A HISTORICAL RECONSTRUCTION OF THE KOMAN LANGUAGE FAMILY by MANUEL ALEJANDRO OTERO A DISSERTATION Presented to the Department of Linguistics and the Graduate School of the University of Oregon in partial fulfillment of the requirements for the degree of Doctor of Philosophy September 2019 DISSERTATION APPROVAL PAGE Student: Manuel Alejandro Otero Title: A Historical Reconstruction of the Koman Language Family This dissertation has been accepted and approved in partial fulfillment of the requirements for the Doctor of Philosophy degree in the Department of Linguistics by: Doris L. Payne Chairperson Spike Gildea Core Member Scott DeLancey Core Member Stephen Dueppen Institutional Representative and Janet Woodruff-Borden Vice Provost and Dean of the Graduate School Original approval signatures are on file with the University of Oregon Graduate School. Degree awarded September 2019 ii © 2019 Manuel Alejandro Otero This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs (United States) License. iii DISSERTATION ABSTRACT Manuel Alejandro Otero Doctor of Philosophy Department of Linguistics September 2019 Title: A Historical Reconstruction of the Koman Language Family This dissertation is a historical-comparative reconstruction of the Koman family, a small group of languages spoken in what now constitutes the borderlands of Ethiopia, Sudan and South Sudan. Koman is comprised five living languages: Gwama, Opo, Komo, Uduk, and the previously unidentified Dana language. The Koman family has been relatively understudied though it has figured prominently in large-scale classifications of the Nilo-Saharan super family. These classifications are radically distinct, given the paucity of research on Koman as a whole at the time. Some current scholars even question Koman’s genetic affiliation to Nilo- Saharan entirely. One main issue in high-level classifications is the lack of low-level reconstructions of families established with verifiable sound correspondences coupled with morphological evidence to support the internal structure of a given family. This dissertation addresses this issue by reconstructing the basic phonology, including segmental and suprasegmental domains, and tracing the evolution from Proto-Koman down through the nodes to the modern-day sound systems. In addition, some of the core lexicon and morphology is reconstructed to Proto-Koman and to the subnodes. iv The data for this dissertation was collected in the field from native speakers of all of the living Koman languages including from previously undocumented varieties. In an effort to make the analyses as faithful to the data as possible, all of the data and all of the correspondence sets employed to reconstruct proto-sounds are provided in the Appendices. Further, an etymological wordlist of lexica reconstructed to distinct nodes within the family is also provided. While Koman’s affiliation to the purported Nilo-Saharan super family is still under debate, the overarching aim of this dissertation is to provide a conservative reconstruction of Proto-Koman which will hopefully serve future Koman scholars as well as those interested in higher-level genetic classifications of East African languages. v CURRICULUM VITAE NAME OF AUTHOR: Manuel Alejandro Otero GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene School of Visual Arts, New York DEGREES AWARDED: Doctor of Philosophy, Linguistics, 2019, University of Oregon Bachelor of Fine Arts, 1994, School of Visual Arts, New York AREAS OF SPECIAL INTEREST: Functional Linguistics Descriptive Linguistics Historical-Comparative Linguistics Nilo-Saharan Languages PROFESSIONAL EXPERIENCE: Graduate Teaching Fellow, University of Oregon, Eugene, OR, 2011–2019 Instructor of Spanish, International Community School of Addis Ababa, Addis Ababa, Ethiopia, 2010 ` TEFL and Spanish Instructor, Madrid, Spain, 1997–2005 vi GRANTS, AWARDS, AND HONORS: Doctoral Dissertation Award: Documenting the endangered Koman languages and their linguistic relationships. National Science Foundation, Documenting Endangered Languages Program, BCS 1628750, 2016 Jaqueline Schachter Outstanding Conference Presentation Award, University of Oregon Department of Linguistics, 2016 Global Oregon International Research Fund, University of Oregon, 2016 Graduate Conference Travel Awards, University of Oregon Department of Linguistics, 2013–2017 PUBLICATIONS: Otero, Manuel A. accepted. Associated motion, direction and (exchoative) aspect in Ethiopian Komo. Studies in Language. Otero, Manuel A. 2018. Directional verb morphology in Ethiopian Komo. In Helga Schröeder, & Prisca Jerono (eds.), Nilo-Saharan issues and perspectives, 165-177. Köln: Rüdiger Köppe Verlag. Otero, Manuel A. 2018. Aspects of Ethiopian Komo (morpho-)phonology. Linguistic Discovery 16(2). 