SG ‘For Jaropolk’s wife was Greek.’ [Laurentian Chronicle, 23b: 15-16] 106 Usually, such a command requests for a certain behavior, which can be broadly interpreted as a mode of acting. In the same way, the Instr can be motivated after the copula inflected for the conditional mood, such as for example, in Esli by ty byl umnic-ej, my by kupili tebe novyj telefon ‘If you-NOM were a good boy-INSTR, we would buy you a new phone.’ The conditional mood is used in Russian to refer to hypothetical situations, that is, situations that are contingent on the realization of other situations. Recall that the Instr always implies its referent contingency with other referents and with the action described by the verb. A purchase of a new phone hinges upon the behavior of the subject referent in the first clause, rather than on his inalienable personal characteristics. The most equal distribution of the Nom and the Instr in predicative constructions in the past tense can be attributed to the fact that because the past tense describes situations that have happened, that is, events that have taken place or states that have held, both properties and capacities are compatible with past tense contexts. Since a property is atemporal and therefore inalienable from the subject referent, it can be envisioned as such in past situations as well. 5.2 Copula in the present tense Although the overt copula byt´ ‘be’ generally does not occur in the present tense, sometimes the defective copula est´/sut´ ‘be-SG/PL’ is used in constructions expressing 107 generalizations, however, the Instr is disallowed in these contexts also.21 Additionally, the defective copula does not agree in number with the grammatical subject, as illustrated by the example below: (117) Det-i est´ det-i. Children-NOM.PL be-3SG children-NOM.PL ‘Children will be (lit. ‘are’) children.’ While East Slavic languages generally do not use an overt copula in the present tense, the zero copula is uncommon outside of the East Slavic branch.22 Thus, except for 21 The copula est´/sut´ ‘be’ is defective because in CSR these old present tense forms of buti ‘be’ are preserved only in the 3SG and 3PL: however, if this copula is used at all, it is mostly used in the 3SG even with the plural subjects. 22 In RChSlav and EES manuscripts, the use of the copula in the present tense was optional, and zero copula was as common a phenomenon as an overt copula (Lomtev 1956: 35). The fact that the zero copula is attested in early manuscripts as often as the overt copula raises the question of whether the copula was simply omitted in some cases or whether it was not structurally present in the EES sentence altogether. Bulaxovskij argues that in the present tense, the copula est´/sut´ ‘be-SG/be-PL’ was underlyingly absent and that in instances where it is attested in RChSlav and secular EES documents, it was influenced by the Old Church Slavonic literary tradition (1950: 378). Conversely, Karskij (1929: 17) and Istrina (1923: 65) argue that the present tense copula was an independent structural development of EES. Istrina (ibid.: 66) also argues that the zero copula in EES predicative constructions was simply omitted rather than being underlyingly absent. That in Russian, Ukrainian, and Belarusian, the copula is barred in the present tense except in certain contexts is thus linked to its gradual loss. Šaxmatov (1941: 179–180) maintains that because zero copula constructions are attested in the EES manuscripts, they must reflect the Proto-Indo-European (PIE) pattern in which constructions with a zero copula coexisted with constructions with the overt present-tense copula *esmi/esti. Šahmatov draws attention to Meillet’s (1906–8) observation that in Ancient Greek and Vedic Sanskrit, a zero copula was much more frequent than an overt one. It should be mentioned in this respect that the question about the structure of the basic nominal sentence has been a matter of debate not only for the East Slavic languages, but also for PIE in general. For example, Benveniste (1950) argued that the zero copula in PIE nominal sentences in the present tense is the result of ellipsis and that the underlying copula *esmi should be assumed. He supported this claim on the grounds that in tenses other than the present, PIE required an overt copula, as illustrated by the following Hittite sentence, ABU. I ̯A genzuu ̯alaš ešta ‘My father was merciful.’ Conversely, Meillet’s (1906–8), Lehmann (1974), Mendoza (1998), and Fritz (2003) maintain that zero copula nominal expressions constituted an independent type of nominal sentences in PIE and that when an overt copula was used, it was done for emphasis. Moreover, Fritz (2003) proposes that an overt 108 Russian, Ukrainian, and Belarusian, the rest of the Slavic languages use a copulative verb in the present tense. At the same time, the zero copula is not a rare phenomenon cross-linguistically. Benveniste (1971: 164) observes that the most common way to construct a predicative expression in languages is by the juxtaposition of two nominals. 5.2.1 No copula, no Instrumental case marking? That in CSR the Instr is disallowed after a zero copula can be explained by the fact that the referent of the Instr NP denotes a capacity, and capacity, as opposed to an inalienable property, requires a spatio-temporal anchor for its realization. The Nom is compatible with a zero copula because Nom referents in predicative constructions denote properties, and properties do not hinge on acting/performance in order to develop and therefore do not need a reference point in space and time. On the other hand, because the referent of the Instr NP denotes a capacity, which can only be realized through acting, which needs a spatio-temporal reference to unfold, when such a reference is not provided, copula in the present tense in PIE nominal sentences emerged as an adjustment to the basic structure of verbal clauses, which had always had a finite verb. Šaxmatov (1941: 179–180) states that following the PIE pattern, zero copula constructions in EES were understood as expressing general truths and unchanging situations, whereas constructions with an overt copula were marked for aspect and indicated that the property expressed by the predicative nominal held true for the present moment only. Šaxmatov further argues that in East Slavic languages, constructions with a zero copula have eventually ousted constructions with the overt copula in the present tense. This is supported by the fact that zero copula constructions were more common as far back as EES, and that in modern East Slavic languages the copula is not used in the present tense (1941: 179–180). Note that although in Russian linguistics scholarship the zero copula is attributed to the PIE inheritance, some studies propose that the zero copula was influenced by Finno-Ugric languages (Veenker 1967, Grenoble 2010). This claim, however, is not uncontroversial, since among the Finnic languages, many require an overt copula, e.g., Mordvin, Votic, Veps, Estonian, Finnish, Karelian, and Livonian (see, e.g., Lees 2015). 109 the occurrence of the Instr is barred. This can be corroborated by the evidence from colloquial Russian, where the Instr, although ungrammatical after a zero copula in CSR, is attested in utterances like this one below: (118) On zdes´ director-om. he-NOM here manager-INSTR.MASC.SG ‘He is a manager here.’ It is noteworthy that the Instr after a zero copula is acceptable only with locative modifiers, such as for example, zdes´ ‘here’ or u nas lit., ‘by us.’ Markman (2008: 205– 206) observes that in the absence of the copula, the Instr is incompatible with temporal modifiers, such as “always”, “usually” or “often”. Interestingly, even though the Instr can occur after some locative modifiers, they need to be strictly deictic or pronominal, as shown in felicitous example (119a) and the corresponding infelicitous variant (119b) below: (119) a. Dima zdes´/tam/ v étom sele predsedatel-em. Dima-NOM here/there/in this village chairman-INSTR.MASC.SG ‘Dima is here/there/in this village a chairman.’ b. ??Dima v Amerike taksist-om. Dima-NOM in America taxi.driver-INSTR.MASC.SG ‘Dima is a taxi driver in America.’ (Markman 2008: 205) 110 Nichols (1981: 125) mentions another restriction, i.e., that the referent of the Instr NP must denote a professional capacity: (120) ??Dima u nas/ zdes´ (naš-im) drug-om. Dima-NOM at our.place/here (our-INSTR.POSS) friend-INSTR.MASC.SG ‘Dima is our friend here.’ (Nichols 1981: 125) A narrow deictic or pronominal locative modifier locates the situation with respect to the speech act and thus establishes a spatio-temporal anchor, which allows the occurrence of the Instr. Note also that constructions like (119a) above describe a purpose that connects a given referent with a particular location through the referent’s specific function in relation to the locality. Usually, if a deictic locative modifier such as “here” or “there” is used, it refers to an institution (e.g, hospital) that provides certain services or an administrative body (e.g, council) that has a designated function in the community. The utterance in (119a) is felicitous because if a village has a local council it also has a chairman, and the latter is associated with specific duties. The semantic unacceptability of (119b) arises from the intuition that while “taxi driver” is Dima’s occupation, it is not an appointed capacity, unlike the occupation “chairman”, and therefore Dima has no particular function in relation to America. The explanation above can also shed light, at least partly, on the restriction that disallows from zero-copula constructions in colloquial Russian referents of the Instr NPs 111 which are not professional capacities. Thus, the marginality of drug-om ‘friend-INSTR’ in a zero-copula sentence may stem from the fact that while “friend” is a capacity, it is not a designated one. 5.2.2 Javljat´sja: The alternative to zero copula? In addition to colloquial expressions, only the Instr, and not the Nom, can occur in the present tense in CSR after the copula javljat´sja ‘be someone/something, serve as someone/something,’ as is shown by the examples below: (121) а. Fotograf-ija javljaetsja predteč-ej photography-NOM.FEM.SG serves-REFL precursor-INSTR.FEM.SG kinematograf-a. cinematography-GEN.MASC.SG ‘Photography is a precursor of cinematography.’ b. *Fotograf-ija javljaetsja predteč-a photography-NOM.FEM.SG serves-REFL precursor-NOM.FEM.SG kinematograf-a. cinematography-GEN.MASC.SG ‘Photography is a precursor of cinematography.’ The copulative status of javljat´sja is somewhat controversial in the literature. While traditionally javljat´sja is treated as a semi-copulative verb, some researchers 112 maintain that it is a true copula on par with the copula byt´ ‘be’. For example, Švedova argues that in predicative constructions, both javljat´sja and byt´ are devoid of any lexical meaning and serve only as a grammatical marker of tense and mood (1960: 414). Similarly, Grudneva (1958: 160) maintains that javljat´sja is a true copula, since like byt´, its function in a predicative construction is to convey an “abstract meaning of general existence” [my translation]. However, dictionary entries for javljat´sja suggest that it is not a true equivalent of byt´, because in addition to expressing existence, it has other, lexical meanings. Thus, Efremova’s (2000) Sovremennyj Slovar´ Russkogo Jazyka lists the following as the meanings of javljat´sja: I. 1) ‘become visible, noticeable’; 2) ‘appear in imagination, mind’; II. 1) ‘arrive, show up somewhere as required’; 2) ‘appear in front of someone’; 3) ‘appear, emerge’; 4) ‘turn out to be, be someone, function as someone.’23 23 In EES, the main meaning of javljat´sja was ‘appear, show up.’ However, it was also used in other senses such as 1) ‘show up as a ghost’; 2) ‘assume an aspect/shape of’; 3) ‘present oneself as, seem’; (Sreznevskij 1963: 1634-1635). Literally, javljat´sja means javlat´ sebja ‘show oneself.’ This original sense is strongly associated with such notions as ‘appearance,’ ‘vision,’ and ‘image,’ which were reflected in its use until the 19th century. The copulative function of javljat´sja was not established until the early 19th century. Akimova (1969: 226) proposes that javljat´sja became a copula due to the gradual disappearance of the copula est´/sut´ from copular constructions. Bulaxovskij (1958: 197) submits that javljat´sja developed a copulative function in constructions with Instr predicative nouns. Note that in the 19th century, the Instr began to be ousted from predicative constructions with third-person singular and plural est´ /sut´ ‘is/are’ and to be replaced in them by the Nom. In instances where the Instr remained in the absence of the copula est´– predominantly with abstract nouns – javljat´sja was used instead. Švedova (1964: 109) notes that the very first instances of javljat´sja as a copulative verb are attested in political and scholarly literature as well as in magazines in the late1820’s. Rudnev (2014: 343) also points out that Dal´’s Tolkovyj Slovar´ Živogo Velikorusskogo Jazyka [Explanatory Dictionary of the Living Great Russian Language], which was published in 1863 and which reflected the use of the spoken language of that time, did not have a dedicated entry for javljat´sja as a copulative verb. This fact supports Švedova’s observations that javljat´sja developed its copulative meaning in the written rather than spoken language. It is also noteworthy that Lönngren’s (1993) frequency dictionary not only names javljat´sja as 113 Moreover, the fact that that the Nom is ungrammatical with javljat´sja, as illustrated by (121b) above, also suggests that javljat´sja and byt´ are not true synonyms. Compare in this respect the examples below: (122) ?a. Len-a javljaetsja babušk-oj. Helen-NOM serves.as grandmother-INSTR.FEM.SG ‘Helen is a grandmother.’ b. Len-a javljaetsja babušk-oj Helen-NOM serves.as grandmother-INSTR.FEM.SG Miš-i Ivan-ova. Michael-GEN Ivanov-GEN ‘Helen is the grandmother of Michael Ivanov.’ While (122a) is grammatical, it is semantically infelicitous, because the Instr implies that “grandmother” is a capacity or role, which is a relational notion, and which, as such, requires a spatio-temporal setting. Although the temporal reference is provided by the verb javljat´sja inflected for present tense, the spatial reference is missing. On the other hand, the Gen phrase Miši Ivanova in (122b) functions as a locative modifier which one of the most frequent lexemes in CSR but also marks it as a sign of specialized (scientific, scholarly, and legal) literature. 114 establishes the required spatial setting and thus enables the instantiation of the capacity “grandmother”. In summary, the restrictions on the occurrence of the Instr after a zero copula, which are supported by the data from colloquial Russian, as well as the semantic requirements imposed on the occurrence of the Instr after javljat´sja substantiate my proposal that the referent of the Instr NP in predicative constructions denotes a capacity and that this capacity is necessarily co-temporal with rather than temporary in relation to the situation described in the construction. Moreover, my proposal is also supported by the fact that javljat´sja developed its copulative meaning in written rather than spoken language, and that in CSR javljat´sja is predominantly used in scientific, scholarly, and legal literature, whose key feature is the description of individuals and/or physical objects in terms of their roles and functions in relation to other individuals or physical objects, that is, to indicate their capacities. 5.3 A diachronic sketch 5.3.1 The first attested instances of the Instr with predicative nouns As has been mentioned earlier in this study, the Instr with predicative nouns is a relatively recent innovation, not having emerged in EES until the mid-13th century. Additionally, as evidenced by RChSlav manuscripts, in the earliest attested instances the Instr almost exclusively occurred on predicative nouns whose referents denote church posts and affiliations such as černecъ ‘monk’, černica ‘nun,’ arxiepiskopъ ‘archbishop,’ popъ ‘pop/priest’, vladyka ‘lord’, knjazь ‘prince,’ mitropolit ‘metropolitan,’ and 115 arximandrit ‘archimandrite’ (Patokova 1929: 3, Lomtev 1956: 100; Bulaxovskij 1958; Mrázek 1964; Moser 1994: 66), as illustrated by the following examples: (123) […] bystь vladyk-oju pjatь lět i tri mesjaca. was lord-INSTR.MASC.SG five years and three months ‘[…] (he) was lord for five years and three months.’ (translation mine – vt) [Komissionnyj Spisok Pervoj Novgorodskoj Letopisi, 411], (124) A černc-om byl na sěněxъ godъ i dvě neděli. and monk-INSTR.MASC.SG had.been at threshold (for) year and two weeks ‘And as a monk he had sat at the threshold for one year and two weeks.’ (translation mine – vt) [ibid., 413] Borkovskij (1949: 198-199, 1978: 82-83) observes that in EES secular documents, the Instr is attested with predicative nouns most frequently in legal texts, whereas in colloquial letters and other non-legal contexts it is rare. 5.3.2 The Instr with predicative nouns from the 15th century onward 5.3.2.1 The 15th-17th centuries While the Instr with predicative nouns is first attested in the second half of the 13th century, as evidenced by RChSlav and EES manuscripts, the Instr with predicative nouns was still a rare phenomenon until the 15th century (Borkovskij & Kuznecov 1963: 116 335). Moreover, although Instr predicative nouns began to occur more frequently in the 15th century, they did not become commonplace until the second half of the 17th century. Significantly, even though by the mid-17th century Instr predicative nouns were frequent in the past tense, they are not attested in the future tense or conditional constructions, where the Nom remained dominant until the late 17th - early 18th century (Bulaxovskij 1958: 301, Borkovskij & Kuznecov 1963: 338). 