Principal aspects of xinaliq phonology and morphosyntax

This dissertation is a description of Xinaliq, a Northeast (Nakh-Daghestanian) Caucasian language spoken primarily in the village also called Xinaliq, which lies at an elevation of 7,000 feet in the Kuba district of Azerbaijan, near the border with Daghestan. Currently there are approximately 1,500 residents in the village. Most of them are bilingual. Use of the Xinaliq language is decreasing rapidly due to many economic and social factors. The aim of this dissertation is to contribute to linguistics scholarship in several ways: (i) Xinaliq offers rich typological traits that have been understudied, due to the relatively sparse linguistic analysis of the Northeast Caucasian languages; (ii) Xinaliq offers many resources for historical linguistics, providing material needed for the study of language change, language contact, and possible genetic relationships among languages in this region; (iii) cultural description of the region will benefit anybody interested in this ancient community, its members and their language. The grammar, although based on a linguistic analysis informed by current linguistic theory and advances in language typology, is theory neutral. An attempt was made to analyze, interpret and synthesize phonological and morphological patterns in formats that will be useful both to linguists and to researchers from other fields, as well as Xinaliq community members. In addition to the grammar, the dissertation describes the historical and cultural background of the language and the speakers of the language. This dissertation is primarily based on data collected during several field trips undertaken by the author between the years 2009 - 2013. It is supplemented with data from the initial visit to Xinaliq village with Dr. Harris in 2009. Limited materials available from previous research on the language have also been researched and studied. The fieldwork consisted of long-term stays in Xinaliq village, eliciting data, collecting texts, making video and audio recordings and participating in daily life activities with the community members. Different socioeconomic groups, including men, women and children of various ages were engaged in the process. This dissertation has been partially supported by a grant from National Science Foundation, DoBes Volkswagen Foundation and The American Research Institute of the South Caucasus.

PRINCIPAL ASPECTS OF XINALIQ PHONOLOGY AND MORPHOSYNTAX by Tamrika Khvtisiashvili A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Linguistics The University of Utah August 2013 Copyright © Tamrika Khvtisiashvili 2013 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Tamrika Khvtisiashvili has been approved by the following supervisory committee members: Mary Ann Christison , Chair 05/24/13 Date Approved Lyle Campbell , Member 05/24/13 Date Approved Alice Harris , Member 06/07/13 Date Approved Maria Polinsky , Member 05/24/13 Date Approved Marianna Di Paolo , Member 05/29/13 Date Approved and by Edward Rubin , Chair of the Department of Linguistics and by Donna M. White, Interim Dean of The Graduate School. ABSTRACT This dissertation is a description of Xinaliq, a Northeast (Nakh-Daghestanian) Caucasian language spoken primarily in the village also called Xinaliq, which lies at an elevation of 7,000 feet in the Kuba district of Azerbaijan, near the border with Daghestan. Currently there are approximately 1,500 residents in the village. Most of them are bilingual. Use of the Xinaliq language is decreasing rapidly due to many economic and social factors. The aim of this dissertation is to contribute to linguistics scholarship in several ways: (i) Xinaliq offers rich typological traits that have been understudied, due to the relatively sparse linguistic analysis of the Northeast Caucasian languages; (ii) Xinaliq offers many resources for historical linguistics, providing material needed for the study of language change, language contact, and possible genetic relationships among languages in this region; (iii) cultural description of the region will benefit anybody interested in this ancient community, its members and their language. The grammar, although based on a linguistic analysis informed by current linguistic theory and advances in language typology, is theory neutral. An attempt was made to analyze, interpret and synthesize phonological and morphological patterns in formats that will be useful both to linguists and to researchers from other fields, as well as Xinaliq community members. In addition to the grammar, the dissertation describes the historical and cultural background of the language and the speakers of the language. iv This dissertation is primarily based on data collected during several field trips undertaken by the author between the years 2009 - 2013. It is supplemented with data from the initial visit to Xinaliq village with Dr. Harris in 2009. Limited materials available from previous research on the language have also been researched and studied. The fieldwork consisted of long-term stays in Xinaliq village, eliciting data, collecting texts, making video and audio recordings and participating in daily life activities with the community members. Different socioeconomic groups, including men, women and children of various ages were engaged in the process. This dissertation has been partially supported by a grant from National Science Foundation, DoBes Volkswagen Foundation and The American Research Institute of the South Caucasus. This grammar is dedicated to the Xinaliq community of Azerbaijan. Thank you. TABLE OF CONTENTS ABSTRACT………………………………………………………………………….…iii LIST OF TABLES………………………………………………………………….…...xi LIST OF FIGURES…………………………………………………………………….xiv LIST OF ABBREVIATIONS……………………………………………..………...…xvi ACKNOWLEDGMENTS………………………………………………………...….....xx 1. INTRODUCTION……………………………………………………………...……..1 1.1. The Aim of this Dissertation………………………….…………………………..1 1.2. Organization of the Grammar…………………………….………………………2 1.3. Typological Overview……………………………………………………………3 1.4. Fieldwork and Methodology……………………………………………………...4 1.5. The Name…………………………………………………….…………………...5 2. XINALIQ: LANGUAGE, PEOPLE AND GEOGRAPHY……………………….…..6 2.1. The Xinaliq Language……………………………………………………………6 2.2. The Xinaliq People…………………………………………………………...…12 2.3. Xinaliq Village…………………………………………………….……….……14 3. PHONOLOGY…………………………………………………...………………….20 3.1. Consonants………………………………………………………..…………….20 3.1.1. Some Relevant Details Concerning Phonemes……………………..……28 3.1.2. Short and Long Voiceless Phonemes…………………………………….29 3.1.3. Allophonic Variations of Xinaliq Consonants……………………….….30 3.1.3.1. Labialization………………………………………………….….30 3.1.3.2. Voicing of /h/……………………………………………….……31 3.1.4. Consonant Distribution…...………………………………….…………..31 3.1.4.1. Word-Initial………………………………………………………31 vii 3.1.4.2. Word-Final……………………...……………….…….…………32 3.1.4.3. Word-Medial Position……………………………………………32 3.2. Vowels……………………………………………………………………….….33 3.2.1. Allophonic Variations in Xinaliq Vowels…………..……………………34 3.2.1.1. Nasalization………………………………………………….…..34 3.2.1.2. Nasal Assimilation……...………………………………………..34 3.2.1.3. Voiceless Vowels……...……………………………..…………..35 3.2.1.4. High Vowel Centralization……..…………..…………………....35 3.2.1.5. Vowel Syncope……...………………..………………...………..36 3.2.1.6. Vowel Harmony……...……………..…………...……........…….36 3.3. Stress…………………………………………………………………...…….38 3.4. Phonological Rules………………………………………………..…………39 3.4.1. Vowel Deletion………………………………………………………..39 3.4.2. Vowel Assimilation………………………………...………………….40 3.4.3. Glide Insertion……………………………………..………………….41 3.4.4. Assimilation………………………………………….………………..42 3.4.5. Free Variation…………………………………………………….…...43 4. THE NOUN………………………………………………………..………………50 4.1. Noun Class……………………………………………………………..…….….50 4.1.1. Class Markers (CM)..…………………………………………..……...…53 4.1.1.1. Series 1 (CM1)……………………………………………...…....56 4.1.1.1.1. Class Markers Series 1 Class MASC. (CM1.MASC.SG).57 4.1.1.1.2. Class Markers Series 1 Class FEM. (CM1.FEM.SG)…....58 4.1.1.1.3. Class Markers Series 1 Class AN (CM1.AN.SG).…….....61 4.1.1.1.4. CM1.MASC.PL and CM1.FEM.PL…………………......62 4.1.1.1.5. CM1.INAN.SG……………………………………......…63 4.1.1.1.6. CM1.AN.PL………………………………………....…...63 4.1.1.1.7. CM1.INAN.PL……………………………………...……64 4.1.1.2. Series 2 (CM2)……………………………………..……..….…..64 4.1.1.2.1. CM2.MASC.SG………………………..…………...……64 4.1.1.2.2. CM2.FEM.SG…………………………….………...……65 4.1.1.2.3. CM2.AN.SG………………………………….……...…..66 4.1.1.2.4. CM2.MASC.PL……………………………………….....67 4.1.1.2.5. CM2.FEM.PL……………………………...……….……67 4.1.1.2.6. CM2.INAN.SG……………………………………....…..67 4.1.1.2.7. CM2.AN.PL………………………………………......….68 4.1.1.3. Series 3 (CM3)……………………………………….……..…....68 4.1.1.3.1. CM3.M.SG…………………………………………..…...68 viii 4.1.1.3.2. CM3.F.SG……………………………..…………..….….68 4.1.1.3.3. CM3.AN.SG……………………………..…………..…..69 4.1.1.3.4. CM3.M.PL………………………….……….………..….69 4.1.1.3.5. CM3.F.PL………………………………………..….…...69 4.1.1.3.6. CM3.INAN.SG……………………...…………….….….69 4.1.1.3.7. CM3.AN.PL……………………………...……….…..….69 4.2. Case System……………………………………………………….…….…...….70 4.2.1. Nominative (Absolutive)………………………………………..………71 4.2.2. Ergative……………………………………………..……………….73 4.2.3. First Genitive..……………..……………………….……………….74 4.2.4. Second Genitive….....………………………………….……………75 4.2.5. Dative………………………………………………………….....….77 4.2.6. Comitative…………………………………………….……...….…..79 4.2.7. Locative………………………………………………...……………81 4.2.7.1. General Locative……..….....………………...………………81 4.2.7.2. Orientational Locative…………...………….…………...…..83 4.2.7.3. Possessive Locative…………...……………………...…...…85 4.2.8. Ablative…………………………………………………………..…..88 4.2.8.1. General Ablative………………………………………..……88 4.2.8.2. Orientational Ablative (Purpose Ablative)………......………91 4.2.8.3. Possessive Ablative………………………...……..….....……92 4.2.9. Comparative…………………..…………………………………...….95 4.3. Nouns--Plurality…...…………………………………………………...…….96 4.3.1. Plural Forms with Noun Declension…..……..……………………….99 4.4. Noun Phrase……………………………………………………………...…103 5. ADJECTIVES…………………………………………………………………....112 5.1. Dependent Adjective…………………………………………………….….112 5.2. Independent Adjective………………………………………………………113 6. PRONOUNS……………………………………………………………………..118 6.1. Personal Pronouns…………………………………………..………….…..…118 6.2. Demonstrative Pronoun Deixis……………………………...………….…...…119 6.2.1. Dependent Demonstrative Pronouns……………………………..……..119 6.2.2. Independent Demonstrative Pronouns……………………………...…..120 6.3. Possessive Pronouns…………………………………………………..….…....121 6.4. Interrogative Pronouns……………………………………………………...…122 6.5. Indefinite Pronouns………………………………………..………………...…123 6.6. Quantifier Affirmative Pronouns…………………………………………..…..123 ix 6.7. Negative Quantifier Pronouns…………………………………….……….…..123 7. THE VERB………………………………………………………………………....138 7.1. Verb Types……………………………………………………………………138 7.1.1. Simple Verbs.……………………………………………………...…..138 7.1.2. Complex Verbs…..……………………………………………………139 7.2. Aspect…………………………………………………………………………141 7.3. Tense………………………………………………………………….……….143 7.4. Modality…………………………………………………….…………………143 7.5. Morphological Structure………………………………………………………144 7.5.1. Area II: Aspect………………………………………….………………145 7.5.2. Area III………………………………………………………………….147 7.5.2.1. Past/Nonpast Indicator……………………………….…………147 7.5.2.1.1. Simple Present= -Ø …………………………………….148 7.5.2.1.2. Aorist= -Ø………………………………………………149 7.5.2.1.3. Future I= -Ø ……………………...………………...…..150 7.5.2.1.4. Neutral Past= -s̆ä…………………………….………….151 7.5.2.2. Other Tenses…………………………………………..………..152 7.5.2.2.1. Present I= -at………………………………..…………..152 7.5.2.2.2. Perfect I= -at……………………………………………153 7.5.2.2.3. Imperfect I= -a……………………..…………………..154 7.5.2.2.4. Pluperfect I= -a……………………………...…………155 7.5.2.2.5. Present Habitual= -(t)ar…………………………...……156 7.5.2.2.6. Past Habitual= -(t)ar……………………………...…….157 7.5.2.3. Tenses with Verb Class Markers……………………..………..157 7.5.2.3.1. Future II…………………………………………...……158 7.5.2.3.2. Irrealis Past……………………………………………..158 7.5.2.3.3. Definite Past…………………………………..………..159 7.5.2.3.4. Indefinite (Long Ago) Past………...…………………..160 7.5.2.4. Tenses with Orientation/Direction Markers……………..……..161 7.5.2.4.1. Present II= -o…………………………………………..161 7.5.2.4.2. Perfect II= -(q/t)o………………………………....……163 7.5.2.4.3. Imperfect II= -qo………………………………....……164 7.5.2.4.4. Pluperfect II= -(q/t)o………………………………..….164 7.5.3. Area IV…………………………………………………………………165 7.5.4. Area V…………………………………………………………………..165 7.5.5. Area VI………………………………………………..………………..166 7.5.5.1. Indicative Mood……………………………………………..…166 7.5.5.2. Irrealis Moods…………………………………………………166 x 7.5.5.2.1. Interrogative Mood……………………………….…….166 7.5.5.2.2. Conditional Mood………………………………………167 7.5.5.2.3. Unexpected Mood…………………………...………….167 7.5.5.2.4. Potential Mood- kwa …………………..……...………..168 7.5.5.2.5. Imperative Mood……………………………..…………168 7.5.5.2.6. Prohibitive Mood……………………………………….170 8. VERB PHRASE………………………………………………………….……..….194 8.1. Auxiliary I……………………………………………………….……..….…..194 8.2. Auxiliary II…………………………………………………….……….…..….195 8.3. Auxiliary III…….…………………………………………....……………...…196 8.4. Auxiliary IV…...…………………………………………………………….…197 9. VERB ALIGNMENT AND AGREEMENT ……………………………………..201 9.1. Verb Alignment……………………………………………………………..…201 9.2. Verb Agreement…………………………………………………...…...............203 10. MOTION ORIENTATION MARKERS………………………………...…….......205 10.1 Point of Reference………………………………………..………...……...…..207 10.1.1. The Speaker as a Point of Reference………………………...…......…..207 10.1.2. The Actant as a Point of Reference………………………..…….....…..209 10.1.3. Location as a Point of Reference……………………………….........…210 11. WORD ORDER……………………………………………………………....…....214 REFERENCES………………...………………………………………………………215 LIST OF TABLES 3.1 Comparison Chart of Xinaliq Phonemes…………………………………...…….….45 3.2 The Xinaliq Consonant Inventory in IPA………….…………………………...…...48 3.3 The Xinaliq Consonant Inventory in "Scientific Transcription".................................49 3.4 Xinaliq Vowel Inventory………………………………………………………….…49 4.1 Xinaliq Class Markers………………………………………………………………105 4.2 Series 1 Class Markers………………………………………...……………………106 4.3 Series 3 Class Markers…………………………………………...…………………106 4.4 Xinaliq Case System Phonological Alternations…………………………….…….107 4.5 Noun Conjugations……………………………………………...………….……...108 4.6 Declension in Plural for Classes AN. and INAN…….………………...…......