136-178. Otero, Manuel A. 2015. [+ATR] dominant vowel harmony except when it’s not? Evidence from Ethiopian Komo. In Ruth Kramer, Elizabeth C. Zsiga, & One Tlale Boyer (eds.), Selected Proceedings of the 44th Annual Conference on African Linguistics, 212-220. Somerville, MA: Cascadilla Proceedings Project. Otero, Manuel A. 2015. Dual number in Ethiopian Komo. In Angelika Mietzner, & Anne Storch (eds.), Nilo-Saharan: Models and descriptions, 123-134. Köln: Rüdiger Köppe Verlag. Otero, Manuel A. 2015. Nominal morphology and ‘topic’ in Ethiopian Komo. In Osamu Hieda (ed.), Information Structure and Nilotic Languages, 19-35. Tokyo: Research Institute for Languages and Cultures of Asia and Africa. vii ACKNOWLEDGMENTS Our families, communities and civilizations are built, nourished and maintained when we all come together. This project has been no exception. I am first and foremost grateful to all of the Koman people who welcomed me into their homes and lives. It has been an unforgettable experience, one which gives me faith in humanity during these disparaging times in which minority communities are singled out and condemned for the excesses of those in power, both at home and abroad. I have worked with so many speakers of Koman languages over the years that I fear there are too many to name individually. I thank them all here. I must also recognize my deep appreciation and mourning for the loss of Somali Pogi, my friend and first Komo teacher who passed unexpectedly at a young age. All of us who have worked on the Komo literacy project will never forget your earnest dedication and warm smile. May you rest in peace. This dissertation could not have been realized without the patience and guidance of my advisor and mentor, Doris Payne. I am truly honored to have been your student and will remember our discussions and trips abroad fondly. Much appreciation must be given to the Linguistics Department of the University of Oregon. It has truly been a life- changing experience. I am ever grateful to Spike Gildea, Scott DeLancey, Melissa Baese- Berk, Tyler Kendall, Charlotte Vaughn and Volya Kapatsinski. I would also like to thank Prof. Stephen Dueppen for serving on my committee. I also could not be writing these words without the support of my colleagues both at the University of Oregon and abroad. I thank Don Killian for so many conversations ans help over so many years. I also thank Paul Olejarczuk for his friendship, viii collaboration, and for all of the epic gentleman’s dinners, which I hope will continue as we part ways from Eugene. I am extremely grateful to Richard Griscom for all of his help- I wish you well! I also thank Jaime Peña, Shahar Shirtz, Amos Teo, Matt Stave, Sara Pacchiarotti, Krishna Boro, Becky Paterson, Justin Walker and so many others that have come and gone. Much love goes out to the Wozny family for their enduring support. I am also grateful to Wendy James, Douglas Johnson and Alfredo González Ruibal. If I have forgotten to mention you, please know that in my true fashion, I have left writing these words until the last minute, my last day in Eugene. I am grateful to SIL Ethiopia for allowing me to work as a consultant on the Komo language literacy project. This was my first foray into the field. I admire the fact that we could set our beliefs aside in order to work for a common goal. I must recognize Andreas Neudorf, Fekadu Derresse, Anne-Christie Hellenthal, Nate Bremer, Joshua and Joelle Goldberg, Joshua Smolders, Andreas Joswig, Joshua Smolders, Colleen and Michael Ahland, among others. Many thanks and blessings to you all. Lastly and most importantly, I dedicate this work to my wife, partner, best friend and hiking partner, mi amor, Becki Quick. I would not be here in this lawn chair on the corner of our house writing the acknowledgments of a dissertation if it weren’t for your enduring love and spirit. I feel blessed to have met you on that bus in Spain so many years ago and since then, my life has only become more enriched when I am with you. We did it! Now onto the next chapter of our lives. Por último, reconozco el apoyo que he recibido por parte de mis padres y de mi familia tras los años. Cariño para los vivos y para los difuntos, que en paz descansen. ix para Becki x TABLE OF CONTENTS Chapter Page I. INTRODUCTION ......................................................................................................................... 1 1.1 The Koman people ................................................................................................................ 4 1.1.1 Issues in nomenclature of Koman groups ........................................................ 10 1.1.2 Population of Koman groups today ................................................................... 18 1.2 Koman classification ........................................................................................................... 21 1.2.1 Koman external classification ............................................................................. 21 1.2.2 Koman internal classification ............................................................................. 23 1.2.3 Current Koman classification ............................................................................. 27 1.3 Methods and the database ................................................................................................. 30 II. TYPOLOGICAL OVERVIEW ................................................................................................... 34 2.1 Phonological comparanda ................................................................................................. 35 2.1.1 Overview of Koman consonants ........................................................................ 36 2.1.2 Overview of Koman vowels and Advanced Tongue Root harmony ......... 37 2.1.3 Overview of Koman tone ..................................................................................... 39 2.1.4 Gwama ..................................................................................................................... 40 2.1.5 Komo phonology ................................................................................................... 47 2.1.6 Uduk phonology .................................................................................................... 56 2.1.7 Opo ............................................................................................................................ 