5.3.2.2 The 18th century In the 18th century Instr predicative nouns began to occur consistently in various grammatical contexts, including in the present tense, with or without the copula.24 Patokova (1929: 363) proposes that the increased frequency of the Instr with predicative nouns was due to the influence of Polish via Ukrainian, where the Instr is required in the present tense after the copula być ‘be’. Note, however, that the frequency should be understood here qualitatively as a wider variety of grammatical and semantic contexts in which the Instr began to occur, such as the future tense, conditional constructions, deverbal nouns, etc., rather than quantitatively as the ratio between Instr and Nom predicative nouns.25 24 An example of the Instr with predicative nouns in the present tense is the following construction: On pričin-oju čto skoree povorotilis´ datčane. he-NOM reason-INSTR.FEM.SG CONJ sooner turned.back Danes ‘He is the reason that the Danes turned back sooner.’ (translation mine – vt) [Xrapovickij, Pamjatnye Zapiski, 191] 25 Interestingly, the distribution of Instr and Nom with predicative nouns unfolded differently in different literary genres in 18th century. Thus, Instr predicative nouns are attested predominantly in comedy and magazine fiction, whereas Nom predicative nouns are attested more frequently in tragedy and epic literature (Bulaxovskij 1958, Lomtev 1956: 153). This distribution is believed to reflect the long-standing 117 In the 18th century the Instr became mandatory in the past tense with referents denoting professional occupations, ranks and posts. Moreover, while before the 18th century the Instr occurred exclusively with nouns of professions and church positions, in the 18th century the semantic scope of Instr predicative nouns expanded to include nouns such as svidetel´ ‘witness’, smotritel´ ‘warden, keeper,’ neprijatel´ ‘foe, adversary’, predvoditel´ ‘leader’, and naslednik ‘heir.’ 5.3.2.3 The 19th century By the 19th century, Instr predicative nouns systematically occurred after the copula in the future tense and in the infinitive, as well as in conditional constructions. The Instr is frequently attested with deverbal nouns, e.g., svidetel´ ‘witness,’ smotritel´ ‘custodian, keeper,’ predvoditel´ ‘leader,’ učitel´ ‘teacher,’ etc. (Mixajlov 2012: 45-47, Krasovitsky et al. 2008:12–13). Additionally, Krasovitsky et al.’s (2008) quantitative study reports that in the first half of the 19th century the distribution of Instr and Nom predicative nouns was uneven. Thus, Instr predicative nouns are attested much more frequently with the infinitive (97%), in the future tense (93%), and in the imperative and subjunctive mood (81%). In the past tense, Instr predicative nouns are insignificantly more frequent (54%) than Nom ones. Because in the past tense the Nom is attested almost as frequently as the Instr, Krasovitsky and colleagues explore possible semantic factors that condition the occurrence of either the Nom or the Instr. They report that the Instr is attested much more competition between colloquial early modern Russian and the bookish language heavily influenced by the Church Slavonic tradition. 118 frequently with predicative inanimate nouns (78% of the cases), whereas with animate predicative nouns the Instr is attested in 38% of the cases. It is also reported that the Instr is frequently attested with temporal modifiers, e.g., uže ‘already’, ečše ‘yet’, v prošlom godu ‘last year’, prežde ‘formerly,’ etc. (2008: 12–13) and in constructions that describe a change of state, as illustrated, respectively, by the following examples: (125) On prežde byl polkov-ym doktor-om. he-NOM formerly was regiment-INSTR.MASC.SG doctor-INSTR.MASC.SG ‘Formerly, he was the regiment’s doctor.’ (Krasovitsky et al. 2008: 10) (126) On krasavc-em byl. he-NOM handsome.man-INSTR.MASC.SG was ‘He was a handsome man.’ (Krasovitsky et al. 2008: 11) The example in (126) does not describe an explicit change of state, however Krasovitsky and colleagues state that the broader context of the sentence, which was not included in their article, indicated that a change of state happened. 5.3.2.4 Intermediate conclusion The goal of this section has been to provide some insights on the evolution of the Instr with predicative nouns. That the Instr was first attested on predicative nouns denoting church posts and positions suggests that the Instr may have occurred as a stylistic innovation. This can be explained at least partly by the intention of EES 119 chroniclers to differentiate initially between the clergy and other occupations. As can be inferred from the further expansion of the Instr to situations in which the referent of the predicative noun denoted affiliations and occupations other than those associated with church, the Instr was developing from a stylistic device to the general marker for capacities. This is supported by the fact that in the 18th century the Instr began to occur with deverbal nouns such as, for example, svidetel´ ‘witness’ or smotritel´ ‘warden, keeper,’ which, strictly speaking, are not professions or occupations but rather capacities and roles. In fact, Nichols (1981: 152) points out that the tendency of deverbal nouns to occur with the Instr increases even further in the 20th century. As will be demonstrated by the analysis of Instr predicative nouns in section 5.5 below, the Instr occurs with these nouns in accordance with the same semantic motivation that accounts for the Instr in similative constructions, in which the referent of the Instr NP denotes a capacity, i.e., a property of the subject referent in relation to the event. In subsection 5.4 I provide an overview of the literature concerning the occurrence of the Instr with predicative nouns in order to point out some controversial issues in the research to date. 5.4 Previous research: the Instr with predicative nouns in CSR It is generally agreed in the scholarship that the difference between the Nom and the Instr with predicative nouns in CSR is semantically motivated (Potebnja 1899, Patokova 1929, Bulaxovskij 1958, Borkovskij and Kuznetsov 1963, Røed 1966, Nichols 1981, Timberlake 2004). It has been proposed that the Nom indicates that the property 120 expressed by the predicative noun is permanent or inherent, whereas the Instr implies that the property is occasional, temporary, or acquired The literature on the Nom-Instr variation with predicative nouns is abundant. However, because of the time and space restrictions I will only discuss several studies. The selection is determined by the scope and relevance of the issues addressed in the study as well as the innovativeness of the study. 5.4.1 Nichols (1981) In her 1981 study, Nichols elaborates on the factors that motivate the distribution of the Instr and the Nom with predicative nouns in CSR. Nichols suggests that the distinction is based on the evaluation of relative temporal features of properties expressed by predicative nouns against a certain reference point, which is always either overtly or covertly present in the context. These relative temporal features are defined in terms of marked relative tense and implicit change of state. More specifically, the Instr occurs with predicative nouns when the situation portrays either “the departure in tense from the immediate context” or a state which emerges as the result of some recent change and which previously did not hold (Nichols 1981: 155–156). Nichols (1981) also posits several lexico-semantic classes of nouns that tend to occur in the predicate with either the Nom or the Instr. If the noun predicate is expressed by any of the following classes of nouns, it is marked by the Nom: i) nationalities; ii) evaluative nouns (e.g., durak ‘fool’, krasavica ‘beauty’, vesel´čak ‘merry person’); and iii) semantically bleached nouns, such as čelovek ‘person’, devuška ‘girl’, mužčina ‘man.’ 121 If the noun predicate is expressed by any of the following nouns, it occurs with the Instr: i) nouns of occupation, status, and function (e.g., učitel´ ‘teacher’, predsedatel´ ‘chairperson’); ii) abstract and deverbal nouns, such as cel´ ‘aim’, pričina ‘reason’, zanjatie ‘occupation’, and the like; iii) kinship terms (Nichols 1981: 155–156). While Nichols (1981) is generally correct in pointing out the semantic and grammatical contexts in which the Instr is likely to occur, she does not explain counterexamples in which predicative nouns denoting occupations, status, and function are marked by the Nom or in which predicative nouns denoting nationalities are marked by the Instr. I will reserve their examination for subsection 5.5. That the variation between the Instr and the Nom with predicative nouns is not motivated by the lexical properties of nouns finds support in Hentschel’s (1991) and Giro-Veber’s (1976, 2007) studies. These analyses examine specifically the Nom and the Instr with predicative nouns denoting professions and temporary posts, character traits, religious and political affiliation, social status, kinship terms, and names for age groups, and do not find direct correlations between the lexical properties of predicative nouns and their case marking. In other words, these studies demonstrate that, for example, nouns denoting traits of character are compatible with the Instr, just as nouns denoting professions are compatible with the Nom. 5.4.2 Timberlake (2004) Like Nichols (1981), Timberlake’s (2004) study states that the Instr with predicative nouns denotes properties restricted in time, as in following example: 122 (127) Brat-Ø dve zimy podrjad byl v Tule repetitor-om. Brother-NOM.MASC.SG two winters in.a.row was in Tula tutor-INSTR.MASC.SG ‘(My) brother was a tutor in Tula two winters in a row.’ (Timberlake 2004: 286) The Nom, on the other hand, is argued to indicate that the property expressed by the predicative noun is envisioned as unrestricted in time (e.g., identification) and that the statement made in the predicative construction is generally true (Timberlake 2004: 286– 288), as illustrated by the example given below: (128) Ved´ on byl člen-Ø politbjuro. after.all he-NOM was member-NOM.MASC.SG political.committee.GEN ‘After all, he was a member of the Political Committee.’ (Timberlake 2004: 288) Note, however, that the definition “generally true” is vague and does not provide any reasonable explanation, because as follows from Timberlake’s example, it is not the same as “general truth” or “generalization,” such as the sentence The Sun is a star. The scope of generalization expressed by the proposition in (128) is not the same as in The Sun is a star. Timberlake (2004: 286–288) also observes that the Instr occurs in situations where the subject referent fits better than anyone else a certain definition expressed by the predicative noun, as illustrated by the following example: 123 (129) On byl sam-ym xrabr-ym rycar-em. he-NOM was most-INSTR.MASC.SG brave-INSTR.MASC.SG knight-INSTR.MASC.SG ‘He was the bravest knight.’ Conversely, according to Timberlake, the Nom occurs when a predicative noun is semantically bleached and the property it expresses contributes little, and thus the communicative load is placed on the adjective, as, for example, in the following sentence: (130) On byl vsestoronne talantliv-yj čelovek-Ø. he-NOM was overall talented-NOM.MASC.SG person-NOM.MASC.SG ‘He was an overall talented person.’ However, Timberlake’s model cannot account for counterexamples such as this one below: (131) On byl vsestoronne razvit-ym he-NOM was overall developed-INSTR.MASC.SG talantliv-ym čelovek-om. talented-INSTR.MASC.SG person-INSTR.MASC.SG ‘He was a well-rounded talented person.’ (translation mine –vt) [Rudakov & Piskunov, Almaznaja kniga Rossii, 305] 124 While Nichols’ (1981) and Timberlake’s (2004) studies define the general grammatical and semantic conditions in which the Instr occurs with predicative nouns, their analyses do not provide any explanation for counterexamples. 5.4.3 Roy (2006, 2013) Although Roy’s (2006, 2013) studies are implemented in a framework that fundamentally differs from the view of language adopted in this dissertation, she makes one important observation that can be applied to the analysis of the Instr with predicative nominals. As noted in subsection 2.5.2 above, Instr nouns, which correspond to predicative nouns without the indefinite article in French, imply involvement in the activity, whereas Nom nouns do not. While this observation is made on syntactic grounds in Roy’s analysis, it is generally congruent with the Russian data and can be explained from a semantic perspective. However, Roy’s definition of the “activity” needs some adjustments. More specifically, as I have proposed in chapter IV, the referent of the Instr NP in predicative constructions denotes a capacity, which is a purposeful rather than any activity, and the latter is directly associated with the capacity. This is supported by the data from colloquial Russian, where the Instr is only acceptable in the present tense with a narrow deictic modifier and a noun referent denoting an occupation. The deictic modifier is required because a capacity is a relational notion, and as such, it needs a spatio-temporal reference to unfold. With this modification in mind, Roy’s intuition can account not only for the Nom with predicative nouns whose referents denote professions, posts, ranks, and the like, but also for the Instr with predicative nouns whose referents denote such inherent properties as character traits, for example. Since I have already 125 discussed Roy’s (2006, 2013) studies in chapter II, I will not review them in further detail in this chapter. 5.4.4 Krasovitsky et al. (2008): A quantitative analysis Krasovitsky et al. (2008) study the variation between the Russian Instr and the Nom with predicative nouns in the 19th–20th centuries. Their analysis is based on 1,853 tokens and covers four fifty-year time periods between 1801 and 2000. Because some of the findings of their quantitative analysis have been discussed in subsection 5.3.2.3 above, below I summarize the points that have not been mentioned earlier in this study. Krasovitsky and colleagues (2008: 13) point out that the distribution of the Instr with animate referents drastically changed in the mid-20th century: while earlier the Instr with animate referents is attested in certain semantic conditions (see the preceding paragraph), in the second half of the 20th century the Instr is attested with animate referents denoting “unspecified, indefinite states” (87% of the cases). In short, Krasovitsky et al.’s (2008) study thus proposes that in CSR, no semantic distinction exists between the Nom and the Instr in predicative constructions. They also state that the Instr has displaced the Nom from virtually all semantic contexts, except for instances where the referent of the predicative noun denotes a nationality. The conclusions drawn in Krasovitsky et al. (2008) raise several concerns. First, while the choice between the Nom or the Instr depends on many factors, such as the meaning of the verb and/or the lexical properties of the noun, the choice of the case marking is strongly influenced by semantic differences. In other words, constructions 126 with Instr predicative nouns express a meaning which cannot be expressed by corresponding Nom predicative constructions. That Instr predicative constructions in CSR are not semantically restricted to nouns denoting professions and occupations indicates that their meaning has broadened considerably since the time when the Instr first occurred with nouns denoting church affiliations. Significantly, because capacity can be understood both literary and metaphorically, and because capacity is associated with related notions such as “function”, “purpose”, “performance”, or “assignment”, nouns from virtually any semantic class can occur with the Instr in predicative constructions. Moreover, Krasovitsky et al.’s data are not confirmed by Kuznecova and Raxilina’s (2014) quantitative study, which was performed on a much larger body of data from the National Corpora of the Russian Language (NCRL). Kuznecova and Raxilina report that in CSR, the Nom with predicative nouns occurs almost twice as frequent as the Instr. Note, however, that this greater frequency is impacted by the fact that Kuznecova and Raxilina included in their analysis instances of the Nom after the copula est´/sut´ ‘is/are’ (see section 5.2 above), whereas such instances were excluded from Krasovitsky et al.’s study. The authors also maintain that Krasovitsky et al.’s (2008) statement about the drastic increase in the use of the Instr in predicative constructions in the mid-20th century conflicts with the data from the NCRL: while Krasovitsky et al.’s analysis reports the Instr in 87% of cases, Kuznecova & Raxilina (2014) report only 50%. In the following subsection 5.5.1 I will demonstrate that contrary to the observations made in the studies reviewed above, the Instr occurs felicitously with nouns 127 denoting inalienable properties, and I will explain why certain semantic classes of nouns tend to occur more frequently with the Instr. 