……109 4.7 Class M and Class F Plural Declension………….……………………....…......…..110 4.8 Irregular Class M and Class F Plural Declension……..…..…………………......…111 4.9 Xinaliq Plurality Paradigm……………………….……………………..…..…...…111 5.1 Sample Declension of the Substantivized Adjective xırıc ̣ "white"…………...…….116 5.2 Conjugated Examples of the Substantivized Adjective ḳok "thick"…………...…..117 6.1 Personal Pronoun Declensions……………………………………………….…..…125 6.2 Forms of the Dependent Demonstrative Pronouns du "this" and hu "that"……………………………………………………………………………………126 6.3 Dependent Orientation/Direction Demonstrative Pronouns…...………………...…127 xii 6.4 Full Paradigm for Independent Demonstrative du "this-he/she/it"…………………128 6.5 Full Paradigm for Independent Demonstrative hu "that-he/she/it"………………...129 6.6 Possessive Pronouns………………………………………………...……….……..130 6.7 Interrogatives…………………………..………………………………….…….….131 6.8 Independent Indefinite Pronouns klaq̇i and yaq̇i………..............................…..…...132 6.9 Independent Indefinite Pronoun sa………………………….........................…...…133 6.10 Dependent Affirmative Quantifiers……………..…………..………………….…134 6.11 Independent Affirmative Quantifiers……………………..........................……….135 6.12 Dependent Negative Quantifiers……………………………..……………………136 6.13 Independent Negative Quantifiers…………………………………………..…….137 7.1 Tense Forms of Xinaliq……………………….……...………………...….…….…184 7.2 Xinaliq Moods………………………………………………………………...……185 7.3 Examples of Morphological Structure of the Xinaliq Verb……..……………….....186 7.4 Verb Inflections for Aspect………………..………..…..…………………………..187 7.5 Examples of Aspect Formations………………………………………......………..187 7.6 Rules for Formation of Perfective I………………………......................………….188 7.7 Rules for Formation of Imperfective I…………………………….......…...……….188 7.8 Past/Nonpast Markers………………………………….....................……….……..188 7.9 Past/Nonpast Indicator Tenses………………………..………………....………….189 7.10 Tenses with at and ar Markers………………………………………...………..…189 7.11 VCM Distribution……………………………………………………..…………..190 7.12 Example of Xinaliq Negative Verb…………………………………..…..………..191 7.13 Past/Nonpast Distinction between Tenses……………..……………….…...…….192 xiii 7.14 Tense-Aspect Markers……………………….…………………...……….………193 8.1 Auxiliary II………………………………………………………………..………..200 10.1 The Orientational/Directional (OD) Verbal Categories…………………………...211 10.2 Direction Markers…………………………………….……………………..…….212 10.3 Orientation Markers…………………………………….………………..………..212 10.4 Orientation/Direction Markers Extended……………………………...………….213 LIST OF FIGURES 2.1 Traditional Nakh-Daghestanian Family Classification………………………...….…17 2.2 Northeast Caucasian Family Classification…………………………………...…..…18 2.3 The Location of Xinaliq Village………………………………………….…..……...19 7.1 Morphological Structure of the Xinaliq Verb….………………………….…….….172 7.2 Perfective Aspect Tenses………………………………………………….………..173 7.3 Imperfective Aspect Tenses…………………….................………………………..174 7.4 Imperfective I Aspect Tenses……………………………………….………….…...174 7.5 Past/Nonpast Indicators…………………………………………...……...……...…175 7.6 Past/Nonpast Perfective Aspect Tenses………………………………...…………..175 7.7 Past/Nonpast Imperfective Aspect Tenses………………..……………….....….….176 7.8 Past/Nonpast Imperfective I Aspect Tenses………………...………………....…...176 7.9 Simple Present Tense Composition……………….…………………..………..…..176 7.10 Aorist Tense Composition……………………………………………...…............177 7.11 Future I Tense Composition…………………………........................……..…….177 7.12 Simple Present Tense Composition…………...……...…………………………...177 7.13 General Tense Location…………………………...…………………………..…..177 7.14 Present I Tense Composition…………………………..…..........................……...178 7.15 Perfect I Tense Composition…………………………………………………...….178 xv 7.16 Imperfect I Tense Composition…………………………………….….....……….178 7.17 Pluperfect I Tense Composition………….…………………………....…..………178 7.18 Simple Present Tense Composition…………………...……………….……….....179 7.19 Past Habitual Tense Composition……………………........................………..…..179 7.20 VCM General Location…………………………...……………….……………....179 7.21 Future II Tense Composition…………………………………………………...…179 7.22 Irrealis Past Tense Composition……………………………………………..……180 7.23 Definite Past Tense Composition…………………………………….………..…..180 7.24 Indefinite Past Tense Composition………………………………….……….……180 7.25 General OD Tense Location………………….………………….…………..……180 7.26 Present II Tense Composition……………………………………….………….…181 7.27 Perfect II Tense Composition………………………………………….……….…181 7.28 Imperfect II Tense Composition……………………..............................................181 7.29 Pluperfect II Tense Composition………………………………….………….…...181 7.30 Moods with Perfective I Aspect…………………………………..…………….....182 7.31 Moods with Imperfective Aspect…………………………………..…….……..…182 7.32 Moods with Imperfective I Aspect……………………………….…………...…..183 LIST OF ABBREVIATIONS - > changes into A agent ABL ablative ABS absolutive ADJ adjective AN animate (class III) AOR aorist APP.AB approaching above APP.BL approaching below APP.LV approaching same level APP.UNS approaching unspecified ASP aspect AUX.I auxiliary I AUX.II auxiliary II AUX.III auxiliary III AUX.IV auxiliary IV C consonant CM class marker COM comitative xvii COMP comparative CON conditional COP copula DAT dative DEF.PST definite past DEP.AB departing above DEP.BL departing below DEP.LV departing same level ERG ergative F feminine (class II) FEM feminine (class II) FUT future FUT.I future I FUT.II future II GEN.1 genitive 1 GEN.2 genitive 2 GEN.LOC general locative HAB.PRES habitual present HAB.PST habitual past IMP imperative IMP.I imperfect I IMP.II imperfect II IMPRFV imperfective xviii IMPRFV.II imperfective II INAN inanimate (class IV) INDEF.PST indefinite past INDIC indicative INST instrumental INTER interrogative IRR.PST irrealis past LOC locative M male (class I) MASC male (class I) NEG negation NOM nominative NP noun phrase NTR.PST neutral past O patient OD orientation/direction marker OR.ABL orientational ablative OR.LOC orientational locative PL plural PLUP.I pluperfect I PLUP.II pluperfect II POS.ABL possessive ablative POS.LOC possessive locative xix POT potential PRF.I perfect I PRF.II perfect II PRFV perfective PRFV.I perfective I PROH prohibitive PRS.I present I PRS.II present II PST past S subject SG singular SMP.PRES simple present SOV subject object verb SVO subject verb object UNEX unexpected V verb V vowel VCM verb class marker ACKNOWLEDGMENTS I would like to acknowledge and express my gratitude to the Xinaliq people for sharing their language with me, for welcoming me to their community and for their utmost dedication to the Xinaliq language, their cultural heritage and revitalization efforts. I specifically would like to thank the late Fexraddin Abdulaev, who in addition to teaching me about his language and culture, taught me about friendship and humor. The lightness of his spirit and his curiosity about life have guided this writing. Many people and organizations must be acknowledged as aiding and supporting the writing of this grammar. First, this grammar would not be possible without my initial fieldwork trip as a research assistant to Alice Harris in 2009, who was kind enough to allow me this opportunity. Discovering the Xinaliq village with her for the first time and making acquaintance with our first language consultants, who have worked with me continuously over the years, was a priceless introduction to Xinaliq and to doing fieldwork lingustics. Her guidance through out the whole process has been unmatched. In the US, Lyle Campbell has been a dedicated committee chair and has guided this research thoughtfully and helpfully. I am grateful for his abundance of comments, his time and support. I owe much of my current thinking about Xinaliq and linguistics in general to him. Without MaryAnn Christison's academic guidance, this work would not have been completed. Her feedback has been invaluable in the writing and revisions of this grammar. I would also like to acknowledge the other members of my graduate xxi committee: Marianna DiPaolo and Maria Polinsky for their helpful feedback and cooperation. I also want to acknowledge the entire faculty in the Department of Linguistics at the University of Utah, who taught the courses that have informed this grammar. Outside of the United States, I want to deeply thank Jost Gippert from Goethe University Frankfurt. Without his initial support this grammar would not be written. It is due to his encouragement that I sought out funding from the DoBes Volkswagen Foundation, which ended up being crucial to writing this grammar. I also want to thank my research partner, Monika Rind-Pawlowski. Her insightful comments during the fieldwork as well as her company was extremely valuable and appreciated. In addition to the Volkswagen Foundation, the National Science Foundation and American Research Institute of South Caucasusa also provided funding for this dissertation. I am grateful for their support. I want to thank my friend and editor Ron Bitton for reading this manuscript and making much needed suggestions. Lastly, I want to thank Dagny and John for being there when it was most needed. Thank you. CHAPTER 1 INTRODUCTION 1.1 The Aim of this Dissertation The aim of this dissertation is to provide an informed grammatical description of the principal aspects of the phonology and morphosyntax of Xinaliq, a Northeast Caucasian language spoken in Azerbaijan. Although informed by current linguistic theory and advances in language typology, this linguistic analysis is theory neutral. The goal for this descriptive grammar is to be clear, accessible, reasonably comprehensive, and to address in depth the structural attributes of Xinaliq phonology and morphosyntax. It is important to note that this dissertation is part of a larger documentation project of Xinaliq, which, in addition to enhancing scholarly understanding of the Xinaliq language, is ultimately designed to address the needs of native speakers in the community for ethnolinguistic documentation and preservation. In the documentation project, in addition to the language, importance was given to recording and describing the cultural heritage of the region. The project included a plan for a long-term, community-based program to sustain and revitalize Xinaliq, with leadership from and involvement of the native speakers in the community. Elicitations, texts and recordings collected during fieldwork (see Section 1.4) were used for linguistic analyses and for the development of this grammar. They are also being 2 used for currently ongoing production of teaching and literacy materials in the Xinaliq language, materials which previously did not exist. 1.2 Organization of the Grammar The contents of the grammar (please see the Table of Contents of the grammar above) were modeled after a number of the best reference grammars. The grammar is organized in eleven chapters. The first two chapters provide the background on this dissertation and place the Xinaliq language in a larger cultural and historical context. Chapter 3 surveys the phoneme inventory of Xinaliq language and its phonology. Chapter 4 discusses the general category of nouns, including the class markers, the case system, plurality and noun phrase. Chapter 5 describes dependent and independent adjectives, understanding of which is crucial for understanding the different types of pronouns discussed in Chapter 6. Chapter 7 provides a detailed description of the Xinaliq verb and its morphology. Aspect, tense and modality are introduced, both separately and as part of Xinaliq morphology. The verb phrase and its components are described in Chapter 8, and verb alignment and agreement in Chapter 9. Xinaliq's orientation-direction markers are analyzed in Chapter 10. Word order, discussed in Chapter 11, completes the morphosyntactic description of the Xinaliq language presented here. The grammatical elements and constructions are in numbered sections and nested subsections, in clear descriptive prose, illustrated with multiple examples, with clear terminology and tables and graphs to further illustrate the point. Each example is given standard interlinear glossing and translation. Extensive effort was made to cross-reference related phenomena elsewhere in the grammar. The Table of Contents should be a reasonably comprehensive aid in locating specific elements. 3 Examples from the previous published work on the Xinaliq language were consulted during the analyses, including unpublished examples collected during the fieldwork trip in 2009 with Dr. Harris. Although not a complete grammar of the Xinaliq language, an attempt was made to describe the key areas of the language. 1.3 Typological Overview The discussion of the typological characteristics of the Xinaliq language in this chapter serves as the the background for the more in-depth analysis of the linguistic patterns in later chapters. The Xinaliq language has several features whose study can significantly contribute to theoretical, typological, and historical-comparative linguistics. Xinaliq has a complex phonology, with forty-three consonants, including stops at four different points of articulation; its verb alignment is ergative; there is four-fold semantic distribution of its noun classes, agreeing with adjectives, verbs, pronouns and adverbs; it has thirteen cases; inclusive/exclusive possessive pronoun contrast with kinship option; complex deictic pronouns; complex class number markers in the verb which can appear multiple times in the verb; and variation in present tense due to the spatial orientation of the speaker in relation to the event. This morphologically integrated orientation-directional system also applies to demonstratives and auxiliaries and rich past tense forms (fourteen tense forms altogether). 