76 xi Chapter Page 2.1.8 Dana .......................................................................................................................... 85 2.2 Morphosyntactic comparanda .......................................................................................... 93 2.2.1 Koman independent pronouns ........................................................................... 97 2.2.2 Koman nominal morphology .............................................................................. 98 2.2.3 Koman verb morphology .................................................................................. 132 III. RECONSTRUCTION OF PROTO-KOMAN PHONOLOGY ........................................... 158 3.1 Proto-Koman (PKMN) tone ............................................................................................. 158 3.1.1 PKMN tone categories based on synchronic tone distribution ................. 160 3.1.2 A historical scenario for the evolution of PKMN tone ................................ 165 3.1.3 PKMN tone and non-stop consonants ............................................................. 168 3.1.4 PKMN tone residue .............................................................................................. 169 3.1.5 Observations on tonogenesis in Pre-Koman ................................................ 173 3.2 Proto-Koman (PKMN) consonants ................................................................................ 174 3.2.1 PKMN bilabial obstruents ................................................................................... 176 3.2.2 PKMN interdental stops ...................................................................................... 183 3.2.3 PKMN alveolar stops and alveolar affricate ejective ................................... 189 3.2.4 Palatal consonants .............................................................................................. 196 3.2.5 PKMN velar stops ................................................................................................. 200 3.2.6 PKMN fricatives .................................................................................................... 205 xii Chapter Page 3.2.7 PKMN *r and *l ..................................................................................................... 210 3.2.8 PKMN nasal consonants ..................................................................................... 211 3.2.9 PKMN glides .......................................................................................................... 213 3.2.10 PKMN consonant residue ................................................................................... 215 3.3 Proto-Koman (PKMN) vowels ........................................................................................ 217 3.3.1 PKMN high vowels .............................................................................................. 219 3.3.2 PKMN [–high] vowels *ɛ, *ɔ and *a ................................................................. 226 3.3.3 Consonant-glide sequences and diphthongs ................................................ 227 IV. PHONOLOGICAL RECONSTRUCTION OF KOMAN SUBNODES ............................ 229 4.1 Gwama ................................................................................................................................ 229 4.1.1 Gwama consonants ............................................................................................ 230 4.1.2 Gwama vowels and tone ................................................................................... 232 4.2 Proto-Central Koman (PCTRL) ....................................................................................... 233 4.2.1 PCTRL consonants ............................................................................................... 234 4.2.2 A note on *T̪ and *D̪ in PCTRL ......................................................................... 237 4.2.3 PCTRL vowels and tone ...................................................................................... 240 4.3 Proto-Komo-Uduk (PKOUD) ........................................................................................... 242 4.3.1 PKOUD consonants ............................................................................................. 242 4.3.2 PKoUd vowels and tone .................................................................................... 245 xiii Chapter Page 4.4 Komo ................................................................................................................................... 245 4.4.1 Komo consonants ............................................................................................... 246 4.4.2 Komo vowels and tone ...................................................................................... 248 4.5 Proto-Uduk (PUD) phonology ........................................................................................ 249 4.5.1 PUD consonants ................................................................................................... 249 4.5.2 PUD vowels and tone ......................................................................................... 251 4.6 Chali Uduk ......................................................................................................................... 252 4.6.1 Chali Uduk consonants ..................................................................................... 252 4.6.2 Chali Uduk vowels and tone ............................................................................ 