5.5 An alternative analysis 5.5.1 Deverbal nouns and nouns denoting occupations As noted earlier in the subsection, it has been suggested for Russian (Švedova 1980: 283, Krasovisky et al. 2008: 3–4) that deverbal nouns, such as pričina ‘reason, cause’ svidetel´ ‘witness’or smotritel´ ‘custodian, keeper’ occur with the Instr because they denote temporary or occasional properties. While it is true that “witness” and “custodian” are acquired rather than inalienable properties, it is equally true that they are typical capacities. Moreover, although “cause” is not a capacity, it is a deverbal noun in Russian, like “witness” or “custodian”. I have proposed in the preceding chapter that referents of the Instr NPs in predicative constructions denote both a capacity and a mode of acting through which the capacity is realized as such. Deverbal nouns are compatible with the Instr because they have an eventive structure. I will elaborate on this point in the following paragraphs. Deverbal nouns are an interesting phenomenon, in that their verbal origin makes them a hybrid category and they retain both nominal and verbal properties. Unlike most common nouns, but like verbs, deverbal nouns can denote events, states, and results, as well as sharing the semantic frame of their corresponding verbs. Moreover, the lexico- semantic properties of deverbal nous are congruent with the notions of contiguity, contingency, and collaterality that are shared by the referents of Instr noun in general (see discussion in chapter IV). Thus, pričina “reason, cause” comes from the verb pričinjat´ 128 ‘cause’ and designates something or someone that makes an event happen, without being involved in it directly. A referent that causes the event is contiguous with it, in that it cooccurs with the event in time and space. The event that unfolds is contingent on the referent that brings it about. In a similar fashion, svidetel´ ‘witness’ derives from the verb videt´ ‘see’ and denotes someone who sees or observes an event take place, without participating in it directly. In this sense, a witness is temporally and spatially contiguous with and contingent on the event. One can be a witness only if there is an event. Both a witness and a cause are collateral to the event, in that they are only indirectly involved in it. Consider the following example: (132) Otec-Ø učastvoval v sraženijax. father-NOM took.part in combats On byl svidetel-em užas-ov vojn-y. he-NOM was-MASC.SG witness-INSTR.MASC.SG horrors-GEN war-GEN ‘Father participated in combats. He was a witness to the horrors of the war.’ (translation mine – vt) [Trojanovskij. Čerez gody i rasstojanija, 1997] One can witness horrors of the war only if one is involved in it in one way or another, which is what the first clause explicitly states. Because the Instr implies that the referent of the Instr noun is contiguous with and contingent on the event, the capacity is instantiated through the father’s participation in the war. 129 Compare in this respect the following examples: (133) V načale dvuxtysjačnyx rabotal v banke, in beginning two.thousands worked in bank, byl načal´nik-om valjutn-ogo otdel-a. was head-INSTR.MASC.SG monetary-GEN department-GEN ‘In the early two thousands (he) worked in a bank, was the head of the monetary department.’ (translation mine – vt) [Molodyx. V poiskax pozitiva, 2014] (134) Roditeli byli: očen´ grustn-aja mamočk-a parents were very sad-NOM.FEM.SG mommy-NOM.FEM.SG i očen´ vesel-yj papočk-a. and very cheerful-NOM.MASC.SG daddy-NOM.FEM.SG26 On byl načal´nik-Ø stanc-ii. he-NOM was master-NOM.MASC.SG station-GEN.FEM.SG ‘My parents were a very sad Mommy and a very cheerful Daddy. He was a stationmaster.’ (translation mine – vt) [Erofeev. Inter´vju, 1989–90] 26 Note that while papočk-a ‘daddy-NOM.FEM.SG’ denotes a male referent, the noun is grammatically feminine. 130 As can be inferred from these examples, načal´nik valjutnogo otdela ‘head of the monetary department’ is the capacity rather than a property of the referent of the Instr NP in (133) because it is contingent on the situation rabotal v banke ‘worked in the bank’ which implies carrying out designated duties. Conversely, the Nom implies that while ‘stationmaster’ in (134) was a property of father, it was not his capacity at the time of the situation described in the second copular clause. Note also that there is no contingency between the two situations described in (134): that is, the father’s being a stationmaster does not follow from his being very cheerful or vice versa. In contrast, the capacity “head of the monetary department” in (133) is contingent on working in the bank. 5.5.2 Nouns denoting nationalities and personal traits 5.5.2.1 Nationalities It has been noted by Nichols (1981: 155–156) and Krasovitsky et al. (2008: 11) that the Instr is disallowed with predicative nouns denoting nationalities. However, the data from the NCRL do not substantiate this statement. Consider the following example in this respect: (135) On byl amerikanc-em Srednego Zapada, he-NOM was American-INSTR.MASC.SG Midwest-GEN každyj den´ čital Bibliju. every day read Bible ‘He was a Midwest American, every day (he) read the Bible.’ (translation mine – vt), [Berberova. Železnaja ženščina,1978–1980] 131 While being an American or any other nationality is not a typical capacity such as being a witness or custodian, it can nonetheless be perceived as such if understood metaphorically as a well-recognized pattern of social behavior characteristic of a given ethnicity or nationality. Thus, the capacity “American” is realized through certain behavioral patterns, just as the capacity “custodian” is realized through performing the duties associated with the position. Note the use of the Gen phrase Srednego Zapada ‘of the Midwest-GEN’ and the proposition “read the Bible every day”, which suggests that the noun “American” does not refer to the referent’s legal status in the state but rather to the stereotypical behavior of the Midwesterners, and therefore is understood as a capacity or role. Note also that the subject referent may not necessarily be an actual Midwesterner but may just behave like one. 5.5.2.2 Personal traits Nichols (1981) points out that nouns denoting personal traits cannot occur with the Instr, because the Instr indicates that the property at issue is temporary, while temporariness is incompatible with the inalienability of personal traits from a human referent. However, as shown by the examples given below, these nouns can in fact occur with the Instr. I propose that in such instances the referent of the Instr NP is discerned metonymically as a behavioral pattern or a way of acting, or performing: 132 (136) Net, on-a byl-a umnic-ej ot prirody. no she-NOM was-FEM.SG clever.person-INSTR.FEM.SG from nature ‘No, she was a clever girl by nature.’ (translation mine – vt) [Iskander. Sandro iz Čegema (Book 3). 1989] Note that umnica ‘clever/good person’ is also used as a praise for a good performance, similar to the English “Good job!” The Instr implies that the female referent acted in a way that is culturally and socially understood as that of a clever/good person. Moreover, while an individual may be clever by nature, they may not in fact practice clever behaviors. Similarly, an individual may be clever and may practice corresponding behaviors. Importantly, acting in a capacity does not mean temporariness, because individuals may exhibit the same patterns of behavior throughout their lives, as illustrated by the following example: (137) On-a provzdyxala do konca rabočego dnja, she-NOM sighed until end work day setuja na sebja, čto kak byl-a dur-oj vsju žisn´, lamenting to herself that as was-FEM.SG fool-INSTR.FEM.SG whole life tak i ostanetsja takov-oj navsegda. so remains-FUT such-INSTR.FEM.SG forever ‘She sighed till the end of the work day, lamenting to herself that as she had been a fool all her life, she would remain so forever.’ (translation mine – vt) [Lipskerov. Poslednij son razuma. 1999] 133 As in (136), the referent of the Instr noun in (137) is perceived metonymically as a particular way of acting. The temporal modifier vsju žisn´ ‘whole life’ suggests that the capacity “fool” endures over a course of time. Compare now the following example in which the Nom signals that the property umnica ‘good girl’ is generally a characteristic of the female referent, but no indication is made as to whether she behaves as such or not: (138) Ona byl-a iz derevni, byl-a umnic-a, she.NOM was-FEM.SG from village was- FEM.SG good/clever.person-NOM.FEM.SG simpatičn-aja […] i daze èlegantn-aja. nice-looking-NOM.FEM.SG and even elegant-NOM.FEM.SG ‘She was a rural girl, was clever, nice-looking […] and even stylish.’ (translation mine – vt) [Katanjan. Lilja Brik. Žizn´. 1999] Note also the discreteness of the situations described in the example above. That the female referent was a good/clever girl is not contingent on her being a rural girl, or her being nice-looking, or stylish. Quite the opposite, the disconnection of the situations from each other, which is discerned as such in part because of the Nom, suggests that the properties umnic-a ‘good/clever.person-NOM,’ simpatičn-aja nice-looking-NOM, and èlegantn-aja ‘elegant-NOM’ are unexpected of a rural girl. 134 While the corpus data suggest that the Instr with predicative nouns denoting personal traits generally refers to a particular way of acting, in some instances the utterance may sound ambiguous and both the Instr and the Nom may seem appropriate. Consider the example below: (139) Ona byla umnic-ej she was good/clever.person-INSTR.FEM.SG i ne mogla pozvolit´ sebe napit´sja. and not could let herself get.drunk ‘She was a good/clever person and could not let herself get drunk.’ (translation mine – vt) [Bykov. Boloto. 2001] Without a broader context, the situation in (139) can be interpreted as such that being a good/clever person was a characteristic feature of the female referent ona ‘she’ and that it is this personal trait that did not let her get drunk in general. At the same time, the situation can yield a reading in which the female referent was practicing behaviors associated with being a good/clever person in a particular instance. Additionally, the corpus data generally confirm the observation made in Nichols (1981) that nouns denoting personal traits such as vesel´čak ‘merry person,’ for example, tend to occur with the Nom. Thus, among 17 instances attested in the NCRL only in two vesel´čak is attested with the Instr. 135 Interestingly, however, contrary to the predictions that the Nom predicative noun designates an inherent and permanent property, vesel´čak is attested in several instances where it clearly refers to a property that no longer holds (these instances are attested in the late-19th century texts): (140) Genrix-Ø prežde byl vesel´čak-Ø […] a kogda on priezžal , Henry previously was merry.person-MASC.SG.NOM and when he came.to.visit to deti zamečali, čto on sil´no peremenilsja. then children would.note that he a.lot changed ‘Henry was a merry person before […] but when he came to visit, the children would note that he had drastically changed.’ (translation mine – vt) [Korolenko. Noč´ju. 1888] 5.6 Conclusion Since the Instr is a relatively late innovation in Russian predicative constructions, section 5.3 has briefly addressed some diachronic issues concerning its emergence on predicative nouns. That the Instr was first attested with predicative nouns denoting church posts and affiliations suggests that it may have first occurred as a stylistic device to differentiate initially between the clergy and other occupations, which is supported by the fact that the Instr was not mandatory with nouns denoting occupations and posts until the 18th century. The examples examined in this chapter by and large corroborate my proposal that the referent of the Instr NP in predicative constructions designates a property of the 136 subject referent in relation to time and space, which is perceived as a capacity, i.e., a role which has its purpose and function, and which is instantiated as such through performance. Conversely, the Nom referent designates a property, that is an inalienable quality or an essential characteristic which, due to its nature, does not presuppose performance. I have demonstrated in this chapter that it is not coincidental that deverbal nouns tend to occur in the predicate with the Instr. Because of their lexico-semantic characteristics, deverbal nouns can denote events, which makes them congruent with the meaning “capacity”. More specifically, a capacity emerges through acting/performance, and acting/performance brings about an event. Contrary to the statement that nouns denoting inherent properties such as nationalities and personal traits cannot occur with the Instr, this chapter shows that these nouns are as compatible with the meaning “capacity” as deverbal nouns. The meaning “capacity” arises with referents denoting nationalities metonymically and refers to a set of the stereotypical patterns of behavior. In a similar way, referents of Instr nouns denoting personal traits are understood as metonyms for a particular way of acting. At the same time, however, the data from the NCRL show that some predicative nouns designating personal traits (e.g., vesel´čak ‘merry person’) have a strong inclination to occur with the Nom and some nouns (e.g., umnica ‘good/clever.person’) can be interpreted as both capacities and properties when marked with the Instr. While the data mentioned in the paragraph above suggest that the choice between the Nom and the Instr is not always influenced by the meaning of the case marking, it is still legitimate to conclude that in general Instr predicative nouns display the same 137 semantic regularities as Instr nouns in non-predicative constructions. These findings are indicative of the semantic link between the meaning of the Instr in predicative constructions and the other meanings of the Russian Instr, which supports the proposal set forth in this study that Instr predicative constructions are not a random phenomenon but a full-fledged member of the family of Instr constructions. 138 CHAPTER VI THE INSTRUMENTAL CASE MARKING WITH PREDICATIVE ADJECTIVES This chapter examines the use of the Instr marking in constructions with predicative adjectives. As with nouns, the synchronic variation between the Nom and the Instr with predicative adjectives is the result of diachronic change. Instr predicative adjectives are an even later innovation than Instr predicative nouns and are not attested in RusChSlav or EES manuscripts (Lomtev 1956: 143–153, Potebnja 1888: 149, Karskij 1929: 20). This chapter demonstrates that like the referents of Instr predicative nouns, properties expressed by Instr predicative adjectives are perceived as capacities rather than as properties. I will also establish that properties expressed by Instr predicative adjectives, like referents of the Instr NPs in other constructions, can be envisioned as stimuli, enablers, conduits, etc. Because Russian makes a distinction between long and short adjectives and because the emergence of the Instr with predicative adjectives is contingent on the emergence of long adjectives, a part of this chapter provides some historical background, as well as discusses briefly the semantic and grammatical properties of short adjectives. 6.1 Short and long adjectives in Contemporary Standard Russian: A grammatical sketch In addition to differential Nom-Instr case marking, predicative adjectives can have short vs. long forms. Until the mid-17th century, only short adjectives could function as predicates. Compare the following examples from CSR: 139 (141) a. Nikolaj-Ø byl spravedliv-yj. Nicholas-NOM was impartial-NOM.MAS.SG ‘Nicholas was impartial.’ b. Nikolaj-Ø byl spravedliv-ym. Nicholas-NOM was impartial-INSTR.MASC.SG ‘Nicholas was impartial.’ c. Nikolaj-Ø byl spravedliv-Ø. Nicholas-NOM was impartial-MASC.SG (short adjective) ‘Nicholas was impartial.’ Most Russian adjectives have two forms: a long form and a short one, as illustrated by the examples here in Figure 2: Long form Short form krasiv-yj/aja/oe/ye krasiv-Ø/a/o/y ‘beautiful-MASC/FEM/NEUT/PL’ beautiful-MASC/FEM/NEUT/PL dobr-yj/aja/oe/ye dobr-Ø/a/o/y kind-MASC/FEM/NEUT/PL kind-MASC/FEM/NEUT/PL interesn-yj/aja/oe/ye interesen-Ø/a/o/y interesting-MASC/FEM/NEUT/PL interesting-MASC/FEM/NEUT/PL Figure 2. Long and short adjectives in CSR 140 Unlike long adjectives, short adjectives in CSR do not decline for case: these adjectives always appear in the default Nom form, although they do follow number and gender agreement, as can be seen from Figure 2 above. Short adjectives have the same declension as nouns, while long forms historically added the 3P personal pronouns to the short forms (including their nominal ending). Additionally, not all adjectives in CSR have both long and short forms. Attributive qualitative (gradable, or scalar) adjectives usually have both forms, e.g., interesnyj ‘interesting’ (long) – interesen ‘interesting’ (short).27 On the other hand, relational adjectives only have a long form: mednyj ‘brass’ (short) – *meden. 6.2 The emergence and development of long adjectives Evidence from OCS texts shows that long adjectives already existed in Proto- Slavic (see, for example, Vaillant 1974, Lunt 1974). They were derived from short adjectives, through the right-attachment of the appropriate gender/number form of the third-person pronoun, with both the short adjective inflectional case morpheme and the pronominal case form declining: 27 In Russian linguistics scholarship, attributive adjectives are divided into two large grammatical classes based on their ability to form degrees of comparison. Qualitative adjectives are gradable, or scalar (e.g., gorjačij ‘hot’ – gorjačee ‘hotter’), while relational adjectives are not (e.g., derevjannyj ‘wooden’ – >*derevjann-ee/ bolee derevjannyj ‘*woodener/ more wooden’). Qualitative adjectives usually denote temporary properties, whereas relational adjectives denote permanent properties. 141 (142) dobr-a- ja sestr-a. good-FEM.SG-3SG.NOM.FEM sister-NOM.FEM.SG ‘the good sister’ (definite referent) Thus, in (142), dobr-a ‘good’ is the short adjective form with the SG.FEM ending -a; the morpheme -ja is the old Proto-Slavic NOM.SG.FEM personal pronoun. The short adjective form had the same declensional paradigm as o- and a-stem nouns, and it declined for case, as it still does in the West Slavic languages, for example, where it is now the only adjectival form.28 When a noun was modified by a short adjective in Proto-Slavic (as evidenced by OCS) without the attachment of the 3 person pronoun, it was considered indefinite:29 28 E.g., Czech krásn-á rek-a ‘beautiful-NOM.FEM.SG river-NOM.FEM.SG (‘a beautiful river’). Note also that Macedonian and Bulgarian use the short form but add a definite clitic derived from the demonstrative pronoun to make them definite. That is very different from what East Slavic languages do. The East Slavic long form does not now make the adjective definite, although that was its original function in ProtoSlavic. 