4 1.4 Fieldwork and Methodology The data for the present work was collected over several fieldwork trips conducted by the author from 2009-2013 with funding from the National Science Foundation, DoBes Volkswagen Foundation and The American Research Institute of the South Caucasus. The fieldwork was conducted according to recommended best practices as outlined by Bird and Simons. The recordings were made to be high quality and long-lasting. Elicitations through questionnaires, direct elicitations, and natural speech collection techniques were incorporated during fieldwork. Naturally occurring conversations in different discourse settings were documented. The goal was to collect several types of texts, ranging from daily dialogues to technical instructions on how to make various items and collected examples of traditional knowledge. Speakers from different social classes were encouraged to participate, including children, elders and females. Elicitations were necessary to get full morphological paradigms and to assess more detailed information that does not always surface in natural speech recordings. Data from those recordings, however, had the advantage of being original and part of natural speech events, without the interference of the contact language used by the author. The two techniques (natural speech recordings and language elicitations) proved to be necessary for different purposes, and both were used during the fieldwork. Elicitations were conducted in Azeri and Russian and recorded on audio and video equipment. Through carefully selected questionnaires (both standard ones and those developed by the author), the language was elicited. Questionnaires, including those developed by Comrie and Smith (1977), Bouquiaux and Thomas (1992), and Payne (1997), were used on numerous occasions. When possible, all the questionnaires were adjusted to the geography and 5 cultural norms of the Xinaliq community. Other stimuli such as photos were also used. Guidelines from Payne (1997) for text collection were followed. Texts were transcribed and translated in the field as much as possible. Extensive metadata was kept on everything recorded. Minimal annotations were included: unique identifier, title, collector, place, target and contact language, consultants' information, date, item type, general description, and miscellaneous comments. Personal metadata included: persons' names, date of birth, first and additional languages spoken, clan, residential history, household information, and more. This information will also be highly useful for a much needed sociolinguistic survey of the Xinaliq village. However, these unique identifiers have not been included in this version of the grammar due to spatial constraints and strict formatting guidelines. 1.5 The Name There are several variations on the name for the Xinaliq language, including Khinalugh, Khinaliq, Khanaluka and Khynalyk, Khinalug, Khinalugi, and Xinalug. All these names and their variants represent either Azerbaijani or Russian pronunciation and titles for this language and people. However, the name that the Xinaliq people use to refer to themselves is Ket, and the language is referred to as the Ketish language. To maintain consistency with the previous work, and to make this research accessible and recognizable by the largest possible audience, the choice was made to stay with the name Xinaliq, which is the name predominantly used by the Azerbaijanis for this language and its speakers. The author intends to introduce the community preferred name Ketish into linguistic discourse by suggesting the change to Ethnologue administrators and incorporating the term into future work. CHAPTER 2 XINALIQ: LANGUAGE, PEOPLE, AND GEOGRAPHY 2.1 The Xinaliq Language The languages of the Caucasus are spoken in the territory extending from the Black Sea to the Caspian Sea in the region of the Caucasus Mountains. This area has been renowned since antiquity for the immense number of languages it hosts. There are three separate language families in this region: the Northwest Caucasian language family, the South Caucasian (Kartvelian) language family, and the Northeast Caucasian (Nakh- Daghestanian) language family. The language families in this region are not related to one another genetically (i.e., they are not descended from a common language ancestor); thus, they represent three separate and distinct language families, although their regional and areal relations include borrowed vocabulary and other factors. Some scholars have speculated on possible relations among these families, or to languages outside the Caucasus region. However, none of the proposals has been proven, and they are not generally accepted by the wider linguistic community. The term "Caucasian Languages" only refers to the geographic proximity of the language families. In addition, none of them has a relationship with any languages outside of the region; all three of the language families are considered indigenous to the Caucasus region. According to Comrie (2009) there is an important distinction between the terms "Caucasian Languages" and 7 "Languages of the Caucasus." The former refers to the three language families that are restricted to the Caucasus. The latter can include other languages that are spoken both in and outside of this region, such as Armenian, Azeri and Russian, among many others. Xinaliq (ISO639-3: kjj) is a Caucasian language, belonging to the Northeast Caucasian (Nakh-Daghestanian) language family. It is spoken primarily in one village in northern Azerbaijan, where it has been spoken for centuries. Although the language has changed over the years and certain expressions, phrases, and grammatical variants have been abandoned by the younger generations, Xinaliq does not show dialectal variation. While some of the Northeast Caucasian languages in the region are spoken both in Daghestan and Azerbaijan, Xinaliq is one of the very few spoken only in Azerbaijan. The border between Daghestan and Azerbaijan was of no great importance during Soviet times, but it is now the national border separating Daghestan, which belongs to Russia, and the independent country of Azerbaijan. Although the languages spoken on the two sides of the border differ, there was strong cultural contact and communication between the communities which is almost nonexistent now, being too difficult to maintain for geographic and sociopolitical reasons. Russians now need a visa to enter the country of Azerbaijan, and vice versa. These political changes and their impacts on the languages spoken in this region have not been studied yet. Almost all Caucasian languages have been significantly influenced by Russian since Czarist times, while the impact of Georgian, Azeri, and, to a lesser extent, also Arabic, Iranian, and Turkish has remained restricted to certain geographic areas. The lexical influence is easiest to detect through the borrowings from Russian, Turkish (Azeri) and Arabic into Xinaliq. However, there are also some indications of phonological and morphological influences that need further 8 study.The influence of Azeri on Xinaliq is observable not only in the vocabulary but also in the phonetics and grammar (Kibrik, 1972). Those influences have continued to grow, especially since the introduction of electricity to Xinaliq village, followed by the quick spread of television sets and the prominence and popularity of Turkish soap operas. The Northeast Caucasian language family is by far the largest and most linguistically diverse of the Caucasian language families, and its age is often compared to that of Indo-European (Nichols, 2010). Although it is difficult to make an exact count, there seem to be thirty-one languages that belong to the Northeast Caucasian language family. Views of the classification of languages within the family tree have varied among language specialists. The family tree is complex, and the documentation of all the languages in this family is not yet complete. Thus, some questions about the family's internal relationships remain unanswered. The traditional classification of the language family splits it into two main branches, a western group consisting of the Nakh languages and an eastern group comprised of the Daghestanian languages. This division is illustrated in Figure 2.1. The dotted lines represent alternative classification options for some of the families and individual languages. The other classifications differ by holding that Nakh and Daghestanian are not the two main branches of the family, but rather the language family has five to seven principal branches (Figure 2.2). According to a 1994 survey (Kibrik, 1994), at that time the three Nakh languages taken as a group had about one million speakers. The same survey numbered all Daghestanian languages taken together at 1.5 million speakers. The Daghestanian languages are divided into four main subgroups: (i) Avar-Andi-Tsezic; (ii) Lak; (iii) Dargi; and (iv) Lezgic. 9 Xinaliq has been classified as a member of the Lezgic subfamily. However, opinions diverge as to the correct classification of Xinaliq in relation to the languages of the Lezgic subgroup. The view that Xinaliq may belong to the Lezgic subgroup (instead of representing a subgroup in its own right) is highly disputed. Although debate continues, most scholars today regard Xinaliq as a separate branch of which Xinaliq is the only member within the Northeast Caucasian family. There is a long history of scholarly interest in the languages of the Caucasus. During Soviet times as well as after, many Russian scholars were interested in the region, and there have also been many publications about the languages in the region in both English and German. For some Caucasian languages there are reference grammars. Thus far, there is little work on aspects of the grammar of Xinaliq and very little textual material has been published. After the first sketch of the language done by Šaumyan (1940), the first grammar was written by Yu.D. Desheriev, entitled Grammatika xinalugskogo jazyka and published in 1959. It was followed by the same author's sketch in 1968 entitled Xinalugskij jazyk. In 1972, A. E. Kibrik and S.V. Kodzasov published Fragmenty grammatiki xinalugskogo jazyka, a brief sketch of Xinaliq, including a score of short text specimens and a concise vocabulary. Although these publications lay some groundwork for research, they are most definitely incomplete and at times are not informed by the insights of modern linguistic theory. A.E. Kibrik's team continued to work on many Daghestanian languages, including Xinaliq, currently designing a software infrastructure to maintain the texts that were collected in Xinaliq by his colleagues over the years (cf. http://www.philol.msu.ru/~languedoc/rus/xin/corpus.php; but note that these links do not 10 always work). A dictionary of Xinaliq (Xinaliq-Russian) was first provided by F.A. Ganieva in 2002 (Xinalugsko-russkij slovar', Xinaliq-Russian Dictionary). Except for a very short sketch of the Xinaliq grammar by Kibrik (1994) that was translated into English, most of the published material on Xinaliq is not available in English. This dissertation utilizes the existing scholarship on the structural traits of Xinaliq. Although the existing literature provides a solid starting point and has been utilized, the new data collected from fieldwork is the main source upon which this reference grammar is based. Initial research was conducted by Alice C. Harris and myself in 2009. Since then I have returned to the village of Xinaliq on two more occasions, spending the summer of 2011 and 2012 collecting data and working with the local community. Most of the indigenous Caucasian languages have never had their own writing. As a matter of fact, no more than twelve of these languages adopted a written standard during Soviet times. Only one of the autochthonous Caucasian languages, Georgian, has a long historical tradition as a written language, persisting uninterrupted since the fifth century AD. For Udi, written attestations of an old form of the language were discovered in the "Caucasian Albanian" palimpsests of the Middle Ages (Gippert, Schulze et al., 2009). Many attempts to provide orthographies for unwritten vernacular languages of the Caucasus have been made in the recent past, partly with emphasis on language maintenance; however, none of them has received official status so far. Since the second half of the twentieth century there have been several attempts to create a writing system for Xinaliq based on the Cyrillic and Latin alphabets. For example, the Xinaliq teacher 11 Rahim Alxas, an enthusiastic proponent of his native language, adapted the Lezgian Cyrillic-based alphabet for Xinaliq and proceeded to publish several books that included both poetry and teaching aids for the Xinaliq school (cf. e.g., http://www.philol.msu.ru/~languedoc/assets/texts/Alhas.zip). Another local poet, Namik, came up with his own orthography, which is the most accurate in terms of reflecting the number of phonemes that actually exist in Xinaliq. However, the system is complicated and was not accepted by most of the school teachers. Also, Namik had only a high school education, so his level of education did not give him enough prestige in the eyes of the teachers in the village to be taken seriously. In 2007, a team of scholars from Moscow University, under the leadership of Dr. Sandro Kodzasov, made a new proposal for a Xinaliq alphabet that closely follows the current Latin-based Azeri alphabet (cf. http://www.philol.msu.ru/~languedoc/assets/texts/xin-abc.pdf; the link does not always work). An alphabet was also introduced by the Summer Institute of Linguistics (SIL) in 2010. Although it is known that this alphabet ignores some phonemic distinctions in Xinaliq, it is currently the one most used, owing its success to the ease with which the orthography can be learned and its similarity to the Latin writing system used for Azeri. According to SIL, the main goal of the new alphabet is ease of learning and use. As a result, the system does not reflect several phonological distinctions that are easy for the native speakers to determine from the context but which might overcomplicate the notation if overtly represented in the writing. I, together with DoBes Volkswagen Foundation researcher Monika Rind- Pawlowski, worked with a team of teachers led by Gurban Abdulaev, a Xinaliq and Azerbaijani language grammar instructor, to create an alphabet that could utilize some of 12 the previous work done by both the locals and linguists. However, neither this alphabet nor its alternatives have become accepted as a standard yet. This work is currently in progress and the hope is for consensus on this issue to be reached within the next few months. There is one class taught on the Xinaliq language that meets irregularly. There are no grammar books available in Xinaliq to assist the instructors in teaching; thus, the class is not taken seriously. All residents use the Latin-based Azeri writing system (being the official writing system of the State of Azerbaijan) today. Although some of the residents still speak Russian, they do not necessarily know how to write in Cyrillic. 