254 4.7 Yabus Uduk ........................................................................................................................ 254 4.7.1 Yabus Uduk consonants .................................................................................... 255 4.7.2 Yabus Uduk vowels and tone ........................................................................... 257 4.8 Proto-Dana-Opo (PDAOP) phonology ......................................................................... 258 4.8.1 PDAOP consonants .............................................................................................. 258 4.8.2 PDAOP vowels and tone .................................................................................... 261 4.9 Dana .................................................................................................................................... 262 4.9.1 Dana consonants ................................................................................................. 262 4.9.2 Dana vowels and tone ....................................................................................... 264 4.10 Proto-Opo (POP) ................................................................................................................ 265 xiv Chapter Page 4.10.1 POP consonants ................................................................................................... 265 4.10.2 POP vowels and tone .......................................................................................... 268 4.11 Opo ...................................................................................................................................... 269 4.11.1 Opo consonants ................................................................................................... 269 4.11.2 Opo vowels and tone ......................................................................................... 272 V. RECONSTRUCTION OF PROTO-KOMAN PRONOMINAL AND DEICTIC MORPHOLOGY ....................................................................................................................... 273 5.1 Reconstruction of Koman pronominal elements ...................................................... 274 5.1.1 PKMN 1SG reconstructed forms ........................................................................ 276 5.1.2 1PL reconstructed forms .................................................................................... 278 5.1.3 PKMN 2SG reconstructed forms ........................................................................ 281 5.1.4 PKMN 2PL *ʊm(a) ................................................................................................. 283 5.1.5 PKMN 3SG reconstructed forms ........................................................................ 284 5.1.6 PKmn 3PL *hʊn(i) ................................................................................................ 288 5.2 Reconstruction of Koman nominal gender/number morphology ......................... 289 5.2.1 PKMN singular nominal morphology ............................................................. 289 5.2.2 PKMN plural nominal gender morphology .................................................... 291 5.3 Reconstruction of PKMN demonstrative elements .................................................... 293 5.4 Reconstruction of Koman Deictic Directional (DD) verb morphology ................. 294 xv Chapter Page VI. CONCLUSION .......................................................................................................................... 299 APPENDICES A. ABBREVIATIONS ............................................................................................................. 301 B. ETYMOLOGICAL WORDLIST ....................................................................................... 303 C. TONE CORRESPONDENCE SETS ................................................................................ 582 D. PKMN CONSONANT CORRESPONDENCE SETS ................................................... 589 E. PKMN VOWEL CORRESPONDENCE SETS ............................................................... 613 F. COMPLETE DATASET ...................................................................................................... 622 REFERENCES CITED .................................................................................................................... 724 xvi LIST OF FIGURES Figure Page 1. Bender’s (1996) Nilo-Saharan classification ............................................................................. 1 2. Ehret’s (2001) Nilo-Saharan classification ................................................................................ 2 3. Map of the living Koman languages ........................................................................................... 8 4. Map of Koman language areas in James (1975:87) .................................................................. 9 5. Distribution of Uduk communities in James (1968:18) ........................................................ 17 6. Distribution of Mao and Komo ethnic terms in the Mao-Komo Special Woreda ............ 20 7. Greenberg’s (1963) Nilo-Saharan classification ..................................................................... 21 8. Bender’s (1983:286) internal classification of Koman .......................................................... 25 9. Koman internal classification from the Global Lexicostatistical Database ..................... 26 10. Proposed Koman internal classification based on results of this study ......................... 28 11. Koman tree generated with Lingpy software distance measurements .......................... 