29 Note, however, that this was not always the case. Marking the definiteness of a given noun was redundant, if the noun was a priori definite. More specifically, short adjectives used with proper names, names of well-known cities and territories, and church holidays did not require a demonstrative pronoun. For example, in EES novъ gorodъ (CSR Novgorod, name of a city, lit., ‘New City’), the adjective novъ ‘new’ was always used in its short form because it referred to a well-known place. Similarly, in OCS velikъ dьnъ, ‘Easter’, lit. ‘great day’, the adjective velikъ ‘great’ was used in the short form because the expression referred to a well-known church holiday. In CSR, as in all the other Slavic languages, the grammatical definiteness-indefiniteness distinction based on attachment of the third person personal pronoun is obsolete (modern Bulgarian and Macedonian right-attach a demonstrative clitic pronoun to nouns and attributive adjectives to mark definiteness, but this was not a Proto-Slavic feature). The major factor in the complete disappearance of the definiteness-indefiniteness opposition was, however, the restrictive use of the personal pronoun ending: they were only required if a short adjective was used attributively. When short adjectives were used predicatively, this ending was not used, as the property expressed by the predicative adjective refers by default to the already-known and thus definite referent. For example, EES teremъ kamjanъ, lit., ‘tower-room (is from) stone’ can only be said about a specific tower-room, and not about any tower-room in general. Note that here the predicative adjective is relational, whereas in CSR relational adjectives do not appear in the short form. . 142 (143) dobr-a sestr-a good-NOM.FEM.SG sister-NOM.FEM.SG ‘good sister’ (does not refer to a specific sister) 6.2.1 Grammatical functions of short and long adjectives Reflecting their origin of inflectional suffixes, although long and short adjective forms existed in EES, their grammatical functions were different from their corresponding functions in CSR and the other modern East Slavic languages, Ukrainian and Belarusian. In EES short adjectives were used both attributively and predicatively, as in OCS and presumably Proto-Slavic, whereas in CSR and the other East Slavic languages, they can function only as predicates. When used predicatively, short adjectives denoted a state of the subject referent. When used attributively, they designated the referent’s most characteristic quality (Lomtev 1956: 134–136). As attributives, short adjectives agreed in gender, number, and case marking with the Nom grammatical subject. However, as short adjectives lost their attributive function, they also lost their case declension. 6.2.1.1 Disappearance of short adjectives from attributive positions Lomtev (1956: 133–134) postulated that the loss of the attributive function by short adjectives was conditioned by the evolution of attributives as a grammatical category, which gave rise to the reanalysis of the construction short adjective + personal pronoun as a long adjectival form, in which the personal pronoun was reanalyzed simply as a grammatical ending. Moreover, these changes affected directly the distribution of 143 grammatical functions of short adjectives. Thus, the reanalyzed adjectival form was used attributively, whereas short adjectives were exclusively used as predicates. Lomtev (1956: 134) observed that the loss of the declension paradigm not only left short adjectives with the Nom as their default predicative form, but also changed the meaning expressed by the short predicative adjective. While the old meaning (e.g., est´ byla dobr-a ‘has been good-NOM.FEM.SG’) was the existence at a specific time of an individual/physical object with a property expressed by the predicative adjective, the new reanalyzed function was to designate the existence at a specific time of the property expressed by the predicative adjective; thus the property was understood as a state of the subject referent. 6.3 Long adjectives in predicate positions Until the beginnings of modern Russian in the 17th century, the distribution of the new grammatical functions carried out by long and short adjectives remained unchanged: long adjectives functioned as attributive modifiers, while short adjectives were used as predicates and attributes. However, in the 17th century, long adjectives began to appear occasionally in predicate positions. Lomtev (1956) and Bulaxovskij (1958) suggest that long predicative adjectives were meant to compensate for the loss of the old meaning expressed by the short adjective, that is, to express the existence of a referent with a property expressed by the adjective. According to Lomtev (1956: 134–136), the distinction between short and long predicative adjectives was clear-cut. The short predicative adjective expressed a state of the subject referent at a specific time and in specific circumstances (e.g., ona dobr-a ‘she 144 (is) good-NOM.FEM.SG’), whereas the long adjective ending (e.g., ona dobr-aja (< -a-ja) ‘she (is) good-NOM.FEM.SG’) was meant to convey the fact of existence of a subject referent with the consistent property “good”, rather than “good” as a quality applicable to the referent at a specific time (but cf. N. V. Švedova (1948: 107–109), who argues that because the original function of long adjectives was to describe qualities rather than states, they denoted inherent, permanent, or essential properties). In the 17th century, long predicative adjectives began to occur with the Instr. This is attributed to the influence of Polish and Ukrainian, where Instr predicative adjectives were already a common phenomenon (Patokova 1929, N.V. Švedova 1948). It is noteworthy that originally the distribution of the Nom and the Instr with predicative adjectives differed between the spoken and written language. Instr predicative adjectives initially appeared in business, legal, and newspaper language, while Nom predicative adjectives were used in the colloquial language and those written genres which were directly associated with the vernacular (N.V. Švedova 1948: 120). Švedova (ibid.) also points out that in the mid-18th century, the Instr with predicative adjectives not only became pervasive in all written genres, but also entered the vernacular. In the mid-17th century, long adjectival forms (e.g., derevjannyj ‘wooden’) completely displaced the short form (e.g., derevjan) of relational adjectives from predicate positions. In the late 17th – early 18th century, Instr predicative adjectives expanded their functional scope: they began to appear after the copula in the future tense, imperative and conditional mood, like Instr predicative nouns, while Nom predicative adjectives became less frequent in these grammatical conditions. 145 Lomtev (1956: 145–155) observes that although throughout the 17th century long predicative adjectives began to appear after lexical and semi-lexical verbs, their distribution was still very limited until the late 18th – early 19th century. During the 19th century, short adjectives were displaced from predicate positions in many grammatical contexts. For example, they could not appear after the infinitive, certain participles, some lexical verbs. Long adjectives, on the contrary, expanded to various genres of the written language, including poetry, in which short adjectives had dominated until the mid-19th century. 6.4 Long and short predicative adjectives in Contemporary Standard Russian The semantic and grammatical distinctions between short and long adjectives in CSR have been discussed in many studies (see, for example, Šaxmatov 1941, Švedova 1946, 1948, Tolstoj 1954, Peškovskij 1956, Škarupo 1957, Mjasnikov 1959, Giro-Veber 1996, Vsevolodova 1972, Nefedova 1978, Kazavčinskaja 1990, Voejkova and Pupynin 1996, Gasparov 1997, Šeljakin 2010). In general, these studies can be divided into two groups.30 In one, it is proposed that the distinction between short and long predicative adjectives is the opposition 30 Note that short adjectives are used in situations when a long adjective cannot be used: 1) propositions that represent general truth, as in the following construction below: (i) Žizn´ prekrasn-a life-NOM beautiful-FEM.SG (short) ‘Life is beautiful.’ 146 between permanent and temporary properties, with temporary properties being bounded in time, whereas permanent properties are not (Šaxmatov 1941, N. V. Švedova 1946, 1948, Tolstoj 1954, Peškovskij 1956, Škarupo 1957, inter alia). In the other group of studies, scholars argue that short and long predicative adjectives can express both permanent and temporary properties, and that the distinction between permanent and temporary properties does not apply across the board (Kazavčinskaja 1990, Voejkova and Pupynin 1996, Gasparov 1997, Boguslavskij 1964, Gvozdev 1962, Kotov 2014). For example, Gvozdev (1962: 229) and Kotov (2014: 236– 237) note that the distinction “permanent–temporary” depends on the properties of the (ii) ?Žizn´ prekrasn-aja. life-NOM beautiful-NOM.FEM.SG (long) ‘Life is beautiful.’ 2) if the short and long form of the adjective have undergone semantic changes: (iii) Ty bol´n-oj? You-2SG.MASC sick-NOM.MASC.SG ‘Are you sick? (implies a mental illness) (iv) Ty bolen-Ø? You-2SG.MASC ill-MASC.SG ‘Are you ill?’ 3) if the subject is expressed by the pronoun èto ‘it’ or by an abstract noun: (v) Èto interesn-o. this interesting-SG.NEUT (short) ‘This is interesting.’ (vi) *Èto interesn-oe. this interesting-SG.NEUT (long) ‘This is interesting.’ A way to test for this in English is to translate the long form predicate adjective as definite ‘the X one’, which does not work for žizn´ or èto: *life is a/the beautiful one, *this is the interesting one. 147 construction, particularly on the presence or absence of temporal lexemes (e.g., overt copula, temporal adverbs, etc). Consider the following constructions: (144) Rek-a tix-aja i spokojn-aja. river-NOM.FEM.SG quiet-NOM.FEM.SG and tranquil-NOM.FEM.SG ‘The river is quiet and tranquil.’ (translation mine – vt) (Kotov 2014: 237) (145) Rek-a v tot moment river-NOM.FEM.SG in that moment byla tix-aja i spokojn-aja. was quiet-NOM.FEM.SG and tranquil-NOM.FEM.SG ‘The river was quiet and tranquil at that moment.’ (translation mine-vt) (Kotov 2014: 237) The interpretation of (144) is that the river is always quiet and tranquil. However, (145) describes the quietness and tranquility of the river at a specific moment and thus conveys the temporariness of the situation. Conversely, short adjectives that generally express temporary properties, e.g., bolen-Ø ‘sick-MASC.SG’, can also express unbounded situations: 148 (146) On vsegda byl bolen-Ø, he-NOM always was ill-MASC.SG xriplo dyšal, kuril vonjučij tabak […]. hoarsely breathed smoked smelly tobacco ‘He was always ill, breathed hoarsely, smoked smelly tobacco [...].’ (translation mine-vt) (Kotov 2014: 238) As can be seen from the examples (144) through (146) above, the semantic difference between long and short predicative adjectives does not necessarily arise because of the lexico-semantic or syntactic properties of predicative adjectives themselves but can be also determined by other factors. The scope of this dissertation does not allow me to investigate short adjectives in more depth, but the overview of the literature in this section contains references that can be consulted for further detail. In the next section, I examine Instr predicative adjectives in CSR. 6.5 Instr and Nom predicative adjectives in Contemporary Standard Russian: An overview of the literature Traditionally, as in the case with predicative nouns, predicative adjectives are studied from the perspective of the Nom-Instr opposition (see Černov 1983, Nichols 1985, Giro-Veber 1996, Šeljakin’s 2010, Zel’dovič 2005, Israeli 2007, Gustavsson 1976, Nikunlassi 1993, Timberlake 2004, Mrázek 1961, 1964, Hentschel and Ueda 1998, 149 among others). Recall that in the present tense, CSR does not require an overt copula and the Nom-Instr opposition does not hold for the present tense but holds for other tenses and moods. Černov (1983: 91–93) maintains that the Instr rather than the Nom is preferred when the property expressed by an adjective predicate is envisioned as distanced from the time of the speech act, as is usually underscored by the use of temporal adverbials (togda ‘(back) then’, kogda ‘when’, v to vremja ‘at that time’, etc.). The Nom, on the other hand, implies that the property expressed by a given predicative adjective reflects the speaker’s subjective evaluation, which is usually emphasized by comparative phrases and intensifiers (e.g., takoj že ‘the same as’, očen´ ‘very’, sovsem ‘completely/totally’, ves´ ‘whole’). Similarly, Nichols (1985: 66–377) juxtaposes the Instr with the Nom as conveying, respectively, an objective versus a subjective evaluation by the speaker. Giro-Veber (1996), Šeljakin (2010), and Zel´dovič (2005) treat the Nom-Instr opposition with predicative adjectives as an opposition between marked versus unmarked, respectively, arguing that the Nom underscores the communicative importance of the property expressed by a given predicative adjective and portrays it as essential, to the extent that it characterizes the subject referent as an individual, while the Instr is the default case marking. Zel´dovič (2005: 25–27) further proposes that the Nom emphasizes the speaker’s involvement in the situation described in the construction, implying “observability” (see also Israeli 2007), and that this explains why Nom predicative adjectives are rarely used in scholarly, business, and legal language, but frequently occur in colloquial speech. Conversely, the Instr indicates that the speaker is not involved in the situation described 150 in the construction, which explains why the Instr occurs more frequently than the Nom in negative and conditional constructions (but cf. Kuznecova and Raxilina 2014 for the opposing view that it is the Instr that implies observability and involvement in the situation). Zel´dovič (2005: 33) also notes that Nom predicative adjectives suggest that the portrayed situation is isolated and distanced from other situations that took place earlier, whereas the Instr emphasizes the connection between a given situation and previous situations. Mrázek (1961,1964), Timberlake (2004), and Israeli (2007) all have postulated instead that the Instr implies that the property expressed by a given predicative adjective has undergone a change or is envisioned as bounded in time. Timberlake specifies that this change can be a change in the observer’s perception or in the real world, and argues that the Instr indicates a contrast between “two polarities of the state in two time-worlds”: that is, in one world where the state holds, and in another where it does not. According to Timberlake, the Instr after the past tense copula also implies that a given state has been cancelled or initiated (2004: 286–287). Nikunlassi (1993) points out that none of the existing analyses of the Nom-Instr variation with predicative adjectives can single out criteria underlying the use of either case marking because the number of potential conditioning factors is infinite (see also Nichols 1985 and Gustavsson 1976). I propose, in contrast, that the controversies over the distribution of the Nom and the Instr arise, in part, from the models themselves. More specifically, the meanings of the Nom and the Instr with predicative adjectives, as well as with predicative nouns, are 151 traditionally studied in relation to each other rather than in relation to the meanings of the Nom or the Instr in other Nom and Instr constructions. While it is true that there exist a number of semantic (and syntactic) factors that trigger the occurrence of the Instr in predicative constructions, we can try to analyze this use of the Instr in a more coherent way, by looking at the other Instr meanings to determine whether the predicative Instr shares the same semantic regularities that underly its use in other Instr constructions in the language. As has been established in the preceding chapters, the Instr with predicative nouns is not only related to the Instr in similative constructions, but also to the Instr in other types of Instr constructions, for example, through the notions “function” and “purpose”. Many observations made in the literature about the Instr in predicative constructions are correct; however, they lack a common denominator or point of reference, because most of the studies examine the Instr in these constructions as a discrete phenomenon, leading to the impression that the Instr with predicative nominals is motivated by unrelated factors. However, many of these factors can be accounted for on the same semantic premises that explain the use of the Instr in other types of Instr constructions. For example, as noted above, Mrázek (1961,1964), Timberlake (2004), and Israeli (2007) state that the Instr implies that the property expressed by a given predicative adjective has undergone a change. The idea of change is common in some other Instr constructions. Thus, in similative Instr constructions the referent of the Instr noun denotes a new capacity which emerges when the subject referent metaphorically adopts a mode of acting associated with the referent of the Instr NP, and which can be interpreted broadly as a metaphorical change of state. In constructions with verbs of 152 naming (e.g., naznačit´ ‘appoint’ or nazvat´ ‘name,’ among others), the change of state happens contingent on naming or appointing and is understood as the name or appointment per se. In depictive constructions (see subsection 4.2.6 and footnote 14 above), the state expressed by the Instr adjective is conceived as non-existent prior to the action described by the verb. 6.6 An alternative analysis 6.6.1 Testing the “subjective–objective characteristic” hypothesis As noted early in this section, it has been suggested in the literature that the Instr signals that the property expressed by the predicative adjective reflects the speaker’s objective characterization, whereas the Nom indicates the speaker’s subjective evaluation (Černov 1983, Nichols 1985, and, indirectly, Zel´dovič 2005). This approach raises several problems. First, the proposed distinction is pragmatic rather than semantic in nature, in that it reflects the properties of the proposition rather than the case marking per se. In philosophical and logical terms, propositions are defined as objective if they concern matters of empirical and mathematical facts, such as “the Earth moves around the sun” or “three times three is nine.” Moreover, what seems an objective characteristic to one person may be perceived differently by another. Note also that since case marking is understood in this dissertation as a set of perceived relations among entities in the situation described by the verb (Vakareliyska 1994), the propositions made in a given sentence pertain to conceptualized scenes, and not to objective reality. In this respect, the distinction “subjective–objective characteristic” has little value for the analysis presented 153 in this study. Also, that the Instr implies an objective characteristic cannot be proven to reflect its meanings in other Instr constructions. Consider the following example: (147) Togda byla kartočnaja sistema, then was card system, xleb-Ø byl ser-yj, a ne bel-yj. bread-NOM.MASC.SG was grey-NOM.MASC.SG and not white-NOM.MASC.SG ‘Then there was a food stamp system, bread was grey and not white.’ (translation mine – vt) [Katanjan. Prikosnovenie k idolam, 1998] If we apply the “subjective–objective characteristic” distinction to the situation described in (147), we will need to assume that the Nom indicates the speaker’s subjective evaluation of the bread’s color. Note, however, that here the predicative adjectives ser-yj ‘grey-NOM.MASC.SG’ and bel-yj ‘white-NOM.MASC.SG’ refer to the quality of the flour that was used to make the bread. In particular, “grey” bread is made from lower-quality flour as opposed to high-quality flour which makes bread white. It is a well-known fact that in times of food shortages in the Soviet Union, bread that was distributed in exchange for food stamps was made from lower-quality flour. Thus, the situation described in (147) reflects the distinction between two sorts of bread rather than the distinction in the speaker’s perception of the color of the bread in question. Moreover, the Nom implies that there is no contingency between the color of the bread and the quality of the flour. 154 While it is ultimately the quality of the flour that makes the bread look either grey or white, it is a natural consequence rather than an accidental circumstance. Consider in this respect another example: (148) Ves´-Ø les-Ø byl ser-ym whole-NOM.MASC.SG forest-NOM.MASC.SG was grey-INSTR.MASC.SG ot soldatskix rubax-Ø. from soldier shirts ‘The whole forest was grey with soldier’s shirts.’ (translation mine – vt) [Krasnov, Na vnutrennem fronte, 1992] Contrary to the predictions made in Černov (1983) and Nichols (1985) that the Instr expresses the speaker’s objective characterization, the example above describes the speaker’s perception of the forest. Here the Instr indicates contingency between its referent and another situation. It is understood that there are many soldiers in the forest and that their traditionally grey uniforms make the forest look grey. Note also that the Gen phrase ot soldatsk-ix rubax-Ø, lit., ‘from soldier-GEN shirts-GEN’, serves as spatio- temporal reference for the property “grey” to become a capacity or a role of the forest in the speaker’s frame of reference. Moreover, by contrast to the property “grey” in (147), the capacity “grey” in (148) emerges in the concourse of circumstances. Similarly, 155 (149) On-i šagnuli vnutr´ étogo vpolovinu vokzala, vpolovinu žilišča they-PL stepped-PL inside this half railway.station half house vozdux-Ø v kotorom air-NOM.MASC.SG in which byl ser-ym ot v´´evšegosja v nego tabačnogo dyma. was grey-INSTR.MASC.SG from ingrained into it tobacco smoke. ‘They stepped inside this half railway station, half house, the air in which was grey from the tobacco smoke ingrained in it.’ (translation mine – vt) [Pavlov, Kazennaja skazka, 1993] In the example above, the color of the air is perceived to be the color of the tobacco smoke. Because the tobacco smoke permeates the air, it makes it look grey. Here again, as in (148), the Instr indicates contingency between the color of the air and another situation. Thus, as with Instr predicative nouns, the capacity ‘grey-INSTR’ in (148) is contingent on a mode of acting, which is understood here as smoking. More specifically, it is implied that somebody has smoked in the house and, most likely, has smoked a great deal, because the tobacco smoke permeated the air, making it look grey. Note also that unlike the scenario in (147), where the property “grey” is an expected outcome, in the example above, as in (148), the capacity “grey” appears in specific circumstances. 156 6.6.2 Testing the “observability” hypothesis Kuznecova & Raxilina (2014: 207–208) propose that the notion of “observability” (nabljudaemost´) is shared by Instr predicative adjectives (but cf. again Zel´dovič 2005 for a different view, in which observability is a feature of the Nom). The situations in (148) and (149) can indeed be explained on these grounds. At the same time, however, observability may also be implied by Nom predicative adjectives, indicating that this notion, at least without a more precise definition, is not unique to the Instr. Consider the following example: (150) Galstuk Dolmatova byl ser-yj, tie.NOM.MASC.SG Dolmatov-GEN was grey-NOM.MASC.SG uzorn-yj vysok-ogo kačestv-a. patterned-NOM.MASC.SG high-GEN.NEUT.SG quality ‘Dolmatov’s tie was grey, patterned, and of a high quality.’ (translation mine – vt) [Šalamov, Kolymskie rasskazy, 1954–1961] The granularity of detail in (150), such as uzorn-yj ‘patterned-NOM.MASC.SG’ and vysokogo kačestva ‘of high quality’, can only be achieved if the speaker is observing the tie. A legitimate question in this respect is why, if the situation is observable in both (150) and in (148) and (149) above, the predicative adjectives are case-marked differently? I propose that the answer to this question lies in the meaning of the Instr and the Nom in predicative constructions, as I have defined them earlier in this study for predicative nouns: i.e., the Instr implies that the referent of the Instr noun is a capacity, 157 and this capacity is always contiguous with and contingent on a mode of acting/performance, whereas the Nom referent denotes a property, and the property does not require an action in order to unfold. Following this definition, I further submit that the differences in the case marking of the predicative adjectives in (150) and (148) and (149) capture the differences in the perception of the scenes portrayed in these examples as non-eventive and eventive, respectively. Thus, the observability of (148) and (149) is enabled by the event unfolding in these scenes, whereas in (150) the observability arises from the visual perception of a stative scene. More specifically, nothing is happening to the holder of the Nom property “grey” in (150), which is Dolmatov’s tie, other than that it is being worn by a human referent. Conversely, the forest in (148) is occupied by soldiers, which is underscored by the intensifier ves´ ‘whole’, and which implies some action unfolding in the forest. Similarly, the air in (149) is permeated by tobacco smoke, which indicates a smoking scene. What follows is that Kuznecova and Raxilina’s thesis about observability being a feature of Instr predicative adjectives referents is intuitively correct, however, as evidenced by the examples above, observability must necessarily arise from an eventive scene in which the holder of the property expressed by the predicative adjective is involved in one way or another. Consider another example with a Nom predicate adjective: 158 (151) Mne udalos´ uxvatit´ moment, me-DAT managed-NEUT.SG catch moment kogda glaza u nego byli golyb-ye. when eyes-NOM.PL by him were blue-NOM.PL ‘I managed to catch a moment when his eyes were blue.’ (translation mine – vt) [Voronel´. Bez prikras. Vospominanija, 1975–2003] Example (151) above clearly implies a certain degree of observability. One can only catch a moment when someone’s eyes look blue if one is observing the eyes. However, the Nom indicates that the property “blue” is not contingent on any other referent or situation, and the scene itself is non-eventive. Note also that example (151) does not conform to the observation that Nom predicate adjectives denote inherent and permanent properties. The clause kogda glaza u nego byli golyb-ye ‘when his eyes were blue-NOM’ suggests that the human referent’s eyes were specifically blue at a given moment and that they may have been of another color at another time. That observability as a feature of Instr adjectives is related to an eventive scene can be illustrated by another example: 159 (152) Počerk ét-ogo xudožnik-a vsegda uznavalsja handwriting-NOM.SG this-GEN artist-GEN always was.recognizable-NEUT.SG po ostro sovremennoj manere s´´emki: — by cutting-edge manner filming-GEN kamer-a v ego rukax byla podvižn-oj, camera-NOM.FEM.SG in his hands was mobile-INSTR.FEM.SG plastičn-oj i otzyvčiv-oj […]. flexible-INSTR.FEM.SG and responsive-INSTR.FEM.SG ‘The style of this artist was always recognizable by the cutting-edge manner of filming. The camera in his hands was mobile, flexible, and responsive […].’ (translation mine – vt) [Kičin. Izvestija, 2002] The broader context in (152) makes explicit that the landmark feature of the artist’s style is the manner of filming. Since manner is an outward behavior or way of acting, it is visible and observable. Moreover, manner is contiguous with and contingent on motion, and motion underlies events. However, even in the absence of the first clause, the Instr predicative adjectives in (152) evoke eventive scenes. Because the Instr adjective is perceived in predicative constructions as a capacity and a mode of acting, kamera byla podvižn-oj, plastičn-oj i otzyvčiv-oj ‘camera was mobile-INSTR, flexible-INSTR, and responsive-INSTR’ refers to the functions of the camera in a given situation rather than to its qualities. 160 As can be seen from the analysis of the examples in this and the preceding subsection, contingency between the Instr predicative adjective and other situations is pervasive in predicative constructions just as it is ubiquitous in other Instr constructions in Russian. The individual meanings of the Instr, such as, for example, instrument, means, or stimulus, all arise from the relation of contingency. In the following sections, I will demonstrate that Instr predicative adjectives can also evoke scenes in which a property expressed by the Instr predicative adjective is perceived as either of the roles mentioned above. 6.6.3 Properties as capacities 6.6.3.1 Stimuli That the Instr on predicative adjectives can be understood as any one of the roles of the referent of the Instr noun in non-predicative constructions can be accounted for by the internal organization of the Instr meanings as a network based on family resemblance. The properties represented by some Instr adjectives are perceived as inadvertently enabling certain behaviors or attitudes. Consider the following: (153) Ona, navernoe, byl-a krasiv-oj togda, she-NOM.FEM.SG certainly was-FEM beautiful-INSTR.FEM.SG then raz ee narisoval xudožnik tri veka nazad […]. since her-ACC painted artist three centuries ago ‘She, certainly, was beautiful then, since an artist painted her three centuries ago.’ (translation mine – vt) , [Černyx. Tri rasskaza. Junost´.1972] 161 In the example above, the Instr implies a contingency relation between the beauty of the female referent of the pronoun ona ‘she’ and the situation in which an artist painted her. Here the meaning of the Instr adjective krasiv-oj ‘beautiful-INSTR’ is perceived as a capacity of the female referent rather than as a property of her, because it has an effect on the referent “artist” and entails another situation, just as stul-om ‘with a chair-INSTR’ in Ivan ubil Petra stul-om ‘John killed Peter with a chair-INSTR’ is interpreted as an instrument rather than a piece of furniture. Note that a capacity is a role, which is a relational notion: that is, it is the function assumed or perceived in a particular circumstance. In the example above, because the property “beautiful” expressed by the Instr predicate adjective motivates the artist to paint the female referent, it can be discerned broadly as a capacity and more specifically as a stimulus. Similarly, the Instr predicate adjective in (154) incentivizes an emotional attitude in the referent of the noun otec ‘father’: (154) Otec javno gordilsja i daze čutočku xvastalsja eju – father obviously was.proud and even a.little boasted she-INSTR nastol´ko mam-a byla krasiv-oj i vesel-oj. so.much mom-NOM was beautiful- INSTR.FEM.SG and cheerful-INSTR.FEM.SG ‘Father was obviously proud of and even a little boastful about her – so beautiful and cheerful Mom was.’ (translation mine – vt) [Borisova. Čelovek i zakon, 1979] 162 Here, Mom’s beauty and cheerfulness inspire Father to be proud of and boastful about her. As in (153) above, a contingency relation holds between Mom’s properties “beautiful and cheerful” and Father’s being proud of and boastful about Mom. Because the properties expressed by the predicative adjectives in (153) and (154) cause an emotional response or attitude only indirectly, they can be defined on semantic grounds as stimuli. Compare now the following example: (155) Mam-a tože byla krasiv-aja, Mom-NOM also was beautiful-NOM.FEM.SG v dlinnyx kudrjax, v novoj koftočke s konfetnymi pugovicami. in long curls in new shirt with candy.like buttons. ‘Mom, too, looked beautiful, with long curls, in a new shirt with candy-like buttons.’ (translation mine – vt) [Slavnikova. Bessmertnyj. Povest´ o nastojaščem čeloveke, 2004] The Nom here implies that krasiv-aja ‘beautiful-NOM.FEM.SG’ is a property of Mom, in the sense of a quality or attribute, rather than a capacity, because, in contrast to (153) and (154) above, there is no contingency between the property “beautiful” and another situation. 163 6.6.3.2 Instruments and conduits In addition to stimuli, contingency can also bring about the role of a conduit, as in the following examples:31 (156) Vozdux-Ø byl prozračn-ym dlja zvukov, air-NOM.MASC.SG was transparent-INSTR.MASC.SG for sounds i poryv vetra prines na sebe and gust wind-GEN.MASC.SG brought on itself šuršanie električk-i. rustle electric.train-GEN.FEM.SG ‘The air was transparent to sounds, and a gust of wind brought a rustle of the train.’ (translation mine – vt) [Azol´skij, Lopušok. Novyj Mir, 1998] In (156) the Instr indicates that prozračn-ym ‘transparent-INSTR.MASC.SG’ is a metaphorical channel for transmitting the sounds and therefore a capacity rather than a property in the situation described in the sentence. As noted earlier, a capacity is a relational notion, while a property is not. Note also that although “transparent” already implies an ability to transmit light or sound waves without scattering, whether or not the property “transparent” becomes a capacity depends on the given situation. In the example below the Instr adjective is perceived as an instrument: 31 The term “conduit” has been suggested by Janda (1993) as a semantic invariant for the Russian Instr. “Conduit” is an umbrella term that covers meanings, such as instrument, means, and pathways. 164 (157) […] kuda Dam-a prošla po konkursu, where Dama-NOM.FEM.SG passed on competitive.basis potomu čto byl-a krasiv-oj because was-FEM.SG beautiful-INSTR.FEM.SG i umela otčetlivo proiznosit´ raznye slova […] and was.able distinctively articulate different words ‘[…] where Dama was admitted on a competitive basis because she was beautiful and could articulate distinctively different words.’ (translation mine – vt) [Klimontovič. Dalee – vezde, 2001] The broader context in (157) describes a situation in which the female referent, Dama, has received a job as a television announcer on a competitive basis. The Instr indicates the contingency between the female subject referent’s appearance and good articulation and her receiving a job. Krasiv-oj ‘beautiful-INSTR.FEM.SG’ in this example is a capacity rather than a property of hers, because it is associated with a specific function of a broadcaster, which is to attract viewers. In this sense, “beautiful” is perceived as an instrument for receiving the job. That properties expressed by Instr predicative adjectives can be perceived as instruments is further supported by the examples given below: 165 (158) a. Lena byla vesel-oj, čtoby ponravit´sja Miše. Helen-NOM was cheerful-INSTR.FEM.SG in.order.to interest Mike-DAT ‘Helen was being cheerful in order to interest Mike.’ b. ?Lena byla vesel-aja, čtoby ponravit´sja Miše. Helen-NOM was cheerful-NOM.FEM.SG in.order.to interest Mike-DAT ?‘Helen was cheerful in order to interest Mike.’ While both (158a) and (158b) are well-formed, (158b) is infelicitous. In both (a) and (b) above, Helen is trying to arouse interest in another person by being cheerful. The Instr adjective form in (158a) suggests that she intentionally displays cheerfulness and thus uses it as an instrument, whereas the Nom predicative adjective in (158b) simply implies that Helen is a holder of the property “cheerful”. Once a property like cheerfulness is used for a specific purpose, it becomes a capacity. Although the second clause in example (158b) explicitly states a purpose, the Nom indicates that that “cheerfulness” as expressed by the adjective veselaja ‘cheerful’, is not used to achieve the goal stated in the second clause, and this is what makes the sentence infelicitous. Recall in this respect, from section 4.2.1.2 above, Instr nouns after verbs of controlled motion such as trjasti pistolet- om ‘brandish gun-INSTR,’ where the manipulation of a hand-held item has the purpose of affecting another noun referent, whereas the manipulation of corresponding Acc noun referents is geared towards the Acc referent itself. Thus, ‘brandish gun-INSTR’ is 166 perceived metonymically as threatening, whereas ‘brandish gun-ACC’ is interpreted literally as manipulation of the gun itself. Note also that if Nom adjectives in predicative constructions denote inherent properties, then Helen’s cheerfulness should in principle also hold in the situation portrayed in (158b). This example sounds awkward, however, because in this situation Helen is not using her property as a capacity. In a similar fashion, Instr predicative nouns are perceived as instrumental conduits: (159) Esli by on, Pilija, byl načal´nik-om aèroport-a, if PRTCL he-NOM Pilija-NOM was manager-INSTR.MASC.SG airport-GEN to posadil by sobačku prjamo […] vozle transporternoj lenty. then put would doggy right by conveyer belt ‘If he, Pilija, was the airport manager, he would put the doggy right by the conveyer belt.’ (translation mine – vt) [Gigolašvili. Čertovo koleso, 2007] It can be inferred from the broader context in (159) that the subject referent, Pilija, is the equivalent of a TSA agent whose duty is to inspect arriving cargos. He thinks that a service dog would facilitate the inspection. However, these dogs are in short supply at the airport, and the only way to get one is by exercising authority, which the Pilija does not have. Here, as in (157) and (158a), the Instr načal´nikom ‘manager’ is understood as a hypothetical instrument for achieving a goal, which here is to obtain a service dog. 167 6.6.3.3 Properties as functions Unlike the properties of the Instr predicative adjectives discussed in the preceding subsections, those in the examples below may not give rise to such Instr meanings as instrument or stimulus, but they are still linked to other Instr meanings through the notion “function”. (160) Gorod doplačival snačala po 100 rublej ežemesjačno [ …] city paid.extra first PREP 100 rubles monthly potom vyplat-y byli ežekvartal´n-ymi […]. then payments-NOM.PL were quarterly-INSTR.PL ‘Initially, the city paid an extra 100 rubles every month, then the payments were quarterly.’ (translation mine – vt) [Elena Kalašnikova. Novorossijskij Rabočij, 2003] While “quarterly payments” already implies the proposition “payments that (generally) are made quarterly,” it does not necessarily mean that the payments are actually dispersed in a given instance. The Instr in (160), however, implies that that the payments were in fact made every quarter. This implication arises from the general meaning of the Instr in predicative constructions, which is a capacity. A capacity in objective reality is a relational notion, which is understood as a role in relation to a specific function. For example, the capacity “doctor” presupposes actual medical practice. If a doctor does not practice medicine, the capacity “doctor” becomes a property, because any capacity implies performing the duties directly arising from the capacity. 168 7. Conclusion This chapter has investigated the meaning of the Instr in constructions with predicative adjectives in CSR. Since Russian makes a distinction between long and short adjectives, and because the emergence of the Instr with predicative adjectives was directly related to the development of long adjectives as a category, part of this chapter has provided historical background. While an analysis of predicative adjectives in Russian is more complicated because they have two oppositions: case marking and long/short forms, I have not intended in this chapter to study the meaning of Instr predicative adjectives in relation to the meaning of Nom predicative adjectives, or to the meaning of short predicative adjectives. Instead, I have endeavored to explore the nature of relationship between the meaning of the Instr in constructions with predicative adjectives and the meanings of the Instr in other Instr constructions in the language. I have demonstrated that like Instr predicative nouns, Instr predicative adjectives are perceived as capacities, while Nom predicative adjectives are discerned as properties. I have also shown that properties expressed by Instr predicative adjectives can also be as envisioned as stimuli, enablers, conduits, just as the referents of Instr nouns in other types of Instr constructions. While the model I have proposed in chapter IV for the semantic organization of the meanings of the Russian Instr may be deemed more suitable for nouns, since I have defined case as perceived physical and cognitive relations among actual people and physical objects, i.e., the referents of the NPs, I believe it has accounted for the data discussed in this chapter. Moreover, because properties expressed by adjectives modify or characterize actual individuals, physical objects, or concepts, they 169 are related to other referents or situations in space and time through their holders, that is, noun referents. 170 CHAPTER VII SUMMARY AND CONCLUSIONS The aim of this dissertation has been twofold: first, to examine the meaning of the Instr in copular predicative constructions in Russian and explain why specifically the Instr, which was a fairy late innovation for Slavic languages, occurs as an alternative to the Nom, and how the meaning of the Instr in predicative constructions resonates with the meanings of the Instr in individual Instr constructions. This investigation has been conducted on the premises that case marking is a polysemous category, and that the meanings it evokes in individual constructions are related to each other. This assumption entails expectations that the Instr in predicative constructions will reflect in one way or another the semantic regularities that underlie its use in other Instr constructions, which this study has undertaken to explore. Second, the study has endeavored to test the cross-linguistic observation that differences in the marking of predicative nominals identically to and differently from the grammatical subject reflect the differences in perception of the properties that they express as inherent vs. temporary. Because cross-linguistic patterns of variation are not uniform, in terms of grammatical means that capture the semantic distinction in question, it can be assumed that the proposed inherent-temporary distinction is only a rough approximation, and that the language-specific grammatical structures that mark predicative nominals may skew this opposition. For example, while the Russian Instr expresses a wide scope of meanings, none of them conveys the concept “temporariness”. 171 In order to answer the questions posed at the beginning of this study and test the implications of the cross-linguistic data, I have used the prototype model for studying the internal organization of the meanings of the Russian Instr. I have also proposed a general meaning of the Russian Instr, which I defined in cognitive terms as relations of contingency and contiguity. Thus, the Instr per se indicates a set of cognitive rather than semantic relations that are perceived to exist among the actual referents, individuals and physical objects, in a situation described by the verb. The individual meanings, such as instrument, cause, and standard for comparison, among others, are realized in individual Instr constructions from the interaction between the verb meaning and the lexico- semantic properties of noun referents. These individual meanings are related to each other through metaphorical and metonymical extensions. The Instr signals how these referents are related to each other and to the action, in space and time, and in what way they affect each other. Contiguity implies that two (or more) referents are spatially adjacent and as such are simultaneously involved in the action described by the verb; the referent of the Instr NP is always contiguous with the action. Contingency means that between two (or more) referents, the emergence of one referent hinges on the intention of the other. I strongly believe that defining the general meaning of the Instr in cognitively enriched spatial terms is psychologically more plausible because it reflects one of the fundamental cognitive abilities, that is, spatial cognition. Additionally, given a broad scope of meanings expressed by the Instr, it seems problematic to define its general meaning in a single semantic notion. While the cognitive definition of case adopted in this study may seem more applicable for nouns because it captures a set of perceived physical and cognitive 172 relations among actual people and physical objects, i.e., the referents of the NPs, it can nonetheless be applied to adjectives. More specifically, because properties expressed by adjectives modify or characterize actual individuals, physical objects, or concepts, they are related to other referents or situations through their holders. All things being equal, I find a radial organization as psychologically more plausible because it reflects the same categorization patterns that underly the work of human mind. For example, while the ability to fly is a less prototypical feature of a mammal, bats can be categorized as mammals based on being vertebrate and having mammary glands. In a similar fashion, while referents of the Instr NPs in constructions with verbs of possession are not prototypical instruments, they are linked to the latter via the notion of “control”. Thus, to the extent that the subject referent exercises control over the instrument in task implementation, subject referents of possession verbs have control over the entities that they own. Similarly, while the referents of the Instr NPs in constructions with verbs of controlled motion are more prototypical than the referents of the Instr NPs in predicative constructions, they nonetheless are related to each other via the notion “purpose.” Manipulation with an entity in constructions with verbs of controlled motion is always done for a purpose, such as, for example, drawing someone’s attention by waving a hand-held item. In predicative constructions, the referent of the Instr NP is perceived as a capacity, and capacity presupposes a purpose, which can be broadly understood as the duty arising from the capacity. Most importantly, the meaning of Instr in predicative constructions is related to the meaning “instrument” through notions such as “function” and “purpose”. It is function that delimitates instruments from physical objects in everyday reality. Similarly, it is function that tells apart Instr 173 capacities from Nom properties in predicative constructions. These one-to-one correspondences also reflect the fact that meaning is not discrete but flexible and open- ended. This study has demonstrated that the prototype model of meaning organization of a complex grammatical category can also account, at least in part, for diachronic change. Because prototype structures do not presuppose clear-cut boundaries, their meanings may be blurred at the edges and therefore may overlap with other grammatical categories. For example, the diachronic change that concerned verbs of naming in Russian can also be better appreciated using the prototype model. As has been pointed out in section 4.2.6.2, originally these verbs required an Acc complement, which was a syntactic calque from Ancient Greek. However, in the mid-13th century they began to take Instr complements, and in CSR only the Instr is grammatical in these constructions. It has been noted in the literature (e.g., Luraghi 2003) that the Acc in Ancient Greek implied total affectedness, which, among other things, can account for its function as direct object marker cross- linguistically, and explain its alternation with the partitive Gen in Russian, which indicates partial affectedness. In constructions with naming verbs in OCS and EES, the noun whose referent denoted an appointment (a post or capacity designated to this referent by another referent) was marked by the Acc, identically to the direct object, which was expressed by a noun denoting a referent being appointed or nominated. The implication of total affectedness does not accurately describe the relation between the three referents in a situation of naming, i.e., the nominator, the nominee, and the naming. While the referent denoting naming is totally contingent on the will of the nominator, it is 174 affected by the latter collaterally or indirectly rather than totally or partially, in that the naming, as a post, emerges parallel with the appointment. Using diachronic and cross-linguistic evidence, I have demonstrated in chapter IV the Instr in predicative constructions has as its semantic source the Instr in similative constructions, which are adverbial expressions and in which the referent of the Instr NP denotes both a mode of acting and a new capacity of the subject referent. I have proposed that the meaning “capacity” is borne out when the subject referent metaphorically adopts some of the most salient features of the referent of the Instr noun and acts in its most prototypical manner. Because this emerging capacity is contiguous with and contingent on a specific mode of acting, it unfolds together with the motion. Moreover, since the capacity develops contiguously with the motion, it is co-temporal with the motion event rather than temporary. Furthermore, I have defined the meaning of the Instr in predicative constructions as capacity, as opposed to property which is implied by the Nom predicative noun. A referent of the Instr noun in predicative construction designates both a capacity and a mode of acting. I have proposed that co-temporality and temporariness are not the same phenomenon, because temporariness refers to the quantity of time for which the property in question lasts, while co-temporality is about the quality or mode of time in which the property holds in relation to event. Thus, the distinction between the referents of Nom and Instr predicative nouns and adjectives is based, respectively, on the distinction between properties and capacities: a property holds of the subject referent itself, while a 175 capacity holds of the subject referent in relation to the event in which it is involved, and events can be either long-lasting or short-lived. The analysis of constructions with Instr nouns and adjectives suggests that while the choice between the Nom and the Instr is not always influenced by the meaning of the case marking, by and large Instr predicative nouns and adjectives display the same semantic regularities shared, to different extents, by all members of the category of the Russian Instr. Thus, properties expressed by Instr nouns and adjectives can be understood as instruments, stimuli, or conduits. The Instr in predicative constructions has a strong tendency to convey contiguity and contingency between situations. These findings are indicative of the semantic link between the meaning of the Instr in predicative constructions and the other meanings of the Russian Instr, which supports the proposal set forth in this study that Instr predicative constructions are not a random phenomenon but a full-fledged member of the family of Instr constructions. This study has also reinforced the importance of delimitating case from case marking. More specifically, while the Instrumental case as a set of perceived cognitive relations between the referents of the NPs, actual individuals and entities, in a situation described by the verb exists in languages irrespective of whether they express these relations morphologically or not, the Instr is language specific. What follows is that the Instr in one language may not convey the same set of relations as it does in another language. For example, while the English with is used to mark instruments, similar to the Russian Instr, it cannot occur with predicative nouns and adjectives. Moreover, the study has shown that case marking can be employed for purposes other than expressing cognitive relations. For example, temporal Instr expressions in Russian (utr-om ‘in the 176 morning-INSTR) do not fit the definition of case adopted in this dissertation, however, nouns denoting parts of the day occur with the Instr. I have also demonstrated that the inherent–temporary semantic opposition proposed in typological literature for alternating case marking strategies in predicative constructions is a rough approximation of a more complex phenomenon that can only be captured by examining individual languages. Even more so, the analysis of the Russian data carried out in this study is evidence that both the Nom and the Instr can denote inherent and temporary properties. These findings are in line with the conclusions made for Lithuanian in Semėnienė (2004). While languages that allow alternating case marking strategies in predicative constructions often employ adverbial marking (e.g., the Ess in Finnic or -gaa in Telugu), the innovativeness of the Slavic languages and Lithuanian is the use of specifically the Instr. This uniqueness of the Balto-Slavic languages is substantiated by the findings discussed in Szucsich (2002) that demonstrate a strong correlation between the use of the Instr in adverbial expressions and the occurrence of the Instr in predicative constructions. These findings corroborate the proposal put forward in this study that the Instr in predicative constructions is not a random phenomenon but instead arises on the same semantic grounds as the Instr in other types of Instr constructions in Russian. It has been my intention to demonstrate convincingly the applicability and comprehensiveness of the proposed model for the semantic organization of the Russian Instr, I believe it would be strengthened by a relevant psycholinguistic study. Additionally, while there is empirical data from infant and children studies indicative of instrument being an uncompromising cognitive category, there is no empirical research to 177 date that tests if the individual Instr meanings (e.g., stimulus, pathway, comparison, etc.) are indeed evoked in event and sentence processing, which is another path to explore. 