2.2 The Xinaliq People Many of the minority languages and people of the Caucasus are heavily endangered as more or less homogeneous communities, much more so now than ever before in their history. With the breakdown of the Soviet Union, the emergence of independent former Soviet states, and the beginning of armed conflicts in the struggle for independence, economic conditions have deteriorated and it has become much harder for smaller communities to survive. Many communities are now dissolving, with their languages being abandoned. The remoteness and isolation of the Xinaliq village have contributed to the survival of its language; however, the number of Xinaliq speakers has been declining more and more rapidly due to many of the changing factors mentioned above. The Xinaliq people were not counted as a separate ethnic group during the Soviet era (Gardanova, 1962) due to the ideological Soviet definition of ethnicity (i.e., natsionalnost) that has persisted up to the present day in Azerbaijan. According to a 1976 census (Clifton, 2005) the Xinaliq population was 2,500, however, the most current survey shows it at the much lower number of 1,000 (Salminen, 2007). The actual number 13 of full-time residents in Xinaliq village is probably even lower today. Xinaliq is the first language of the residents; however, almost everyone in the village also speaks Azerbaijani (Azeri), and some speak Russian. Although many children still spend their first few years speaking Xinaliq, they are obliged to speak Azeri as soon as they enter the school system. There is no schooling in the Xinaliq language, and Russian is only spoken by elderly men who served in the military during the Soviet era. Even though the residents have been used to the presence of those dominant languages for quite some time, they have maintained Xinaliq as the primary language in most home and family domains. Yet recently more and more domains are being lost to Azeri (Clifton, 2005). Although limited trade with the outside world has always existed, there have been growing work opportunities extraneous to Xinaliq due to more access to transportation and better roads. As a result, there has been a change in the character of those with whom the Xinaliq people work and for whom they work. A growing portion of the Xinaliq population conducts business with the residents of the nearby industrial town of Guba, where the language spoken is Azerbaijani. Thus, Azerbaijani is slowly becoming the language of commerce among Xinaliqi. Currently there are very few monolingual Xinaliq speakers left, mostly women, who are the members of the population least likely to attend school. More and more Xinaliq speakers find it necessary to conduct business and to communicate with others outside of their village. The construction of a better road connecting Xinaliq to the larger industrial town of Quba has contributed to more Xinaliq speakers finding it tempting and at times necessary to conduct business outside of their village. The road has also led to changes in many other aspects of traditional life, such as the introduction of electricity and an improved diet. With all of these changes in recent 14 years, there is a serious danger that Xinaliq may die out completely within the next two generations. Although the number of speakers is declining, most residents still consider Xinaliq to be their mother tongue and take an enormous pride in their language and heritage. A strong component of the documentation project that extends beyond the dissertation research is the training of the willing native Xinaliq speakers in transcription, recording, and creation of teaching materials. As part of the process of creating a descriptive grammar for Xinaliq, a range of data for a variety of language phenomena was recorded and analyzed. These texts are being used for the development of language teaching and literacy materials. By training locals in linguistic fieldwork methods, the project will be ensured a long livelihood beyond this dissertation research. 2.3. The Xinaliq Village Xinaliq village can be seen in Figure 2.3 and is identified as No.4 on the map. The village lies at an altitude of about 2,000 meters in the Quba district of Azerbaijan, in the middle of the Greater Caucasus mountain range that divides Russia and the Southern Caucasus region. Xinaliq is separated from most Daghestanian languages by the Caucasus chain. Its closest Daghestani language neighbor is Kryts (labeled on the map as Kryz), spoken in the villages of Alik, Djek and Kryts. Xinaliq is the highest and most remote and isolated village in Azerbaijan, and it is among those villages with the highest altitude in the Caucasus region. In fact, because of its remoteness, the government of Azerbaijan subsidizes teachers willing to teach in Xinaliq. Although they are citizens of Azerbaijan, Xinaliq residents are considered to be of a different ethnicity by Azerbaijanian people, a consideration that has negative social implications and that needs 15 to be studied further. Xinaliq is believed to be an ancient Caucasian village going back to the Caucasian Albanian period. According to Schulze (1994), both the local history and the linguistics of Xinaliq clearly indicate that the early speakers of Xinaliq migrated into their present location at some point during the period from 1000 BC to AD 300. It is believed by the Xinaliq residents that the ancestors of the Xinaliq people were followers of Zoroastrianism. In the third century they converted to Christianity, and then to Islam in the seventh century. All residents are Muslim. Because of the high altitude and its remoteness, the Xinaliq village and its residents have managed to survive and withstand the many invasions the region has witnessed. The area has many historical sites including ancient holy caves. Tellingly, the village of Xinaliq was included on the 2008 World Monuments Fund Watch List of the 100 Most Endangered Sites in the World. In the last twenty years a new settlement of Xinaliq speakers was established about eight kilometers from Xinaliq village. The settlement is called Boston Kesh and has a population of approximately three hundred people. Because of the lower altitude, residents can grow potatoes in Boston Kesh and life there is a little easier. Xinaliq village remains the center of the Xinaliq people and their cultural activities. With all the best intentions and efforts, there is only so much work that outsiders can do when it comes to language maintenance; ultimately, it is up to the people of the community to preserve their language. If the community members are encouraged to be involved, the chances of Xinaliq language survival increase significantly. The hope is that 16 this grammar will add to the necessary framework needed for future preservation and revitalization efforts by the Xinaliq community. 17 Figure 2.1 Traditional Nakh-Daghestanian Family Classification. (Nichols, 2003) (Nakho-Dagestanian=Nakh-Daghestanian; Dargi=Dargwi; Lezgic=Lezgian; Khinalugh=Xinaliq) Figure 2.2 Northeast Caucasian Family Classification. (Klimov, 1994) 18 Figure 2.3 The Location of Xinaliq Village (Adapted from The Red Book of the Peoples of the Russian Empire. www.eki.ee/books/redbook/khinalugs.shtml) 19 CHAPTER 3 PHONOLOGY This chapter is dedicated to a description of the phonology of Xinaliq, with emphasis on the phonemes and phonological processes of the language. 3.1 Consonants Many languages of the Caucasus are known for their large numbers of consonants. Xinaliq is no exception. Although the number of Xinaliq phonemes proposed in different analyses has varied, after looking at previous work by other scholars, speaking with a number of Xinaliq school teachers and various language consultants and after completing extensive fieldwork, it seems clear that there are forty-four contrasting consonants in Xinaliq. This number of phonemes is based on initial work by Desheriev (1959) and later Kibrik (1972), as well as on thorough research conducted in 2007, when a team of scholars from Moscow University, under the leadership of Dr. Sandro Kodzasov, created a new proposal for a Xinaliq alphabet that closely followed Kibrik's 1972 proposal (cf. http://www.philol.msu.ru/~languedoc/assets/texts/xin-abc.pdf). The proposal was developed in cooperation with the language teachers of the Xinaliq school. The phoneme inventory presented in this dissertation differs from that original work by including three extra phonemes as part of the Xinaliq phonemic inventory. They are: (i) glottal stop ʔ, (ii) velar glide w, and (iii) uvular stop q. Yet when Kibrik's dissertation was published in 21 1972, he also included the glottal stop ʔ and uvular stop q in his description of the Xinaliq phonemic inventory. Between the years of 1995 and 2000, Alexander Nakhimovsky from Colgate University worked in conjunction with A. Kibrik from Moscow University on the development of a database of five Eurasian languages including Xinaliq (cf. http://www.philol.msu.ru/~languedoc/eng/xin/index.php/). In this newer orthography, these phonemes were also recognized. Thus, my analysis of the Xinaliq phonemic inventory diverges from theirs only by having one more phoneme: the velar glide w. The first year of my fieldwork was spent on studying and analyzing the Xinaliq phonemic inventory. Considerable amount of time was spent on making the recordings and learning the sounds of Xinaliq. Because the phonemic inventory has been established and accepted by many scholars before me, minimal pairs were not always collected. Instead, the main effort was spent on learning the phonemes and working with the residents on creating an orthography that would be most useful. A list of minimal pairs with phonemes that most likely be mistaken for allophones is given in this section, after the phonemic inventory is introduced. The Xinaliq consonants are distinguished by place and manner of articulation. The places of articulation of Xinaliq phonemes are: bilabial, dental, alveolar, palato-alveolar, palatal, velar, uvular, pharyngeal and glottal. The manners of articulation are: stop, fricative, affricate, approximant, trill, lateral, nasal and ejective. Consonants can also contrast with respect to voicing (voiced/voiceless) and length (short/long distinction, i.e., geminates), although length applies only to plain voiceless stops and affricates. There is only one voiced affricate. It seems reasonable to assume that the Xinaliq phoneme inventory used to be larger and has continually decreased under the impact of increased 22 contact with outside groups. This would account for some "gaps" in the current inventory, such as lack of length distinction in all consonants and the small number of voiced affricates. The complete list of Xinaliq phonemes is given below, together with a few example words containing the given phoneme. Transcription is in standard International Phonetic Alphabet (IPA) notation. 1. /p/ voiceless bilabial stop paga ‘tomorrow' pan ‘hundred' 2. /p:/ long voiceless bilabial stop qup:a ‘belt buckle' p:ejram ‘shirt' 3. /p'/ voiceless bilabial ejective stop p'a ‘a kiss' p'ap'aχ ‘Xinaliq style hat' 4. /b/ voiced bilabial stop bıj ‘father' bemb ‘fly' 5. /f/ voiceless bilabial fricative fara ‘warm' fitʃ'æ ‘wet' 6. /v/ voiced bilabial fricative vats:ɯnɯ ‘to stay' dɯvar ‘wall' 7. /m/ voiced bilabial nasal tsɯma ‘red' q'ami ‘spoon' 8. /t/ voiceless dental stop 23 toz ‘dust' dostır ‘friends' 9. /t:/ long voiceless dental stop bat:a ‘short' t:oz ‘door' 10. /r/ voiced dental trill rang ‘color' dahari ‘rock' (ergative) 11. /l/ voiced dental lateral approximant lap ‘very' t'ali ‘lip' (ergative) 12. /d/ voiced alveolar stop dost ‘friend' mɯde ‘mountain' 13. /t'/ voiceless alveolar ejective stop t'al ‘a lip' pʃit' ‘kitten' 14. /ts/ voiceless alveolar affricate tsɯma ‘red' tsuloz ‘tooth' 15. /ts:/ long voiceless alveolar affricate jets:ini ‘to stay back' 16. /ts'/ voiceless alveolar ejective affricate ts'u ‘name' ts'in ‘sweet' 17. /s/ voiceless alveolar fricative san ‘night' varɯsi ‘relative' 24 18. /z/ voiced alveolar fricative zɯ ‘I' bzi ‘pear' (ergative) 19. /n/ voiced alveolar nasal nuk' ‘hail' kona ‘old' 20. /tʃ/ voiceless palato-alveolar affricate tʃoz ‘spring' tʃe ‘tea' 21. /tʃ:/ long voiceless palato-alveolar affricate k'utʃ:ɯ ‘heel' 22. /dʒ/ voiced palato-alveolar affricate dʒydʒæ ‘chick' kundʒ ‘corner' 23. /tʃ '/ voiceless palato-alveolar ejective affricate tʃ 'æ ‘fire' tʃ 'utʃ 'on ‘whip' 24. /ʃ/ voiceless palato-alveolar fricative ʃum ‘sand' q'amaʃ ‘secret' 25. /ʒ/ voiced palato-alveolar fricative bʒuvɯ ‘to leave' q'aʒ ‘tail' 26. /y/ voiced palatal glide jetmæ ‘want' (Class M.) majlɯn ‘a fool' 27. /k/ voiceless velar stop kul ‘hill' 25 akuval ‘to go away' 28. /k:/ long voiceless velar stop bok:u ‘Baku' k:ɯza ‘snow' 29. /g/ voiced velar stop gada ‘boy' dalɯg ‘work' 30. /k'/ voiceless velar ejective stop k'ıl ‘arm' t'uk'un ‘cheek' 31. /kx/ voiceless velar affricate mokx ‘worm' kxe ‘wool' 32. /w/ voiced velar glide wa ‘it's there' (on top) swa ‘village' 33. /x/ voiceless velar fricative xu ‘water' lɯx ‘shepherd's crook' 34. /ɣ/ voiced velar fricative k'ɯɣ ‘eyebrow' joɣi ‘to shake' 35. /q/ voiceless uvular stop qi ‘cold' 36. /q:/ long voiceless uvular stop toqud ‘lightning' 37. /q'/ voiceless uvular ejective stop 26 q'aq'al ‘back' (body) tilq'wi ‘to milk' 38. /qX/ voiceless uvular affricate qXɯtsɯz ‘nut' ɯnqXa ‘field' 39. /χ/ voiceless uvular fricative χolɯ ‘uncle' miχ ‘summer' 40. /ʁ/ voiced uvular fricative ʁutsar ‘god' utaʁ ‘room' 41. /ħ/ voiceless pharyngeal fricative ħæjæl ‘child' mæħsul ‘harvest' 42. /ʕ/ voiced pharyngeal fricative næʕnæ ‘saliva' mæʕni ‘song' 43. /ʔ/ glottal stop vaʔ ‘no' sæʔil ‘here' 44. /h/ voiceless glottal fricative hu ‘he' k'adah ‘pitcher' The village of Xinaliq has had a long history of different