29 12. Gwama contrastive vowel inventory ..................................................................................... 44 13. Ethiopian Komo vowel system ................................................................................................ 52 14. Uduk contrastive vowel inventory ........................................................................................ 71 15. Opo contrastive vowel inventory ........................................................................................... 83 16. Dana contrastive vowel inventory ......................................................................................... 91 xvii LIST OF TABLES Table Page 1. Cognacy rates from the current database calculated with LingPy ................................... 30 2. Gwama contrastive consonant inventory ............................................................................... 41 3. Distribution and allophones of Gwama obstruents .............................................................. 42 4. Distribution and allophones of Gwama sonorants ............................................................... 43 5. Ethiopian Komo contrastive consonant inventory ............................................................... 49 6. Distribution and allophones of Komo stops and sibilants .................................................. 51 7. Chali Uduk contrastive consonant inventory ........................................................................ 59 8. Killian’s (2015:20) Chali Uduk contrastive consonant inventory ...................................... 60 9. Distribution and allophones of Chali Uduk voiceless plosives .......................................... 62 10. Distribution and allophones of Chali Uduk voiced, ejective and implosive obstruents ............................................................................................................................................................... 62 11. Yabus Uduk contrastive consonant inventory ..................................................................... 65 12. Distribution and allophones of Yabus Uduk voiceless plosives ....................................... 66 13. Distribution and allophones of Yabus Uduk implosives, ejectives and voiced plosives ............................................................................................................................................................... 68 14. Distribution and allophones of Yabus Uduk ........................................................................ 70 15. Opo contrastive consonant inventories ................................................................................ 78 16. Distribution and allophones of Opo plosives, implosives and affricates ....................... 80 17. Dana contrastive consonant inventory ................................................................................. 87 18. Distribution and allophones of Dana plosives ..................................................................... 89 19. Koman independent personal pronominals ......................................................................... 97 20. Lowland Gwama independent and possessive pronouns .............................................. 100 xviii Table Page 21. Lowland Gwama pronominal demonstrative enclitics ................................................... 102 22. Lowland Gwama nominal gender/number proclitics ..................................................... 103 23. Ethiopian Komo independent and possessive pronouns ................................................ 107 24. Komo pronominal demonstrative enclitics ....................................................................... 109 25. Ethiopian Komo nominal gender/number proclitics ....................................................... 111 26. Sudanese Komo nominal gender/number proclitics (Burns 1947:12) .......................... 115 27. Chali Uduk and Yabus Uduk independent and possessive pronouns ......................... 116 28. Chali Uduk pronominal demonstrative bases ................................................................... 117 29. Chali Uduk pronominal demonstrative configurations .................................................. 118 30. Dana independent, possessive and bound pronominals ................................................. 122 31. Dana demonstrative enclitics ................................................................................................ 123 32. Dana nominal gender/number proclitics ........................................................................... 125 33. Independent pronouns in four Opo varieties .................................................................... 127 34. Possessive pronominal enclitics in four Opo varieties ................................................... 128 35. Bilugu Opo demonstrative enclitics .................................................................................... 129 36. Opo third person demonstrative pronouns ....................................................................... 130 37. Bilugu Opo human nominal gender/number proclitics .................................................. 131 38. Lowland Gwama independent and bound pronominals ................................................ 134 39. Ethiopian Komo independent and bound pronominals .................................................. 137 40. Chali Uduk independent pronouns and argument indexing verb morphology ........ 140 41. Yabus Uduk independent pronouns and argument indexing verb morphology ...... 