178 AAPPENDIX A LIST OF ABBREVIATIONS 1 first person IFX infix 2 second person INESS inessive 3 third person INF infinitive II second gender INSTR instrumental ABS absolutive LOC locative ACC accusative MASC masculine ADV adverbial NEUT neuter CONJ conjunction NOM nominative COP copula PAST past DAT dative PFV perfective DEM demonstrative PL plural DIR directive PREP preposition DU dual PRES present EX existential PRTCL particle ESS essive POSS possessive ERG ergative REFL reflexive FEM feminine SG singular FUT future TRANSL translative GEN genitive VN verbal noun IFX infix 179 APPENDIX B TRANSLITERATION TABLE OCS/ EES CSR TRANSLITERATION ě ja (close to English a in man) é ė ǫ õ (nasalized o) jǫ j (as in yoga) + õ ъ front jer (a reduced vowel, similar to u in English put) ь back jer (a reduced vowel, similar to i in English pit) ž ž zh (as in beige) c c ts x x kh č č ch (as in church) š š sh (as in shake) šč shch ´´ hard sign ´ soft sign 180 APPENDIX C OLD CHURCH SLAVONIC AND EARLY EAST SLAVIC SOURCES Gramoty Velikogo Novgoroda i Pskova. (1949). Valka S. (Ed.) M. Evangelie ot Matfeja.[Matthew]. Accessed at Evangelie ot Marka.[Mark]. Accessed at < http://www.my-bible.info/biblio/bib_tsek/ev_mark.html> Žitie protopopa Avvakuma. Accessed at http://feb-web.ru/feb/avvakum/default.asp?/feb/avvakum/texts/jag/jag.html Zografskoe Evangelie. (11th–13th centuries). [Codex Zographensis]. Accessed at Ipat´evskaja Letopis´. (late 13th – early 14th centuries). [Hypatian Chronicle]. Accessed at http://litopys.org.ua/ipatlet/ipat.htm Komissionnyj spisok pervoj Novgorodskoj letopisi. (1440–1451). [Comission Scroll of Novgorod First Chronicle]. Accessed at http://yakov.works/acts/12/pvl/novg.htm Lavrent´evskaja letopis´. (1377). [Laurentian Chronicle]. Accessed at Ostromirovo Evangelie: [cifrovaja kopija]. [Ostromir Gospels]. Sankt-Peterburg: Rossijskaja nacional´naja biblioteka. Accessed at Pamjatnye zapiski A.V. Xrapovickogo, stats-sekretarja Imperatricy Ekateriny Vtoroj. (1862). [1990]. Accessed at https://imwerden.de/pdf/khrapovitsky_pamyatnye_zapiski_1862_1990__ocr.pdf Sinodal´nyj spisok pervoj Novgorodskoj letopisi. (1331–1352). [Synod Scroll of Novgorod First Chronicle]. Accessed at Suprasal´skaja rukopis´: Тrud [po publikacii i kommentarijam Sergeja Sever´ janova (Ed.) Vol. 1. (1904). [Codex Supraslienis]. Accessed at https://www.prlib.ru/node/372615/source 181 Smolenskaja Gramota. Spisok A (Gotlandskaja redakcija). [Smolensk Treaty of 1229]. Accessed at 182 APPENDIX D CONTEMPORARY STANDARD RUSSIAN SOURCES (from the National Corpora of the Russian Language) Aleksandrova, Natal´ja. (2010). Poslednij učenik da Vinči. Azol´skij, Anatolij. (1998). Lopušok. Novyj Mir. Berberova, Nina. (1978–1980). Železnaja ženščina. Berežkov, Valentin. (1971-1998). Rjadom so Stalinym. Borisova, S. (1979). Pri roditeljax – bez roditelej. Čelovek i zakon. Bykov Vasil´. (2001). Boloto. Voronel´, Nina. (1975–2003). Bez prikras. Vospominanija. Gigolašvili, Mixail. (2007). Čertovo koleso Evtušenko, Evgenij. (2015) [1999]. Volčij passport. Žigulin, Anatolij V. (1988). Černye kamni. Iskander, Fazil´. (1989). Sandro iz Čegema (book 3). Kalašnikova, Elena. (2003.02.18). Zarplatu povysjat zavtra, a kušat´ xočetsja sejčas. Novorossijskij Rabočij. Katanjan, Vasilij. (1998). Prikosnovenie k idolam. Katanjan, Vasilij. (1999) Lilja Brik. Žizn´. Kičin, Valerij. (2002.03.10). Operator ljubimyx fil´mov. Umer Vladimir Naxabcev. Izvestija. Klimontovič, Nikolaj. (2001). Dalee – vezde. Korolenko V. (1888). Noč´ju. 183 Krasnov, P. (1922). Na vnutrennem fronte. Lipskerov, Dmitrij. (1999). Poslednij son razuma. Molodyx, Andrej. (2014). V poiskax pozitiva. Russkij Reporter. Pavlov, Oleg. (1993). Kazennaja skazka. Rasputin, Valentin. (1970). Poslednij srok. Slavnikova, Ol´ga. (2004). Bessmertnyj. Povest´ o nastojaščem čeloveke. Solženicyn, Aleksandr I. (2016). V Kruge pervom. Strel´nikov, Boris, Vasilij Peskov. (1977). Zemlja za okeanom. Traub, Maša. (2011). Nam vyxodit´ na sledujuščej. Trifonov, Jurij. (2011) [1976]. Dom na naberežnoj. Trojanovskij, Oleg. (1997). Čerez gody i rasstojanija. Uspenskij, V. (2002) Vitgenštejn i osnovanija matematiki. Černyx, Valentin. (1972). Tri rasskaza. Junost´. Šalamov, Varlam. (1954-1961). Kolymskie rasskazy. 184 REFERENCES CITED Akimova, Galina N. (1969). Poluznamenatel´nye i znamenatel´nye svjazki v jazyke Lomonosova. Očerki po istorii russkogo jazyka i literatury 18 veka [The semi- lexical and lexical copulas in the language of Lomonosov. Essays on the history of the Russian language]. Lomonosovskie čtenija 2–3. Kazan´. Izdatel´stvo Kazanskogo universiteta, cc. 223–233. Akmajian, Adrian. (1979). Aspects of the grammar of focus in English. New York: Garland. Anderson, John M. (1971). The grammar of case: Towards a localistic theory. Cambridge: Cambridge University Press. Babby, Leonard H. (1987). Case, prequantifiers, and discontinuous agreement in Russian. Natural Language and Linguistic Theory 5. pp. 91–138. Babby, Leonard H. (1991). Noncanonical configurational case assignment strategies. Cornell Working Papers in Linguistics 9, pp. 1–55. Bacz, B. (1993). On the Instrumental in Polish. The Twentieth Lacus Forum, Lake Bluff, Illinois, pp. 418–427. Bailyn, John F. (2001). The syntax of Slavic predicate case. ZAS Papers in Linguistics 22, pp. 1–23. Bailyn, John F. (2011). The syntax of Russian. Cambridge: Cambridge University Press. Baker, Mark. (2003). Lexical categories. Verbs, nouns, and adjectives. Cambridge, UK: Cambridge University Press. Balusu, Rahul. (2016). The eventive predicator -gaa in Telugu. Linguistic Analysis, Vol. 40, 3–4. Benveniste, Émile. (1950). Actif et moyen dans le verbe. Journal de Psychology Normale et Pathologique 43, pp. 121–129. Benveniste, Émile. (1971). Problemes in general linguistics. [Translated by Mary E. Meek]. University of Miami Press. Bernštejn, Samuil B, (Ed.). (1958). Tvoritel´nyj padež v slavjanskix jazykax [The Instrumental case in the Slavic languages]. Moskva: Akademija nauk. 185 Bhat, Darbhe N. Shankara and M.S. Ningomba. (1997). Manipuri Grammar. Munich/Newcastle: Lincoln. Biro, S., A. M. Leslie. (2007). Infant’s perception of goal-directed actions: development through cue-based bootstrapping. Developmental Science, 10(3), pp. 379–398. Bondaruk, Anna. (2012). Predicational, specificational, equative and identificational copular clauses in English and Polish. Accessed online at Bondaruk, Anna. (2013). Copular clauses in English and Polish: Structure, derivation and interpretation. Lublin. Wydawnictwo KUL. Bondaruk, Anna. (2014). Characterizing and defining predicational clauses in Polish. Accessed online at Borkovskij V.I., P. S. Kuznecov. (1963). Istoričeskaja grammatika russkogo jazyka [A historical grammar of the Russian language]. Moskva. Borkovskij V.I. (1978). (ed.). Istoričeskaja grammatika russkogo jazyka. Sintaksis. Prostoe predloženie. [A historical grammar of the Russian language. Syntax. Simple clause]. Moskva. Brugman, Claudia, Lakoff, George. (1988). Cognitive topology and lexical networks. In Cottrell, G. W., S. Small, and M. K. Tannenhaus (Eds.), Lexical ambiguity resolution: perspectives from psycholinguistics, neuropsychology and artificial intelligence. San Mateo, CA: Morgan Kaufman Publishers. Brugman, Claudia M. (1981). The story of over. M.A. Thesis, University of California, Berkeley. Brugmann, Karl. (1911). Grundriss des vergleichenden Grammatik der indogermanischeen Sprachen II.2. Strassburg: Trübner. Bulaxovskij, Leonid A. (1958). Istoričeskij kommentarij k literaturnomu russkomu jazyku [A historical commentary to Contemporary Standard Russian]. Xar´kov; Kiev. Carlson, Gregory N. 1977. Reference to Kinds in English. Unpublished Ph.D thesis. University of Massachusetts at Amherst. Published 1980 by Garland Press, New York. 186 Carlson, Gregory N. (1980). Reference to Kinds in English. Garland Publishing, NY. Černov, V. I. (1983). Funkcional´nye i semantičeskie svojstva predikativnyx prilagatel´nyx v forme imenitel´nogo i tvoritel´nogo padeža [The functional and semantic properties of predicative adjectives in the Nominative and the Instrumental]. Russkij Jazyk v Škole 6, cc. 9096. Chomsky, Noam. (1980). On Binding. In Linguistic Inquiry 11. pp. 1–46. Chomsky, Noam. (1981). Lectures on Government and Binding: the Pisa lectures. Dordrecht: Foris. Chomsky, Noam. (1995). The Minimalist program. Cambridge, MA: MIT Press. Comrie, Bernard. (1976). Aspect. Cambridge: Cambridge University Press. Comrie, Bernard. (1997). The typology of predicate case encoding. In Bybee, Joan, John Haiman, and Sandra Thompson Essays on Language Function and Language Type Dedicated to T. Givon, pp. 39–50. Amsterdam-Philadelphia: John Benjamins. de Groot, Casper. (2017). Discovering the assignment: A Uralic essive typological questionnaire. In C. de Groot (Ed.), Uralic essive and the expression of impermanent state (pp. 1–28). (Typological Studies in Language; Vol. 119). Amsterdam. Cruse, Alan D. (1973). Some thoughts on agentivity. Journal of Linguistics, 9(1), pp. 11–23. Dąbrowska, Ewa. (1997). Cognitive semantics and the Polish Dative. Berlin/New York: Mouton de Gruyter. Delbrück, Berthold. (1893). Vergleichende Syntax der Indogermanischen Sprachen. Erster Theil. Strassburg: Karl J. Trübner. Demjjanow, Assinja, Anatoli Strigin. (2000). Real adjuncts in Instrumental in Russian. ZAS Papers in Linguistics 17, pp. 1–26. Demjjanow, Assinja, Anatoli Strigin (2001). Measure Instrumental in Russian. ZAS Papers in Linguistics 22, pp. 69–97. Dewell, Robert. (1994). Over again: Image schema transformations in semantic analysis. Cognitive Linguistics 5(4), pp. 351–380. 187 Dowty, David R. (1991). Thematic proto-roles and argument selection. In Language 67, pp. 547–619. Efremova, Tat´jana F. (2000). Novyj slovar´ russkogo jazyka. Tolkovo- slovoobrazovatel´nyj. [A new explanatory and word-formation dictionary of the Russian language]. M.: Russkij jazyk. Endzelīns, Jānis. (1951). Latviešu valodas gramatika. (Translated and expanded edition of Endzelīns 1922.) Riga: Latvijas valsts izdevniecība. Filip, Hana. (2001). The semantics of case in Russian secondary predication. SALT XI Proceedings, (Eds.), Rachel Hasting, Brendan Jackson i Zsofia Zvolenszky, pp. 192–211. Ithaca: CLC Publications. Fillmore, Charles. (1968). The case for case. In E. Bach and R.T. Harms (eds.). Universals of Linguistic Theory. New York: Holt, Rinehart and Winston, pp. 1– 90. Fillmore, Charles. (1977). The case for case reopened. In P. Cole and J. Sadock (eds.). Syntax and semantics: Grammatical relations, Volume 8. New York: Academic Press, pp. 59–91. Fraenkel, Ernst. (1926). Das prädikative Instrumental im Slavischen und Baltischen und seine syntaktischen Grundlagen. Archiv für slavische Philologie 40, pp. 77–117. Fraenkel, Ernst. (1928). Syntax der Litauischen Kasus. Kaunas. Freese, J. (Ed.). (1959). A Polish manual for self-tuition containing a concise grammar with exercises. London: Kegan Paul, Trench, Trübner. Freidin, Robert, Rex A. Sprouse. (1991). Lexical Case phenomena. In Principles and parameters in comparative grammar. R. Freidin (Ed.). Cambridge: MIT Press. Fritz, Matthias. (2003). Altgriechisch. In Roelcke, Thorsten (Ed.), Variationstypologie / Variation Typology. Ein sprachtypologisches Handbuch der europäischen Sprachen in Geschichte und Gegenwart, pp. 798–822. Berlin, New York: Walter de Gruyter. Ganenkov, Dmitrij S. (2002). Tipologija padežnyx značenij: Semantičeskaja zona prolativa. Issledovanija po teorii grammatiki. Vol. 2: Grammatikalizacija prostranstvennyx značenij v jazykax mira. [A typology of case meanings: A semantic zone of prolative. Research on theoretical grammar. Vol. 2: Grammaticalization of spatial meanings in languages of the world]. Moscow: Russkie slovari, cc. 35–56. 188 Gasparov, Bоris M. (1997). Upotreblenie kratkix i polnyx form prilagatel´nogo. Trudy po russkoj i slavjanskoj filologii. [On the use of short and long adjectives. Works on Russian and Slavic philology]. Lingvistika. Novaja serija. I. Tartu. Giro-Veber, Margaret. (1996). Bisinxronnyj metod opisanija prilagatel´nogo v predikativnoj pozicii v sovremennom russkom jazyke. Teorija funkcional´noj grammatiki: Količestvennost´. Kačestvennost´. [Bisynchronic method of a description of predicative adjectives in Contemporary Standard Russian. A theory of functional grammar. Quantity. Quality]. SPb. Nauka, cc. 65–79. Giro-Veber, Margaret. (2007). Suščestvitel´noe v funkcii imennogo skazuemogo v sovremennom russkom jazyke: Možno li ešče govorit´ o semanticeskom protivopostavlenii “imenitel´nyj vs. tvoritel´nyj”? [Nouns as nominal predicates in Contemporary Standard Russian: Can one still talk about the semantic distinction “the Nominative vs. the Instrumental?”].Voprosy Jazykoznanija 1, cc. 18–26. Grenoble, Lenore A. (2010). Contact and the development of the Slavic languages. In Hickey, Raymond (ed.), The Handbook of language contact, Wiley Online Library, pp. 581–597. Gruber, Jeffrey S. (1965). Studies in lexical relations. Doctoral dissertation, MIT. [Published as Lexical structures in syntax and semantics, North-Holland, Amsterdam, 1976.] Grudneva, Galina N. (1958). Sostavnoe i složnoe skazuemoe v sovremennom russkom jazyke [The nominal and complex predication in Contemporary Standard Russian]. Voprosy Russkogo Jazykoznanija 3. L’vov, cc. 59–180. Gustavsson, Sven. (1976). Predicative adjectives with copula “byt´” in Modern Russian. Stockholm. Halliday, Michael A. K. (1967). Notes on transitivity and theme in English, part I. Journal of Linguistics 3, pp. 37–81. Haspelmath, Martin, Oda Buchholz. (1998). Equative and similative constructions in the languages of Europe. In Eurotyp, 3, Adverbial Constructions in the Languages of Europe. Berlin, Boston: De Gruyter Mouton, pp. 277–334. Heine, Berndt. (1997). The cognitive foundations of grammar. Oxford: Oxford University Press. Hentschel, Gerd, Masako Ueda. (1998). The interaction between clause-level parameters and context in Russian morphosyntax: Genitive of negation and predicate adjectives. Russian Linguistics 22, pp. 333–340. 189 Hernik, M., G. Csibra. (2015). Infants learn enduring functions of novel tools from action demonstrations. Journal of Experimental Child Psychology, 130, pp. 176–192. Heycock, Caroline, Anthony Kroch. (1999). Pseudoclefts connectedness: Implications for the LF interface level. Linguistic Inquiry, 30 (3), pp. 365–397. Higgins, Roger. (1979). The pseudo-cleft construction in English. New York: Garland. Hinterhölzl, Roland. (2001). Semantic constraints on case assignment in secondary adjectival predicates in Russian. ZAS Papers in Linguistics 22, pp. 99–112. Hjelmslev, Louis. (1935). La catégorie des cas: Étude de grammaire générale. Acta Jutlandica VII,1. Århus: Universitetsforlaget i Aarhus. Reprinted in (1972) La catégorie des cas: Étude de grammaire générale. Munich: Wilhelm Fink Verlag. Hofer, T., P. Hauf, and G. Aschersleben. (2005). Infants’ perception of goal-directed actions performed by a mechanical device. Infant Behavior and Development, 28(4), pp. 466–480. Holvoet, Axel. (2004). On the marking of predicate nominals in Baltic. In Baldi, Philip and Pietro U. Dini (Eds.). Studies in Baltic and Indo-European Linguistics in Honor of William R. Schmalstieg, Current Issues in Linguistic Theory vol. 254. Amsterdam/Philadelphia: John Benjamins, pp. 75–90. Israeli, Alina. (2007). Nominative and instrumental variation of adjectival predicates with Russian copula byt´: reference, limitation, and focalization. Cognitive Paths into the Slavic Domain (Cognitive Linguistics Research 38). Berlin, New York, pp. 21–53. Jackendoff, Ray. (1972). Semantic interpretation in Generative grammar. Cambridge, MA: MIT Press. Jackendoff, Ray. (1976). Toward an explanatory semantic representation. In Linguistic Inquiry 7(1), pp. 89–150. Jackendoff, Ray. (1983). Semantics and cognition. Cambridge, MA: MIT Press. Jakobson, Roman. (1936). Bieträge zur allgemeinen Kasuslehre: Gesamtbedeutungen der russischen Kasus. In Travaux du Cercle Linguistique de Prague 6, pp. 240–88. [Reprinted in 1966 in E.P. Hamp, F.W. Householder, and R. Austerlitz (eds.). Readings in linguistics II. Chicago: The University of Chicago Press, pp. 51–89. Reprinted in 1971 in Selected Writings II. Den Haag: Mouton, pp. 23–72.] 