writing systems being introduced by outside researchers as well as by some of the Xinaliq speakers themselves (see Chapter 2). Different scholars have disagreed on the number of phonemes in Xinaliq, as well as how best to represent them. After conducting original fieldwork (2009, 2011 27 and 2012), the number of consonants has been established based on previous work mentioned above and in agreement with the team of Xinaliq residents, which included language consultants, teachers and interested residents. A transcription orthography is being debated to this day. The "scientific transcription" was created which retained the positive points of the other alphabets but was new to all language consultants. Therefore there was consensus on its use. (Otherwise each consultant would want to use his own orthography or the one they favored.) This orthography was based on orthographies used by other languages in the region. It was created in collaboration with researchers from the Summer Institute of Linguistics (SIL), Xinaliq language consultants and the Volkswagen research team. It was used during the fieldwork and will be used for the rest of this dissertation. It was used during the fieldwork and will be used for the rest of this dissertation. This orthography was created with certain considerations in mind, including the following: letters known from the Azeri alphabet should keep their phonetic values; two letters (digraphs) for one sound should be avoided (velar affricate being an exception); ejective consonants should not be marked with apostrophe, to avoid confusion with quotation marks in direct speech, rather they should be designated by a point underneath or above; sounds which are similar to Azeri sounds but not identical should take the appropriate letter from the Azeri alphabet marked with an additional diacritic symbol. This orthography was welcomed, and made the workflow during fieldwork much more efficient. However, as the conversation about orthographies continues between researchers in Xinaliq and the native speakers, this "scientific orthography" will most likely continue to change. It was used for this dissertation as the best and most useful alternative for now, not as the final choice. The orthography created 28 and used throughout this dissertation with its equivalents to some of the more common other orthographies previously used in Xinaliq is given in Table 3.1. The different categories discussed in Table 3.1 are explained here. "Scientific" refers to the orthography used hereafter; "Kibrik" - to the alphabet introduced in the late 1970s and used on and off in the village most frequently; "Namik's alphabet" was created by the young poet himself (Namik), but is not accepted by many of the school teachers; "The Village Alphabet" was originally created by the Summer Institute of Linguistics and is most liked and used by the community, however, it fails to represent some of the phonemes; "Ganieva's orthography" is used in the only dictionary in existence for the Xinaliq language (Xinaliq-Russian, Russian-Xinaliq) written by Ganieva, thus her orthography is presented; IPA is there to make easy comparisons. The Georgian alphabet is presented in case of comparisons to be made with other Caucasian languages in the future, particularly Georgian. Scientific orthography matches and represents all the phonemes discussed above. Although the rest of the document is in scientific transcription, Table 3.2 displays the Xinaliq phonetic inventory in IPA. Table 3.3 does the same in scientific transcription. Although various orthographies have been proposed by various people, to this day Xinaliq is only rarely written; for the most part it remains an oral language only. I turn now to a discussion of some details of certain specific phonemes that may need closer attention. 29 3.1.1 Some Relevant Details Concerning Specific Phonemes Xinaliq bilabial, dental and velar stops show variation on three parameters - short vs. long, ejective vs. nonejective, and voiced vs. voiceless. Uvular's treatment falls outside of this pattern in that it has no voiced stop. However, both velar and uvular have affricates. Both affricates can occur after nasals, which voiceless bilabial, dental, velar and uvular stops cannot do (Ex. 3.1). 3.1 ınx̑a ‘field' In several situations Xinaliq voiced fricatives commonly become ejectives, as in Examples 3.2 and 3.3. 3.2. /z/ - > [c]̣ /ḳaz/ - > [ḳıcị] ‘snake' - > ‘snake' 3.3. /ž/ - > [č]̣ /q̇ až/ - > [q̇ ačị] ‘tail' - > ‘tail' However, in 3.4 instead of the expected /q̇/, we get /x̑/ because of the nasal /n/ preceeding it: 3.4 /ǧ/ - > [x̑] /enǧ/ - > [inx̑i] ‘wedding' - > ‘wedding' 3.1.2 Short and Long Voiceless Phonemes Long voiceless stops and affricates occur only intervocalically: 3.5 qup̄a ‘buckle' 3.6 bat̄a ‘short' 30 3.7 yec̄ini ‘to stay' 3.8 yäk̄il ‘mountains' 3.9 toq̄ud ‘lightning' The uvular stop is only long intervocalically in native words. In Azeri loanwords, voiceless uvular stops are not long intervocalically. Short voiceless consonants occur either word initially or word internally. When occurring word internally, they precede or follow another consonant: 3.10 patram ‘patron' 3.11 erpi ‘to melt' 3.12 toz ‘door' 3.13 dostır ‘friends' 3.14 cıma ‘red' 3.15 ki ‘to burn' 3.16 orkiri ‘to send' 3.17 kul ‘hill' 3.18 qi ‘cold' There seem to be no exceptions to this pattern of distribution for these consonants. 3.1.3 Allophonic Variations of Xinaliq Consonants 3.1.3.1 Labialization Velar stops are labialized after /u/ or /o/ followed by /n/. For example: ḳ - > ḳʷ/ [+round V] n___ 3.19 /gongaz/ - > [goŋgʷaz] ‘circle' 3.20 /ung/ - > [ũŋgʷ] ‘heart' 31 3.21 /unḳ/ - > [ũŋḳʷ] ‘fog' In this process, n - > ŋ 3.1.3.2 Voicing of /h/ The voiceless glottal fricative /h/ becomes voiced intervocalically, as in the following examples: 3.22 /ǯähil/ - > [ǯäɦ̑ il] ‘young' /h/ is also voiced before sonorants: 3.23 /dähnä/ - > [däɦ̑ nä] ‘ravine' 3.1.4 Consonant Distribution Xinaliq has a rich and complex consonant inventory. Consonant clusters are common among Northeast Caucasian Languages in general. Only certain combinations of consonants are permissible in different parts of the word, however. Below are the permitted consonant clusters in Xinaliq, for various positions within the word. 3.1.4.1 Word-Initial Combination of Stops + Fricatives (bzı ‘pear', pšä ‘bread') + Lateral (kla ‘who') Affricate + Fricative (čxạ r ‘buckweat') Stops + Fricatives + Liquids (pšlä ‘fox', pxṛ a ‘dog') Example 3.24 shows three consonants in a row (stop, nasal, stop), which rarely appeared in the data. Thus the rules for word-initial three-consonant clusters need to be 32 investigated in more depth, as well as possible variations and patterns involving stops, fricatives, nasals and liquids. 3.24 tnka ‘river' 3.1.4.2 Word-Final No combinations of three consonants were found word-finally. Permissible clusters of two consonants include: Sonorant + Obstruents (lıgıld ‘man', borc ‘father's sister', ant ‘earth') Fricative + Stop (vaxṭ ‘time', dost ‘friend') 3.1.4.3 Word-Medial Position Permissible combinations of consonants include: Sonorant + Obstruent (kulga ‘shadow') Fricative + Stop (äsḳi ‘towel') Approximant + Consonant (ayvan ‘balcony') Glottal stop + Consonant (säʔbi ‘master', näʔnä ‘saliva') Obstruent + Nasal (ägni clothing, häfmi ‘human') With clusters involving obstruents, both obstruents must be either voiced or both voiceless, unless nasals are involved: 3.25 bzı ‘pear' 3.26 aǧzı ‘mouth' 3.27 koksı ‘breast' 3.28 axṭ a ‘to castrate' 3.29 häfmi ‘human' 33 Ejectives can occur generally in clusters with sonorants, but not with the other obstruents: 3.30 ḳarṭa ‘nail' 3.31 ḳli ‘to die' 3.32 ınḳa ‘flat rock' 3.33 ınq̇ ‘sun' However there are some exceptions to this: 3.34 čḳas ‘bring' 3.35 ısṭarmeyram ‘sorceress' 3.2 Vowels The vowel system of Xinaliq is less complex and there is no disagreement about the number of vowel phonemes in the language. Its set of vowels is similar to that of Azerbaijani and might have been influenced by Azerbaijani. Each vowel can be described in terms of three properties: the horizontal dimension (front vs. back), height (high, mid, low), and roundness (round vs. plain [nonround]), as seen in Table 3.4. Below are some examples of these vowel phonemes in different positions within words: /i/: iži ‘face' ninig ‘eyelash' ki ‘ram' /e/: eg ‘wedding' če ‘tea' /ä/: ägni ‘clothing' dädä ‘mother' /ü/:* üs ‘year' nüc ̣ ‘honey' 34 sürü ‘herd' /ö/:** dövlät ‘wealth' örpüri ‘melt' /u/: urta ‘middle' ḳun ‘cooking flower' xu ‘water' /o/: odul ‘handful' kona ‘old' /ı/: ınka ‘river' bıǯ ‘father' bzı ‘pear' /a/: aba ‘grandfather' ṭal ‘lip' gada ‘boy' *The /ü/ occurs in native and borrowed words with the same frequency. **The /ö/ occurs primarily in borrowing from Azerbaijani. 3.2.1 Allophonic Variations in Xinaliq Vowels There are a number of rules that affect Xinaliq vowels. These are described below. 3.2.1.1 Nasalization Vowels before /n/ undergo nasalization, which is a common process of assimilation that happens when vowels are adjacent to nasal consonants. In Xinaliq this does not create a phonemic distinction between vowels. In fact this nasalization might go unnoticed in rapidly spoken Xinaliq. 35 3.2.1.2 Nasal Assimilation /n/ gets realized as [ŋ] when followed by velar or uvular, as in: 3.36 /ank/ - > [aŋk] ‘leg' 3.37 /eng/ - > [eŋg] ‘cheese' 3.38 /ınq̇/ - > [ıŋq̇] ‘sun' As compared to: 3.39 /ans/ - > [ans] ‘game' 3.40 /ant/ - > [ant] ‘earth' 3.41 /anšxẉ i/ - > [anšxẉ i] ‘to brush' 3.42 /anxṿ iä/ - > [anxṿ iä] ‘to throw' 3.43 /inǯitmiškwi/ - > [inǯitmiškwi] ‘to bother', ‘to torture' 3.2.1.3 Voiceless Vowels Underlyingly Xinaliq only has voiced vowels. However, in some phonological situations they experience devoicing. If a vowel is not stressed and is placed between two voiceless consonants, it will be realized as a voiceless vowel. 3.44 /kukác/̣ - > [ku̥ kác]̣ ‘chicken' 3.45 /sü̥ftä́/ - > [sü̥ftä́] ‘at first' 3.46 /qı̥cı́z/ - > [qı̥cı́z] ‘nut' 3.2.1.4 High Vowel Centralization /i/ and [ı] are both centralized when in certain phonological environments. /i/ - > [ɨ] (front high vowel moves to the center) after affricates and uvular consonants: 3.47 /xị nimḳir/ - > [xɨ̣ nimḳir] ‘woman' 36 3.48 /qim/ - > [qɨm] ‘hook' 3.49 /cịn/ - > [cɨ̣n] ‘sweet' /ı / - > [ɨ] (back high vowel moves to the center) after hissing consonants 3.50 /čıḳıri]/ - > [čɨḳıri] ‘to talk' 3.51 /šıdıl/ - > [šɨdıl] ‘great-grandson' 3.2.1.5 Vowel Syncope Vowel syncope in Xinaliq is a loss of an unstressed vowel from the interior of the word. It only occurs with two of the vowels, the high unrounded vowels, and only in informal/rapid speech. /ı/ - > Ø /i/ - > Ø It is a common occurrence, especially with /ı/. 3.52 [mıda] - > [mda] ‘mountain' 3.53 [cıma] - > [cma] ‘red' 3.54 [lıka] - > [lka] ‘meat' 3.55 [ǯigä] - > [ǯgä] ‘place' 3.2.1.6 Vowel Harmony Vowel harmony is a long-distance assimilatory phonological process. In languages that exhibit vowel harmony, there are constraints on which vowels may be found in combination with others, and which combinations within the word are prohibited. Xinaliq demonstrates front/back vowel harmony, which means that words may not contain both front and back vowels. Therefore a grammatical suffix changes to harmonize with other vowels in the root. Most grammatical suffixes come in front and 37 back forms. The words may not contain both front and back vowels, meaning all vowels must harmonize. 3.56 ägni ‘clothes' 3.57 dövlät ‘wealth' 3.58 ǯıbın ‘pocket' (SG) - > ǯıbındır ‘pockets' (PL) 3.59 gardan ‘neck' (SG) - > gardandır ‘necks' (PL) We see that both vowels in Examples 3.56 and 3.57 are front vowels; both vowels in the plural morphemes of Examples 3.58 and 3.59 harmonize to the vowels in the stem. These examples exhibit (front/back) vowel harmony. Xinaliq does not exhibit roundness vowel harmony. For instance, Examples 3.57 and 3.58 have both a round vowel and a nonround vowel. Azerbaijani (which has influenced Xinaliq) also exhibits similar vowel harmony. A kind of labial harmony exists in Xinaliq as well. The term "labial harmony" is normally restricted to consonants; when used with vowels, it refers to roundness. In Xinaliq, roundness vowel harmony is less consistent and less predictable. The roundness of a vowel in the root is dependent partially upon stress. In Examples 3.60 and 3.61 the stressed vowel is round and so obligatorily the preceding unstressed vowel is also round. 3.60 kotùk ‘stump' 3.61 culòz ‘tooth' This exemplifies a kind of labial harmony. However, if the following stressed vowel is not round, it is still possible for an unstressed vowel to be round, as seen in Example 3.62. 3.62 čụ vàl ‘sheep' 38 The rules for this can be stated as follows: ‐ If the stressed vowel is /a/ or /ı/ in the root, then the preceding unstressed vowel becomes [ı] ‐ If the stressed vowel is /o/ or /u/ in the root, then the preceding unstressed vowel becomes [u] or [ı]* ‐ If the stressed vowel is /ä/, /e/ or /i/ n the root, then the preceding unstressed vowel becomes [i] ‐ If the stressed vowel is /ö/ or /ü/ in the root, then the following vowel takes the form [ı] *The distribution alternates freely. Thus both variations in the example below (3.63) are possible. The first one shows labial harmony, while the second variation of the same word does not follow the rules of vowel labial harmony. 3.63 güldür / güldir ‘flowers' (PL) 3.3 Stress Not all languages have strictly phonemic stress; some exhibit a combination of types of stresses. It is said (Kibrik, 1994) that Xinaliq has dynamic stress. Dynamic stress is achieved through intensified muscle activity during articulation, interpreted as loudness. The stressed syllable is generally more powerful, intensive and loud in pronunciation. The stress in a word in Xinaliq depends on the morphology of the word. Generally stress falls on the last vowel of the root of the word: 3.64 gadá ‘boy' 3.65 q̇aqál ‘back' 39 If a suffix beginning with a vowel is attached to a root ending in a vowel, it replaces the final vowel and takes the stress: 3.66 ḳilí-u - > ḳilú ‘friend' -> friend (DAT.) 3.67 kizí-at-mä -> kizétmä ‘is burning' In derivative words, if the derivation is created by a suffix attached to the root, the stress remains on the final vowel of the root: 3.68 q̇aqál -> q̇aq̇álır ‘back' -> ‘on the back' 3.69 dáb - > dábxẹ r ‘a lie' - > ‘a liar' Listed below are exceptions to the rules for stress, which can probably be explained either as borrowings or now extinct previous morphological boundaries: 3.70 mátıška ‘Russian woman' 3.71 bášqap ‘plate' 3.72 tóq̄ud ‘lightning' 3.73 hána ‘after' 3.74 qárı ‘old lady' 3.75 qíni ‘good' 3.76 ḳä́vi ‘hard' 3.77 láǧın ‘yesterday' 3.4 Phonological Rules 3.4.1 Vowel Deletion The combination of two vowels next to each other is very rare. When a suffix begins in a vowel, the root's final vowel generally drops. 