143 42. Dana independent and bound pronominals ...................................................................... 145 xix Table Page 43. Bilugu Opo independent and bound pronominals ........................................................... 147 44. Koman Deictic Directional morphemes ............................................................................. 149 45. Tone patterns in PKMN correspondence sets .................................................................... 161 46. PKMN bilabial stops correspondence sets ........................................................................... 164 47. Evolution of Koman tone in PKMN cognates ..................................................................... 167 48. Proto-Koman (PKMN) consonant inventory ...................................................................... 175 49. Schematic for PKMN correspondence sets ......................................................................... 176 50. PKMN *pʰ correspondence set ................................................................................................ 177 51. PKMN *p correspondence set ................................................................................................. 178 52. PKMN *b correspondence set ................................................................................................. 179 53. PKMN *ɓ correspondence set ................................................................................................. 181 54. PKMN *pʼ correspondence set ................................................................................................ 182 55. PKMN *t ̪h correspondence set ................................................................................................ 184 56. PKMN *t ̪correspondence set .................................................................................................. 185 57. PKMN *d̪ correspondence sets ............................................................................................... 186 58. PKMN *t ̪̓ correspondence sets ............................................................................................... 188 59. PKMN *tʰ word-final correspondence set ............................................................................ 190 60. PKMN *t correspondence set .................................................................................................. 191 61. PKMN *d correspondence sets ............................................................................................... 191 62. PKMN *ɗ correspondence set ................................................................................................. 193 63. PKMN *tʼ correspondence set ................................................................................................. 194 64. PKMN *sʼ correspondence set ................................................................................................. 195 65. PKMN *c correspondence set ................................................................................................. 196 xx Table Page 66. PKMN *ɟ correspondence set .................................................................................................. 197 67. PKMN *cʼ correspondence set ................................................................................................. 200 68. PKMN *kʰ correspondence set ................................................................................................ 201 69. PKMN *k correspondence sets ............................................................................................... 202 70. PKMN *g correspondence sets ............................................................................................... 203 71. PKMN *kʼ correspondence sets .............................................................................................. 205 72. PKMN *s correspondence sets ................................................................................................ 206 73. PKMN *ʃ correspondence set .................................................................................................. 208 74. PKMN *h correspondence set ................................................................................................. 210 75. PKMN *r and *l correspondence sets .................................................................................... 210 76. PKMN *m and *n correspondence sets ................................................................................. 211 77. PKMN *ɲ correspondence set ................................................................................................. 212 78. PKMN *ŋ correspondence set ................................................................................................. 213 79. PKMN *w correspondence set ................................................................................................ 214 80. PKMN *j correspondence set .................................................................................................. 214 81. PKMN *T̪ and D̪ correspondence set .................................................................................... 215 82. PKMN *i and *ɪ correspondence sets .................................................................................... 220 83. PKMN *u correspondence set ................................................................................................. 222 84. PKMN *ʊ correspondence set ................................................................................................. 225 85. PKMN *ɛ, *ɔ and *a correspondence sets ............................................................................. 