190 Jakobson, Roman. (1958) [1971]. Morphological inquiry into Slavic declension: Structure of Russian case forms [text in Russian, summary in English in 1971 version]. In R. Jakobson, Selected Writings, Vol. II: Words and Language. The Hague: Mouton, pp. 154–183. Jakobson, Roman O. (1964). Towards a linguistic typology of aphasic impairments. CIBA Foundation Symposium on Disorders of Language, (Eds.) A. V. S. de Reuck, Maeve O’Connor, pp. 21–42. Janda, Laura. (1990). The radial network of a grammatical category – its genesis and dynamic. structure. Cognitive Linguistics 1-3: 269–288. Janda, Laura. (1993). A geography of case semantics. The Czech Dative and the Russian Instrumental. Berlin – New York. Jovanovic, B., I. Kiraly, B. Eslner, B. Gergely, G. Prinz & G. Aschersleben. (2007). The role of effects for infants’ perception of action goals. Psychologia, 50(4), pp. 273– 290. Istrina, Еvgenija S. (1923). Sintaksičeskie javlenija Sinodal´nogo spiska I Novgorodskoj letopisi. [Syntactic phenomena of the Synod Scroll of the Novgorod First Chronicle]. Izvestija Otdelenija Russkogo Jazyka i Slovesnosti 24 (2), cc. 207– 239. Kalinina, Elena. (1993). Sentences with non-verbal predicates in the Sogratl dialect of Avar. In Kibrik, A. E. (Ed.), The noun phrase in the Andalal dialect of Avar as spoken at Sogratl. Strassbourg: EUROTYP Working Papers, Theme 7: Noun Phrase Structure, Working Paper No. 18, pp. 90–104. Karskij, Evfimij F. (1929). Nabljudenija v oblasti sintaksisa Lavrent´evskogo spiska letopisi [Obseravtions on the syntax of the Laurentius Chronicle]. Kazavčinskaja, Natalija A. (1990). Funkcii kratkix i polnyx form prilagatel´nyx v pozicii predikata v sovremennom russkom jazyke. [The functions of short and long adjective forms in predicate positions in Contemporary Standard Russian]. Aftorefereat dissertacii. L. Klein, Jared, Brian Joseph, Matthias Fritz (Eds). (2017). Handbook of comparative and historical Indo-European Linguistics. Mouton Korotaeva, M. V. (1996). Poluznamenatel´nye svjazočnye glagoly v sovremennom russkom jazyke. Učebnoe posobie. [Semi-lexical copular verbs in Contemporary Standard Russian. A manual]. Kirov: Izdatel´stvo VGPU. 191 Krasovitsky et al. (2008). Alexander Krasovitsky, Alison Long, Matthew Baerman, Dunstan Brown, Greville Corbett. Predicate nouns in Russian. Russian Linguistics 32:2, pp. 99–113. Kreitzer, Anatol. (1997). Multiple levels of schematization: A study in the conceptualization of space. Cognitive Linguistics 8(4), pp. 291–325. Krishnamurti, Bhadriraju, Gwynn, John Peter Lucius. (1985). A grammar of modern Telugu. Oxford: Oxford University Press. Kustova, Galina I. (2005). Situativnaja norma v značenii prilagatel´nyx, narečij, predikativov: Semantika i konstrukcii. [Contextual norm in the meaning of adjectives, adverbs, predicates: Semantics and constructions]. Komp´juternaja lingvistika i intellektual´nye texnologii. Trudy meždunarodnoj konferencii “Dialog–2005” [Computational linguistics and intellectual technologies. The working papres of the international conference “Dialogue”] cc. 316–320. Kustova, Galina I. (2017). Russkie prilagatel´nye: Semantika, konstrukcii, tekst. [Russian adjectives. Semantics, constructions, text]. M., SPb. Lakoff, George. (1968). Instrumental adverbs and the concept of deep structure. Foundations of Language 4 (1), pp. 4–29. Lakoff, George. (1977). Linguistic gestalts. Papers from the Thirteen Regional Meeting of the Chicago Linguistic Society. Chicago: Chicago Linguistic Society, pp. 236- 287. Langacker, Ronald W. (1987). Foundations of cognitive grammar, vol. 1: Theoretical prerequisites. Stanford, CA: Stanford University Press. Langacker, Ronald W. (1991). Concept, image, and symbol. Berlin/New York: Mouton de Gruyter. Langacker, Ronald W. (2008). Cognitive grammar: A basic introduction. Oxford: Oxford University Press. Lees, Aet. (2015). Case alternations in five Finnic languages: Estonian, Finnish, Karelian, Livonian and Veps. Volume 13, Brill’s Studies in Language, Cognition and Culture, Brill. Lehiste, Ilse. (1972). “Being” and “having” in Estonian. In: Verhaar, J.W.M. (Ed.), The verb “be” and its synonyms. Part 5. Dordrecht: Reidel, pp. 207–224. 192 Lehmann, Christian and Yong-Min Shin. (2005). The functional domain of concomitance. A typological study of Instrumental and Comitative relations. In Studia Typologica 7, pp. 9–104. Lehmann, Winfred P. (1974). Proto-Indo-European syntax. Austin: University of Texas Press. Lehtinen, Mari. (1963). Basic course in Finnish. Bloomington: Indiana University/The Hague: Mouton. Lomtev, Timofej P. (1956). Očerki po istoričeskomu sintaksisu russkogo jazyka. [The essays on the historical syntax of the Russian language]. M.: Izdatel´stvo Moskovskogo universiteta. Lönngren, Lennart. (1993). Častotnyj slovar´ sovremennogo russkogo jazyka. [A frequency dictionary of Contemporary Russian language]. Uppsala 1993. Acta Universitatis Upsaliensis, Studia Slavica Upsaliensia 32. Lunt, Horace G. (1974). Old Church Slavonic Grammar. Mouton. Luraghi, Sylvia. (2001). Some Remarks on Instrument, Comitative, and Agent in Indo- European. In Sprachtypologie und Universalienforshung 54, pp. 385–401. Luraghi, Sylvia. (2003). On the meaning of prepositions and cases. A study of the expression of semantic roles in Ancient Greek. Amsterdam: Benjamins. Lutkat, Florence-Silvia, Cornelius Hasselblatt. (1993). Estnisch intensive. Hamburg: Bibliotheca Baltica. Mal´chukov, Andrej L. (2005). Case patterns splits, verb types and construction competition. In M. Amberber, and H. de Hoop (eds.). Competition and variation in natural languages: The case for case. Amsterdam: Elsevier Sciences, pp. 73– 119. Markman, Vita G. (2008). The case of predicates (revisited): Predicate Instrumental in Russian and its restrictions. Journal of Slavic Linguistics 16 (2), pp. 187–246. Matushansky, Ora. (2012). On the internal structure of case in Finno-Ugric small clauses. Finno-Ugric Languages and Linguistics, Vol.1, 1-2, pp. 3–43. Meillet, Antoine. (1906–8). La phrase nominale en indo-européen. Memoires de la Société Linguistique de Paris 14, pp. 1–26. 193 Meira, Sérgio, Spike Gildea. (2009). Property concepts in the Cariban family: Adjectives, adverbs and/or nouns?’ In: Wetzels, W.L. (ed.) ThelLinguistics of endangered languages – contributions to morphology and morphosyntax, pp. 95–133. Utrecht: LOT Occasional Series. Mendoza, Julia. (1998). Syntaxis. Manual de lingüística Indoeuropea III. In Adrados, Francisco R., Alberto Bernabé, and Julia Mendoza (Eds.). Madrid: Ediciones Clásicas, pp. 141–246. Mikkelsen, Line. (2011). Copular clauses. In Maienborn, Klaus, and Paul Portner (Eds.) Semantics: An international handbook of natural language meaning, volume 2, pp. 1805–1829. Berlin: Mouton de Gruyter. Milsark, Gary L. (1974). Existential Sentences in English. Unpublished Ph.D thesis, Massachusetts Institute of Technology. Mixajlov, Nikita. (2012). Tvoritel´nyj padež v russkom jazyke 18 veka [The Instrumental case in the Russian language of the 18th century]. Published doctoral thesis. Acta Universitatis Upsaliensis. Studia Slavica Upsaliensia 47. Uppsala. Mjasnikov, Georgij F. (1959). Predikativnoe ispol´zovanie polnyx i kratkix prilagatel´nyx v sovremennyx govorax russkogo jazyka [The predicative use of long and short adjectives in contemporary dialects of the Russian language]. Učebnye zapiski Kurganskogo pedagogičeskogo instituta II, cc. 55–83. Moser, Michael. (1994). Der prädikative Instrumental. Aus der historischen Syntax der Nordostslavischen. Von den Anfängen bis zur petrinischen Epoche. Frankfurtam Main. Mrázek, Roman. (1961). Užívání prostého instrumentálu v současné ruštině [On the use of prepositionless Instrumental in the Russian language]. Kapitoly ze srovnávací mluvnice ruské a české, 2: Studie syntaktické, CSAV [Working papers on the comparative grammar of Russian and Czech, 2. Syntactic studies, CSAV], Prague, cc. 153–256. Mrázek, Roman. (1964). Sintaksis russkogo tvoritel´nogo. (Strukturno-sravnitel´noe issledovanie) [The Syntax of the Russian Instrumental. (A structural and comparative research)]. Prag: Státni pedagogicke nakladatelstvi. Narrog, Heiko. (2009). Varieties of Instrumental. Malchukov Andrej and Andrew Spencer (Eds.), The Oxford handbook of case. Oxford University Press. pp. 593– 600. 194 Nefedova, Tat´jana P. (1978). Semantiko-sintaksičeskie svojstva i stilističeskoe naznačenie kratkix prilagatel´nyx v sovremennom russkom jazyke (na materiale jazyka gazety) [The semantic and syntactic properties and stylistic function of short adjectives in Contemporary Standard Russian (on the language of newspapers)]. Aftoreferat dissertacii. Rostov-na-Donu. Nichols, Johanna. (1981). Predicate nominals: A partial surface syntax of Russian. Berkeley. Nichols, Johanna. (1985). Padežnye varianty predikativnyx imen i ix otraženie v russkoj grammatike [The case variants of predicative nominals and their reflection in the Russian grammar]. Novoe v Zarubežnoj Lingvistike 14. M., cc. 342–387. Nikiforidou, Kiki. (1991). The meanings of the genitive: A case study in semantic structure and semantic change. Cognitive Linguistics 22, 149–205. Nikunlassi, Ahti. (1993). Imenitel´nyj ili tvoritel´nyj? Sintaksičeskie prilagatel´nye pri poluznamenatel´nyx glagolax v russkom jazyke: Problemy vybora padeža [The Nominative or the Instrumental? Syntactic adjectives after semi-lexical verbs in the Russian language: The problems of case choice]. Helsinki: University of Helsinki. Palancar, Enrique L. (1999). Instrumental prefixes in Amerindian languages: An overview. To their meanings, origins, and functions. In Sprachtypologie und Universalienforshung 52, pp. 151–166. Palancar, Enrique L. (2001). Emergent markers: The relations of instruments, causes, and agents in grammatical constructions. In Sprachtypologie und Universalienforshung 52, pp. 365–384. Patokova, O. V. (1929). K istorii razvitija tvoritel´nogo predikativnogo v russkom literaturnom jazyke [On the history of the development of the predicative Instrumental in the Russian Literary language]. Slavia 8, cc. 1–37. Pavlov, V. M. (1960). O razrjadax imen prilagatel´nyx v russkom jazyke [On the classes of adjectives in the Russian language]. Voprosy Jazykoznanija 2, cc. 65–70. Pečenyj, Aleksandr P. (2012). Tvoritel´nyj padež. Materialy dlja proekta korpusnogo opisanija russkoj grammatiki (http://rusgram.ru) [The Instrumental case. Materials for the project of the corpus description of the Russian grammar]. Na pravax rukopisi. M. Pollard, Carl, and Ivan A. Sag. (1994). Head-Driven phrase structure grammar. Chicago and London: University of Chicago Press. 195 Popova, Larisa V. (2011). O meste svjazki v grammatičeskoj sisteme russkogo jazyka [On the place of a copula in the grammatical system of the Russian language]. Izvestija Penzenskogo Gosudarstvennogo Pedagogičeskogo Universiteta im. V. G. Belinskogo. Gumanitarnye nauki 23, cc. 215–218. Potebnja, Aleksandr A. (1888) [1954]. Iz zapisok po russkoj grammatike [From the notes on the Russian grammar]. Moskva. Raxilina, Ekaterina V. (1998). Semantika otymennyx prilagatel´nyx [The semantics of denominative adjectives]. Liki Jazyka: K 45-letiju naučnoj dejatel´nosti E. A. Zemskoj. M. Raxilina Ekaterina, Julija Kuznecova. (2010). Grammatika konstrukcij: teorii, storonniki, blizkie idei [Construction grammar: theories, proponents, relative ideas]. Elena V. Raxilina (red.). Lingvistika konstrukcij. M.: Azbukovnik. cc. 18–82. Raxilina Ekaterina, Elena Tribušinina. (2011). The Russian instrumental-of-comparison: Constructional approach. Slavic Linguistics in a Cognitive Framework. (Eds.) M. Grygiel, L. Janda. Frankfurt am Main etc., pp. 145–174. Richardson, Kylie R. (2001). What secondary predicates in Russian tell us about the link between tense, aspect and case. ZAS Papers in Linguistics 26, pp. 1–25. Røed, Ragnar. (1966). Zwei Studien über den prädikativen Instrumental im Russischen. Oslo. Rogava, Gеorgij V., Keraševa, Zajnab I. (1966). Grammatika adygejskogo jazyka [A grammar of the Adyghe language]. Krasnodar: Krasnodarskoje knižnoe izdatelstvo. Roy, Isabelle A. (2006). Non-verbal predications: A syntactic analysis of predicational copular sentences. Doctoral dissertation. University of Southern California. Roy, Isabelle A. (2013). Nonverbal predication: Copular sentences at the syntax- semantics interface. Oxford: Oxford University Press. Rudnev, Dmitrij V. (2014). Svjazočnye glagoly v russkom jazyke 17-18 vekov [Copulative verbs in the Russian language of the 17th-18th centuries. PhD dissertation]. Kandidatskaja dissertacija. SPb. Saltarelli, Mario. (1988). Basque. London: Croom Helm. Šaxmatov, Aleksej A. (1941). Sintaksis russkogo jazyka [The syntax of the Russian language]. 2nd edition. Leningrad. 196 Schuchardt, Hugo. (1923). Primitiae linguae Vasconum: Einführung ins Baskische. Halle (Saale): Niemeyer. Šeljakin, Mixail A. (2010). Očerki po pragmatike russkogo jazyka [Essays on the grammar of the Russian language]. Moskva: Russkij jazyk Media: Drofa. Semėnienė, Loreta. (2004). Predikatinio daiktavardžio vartojimas: vardininkas vs. įnagininkas [The predicative use of nouns in Lithuanian: nominative vs. instrumental]. Acta Linguistica Lithuanica, 50, pp. 99–125. Senn, Alfred. (1966). Handbuch der litauischen Sprache. Heidelberg:Winter. Senn, Alfred. (1974). Kleine litauische Sprachlehre. Hildesheim/New York: Georg Olms. Serdobolskaya, Nataliya, Juliya Kuznecova. (2007). The semantics of Adyghe instrumental case. Conference on the Languages of the Caucasus Max Planck Institute for Evolutionary Anthropology. Leipzig. Škarupo, Z. V. (1957). Kratkie i polnye prilagatel´nye s grammatičeskim značeniem postojannogo priznaka (na materiale proizvedenij A. M. Gor´kogo) [Short and long adjectives with the meaning of a permanent property (based on the corpus from the literary works of A.M. Gor´kij)]. Naučnye zapiski Odesskogo gosudarstvennogo pedagogičeskogo instituta. T 18, cc. 63–73. Šmelev, Dmitrij N. (1976). Sintaksičeskaja členimost´ vyskazyvanija v sovremennom russkom jazyke [The syntactic dividedness of the utterance in Contemporary Standard Russian]. M.: Nauka. Stassen, Leon. (2001). Nonverbal predication in the Circum-Baltic languages. In Dahl, Östen and Marija Koptjevskaja-Tamm (Eds.), Circum-Baltic Languages: Volume 2: Grammar and Typology. Studies in Language Companion Series 55, pp. 569– 590. Stolz, Thomas. (1996). Some instruments are really good companions – some are not: On syncretism and the typology of instrumentals and comitatives. In Theoretical Linguistics 23, pp. 113–200. Stolz, Thomas. (2001). Comitatives vs. instrumentals vs. agents. In W. Bisang (ed.). Aspects of Typology. Studio Typological. Berlin: Akademie Verlag, pp. 153– 174. Sreznevskij, Izmail I. (1963). Slovar´ drevnerusskogo jazyka [A dictionary of the Old Russian language]. Reprintnoe izdanie. Moscow. 197 Švedova N. V. (1948). Vozniknovenie i rasprostranenie predikativnogo upotreblenija člennyx prilagatel´nyx v russkom literaturnom jazyke 15–18 vv [The emergence and distribution of long adjectives in the Russian Literary language of the 15th- 18th centuries]. Doklady i soobčšenija Instituta russkogo jazyka. M.: L. Vypusk 1, cc. 102–126. Švedova N. Ju. (1980). Russkaja grammatika [The Russian grammar]. Vol. I-II. (1980). M.: Nauka. Švedova N. Ju. (1980). Podčinitel´nye svjazi slov i slovosočetanija. Prostoe predloženie [The subordinate relations of words and expressions. A simple clause]. Russkaja grammatika, vol. 2. Sintaksis [The Russian grammar, vol. 2. Syntax]. M.: Nauka, cc. 13–92. Sweetser, Eve. (1990). From etymology to pragmatics. Cambridge: Cambridge University Press. Szucsich, Luka. (2002). Nominale Adverbiale im Russischen. Syntax, Semantik und Informationsstruktur. München. Taylor, John R. (1989). Linguistic categorization. Prototypes in linguistic theory. Oxford: Oxford University Press. Taylor, John R. (1995). Linguistic Categorization. Second Edition. Oxford: Oxford University Press. Timberlake, Alan. (1986). The semantics of case in Russian predicate complements. Russian Linguistics 10:2, pp. 137–165. Timberlake, Alan. (2004). A Reference Grammar of Russian. Cambridge. Tolstoj, Nikita I. (1954). Kratkie i polnye prilagatel´nye v staroslavjanskom jazyke [Short and long adjectives in Early East Slavic]. Avtorefereat dissertacii. Moskva. Tomasello, Michael. (1987). Learning to use prepositions: a case study. Journal of Child Language 14(1), pp. 79–98. Träuble, B, S. Pauen. (2007). The role of functional information for infant categorization. Cognition, 105(2), pp. 362–379. Tyler, Andreas, and Evans, Vyvyan. (2003). The semantics of English prepositions. Cambridge: Cambridge University Press. 198 Vakareliyska, Cynthia M. (1994). A model of the dative/ accusative opposition for Slavic languages, based on the data from aphasia. S´´postavitelno Jezikoznanie XIX (3- 4), pp. 7–15. Vaillant, André. (1974). Grammaire comparée des langues slaves. La Formation des noms. Paris : Klincksieck. Van Valin, Robert D. Jr. (1990). Semantic parameters of split intransitivity. In Language 66, pp. 221–260. Veenker, Wolfgang. (1967). Die Frage des finnougrischen Substrats in der russischen Sprache. Bloomington: Indiana University/The Hague: Mouton. Verhagen, Arie. (1992). Praxis of linguistics: Passives in Dutch. Cognitive Linguistics 3: 301–342. Vojejkova, Marija D., Pupynin Jurij A. (1996). Predikativnaja kačestvennost´. Teorija funkcional´noj grammatiki. Količestvennost´. Kačestvennost´ [Predicative quality. A theory of functional grammar. Quantity. Quality]. SPb.: Nauka, cc. 53–65. Vol´f, E. M. (1978). Grammatika i semantika prilagatel´nogo [The grammar and semantics of adjectives]. M. Vsevolodova, Maja V. (1972). Upotreblenie polnyx i kratkix prilagatel´nyx. Stat´ja vtoraja. [On the use of long and short adjectives. Second article]. Russkij Jazyk za Rubežom 1, cc. 59–64. Wierzbicka, Anna. (1980). The case for surface case. Ann Arbor: Karoma. Wittgenstein, Ludwig. (2001) [1953]. Philosophical investigations. Blackwell Publishing. Xalbad, Taras. X. (1975). Vyraženie kategorij opredelennosti i neopredelennosti v abxazo-adygskix jazykax [The representation of definiteness and indefiniteness in the Abkhaz and Adyghe languages]. Tbilisi: Mecniereba. Xodova, K. I. (1958). Tvoritel´nyj prevraščenija i sravnenija [The Instrumental of transformation and comparison]. In Bernštejn, Samuil B. (Ed.). Tvoritel´nyj padež v slavjanskix jazykax [The Instrumental case in the Slavic languages]. Moskva: Akademija nauk SSSR. cc. 181–192. Xodova K. I. (1960). Značenija tvoritel´nogo bespredložnogo v staroslavjanskom jazyke [The meanings of the prepositionless Instrumental in Early East Slavic]. Učenye zapiski Instituta slavjanovedenija AN SSSR, cc. 101–158. 199 Zekox, Učužuk S. (2002). Adygejskaja grammatika. [A grammar of Adyghe]. Majkop: Adygeja. Zel´dovič, Gennadij M. (2005). Russkoe predikativnoe imja: Soglasovannaja forma, tvoritel´nyj padež. [Russian predicative nominals: Agreement, Instrumental case]. Voprosy Jazykoznanija 4, cc. 21–38. 200