3.78 ki+i - > ki ‘ram' (ERG, GEN.I and GEN. II) 40 3.79 ki+u - > ku ‘ram?' (DAT. and Question) This is true for verbal morphology as well: 3.80 kui-dä-i-mä - > kuidimä ‘will not go' (Class II) 3.81 kui-dä-u -> kuidu ‘Will she go? (Class II) 3.82 kizi-a-šä-u -> kizešu ‘Was it on fire?' The examples below go through a phonological process of vowel deletion: 3.83 swa+u -> su ‘village?' 3.84 cẉ a+u -> cụ ‘house?' /y/ drops at the end of words when preceded by a front vowel. 3.85 liy - > li compare with liy-ir the root of ‘seed' - > ‘seed' compare with ‘in the seed' 3.86 bıniy - > bıni compare with bıniyir the root of ‘pasture' - > ‘pasture' compare with ‘in the pasture' 3.87 čey - > če compare with čeyir the root of ‘tea' - > ‘tea' compare with ‘in the tea' /y/ cannot be considered as an insertion, because there are examples of words that end with a front vowel where this does not occur: 3.88 ṭäsir - > ṭäsi - > ṭäsi ‘on the skin' - > the root of ‘skin' - > ‘skin' 3.4.2 Vowel Assimilation The combination of the two vowels [ia] coalesces to [ä] or [e]. 3.89 ki-a - > kä ‘do' 3.90 li-at-i-šä-mä - > letišämä ‘didn't say' 3.91 kizi-a-šämä - > kizešämä ‘burned' 41 The combination of the two vowels [ua] or [wa] coalesces to [o] if a suffix begins with a consonant: 3.92 swa+r -> sor ‘in the village' 3.93 cẉ a+r -> cọ r ‘in the house' 3.4.3 Glide Insertion The combination of two vowels [iu] [io] or [oi] [ui] inserts /y/, as in: 3.94 zaǧi-u -> zaǧi-y-u ‘you see' 3.95 q̇i-u - > q̇i-y-u ‘it became' 3.96 misi-orun -> misi-y-orun ‘when he was little' 3.97 yec̄ini-o-i-u -> yec̄ino-y-i-y-u ‘left?' 3.98 du-i-u -> du-y-i-y-u ‘not this one?' /v/ is inserted intervocalically after round back vowels /u/ or /o/: 3.99 cụ ‘name' 3.100 cụ -v-i ‘name' (erg.) 3.101 cụ -v-ol ‘names' (pl.) 3.102 cụ -v-or ‘in the name' 3.103 kino ‘movie' 3.104 kino-v-ır 'in the movie‘ A vowel is inserted to keep nonpermitted consonant clusters from appearing. Because of the morphological borders, when two consonants that are not permitted adjoin each other, an unrounded high vowel /i/ or /ı/ is inserted. 3.105 tal+r -> talır ‘on the lip' 3.106 gis+r -> gisir ‘on the roof' 42 3.107 z+ḳišämä -> ziḳišämä ‘she died' (FEM.class) 3.108 k+a+z+ǧi+šä+mä -> kazıǧsämä ‘came' Similarly to above, high unrounded vowels have a tendency to be dropped when this results in a permissible combination of two consonants at the morpheme boundary: 3.109 ni-ši-ri-at-mä -> nišrätmä ‘am putting on' 3.110 tu-vı-ri -> tuvri ‘takes' (INAN.class) 3.111 q̇i-du-mä -> q̇idmä ‘became' (MAS.class) 3.4.4 Assimilation Consonants assimilate to agree in voicing with the following consonant. In the examples below, a consonant becomes voiceless when followed by other voiceless phonemes: 3.112 xạ̈ -z-kindä - > xạ̈ -s-kindä ‘the laughing one' FEM. class 3.113 xạ̈ -b-kindä- > xạ̈ -p-kindä ‘the laughing one' ANIM. class In the following example, voiceless consonants become voiced when they are followed by a voiced consonant: 3.114 q̇amašbq̇i ->q̇amažbiq̇i ‘hid away' ANIM. class Assimilation generally does not occur across word borders which are now a single word, but were originally separate words: 3.115 dyz-kui - > dyzkui ‘to fix' from the word /dyz/ ‘correct', /kui/ - ‘to do' 3.116 dad-kui - > dadkui ‘to complain' from the word /dad/ ‘a complaint', /kui/ - ‘to do' If the word begins with two fricatives, the first one becomes an affricate. In the example below, /z/ becomes both voiceless and an affricate. 3.117 z-xị - > s-xị - > c-xị ‘go' FEM. class 43 3.118 š-x-a-bır - > čxabır ‘bellies' /ž/ - > /y/ at the beginning of the word. 3.119 ži-at-mä - > yetmä ‘want' MASC. class Ejectives sometimes lose their ejective quality before voiced obstruents: 3.120 čị-q-a-l-ǧi - > č-̣ q-a-l-ǧi - > č-q-a-l-ǧi ‘brought' (from below) 3.121 čị-t-a-l-ǧi - > č-̣ t-a-l-ǧi - > č-t-a-l-ǧi ‘brought' (horizontally) 3.122 čị-q̇ ui-šä-mä - > čị-q̇ u-šä-mä - > čụ -q̇ -šä-mä - > čụ -q-šämä ‘tore out' 3.123 enǯi-q̇ui-dä-mä - > enǯi-q̇u-dä-mä - > enǯu-q̇ui-dämä - > enǯu-q-dä-mä ‘take' 3.4.5 Free Variation /a/ and /ä/ can freely alternate in some environments. 3.124 taza / täzä ‘new' 3.125 pänäd / panad ‘nostrils' There is a tendency for /a/ -> /ä/ when in close proximity to /y/, pharyngeals or fricatives. 3.126 ač̄uvuri / äč̄uvuri ‘to lay' 3.127 yaxị ri / yäxị ri ‘to leave' 3.128 ḥayvan / ḥäyvan ‘animal' /wi/ /u/ and /uy/ can often be used interchangeably without changing the meaning. 3.129 enǯikuval / enǯikuyval ‘to come down' 3.130 kakwidmä / kakudmä ‘will come' MASC.class 3.131 xạ̈ č̄ ulq̇ utomä / xạ̈ č̄ ulq̇ witomä ‘hangs' 3.131 kuidämä / kudämä / kuydämä ‘will be' FEM.class Some consonants can be used interchangeably in some isolated words: 44 /c/ // /s/: 3.133 gic / gis ‘roof' /š/ // /s/: 3.134 šicị / sicị ‘to write' /t/ // /d/ and /č/ // /ǯ/ after /n/: 3.135 antırǧi / andırǧi ‘to fall' 3.136 enǯikui / enčikui ‘to come down' 45 Table 3.1 Comparison Chart of Xinaliq Phonemes Scientific orthograph y Kibrik Namik Villag e Ganiev a IPA Georgia n pača ‘leg' a a a A а a ა sıb ‘green' b b b B б b ბ lucoz ‘cow' borc ‘father's sister' c c ŝ Ts ц ts ც yec̄ini ‘to remain' c̄ cc ẑ T́s цц ts: წწ bacịz ‘small goat' c ̣ c' s̆ ts' цӏ ts' წ mıčäš ‘tea' mıč ‘apple' č ç ç Ç ч tʃ ჩ ınč̄ i ‘earth' ḳuč̄ i ‘heel' č̄ çç ĉ Ḉ чч tʃ: ჭჭ fičạ̈ ‘wet' č ̣ ç' ç̆ ç' чӏ tʃ' ჭ dädä ‘mother' d d d D д d დ q̇ilez ‘salty' e e e E е e ე äjä ‘grandmother' ä ə ə ə аь æ ა̈ čạ f ‘untouched grazing' f f f F ф f ჶ ninig ‘eyelash' g g g g г g გ ḳıγ ‘eyebrow' γ gh ĝ gh гг γ ღ̃ aliğ ‘horse covering' zeeğ ‘chicken broth' ǧ ğ ğ ğ гъ ʁ ღ dahar ‘a big rock' h h h h гь h ჰ ḥädmi ‘human' ḥ hh ħ ħ хӏ ħ ჰ ninig ‘eyelash' i i i i и i ი mıč ‘apple' ı ı ı ı ы ɯ ჷ kunǯ ‘corner' ǯ j c c дж dʒ ჯ 46 Table 3.1 Continued Scientific orthograph y Kibrik Namik Villag e Ganiev a IPA Georgia n kukac ̣‘chicken' mok ‘worm' k k k k к kʰ ქ muk̄ ur ‘dustpan' k̄ kk ҟ ḱ кк k: კკ ḳuḳa ‘knot' näḳid ‘back' (on a body) ḳ k' қ k' кӏ k' კ lıx ‘herder's stick' x kh ĥ kh хь x ხ mokx ‘worm' kxe ‘wool' kx kx kx ქხ ḳulla ‘hammer' cal ‘sheaf' l l l l л l ლ hilam ‘donkey' m m m m м m მ ḳind ‘bridge' n n n n н n ნ culoz ‘tooth' o o o o о o ო söbät ‘conversation' ö ö ö ö оь ø ო ṭop ‘ear' p p p p п pʰ ფ k̄ arp̄a ‘lamb' p̄ pp ḇ ṕ пп p: პპ ṗaṗax ‘fur hat' ṗ p' ҏ p' пӏ p' პ ınx̑a ‘field' läx̑iz ‘calf' x̑ q ҳ qx хъ qχ ჴ toq̄ ud ‘lightning' q̄ qq q q къ q: ყყ ınq̇ ‘sun' q̇ q' ǭ q' кь q' ყ p̄eyram ‘shirt' r r r r р r რ ans ‘game' s s s s с s ს pšä ‘bread' š ş ş ş ш ʃ შ ant ‘floor' t t t t т tʰ თ 47 Table 3.1 Continued Scientific orthograph y Kibrik Namik Villag e Ganiev a IPA Georgia n kät̄id ‘Xinaliq person' bat̄a ‘short' t̄ tt đ t́ тт t: ტტ xurṭum ‘throat' ṭ t' t̆ t' тӏ t' ტ culoz ‘tooth' u u u u у u უ nüc ̣‘honey' ü ü ü ü уь y უ̈ čụ val ‘ram' v v v v в v ვ arxạ c ‘yard' (?) mixẹ z ‘needle' x ̣ x x x х χ ხ qoyu ‘water well' y y y y й j ჲ mixez ‘needle' z z z z з z ზ až ‘side' (on a body) ḳinaž ‘winter' Ž z̧ j j ж ʒ ჟ ʕaza ‘moaning' ʕ ʕ ̀ ʼ гӏ ˁ ჺ vaʔ ‘no' sæʔil ‘here' ʔ ʔ ˀ ჸ Table 3.2 The Xinaliq Consonant Inventory in IPA. Bilabial Dental Alveolar Palato- Alveolar Palatal Velar Uvular Pharyngeal Glottal Stop p p: b t t: D k k: g q q: ʔ Ejective p' t' ts' tʃ ' k' q' Fricative f v s z ʃ ʒ x ɣ X ʁ ħ ʕ h Affricate ts ts: tʃ tʃ : dʒ Kx qX Nasal m N Trill r Lateral approx. l Glides j w 48 Table 3.3 The Xinaliq Consonant Inventory in "Scientific Transcription" Bilabial Dental Alveolar Palato- Alveolar Palatal Velar Uvular Pharyngeal Glottal Stop p p̄ b t t̄ d k k̄ g q q̄ ʔ Ejective ṗ ṭ c ̣ č ̣ ḳ q̇ Fricative f v s z š ž x ɣ x ̣ ǧ ḥ ʕ h Affricate c c̄ č č̄ ǯ Kx x̑ Nasal m n Trill r Lateral approx. l Glide y w Table 3.4 Xinaliq Vowel Inventory Front Back Plain (nonround) Round Plain (nonround) Round High i ü ı * u Mid e ö o Low ä a *This is what is traditionally transcribed as [ɯ] phoneme, back high vowel. 49 CHAPTER 4 THE NOUN Although often straightforward, in some languages it is not always possible to determine clean cut boundaries between different parts of speech. Even with the best guidelines and grammatical tests for determining category membership for different parts of speech, in certain instances some parts of speech bleed into different categories on the continuum, creating exceptions and variations. The discourse context can have that effect on different parts of speech in Xinaliq. That cautionary note being stated, the patterns for the most part can be readily identified, and the prototypes of Xinaliq's parts of speech are presented here beginning with the noun. Xinaliq nouns have grammatical gender, case, and number. 4.1 Noun Class In different languages nouns can be categorized by classes in different ways, for instance by the gender or animacy of the noun. Noun classes generally form a system of grammatical agreement, often referred to as grammatical gender. Most Northeast Caucasian languages manifest noun class distinctions, varying from none up to eight. The Xinaliq noun class system has four classes. A fourfold class division is found in many of the Lezgic languages, as well as in Lak, Tsez, Hinukh and Bezhta and some dialects of Andi. However, there is variation among the fourfold class systems. Generally the 51 Xinaliq system distinguishes animate from inanimate, and human from nonhuman categories. The traditional Classes I, II and III all refer to animate nouns, while Class IV includes the inanimate nouns. More specifically, only human nouns belong to Classes I and II, while nonhuman animates go into Class III. Within the human nouns there is a distinction based on biological gender: masculine (Class I) and feminine (Class II). The nonhuman nouns, instead of taking some single neuter form, are further divided into animate nonhuman and inanimate nouns. Class I and II are more consistent and straightforward; however, Class III and Class IV distinctions are less obvious at times and require further detailed investigation. Henceforth, different classes will be referred to by their semantic associations rather than numbers, thus Masculine (Class I), Feminine (Class II), Animate (Class III) and Inanimate (Class IV). Noun classes are not overtly visible in the nouns themselves. The main indicator of class membership and hence of class distinctions is verb agreement markers, the morphemes which mark agreement on verbs with the class and number of the noun. Some examples of masculine class and feminine class nouns are presented below: Class MASC - human males: 4.1 aba ‘grandfather' 4.2 bäg ‘bridegroom' 4.3 ši ‘son' Class FEM - human females: 4.4 äǯä ‘grandmother' 4.5 dädä ‘mother' 4.6 ris̆i ‘daughter' 52 The nouns that are less straightforward, for instance those denoting professions and social or moral features, depend on the assumed biological gender of the person referred to. For instance, the nouns below can be either Class MASC. or Class FEM.: 4.7 gons̆u 'neighbor' 4.8 dost 'friend' 4.9 dabxer 'deceiver' Membership of a noun in Class AN. or Class INAN. is determined by semantic criteria. Both classes are for nonhuman nouns. Animate nouns fall into Class III, while Class IV receives inanimate, nonhuman nouns: Class AN - animate nonhumans: 4.10 gra ‘wolf' 4.11 hilam ‘donkey' 4.12 ḳaḳıd ‘partridge' 4.13 mokx ̣ ‘worm' Class INAN - other nouns (inanimate, nonhuman): 4.14 cı̣za ‘fear' 4.15 ula ‘smell' 4.16 fikir ‘thought' 4.17 ḳaḳa ‘bite' (noun) 4.18 čee ‘tea' 4.19 gis ‘roof' From the examples given, it is clear that the INAN. Class includes concrete things as well as words denoting states, actions, properties, etc. 53 4.1.1 Class Markers (CM) Class Markers are affixes, the main function of class markers is to make agreement between the verb and the different classes and numbers of the nouns. More specifically, they indicate different class membership and the number of the Xinaliq noun. Almost all Northeast Caucasian languages have the category of class, but their distribution of markers differs drastically. Xinaliq nouns themselves do not bear explicit class marking. Similarly to other Northeast Caucasian languages, class markers in Xinaliq are generally determined by looking at the dependent verbs, adjectives, pronouns and prepositions. Some classes have zero as the class morpheme marker, and some differentiate between singular and plural, which means that their class membership when they are singular differs from the class to which they belong when they are plural. All and all there are three series of noun class-number markers. The phonological environments they occur in determine the series. In other words, different series are designated for different forms of the class-marker affixes, which can vary according to the phonological environment they occur in, which is also what determines its series. If the verb has more than one stem, then the appropriate class-number markers can repeat, attaching themselves to each stem. This feature seems to be becoming obsolete as Xinaliq loses speakers and falls more heavily under the influence of Azeri. Many sentences can be uttered with or without repeated class markers, only some of them. Some language consultants insist the class markers are needed; while others think they are not mandatory. This is likely indicative of a language feature in the process of change. Noun-verb agreement in Xinaliq depends on verb transitivity. Transitive verbs agree with the direct object and therefore take class-number markers that agree with 54 direct object nouns, while intransitive verbs agree with their subjects and therefore take class-number markers that agree with those nouns. Thus the verbs indicate the class membership of the subject or direct object, depending what kind of verb it is. The future tense tends to be more flexible