226 86. Gwama phonological innovations ....................................................................................... 231 87. Proto-Central Koman (PCTRL) consonant inventory ...................................................... 234 88. Proto-Central Koman (PCtrl) phonological innovations ................................................ 235 xxi Table Page 89. PCTRL *T̪ and D̪ correspondence set .................................................................................... 237 90. Tone categories in PCTRL correspondence sets and the distribution with stop onsets ............................................................................................................................................................ 242 91. Proto-Komo-Uduk (PKOUD) consonant inventory ........................................................... 243 92. Proto-Komo-Uduk (PKoUd) phonological innovations .................................................. 243 93. Komo phonological innovations .......................................................................................... 246 94. Proto-Uduk (PUD) consonant inventory ............................................................................ 249 95. Proto-Uduk (PUd) phonological innovations .................................................................... 250 96. Chali Uduk phonological innovations ................................................................................ 253 97. Yabus Uduk phonological innovations ............................................................................... 255 98. Proto-Dana-Opo (PDAOP) consonant inventory .............................................................. 259 99. Proto-Dana-Opo (PDaOp) phonological innovations ..................................................... 259 100. Dana phonological retentions ............................................................................................ 263 101. Proto-Opo (POP) consonant inventory ............................................................................. 265 102. Proto-Opo (POp) phonological innovations .................................................................... 266 103. Opo phonological innovations ........................................................................................... 270 104. PKMN reconstructable independent personal pronouns ............................................... 275 105. Additional reconstructable pronominal elements ......................................................... 276 106. Reflexes of PKMN 1SG *aGa .................................................................................................. 276 107. Reflexes of PKMN 1SG *na ..................................................................................................... 277 108. Reflexes of PCTRL 1SG *má ~ *ám ....................................................................................... 278 109. Koman 1PL independent pronouns .................................................................................... 279 xxii Table Page 110. Reflexes of 1PL in Komo-Uduk (KOUD) ............................................................................. 280 111. Reflexes of 1PL in Gwama, Dana and Opo ....................................................................... 281 112. Reflexes of PCTRL 2SG *aj ..................................................................................................... 281 113. Reflexes of a questionable PKMN 2SG ?*aɪk ...................................................................... 282 114. Reflexes of PKMN 2SG *mini ................................................................................................. 283 115. Reflexes of PKMN 2PL *ʊm(a) ............................................................................................... 284 116. Reflexes of PKMN 3N *han ~ hɪn .......................................................................................... 285 117. Reflexes of PKMN 3SG.F *haɓ ................................................................................................ 286 118. Reflexes of PKMN 3SG.M *haɗ(i) .......................................................................................... 287 119. Reflexes of PKMN 3PL *hʊn .................................................................................................. 289 120. PKMN reconstructed singular nominal morphology ...................................................... 290 121. PCTRL singular nominal morphology? ............................................................................. 291 122. Koman plural nominal (gender) morphology ................................................................. 291 123. PKMN plural nominal (gender) morphology ................................................................... 292 124. Koman proximal demonstrative enclitics ........................................................................ 293 125. Koman medial demonstrative enclitics ............................................................................ 293 126. Koman Deictic Directional 1 (DD1) morphemes ............................................................. 295 127. Koman Deictic Directional DD∅ and DD2 morphemes .................................................. 