about the use of class markers, while the present and past tenses bear them most of the time. Most Xinaliq verbs take agreement markers that indicate both noun class and number, but there are some exceptions. For instance some of the non-resultative verbs do not require these agreement markers, such as: küi "to go," küi "to be," daxị "to see," ḳli "to die," as well as some compound verbs that include küi as one element of the compound (e.g., ansküi "to play," dalıgküi "to work," paküi "to kiss"). The three series of class markers are discussed below as can be seen in Table 4.1. In AN class (discussed below) b> Ø / [+segment]__ k (see Rule 7 below), and in INAN class it is Ø. Thus, an overt class marker does not appear in these instances. The phonetic variation in Series 1 among FEM sing., AN sing., MASC pl. and FEM pl. is based on certain phonological restrictions. The phonological rules are presented below, and examples with all possible affixes will be given under each appropriate subsection to follow. Discussion of permitted and not permitted consonant clusters follows the rules below. Rule 1: (CM1.F) z → z/___ [+voice]C (z stays z before voiced consonant) Rule 2: (CM1.F) z → s/__ [-voice]C (z becomes s before voiceless consonants) Rule 3: (CM1.F) 55 z → c/ __ x (z becomes c before x in prefix-initial [word-initial only] position) Rule 4: (CM1.AN) Ø →i/z__C that creates unpermitted CC cluster therefore CC →CiC (i is inserted between two CC that would otherwise result in a cluster that is not permitted) Rule 5: (CM1.AN) b → b/___ [+voice]C (b stays b before the voiced consonants) Rule 6: (CM1.AN) b → p/__ [-voice]C with the exception of Rule 7 (b becomes p before voiceless consonants) Rule 7: (CM1.F) b → Ø / [+segment]__ k, p (b becomes Ø before k or p in non-word-initial position) Rule 8: (CM1.AN) Ø → (i)/b__C that creates unpermitted CC cluster therefore CC →CiC (i is inserted to break up a consonant cluster which is not permitted in the language) Rule 9: (CM2.F) z -> r/___syllable with any sibilant (c, č, z, ž, ǯ, š) Rule 10: (CM3.M) h - >Ø/__r Each series will be discussed separately for each of the four noun classes. Examples with all possible affixes are presented. Permitted consonant clusters in Xinaliq: 56 Word initial: Combination of Stop + Affricate (čxạ r ‘buckwheat') + Fricative(bzı ‘pear', pšä ‘bread') + Liquid (kla ‘who') (not necessary for stops to be cluster initial) Stop + Fricative + Liquid (pšlä ‘fox', pxṛ a ‘dog') Word final: No combinations of three phonemes were found. Sonorant + Obstruent (lıgıld ‘man', borc ‘father's sister', ant ‘earth') Fricative + Stop (vaxṭ ‘time', dost ‘friend') Medial: Sonorant + Obstruent (kulga ‘shadow',) Fricative + Stop (äsḳi ‘towel,) Approximant + consonant (ayvan ‘balcony') Glottal + consonant (säʕbi ‘master', näʕnä ‘saliva') Obstruent + Sonorant (ägni ‘clothing', häfmi ‘human') In all consonant clusters, both obstruents must be either voiced or both voiceless. 4.1.1.1 Series 1 (CM1) Series 1 includes the consonant-initial words. The phonemic variations within FEM sing., as well as among AN sing., MASC pl. and FEM pl., arise from the avoidance of nonpermitted consonant clusters. The Series 1 class-number markers are attached to the beginning of the verb stem. 57 4.1.1.1.1 Class Markers Series 1 Class MASC. (CM1.MASC.SG) In the example below ‘boy' belongs to Class MASC, resulting in a CM1.MASC.SG marker on the verb, which is Ø. 4.20 gada Ø-ḳi-šä-mä boy.ABSL CM1.M.SG-die.PRFV-PST-INDIC ‘The boy died' In Example 4.21, because "child" here is assumed to be male and therefore to belong to Class MASC as a male human, there is a CM1.MASC.SG marker, which is Ø, on the verb. 4.21 lıgıld-i ḥäyäl ṭuv-Ø-šä-mä man-ABS child buy-CM1.M.SG-PST-INDIC ‘The man is buying a child' 4.22 as lıgıld zaǧ-Ø-šä-mä I man saw-CM1.M.SG-PST-INDIC ‘I saw a man' 4.23 lıgıld-u hine ši yiq̇-Ø-šä-mä man-ERG his son want-CM1.M.SG-PST-INDIC ‘The man wants his son' 4.24 zar-Ø-k-ir-dü-mä OD-CM1.M.SG-send-IMPRFV-VCMII.SG-INDIC ‘Someone will send him down….' 4.25 yec̄ın-Ø-däd-i-mä stay-CM1.M.SG-VCM-NEG-INDI ‘[the man] will not stay' 4.26 gada yec̄ in-Ø-i-šä-mä cẉ a boy stay- CM1.M.SG-PRFV-PST-INDIC home ‘The boy stayed at home' 58 4.27 Ø-xä q̇inä CM1.M.SG-go eat ‘Go eat' (said to a man) 4.28 mäʔälimmdar laǧın Ø-xi-d-mä teacher mountain yesterday CM1.M.SG-go-VCM-INDIC ‘The teacher (male) went to the mountain yesterday' 4.1.1.1.2 Class Markers Series 1 Class FEM. (CM1.FEM.SG) In Example 4.29 below, because "girl" belongs to FEM Class, there is a CM1.FEM.SG marker on the verb. We know that it is part of Series 1, because the verb begins with a consonant. Because z and ḳ together would be a consonant cluster that is not permitted in Xinaliq, i is inserted to break up the cluster, as in Rule 4 above. The same sentence in the future tense (4.30) does not take class marking. 4.29 riši zi-ḳ-i-šä-mä girl.ABS CM1.F.SG-die-PRFV-PST-INDIC ‘The girl died' 4.30 riši ḳ-ıli-dä-mä girl.ABS die-IMPRFV-VCMII.SG-INDIC ‘The girl will die' Another example of Rule 4 to break up a forbidden cluster: 4.31 č̣e-zi-v-šä-mä sit-CM1.F.SG-sit-PST-INDIC ‘The woman sat down' 4.32 lıgıldi attı-zi-q̇-dä-z-i-šä-mä xinimḳir MAN have-CM1.F.SG-have-VCMII.SG-CM1.F.SG-NEG-PST-INDIC wife ‘The man didn't have a wife' In Example 4.33, because "fiancée" refers to a female it belongs to FEM Class; as 59 a female animate human noun, there is a CM1.F.SG marker, z, on the verb. Example 4.35 is future tense and is missing the class marker. 4.33 lıgıldi cnas ṭu-zi-v-šä-mä man.ABS fiancée buy-CM1.F.SG-buy-PST-INDIC ‘The man bought a fiancée' 4.34 as e rıtsı za-zı-ǧ-d-i-mä I.DAT my sister see-CM1.F.SG-see-VCMII-NEG-INDIC ‘I have not seen my sister' 4.35 lägeldu at-ku-dä-mä riše man.ERG have-FUT-VCMII.SG-INDIC daughter ‘The man will have a daughter' 4.36 uxur zı za-z-ɣ-i-du you me see-CM1.F.SG-see-PRFV-VCM.I.SG ‘Did you (to a woman) see me?' 4.37 uxur dädä za-z-ɣ-i-du you mom see-CM1.F.SG-see-PRFV-VCM.I.SG ‘Did you see Mom?' 4.38 hä bičẹ q̇ o-z-i-mä she fat be-CM1.F.SG-NEG-INDIC ‘She is not fat' 4.39 hä hundur q̇o-z-i-mä she tall be-CM1.F.SG-NEG-INDIC ‘She is not tall' In the examples below z → s/__[-voiced]C (z becomes s before voiceless consonants). 4.40 Xeyraddini xinimḳir ḳur-s-qen-šä-mä Xeyraddine wife forget-CM1.F.SG-forget-PST-INDIC 60 ‘Xeyraddine forgot the wife' 4.41 Xeyraddini ḳur-s-qink-ir-dä-mä xinimḳir Xeyraddine forget-CM1.F.SG-forget-IMPRFV-VCM.II.SG-INDIC wife ‘Xeyraddine will forget the wife' 4.42 kur-s-qin-s-k-ir-dä-z-i-mä forget-CM1.F.SG-forget-CM1.F-forget-IMPRFV-VCMII.SG-CM2.F.SG-NEG-INDIC ‘I will not forget (woman)' 4.43 dä xinimḳir al-s-kır-dä-mä this wife OD-CM1.F.SG-fall-VCMII.SING-INDIC ‘This wife will fall from above' 4.44 lıgıldi xinimḳir or-s-k-ir-mä tukanır man.ABS wife send-CM1.F.SG-send-IMPRV-INDIC store ‘The man will certainly send the wife to the store' z → c/ __ x 4.45 dä xinimḳir za-c-xir-dä-mä this wife OD-CM1.F.SG-fall-VCM.II.SING.-INDIC ‘This woman will fall forward‘ 4.46 dä xinimḳer la-c-xır-dä-mä this wife OD-CM1.F.SG-fall-VCMII.SING-INDIC ‘This woman will fall to the side' 4.47 dä xinimḳer a-c-xir-dä-mä this wife OD-CM1.F.SG-fall-VCMII.SING-INDIC ‘This woman will fall back' 4.48 c-xä q̇inä CM1.F.SG-go eat ‘Go eat' (to a woman) 4.49 mäʔälim mdar laǧın c-xi-dä-mä 61 teacher mountain yesterday CM1.F.SG-go-VCMII.SG-INDIC ‘The teacher (female) went to the mountain' 4.1.1.1.3 Class Markers Series 1 Class AN (CM 1.AN.SG.) In Example 4.50, because "donkey" belongs to Class AN as an animate nonhuman noun, there is a CM1.AN.SG marker on the verb. Because b and ḳ are not permitted together in a consonant cluster in Xinaliq, i is inserted to break up the cluster. 4.50 hilam bi-ḳ-i-šä-mä donkey.ABSL CM1.IN.SG-die-PRFV-PST-INDIC ‘The donkey died' In Example 4.51, because "horse" is nonhuman animate, it belongs to Class AN and carries the CM1.AN.SG marker. The sentence in (4.51) exemplifies the phonological process b→ p before the voiceless consonant, in this case before q̇, of the verb stem. 4.51 Xeyraddine kur-p-q̇ink-ir-dä-mä pši Xeyraddine forget-CM1.AN.SG-forget-IMPRFV-VCM.III.SG-INDIC horse ‘Xeyraddine will forget the horse' In the examples below "mountain" takes Class AN marker even though it is an inanimate object. These inconsistencies in the noun class categories need to be investigated further. 4.52. Tom-u za-b-ɣ-dä-mä mıda Toma-DAT. OD-CM1.AN.SG.-saw-VCMIII.SG-INDIC mountain ‘Toma saw a mountain' 4.53 Xeyraddine kur-p-qink-ir-dä-mä pši Xeyraddine forget-CM1.AN.SG-FORGET-IMPRFV-VCM.III.SG-INDIC horse ‘Xeyradinne will forget the horse' 4.54 lägäld-i att-bı-q̇-dä-šä-mä pši 62 man-GEN have-CM1.AN.SG-have-VCMIII.SG-PST-INDIC horse ‘The man had a horse' 4.55 lıgıld-i att-bı-q̇-dä-v-i-šä-mä pši man-GEN have-CM1.AN.SG-have-VCMIII.SG-CM2.AN.SG-NEG-PST-INDIC horse ‘The man did not have a horse' 4.56 ḳur-p-q̇ink-ir-dä-v-i-mä forget-CM1.AN.SG-forget- IMPRFV-VCMIII.SG-CM2.AN.SG-NEG-INDIC ‘I didn't forget (animal)' 4.57 uxur za-b-ɣ-i-d-u? you OD-CM1.AN.SG-see-PRFV-VCM.III.SG-INTER. ‘Did you see me?' (animal speaking) 4.58 pıše antır č̣e-bi-šä-mä horse floor sit-CM1.AN.SG-PST.INDIC ‘The horse sat down on the floor' 4.59 p-xä q̇inä CM1.AN.SG-go eat ‘Go eat!' (to an animal) 4.1.1.1.4 CM1.MASC.PL and CM1.FEM.PL In Example 4.60, because "them" is plural, it belongs to Class MASC.PL or Class FEM.PL. The sentence again exemplifies the phonological process b → p before the voiceless consonant, in this case before q̇, of the verb stem. 4.60 Kur-p-q̇ink-ir-dur-v-i-mä forget-CM1.M.PL-forget-IMPRFV-VCMI.PL-CM2.M.PL-NEG-INDIC ‘I will not forget them' (males of females) 4.61 mäʔälımir mıdar laǧın p-xi-dur-mä teachers mountain yesterday CM1.M.PL-went-VCMI.PL-INDIC ‘Teachers went to the mountain yesterday' 63 4.1.1.1.5 CM1.INAN.SG In Example 4.62, because "book" is inanimate, it belongs to Class INAN and carries the CM1.INAN.SG marker Ø. 4.62 kitab ustuli üstür č̄äf-Ø-q̇o-mä book table top laying- CM1.INAN.SG-is-INDIC ‘The book is lying on the table' 4.63 kätši ǯämäʔätiixer če cular- Ø-mä inxinaliq people a lot tea drink- CM1.INAN.SG-INDIC ‘People in Xinaliq drink a lot of tea' 4.64 dädänziketišir ixer če cular-Ø-mä mothers in Xinaliq a lot tea drink- CM1.INAN.SG-INDIC ‘Mothers in Xinaliq drink a lot of tea' 4.1.1.1.6 CM1.AN.PL. In Example 4.65, because "frogs" is animate and in plural, it takes Class AN plural marker Ø. 4.65 q̇urq̇ordri ınk̄iüstüllü ṭäxirkitar-Ø-mä frog lake above jump-CM1.AN.PL-INDIC ‘Frogs jump above the lake' 4.66 ḳaz al buǧrur ṭaḳoar-Ø-mä snake water near are- CM1.AN.PL-INDIC ‘Snakes are near the water' 4.67 c̄xıbaalıǧi mısi balıǧ q̇antar-Ø-mä big fish small fish eats-CM1.AN.PL-INDIC ‘Big fish eat little fish' 64 4.1.1.1.7 CM1.INAN.PL. In Example 4.68, because "flowers" is inanimate and in plural, it takes Class INAN plural marker Ø. 4.68 kätišir iüni vcır gül äc̄mıškitar-Ø-mä inxinaliq in june month flowers grow-CM1.INAN.PL-INDIC ‘Flowers grow in June in Xinaliq' 4.69 misi fikir sin-sin c̄xi fikir ḳoar-Ø-mä small ideas often big idea become- CM1.INAN.PL-INDIC ‘Small ideas often become big ideas' 4.70 ınq̇ här q̇ä qaltıxḳoar-Ø-mä sun every day rise-CM1.INAN.PL-INDIC ‘The sun rises every day‘ Table 4.2 shows more examples of Series 1 with regular verbs. 4.1.1.2 Series 2 (CM 2) Series 2 markers are used before the vowel-initial verbs. They can be used in combination with Series 1 markers. 4.1.1.2.1 CM2.MASC.SG In the example below because "man" is a human male, it belongs to Class MASC. Because the verb "kill" begins with a vowel, it is in Series 2. Therefore it carries the CM2.MASC.SG marker y on the verb. 4.71 lıgıldı lıgıld y-ib-šä-mä man man CM2.MASC.SG-kill-PST-INDIC ‘Yesterday a man killed a man' 4.72 y-eḳḳäk̄u-i CM2.MASC.SG-burn-NEG 65 ‘Don't burn!' (talking to a man) 4.1.1.2.2 CM2.FEM.SG In the example below, because "the enemy" is assumed to be a woman by the speaker, it takes Class FEM. As determined above, the verb "kill" is in Series 2. Therefore it carries the CM2.FEM.SG marker z on the verb. 4.73 xisansan lıgalde z-ib-šä-mä dušman yesterday man CM2.FEM.SG-kill-PST-INDIC enemy ‘Yesterday a man killed his enemy' (of woman) 4.74 xisansan lıgalde z-i-z-b-šä-mä dušman yesterday man CM2.FEM.SG-kill-CM1.FEM.SG-kill-PST-INDIC enemy ‘Yesterday a man killed his enemy' (of woman) maybe more certainty 4.75 paga lıgılde z-i-z-b-ir-dä-z-i-mä xinimḳer tomorrow man CM2.FEM.SG-kill-CM1.FEM.SG-kill-IMPERF-VCMII-CM2. FEM.SG-NEG-INDIC woman ‘Tomorrow a man will not kill a woman' 4.76 riše r-ac̄ in-ı-šä-mä cẉ a girl CM2.FEM.SG-stay-PRFV-PST-INDIC home ‘The girl stayed home' 4.77 räše r-ac̄ in-ı-nätt-i-šä-mä cẉ a girl CM2.FEM.SG-stay-PRFV-VCMII.SG-NEG-PST-INDIC home ‘The girl didn't stay home' 4.78 z-ek-s-käku-z-u! CM2.FEM.SG-burn-CM2.FEM.SG-burn-CM1.FEM.SG-INTER ‘Don't burn!" (to a woman) Examples 4.76 and 4.77 show the phonological variation that sometimes occurs with Series 2, Class FEM markers, z -> r/___syllable with a sibilant. 66 4.1.1.2.3 CM2.AN.SG In the example below, because "horse" is a nonhuman animate, it belongs to Class AN. Because the verb ‘stay' begins with a vowel, it is in Series 2. Therefore it carries the CM2.AN.SG marker v on the verb. 4.79 pši v-ac̄ in-i-šä-mä cẉ a horse CM2.AN.SG-stay-PRFV-PST-INDIC home ‘The horse stayed home' (own will) 4.80 pši v-ac̄ in-i-d-i-šä-mä cẉ a horse CM2.AN.SG-PRFV-VCM.III.SG-NEG-PST-INDIC home ‘The horse didn't stay home' In the example below, because "lamb" is a nonhuman animate, it belongs to Class AN. Because the verb "kill" begins with a vowel, it is in Series 2. Therefore it carries the CM2.AN.SG marker v on the verb. 4.81 xisansan lıgılde v-i-b-d-i-mä čụ val yesterday man CM2.AN.SG-kill-VCMIII.SING-NEG-INDIC lamb ‘Yesterday a man did not kill a lamb' 4.82 xisansan lıgılde v-i-b-dä-v-i-mä čụ val yesterday man CM2.AN.SG-kill-VCMIII.SING-CM2.AN.SG-NEG-INDIC lamb ‘Yesterday a man did not kill a lamb' 4.83 pši v-ac̄in-d-i-šä-mä horse CM2.AN.SG-stay-VCMIII.SG-NEG-PST-INDIC ‘The horse did not stay' 4.84 pši v-ac̄in-dä-dä-v-i-šä-mä horse CM2.AN.SG.-stay-VCMIII.SG-VCMIII.SG-CM2.AN.SG-NEG-PST-INDIC ‘The horse will not stay' 4.85 hä hundur qo-v-i-mä it tall is-CM2.AN.SG-NEG-INDIC 67 ‘It (horse) is not tall' 4.86 v-ekkäku-i! CM.2.AN.SG-burn-NEG ‘Don't burn yourself!' (animal) 4.87 v-ekk-ir-va-l-v-i-mä CM2.AN.SG-burn-IMPRFV-CM2.AN.SG-?-CM2.AN.SG-NEG-INDIC ‘I am not burning (myself)' (animal) 4.88 v-ibe pši CM2.AN.SG-kill horse ‘Kill the horse' 4.1.1.2.4 CM2.MASC.PL 4.89 lıgılır q̇ätmiḳilanttın islanmıšbiq̇i-šä-mä men all day sweated-PST-INDIC ‘Men sweated all day' 4.90 gadadiṭuzaxanskirto-mä boys yard play-INDIC ‘Boys are playing in the yard' 4.1.1.2.5 CM2.FEM.PL 4.91 xisansan lıgılde v-ib-šä-mä dušman yesterday man CM2.FEM.PLU-kill-PST-INDIC enemy ‘Yesterday a man killed his enemies' (of many women) 4.1.1.2.6 CM2.INAN.SG 4.92 I putu tämizval y-uḳoar-mä my hair cleanliness CM2.INAN.SG-love-INDIC ‘My hair loves cleanliness' 68 4.1.1.2.7 CM2.AN.PL 4.93 pšor cẉ a y-etstsin-šä-mä horses home CM2.AN.PL-stay-PST-INDIC ‘Horses stayed home' (own will) 4.1.1.3 Series 3 (CM3) Series 3 seems to be dominated by imperative constructions. Most of the Series 3 markers are used with the verb "to be" q̇i and other verbs that are formed from this verb. The imperative suffix in Xinaliq is r, and class markers in Series 3 are always placed before it as demonstrated in Table 4.3. 4.1.1.3.1 CM3.M.SG 4.94 yi-h-ar vavegada love-CM3.M.SG-IMP your son ‘Love your son!' 4.95 il ka-Ø-r here come-CM3.M.SG-IMP ‘Come here!' (to a man) 4.96 il k-enǯi-Ø-r here OD-come.down-CM3.M.SG-IMP ‘Come down from above!' (to a man) 4.1.1.3.2 CM3.F.SG 4.97 il ka-s here OD-come-CM3.FEM.SG ‘Come here!' (to woman) 4.98 il k-enǯi-s here OD-come.down- CM3.FEM.SG 69 ‘Come down from above!' (to woman) 4.1.1.3.3 CM3.AN.SG 4.99 bi-yi-f-ar pši CM1.AN.SG-love-CM3.AN.SG-IMP horse ‘Love your horse!' 4.100 ṭoč-f-är stand-CM3.AN.SG-IMP ‘Stand up!' (animal) 4.1.1.3.4 CM3.M.PL 4.101 il k-a-f-un here OD-come-CM3.M.PL-PL.M. ‘Come here!' (to many) 4.1.1.3.5 CM3.F.PL 4.102 il k-a-f-ir-uz here OD-come-CM3.FEM.PL-IMP-PL.M ‘Come down here!' (to women) 4.1.1.3.6 CM3.INAN.SG 4.103 dä fikir yi-h-ar this idea love-CM3.INAN.SG-PL.M ‘Love this idea!' 4.1.1.3.7 CM3.AN.PL 4.104 il kaf-r-in here idea-CM3.AN.PL-PL.M ‘Come here!' (to many animals) 70 4.2 Case System A language has a case system if its nouns (and sometimes other parts of the sentence) inflect (i.e., change their form) to define their relationship to the other morphemes in a statement. There are many ways to perform this function, the inflection of nouns being one of them. Languages that do have cases can be ranked according to a hierarchy. If a given case at the far left of the hierarchy does not exist in a language, none of the remaining cases to the right will either. Nominative > accusative or ergative > genitive > dative > locative > ablative >instrumental > prepositional > others (Blake, 2001) Languages can also be categorized according to their treatment of agents and patients, referred to as morphosyntactic alignment. Distinctions may be made morphologically through grammatical case and verbal agreement, or syntactically through word order. Xinaliq's use of grammatical case indicates that it is an ergative-absolutive language (usually simply referred to as ergative). In ergative languages, the argument (i.e., the subject) of an intransitive verb is in the same case as the patient (i.e., direct object) of a transitive verb. This case is then called the absolutive (or nominative), with the agent (i.e., subject) of a transitive verb being in the ergative case (see Chapter 9). Languages that use case systems vary widely in the number of cases they have. For instance, the most conservative Indo-European languages have approximately eight cases; however, in Xinaliq there are thirteen cases: 1. Nominative 2. Ergative 3. First Genitive 4. Second Genitive 71 5. Dative 6. Comitative 7. Locative (General) 8. Orientational locative 9. Possessive locative 10. Ablative (General) 11. Orientational ablative 12. Possessive ablative 13. Comparative All Northeast Caucasian languages tend to have nominative, ergative, first genitive, second genitive, dative and comitative cases, which show the direct relationship with the subject or object of the clause. Locative and motion cases give information about the orientation in space or time of the subject or the object. Table 4.4 lists the affixes and phonological alternations for all the Xinaliq cases (vowel harmony applies). Table 4.5 shows examples of nouns ending in a high vowel, a low vowel, and a consonant. The descriptions provided below follow traditional analyses and consider each case separately. It is possible to think of some cases as combinations of different morphemes from other cases. For instance, possessive ablative -s̆illi may be regarded not as an independent case but as the combination of two morphemes, marker -s̆ (possessive locative) and -illi (ablative). This is also true for the other combinations above, although in the traditional analysis they are treated as distinct cases and separate morphemes. 4.2.1 Nominative (Absolutive) What is called "nominative" case in Caucasian terminology is actually an "absolutive" case in modern general linguistic terminology when speaking about ergative languages (the case that signals the subject of an intransitive verb and the object of a 72 transitive verb). The absolutive in Xinaliq has the same form as the singular noun stem; that is, it has Ø marking. It is common cross-linguistically for the absolutive case to have no overt marking, that is, to be marked by Ø. Kibrik (1994) describes cases according to their semantic roles in addition to their grammatical functions. He describes the nominative case using the term "actant," defined as a noun phrase functioning as the agent immediately associated with an intransitive verb. When associated with an intransitive verb, its nominative function changes. When the actant takes the nominative case, in other words, it expresses either the agent of an intransitive verb or main actor. Examples 4.105-4.106 show nonaccusative intransitive verbs, while (4.107) and (4.108) are examples of nonergative intransitive verbs. 4.105 cạ̈ -ø yavas̆ ḳl-i-qo-mä fire-NOM slowly go.out-PRF.ASP-PRF.II-INDIC ‘The fire is slowly going out' 4.106 gada-ø ḳ-i-šä-mä boy.NOM die-PRFV-PST-INDIC ‘The boy died' 4.107 bu lägäld-ø latkär-mæ this man.NOM fall-INDIC ‘This man will fall' 4.108 xisansan lägald-ø z-ib-šä-mä dušman yesterday man CM2.F.SG-kill-PST-INDIC enemy ‘Yesterday a man killed his enemy' (of woman) 4.109 rišɛ-ø r-accin-i-šä-mä girl CM2.F.SG-stay-PRFV-PST-INDIC ‘The girl stayed' 73 When the verb is transitive, the nominative case is expressed on the direct object, while the subject takes an ergative case. This is true in what are traditionally called dative constructions as well (i.e., constructions where the direct object also takes the so-called nominative case). This construction will be discussed below. The following examples illustrate the case marking of subjects of transitive clauses, including Ø "absolutive" markers of direct objects. 4.110 yä xın-ır ḳač̣ın-ø fatku-s̆ä-mä I water-LOC stone-NOM threw-PST- INDIC ‘I threw a stone into the water' 4.111 dämirc̄-i ura-ø antk-ir-qo-mä smith-ERG metal-NOM make- IMPRF.ASP-PRS.II-INDIC ‘The smith forges metal' 4.2.2 Ergative The ergative case signals the subject of a transitive verb. It is marked with -i. 4.112 pxṛ -i zı-Ø cụ x-̣ s̆ ä-mä dog-ERG me-NOM bit-PST-INDIC ‘The dog bit me' 4.113 pšor-i ink-Ø q̇andäto-mä horses-ERG grass-NOM eat-INDIC ‘Horses eat grass' The ergative case can also have the same connotation as the instrumental. The ergative and instrumental cases are homophonous. This feature is not unusual cross-linguistically. There are many languages in which the marker for the ergative case is the same in form as the marker for some other case, here the instrumental. 4.114 yä top-Ø ink-i fatku-s̆ä-mä 74 I.ERG ball-NOM foot-INST hit-PST-INDIC ‘I hit the ball with my foot' 4.115 ḳütc-̣ i q̇ až-i latürku-šä-mä snake-ERG tail- INST push-PST-INDIC ‘The snake pushed it with its tail' The case marking for personal nouns is discussed in Chapter 6. 4.2.3 First Genitive Alienable vs. inalienable possession is grammatically distinct in Xinaliq and expressed through cases. The first genitive case generally is used with animate nouns to denote inalienable possession. "Inalienable" possession is generally defined as that which is always possessed. In Xinaliq this category is less straightforward, as it does not include the category of kinship. In the interests of accuracy, the term "organic" will be used to indicate inalienable possession in Xinaliq. The exact distinction between alienable and inalienable possession in Xinaliq will be discussed below. The first genitive is marked with -i, just as the ergative case is. It is used independently as well as with postpositions, which will be discussed later. As an independent case, first genitive is used with animate nouns only, therefore only with nouns of the noun classes I, II, and III, as well as with personal nouns, expressing inalienable possession. 4.116 pxṛ -i q̇ az̆ dog-GEN.1 tail ‘The dog's tail' 4.117 gad-i cụ boy-GEN.1 name ‘Boy's name' 4.118 gad-i kalla 75 boy-GEN.1 head ‘Boy's head' 4.119 dä bayilaǧ hadm-i kul-mä this blind person-GEN.1 hand-INDIC ‘This is the blind person's hand' 4.120 hin-i xürüc ̣ mitšäš attidä-mä he-GEN.1 black beard has-INDIC ‘He has a black beard' Exceptions exist so that the first genitive is sometimes used with inanimate nouns, although it is always used in certain fixed expressions. Currently, there are not enough examples to determine if the use of first genitive with inanimate nouns always denotes whole-part concepts. 4.121 ınk-i ayağ river-GEN.1 mouth ‘The mouth of the river' 4.2.4 Second Genitive The second genitive case is generally used with nouns to denote "inorganic" (alienable) possession. This contrasts with the first genitive case, which is restricted mostly to inalienable possession. Kinship falls into alienable "inorganic" possession in Xinaliq. The underlying form of the second genitive affix is /e/, but it can vary. After low vowels a and ä in the noun stem, the marking is -e. With high vowels as well as with consonants, the marking is /i/. In Example 4.121 we know it is first genitive, because the final vowel in the stem of ınka is -a, therefore /i/ suffix would indicate first genitive. The examples below are of noun classes I, II and III using the second genitive. Example 4.124 76 uses /i/ form for second genitive because the stem vowel is high vowel /ı/. In comparison with first genitive, second genitive has more functions and a higher frequency of usage. 4.122 gad-e cẉ a boy-GEN.2 house ‘The boy's house' 4.123 pxṛ -e yuva dog-GEN.2 kennel ‘The dog's kennel' 4.124 qar-i cẉ a grandma-GEN.2 house ‘The grandma's house' In addition to describing alienable possession, second genitive is also used for describing part of a whole, the material an object is composed of, measurements and ingredients of objects, and, interestingly, for some kinship relationships. In most other languages with an alienable-inalienable distinction, kinship terms are inalienably possessed. Relationship: 4.125 gad-e bıy boy-GEN.2 father ‘The boy's father' 4.126 e csi xị nimḳir attidä-mä my-GEN.2 brother wife have-INDIC ‘My brother has a wife' Part of a whole: 4.127 bädr-e kulp bucket-GEN.2 handle 77 ‘The handle of a bucket' 4.128 ḳıc-̣ e qabuǧ egg-GEN.2 shell ‘Eggshell' (literally ‘the shell of an egg') Material of which an object is composed: 4.129 ur-e ṭoz iron-GEN.2 door ‘An iron door' Measures: 4.130 ps̆-e tikä bread-GEN.2 piece ‘A piece of bread' Ingredients: 4.131 paxl-e sup bean-GEN.2 soup ‘Bean soup' 4.2.5 Dative The case traditionally called "dative" is used with subject, direct object and indirect object. When an act expresses a sensory perception, feeling, or emotional and intellectual activity, the subject takes the dative. These forms are sometimes called the dative-experiencer (or dative subject). Thus in some languages, including Xinaliq, the experiencer (i.e., the noun phrase or NP that is also the subject by the subject criteria of some languages) takes the dative case, not the case normally expected for subjects. The dative is marked by -u. 4.132 mäʕälim-u buto muxiž-mä 78 teacher-DAT everything know-INDIC ‘The teacher knows everything' 4.133 bıy-u häne ši- Ø muxuq̇äṭṭ-i-šä-mä father-DAT his son-NOM recognize-NEG-PST-INDIC ‘The father didn't recognize his son' 4.134 gad-u yolu-Ø zabüǧ-šä-mä boy-DAT flames-NOM notice-PST-INDIC ‘The boy noticed the flames' 4.135 šä büy-u sas-Ø kl-et-mä my father-DAT noise-NOM hear-PRF.I-INDIC ‘My father hears noise' 4.136 as qävil-Ø at-mä I-DAT cold-NOM to.be.PRS.I-INDIC ‘I am cold' 4.137 bua ḥädm-u qäläm-Ø biži z-et-mä this person-DAT pencil-NOM one want-PRF.I-INDIC ‘This person wants a pencil' 4.138 gad-u kičeb liḳuvri yukuar-mä boy-DAT book read love-INDIC ‘A boy loves to read a book' 4.139 asür ümza at-mä I-DAT hungry to.be.PRF.I-INDIC ‘I am hungry' 4.140 misi ḥäyälirz-u pxṛ a-šilli ünqkuar-mä little children-DAT dog-ABL afraid-INDIC ‘Little children are afraid of dogs' In other instances the dative marks indirect objects. 4.141 yä e ḳil-u kağaz-Ø s̆ icịdä-mä 79 I-ERG my friend-DAT letter-NOM write-INDIC ‘I wrote a letter to my friend' 4.142 yä puxṛ -u inḳ läḳ-ir-qo-mä I dog-DAT bone give- IMPRF.ASP-PRS.II-INDIC ‘I give a bone to the dog' 4.143 pšor-u ink täḳä horse- DAT hay give (synonyms) ‘Give hay to the horses!' Dative can also be attached to a recipient of the experience, the person or thing to whom the event is occurring, the beneficiary. Thus: 4.144 hin-i hine dost-u latürku-šä-mä he-ERG his friend- DAT hit-PST-INDIC ‘He hit his friend' 4.145 jir-i inqxẹ rq̇ oy-u ḥyrmäthitar-mä we-ERG elderly-DAT respect-INDIC ‘We respect the elderly' 4.146 bıy s̆-u insaflıdu-mä father son- DAT fair-INDIC ‘The father is fair to his son' 4.2.6 Comitative The comitative case signifies accompaniment, companionship. It signifies being together, accompanying someone. That could be a subject or an object. It is marked by - is̆ḳili if the stem ends with a consonant, and by -s̆ḳili if the stem ends with a vowel. When used with the subject and signifying joint unidirectionality, the meaning is similar to "and," "in the company of," "together with," as in the examples below. 4.147 zı ps̆-is̆ḳili kw-et-mä 80 with horse-COM I.come-PRF.I-INDIC ‘I am coming with a horse' 4.148 gada riš-iškili kaǧ-šä-mä boy-ABS girl-COM came-PST-INDIC ‘The boy c