296 xxiii CHAPTER I INTRODUCTION 1 Introduction This dissertation is a historical-comparative reconstruction of the Koman family, a small group of languages spoken in the borderlands of Ethiopia, Sudan and South Sudan. This language family has been relatively understudied though it has figured prominently in large-scale classifications of the Nilo-Saharan super family (Bender 1971, 1989, 1996; Ehret 2001). One of the motivations for this project was comparing how prior scholars situated Koman within Nilo-Saharan (NS). Consider the following Nilo-Saharan classifications. In Figure 1, Bender’s (1996) classification situates Koman in what he refers to as the “core” of Nilo-Saharan. In Figure 2, Ehret’s (2001) classification situates Koman as one branch in a binary split from Proto-Koman. Figure 1 Bender’s (1996) Nilo-Saharan classification (adapted from C. Ahland 2012:24) 1 Figure 2 Ehret’s (2001) Nilo-Saharan classification (adapted from C. Ahland 2012:25) These classifications and other attempts at constructing a genealogical tree of Nilo- Saharan employed very little data on Koman, given the paucity of research on Koman as a whole at the time. Our understanding of Koman has drastically evolved since the early 2010’s with ongoing research continuing to the time of this writing. One of the issues with prior reconstructions of Koman was the fact that it was assumed that Koman is of Nilo-Saharan stock. This is still under debate and some prominent scholars currently consider Koman a family that is independent from Nilo- Saharan (e.g. Dimmendaal 2011, 2018, to appear) while others question the validity of Nilo-Saharan as a whole (Güldemann 2018). Nevertheless, the status of Koman as a genetic unit has not been disputed, though possible peripheral members, such as Gule and Gumuz have been questioned. The main arguments for classifying Koman as an independent family (i.e. not within Nilo-Saharan) has been the relative lack of data and analyses on individual Koman languages. This is coupled with classifications which a priori assumed Koman to fit 2 within Nilo-Saharan, which often led to cherry-picking supposed cognates from particular Koman languages in order to validate proposed high-level sound correspondences at distinct nodes within the Nilo-Saharan family tree. The field of historical linguistics in Africa in general lacks low-level reconstructions of families established with verifiable sound correspondences as well as morphological evidence to support the internal structure of a given family. For instance, Vossen’s (1982) reconstruction of Eastern Nilotic is one prominent example of a substantive historical work on a tight-knit language family. In light of these issues, this project aims to provide phonological and some lexical and morphological reconstruction of Koman. I employ firsthand data collected in the field from native speakers of all of the living Koman languages including from previously undocumented varieties. I also incorporate and synthesize prior research from Koman scholars. In terms of the reconstruction, I take an ultra-conservative methodological stance and do not assume Koman to be related to any other language or language family. I focus only on “nuclear” Koman; i.e. the living languages and varieties that undisputedly comprise the family. One main aim is to provide a reconstruction of the basic phonology, including segmental and suprasegmental domains, and trace the evolution from Proto-Koman down through the nodes to the modern-day sound systems. I also reconstruct some of the core lexicon and morphology to Proto-Koman and to the subnodes. I recognize that the comparative method employed here can make some controversial assumptions about the nature of family trees (e.g. that clean splits necessarily represent a separation of people groups). The historical evolution of people groups can take place gradually and linguistic innovations can spread across these groups over time. Thus, the family 3 tree of Koman proposed here is a formalization or schematization of the linguistic innovations that defined and separated the Koman people groups. In this dissertation, I strive to make my analyses as faithful to the data as possible and I also try to make the all of my data available to the reader. I provide all of the data and all of the correspondence sets employed to reconstruct proto-sounds in the Appendices. I also provide an etymological wordlist of lexica reconstructed to distinct nodes within the family. Before embarking on the reconstruction, in §1.1 I discuss some of the history of the Koman people and some of the issues in identification and classification of speakers of Koman languages in the literature. In §1.2 I discuss the classification of the Koman family as a linguistic unit. The methods and the database employed for this study are discussed in §1.3. Chapter II provides phonological and morphological sketches of all of the languages and varieties employed in this study. Chapter III is a reconstruction of Proto-Koman segmental and suprasegmental phonology and lexica. Chapter IV reconstructs the phonology and lexica at each subnode down to the modern-day language varieties. Chapter V reconstructs some core-Koman morphology. An etymological wordlist of the reconstructed lexicon is provided in Appendix B and all of the correspondence sets employed in this study are presented in Appendix C-E. Appendix F contains the full dataset employed in this study organized alphabetically by meaning. 1.1 The Koman people This section provides a brief overview of the history of the people groups who speak what are now considered to be the living languages of the Koman family. For 4 clarity, I employ the term “people” for groups who speak the same language and share similar socio-cultural principles and practices (Johnson 2016:20). The term “people” here is akin to an ethnic group, though I recognize that the concept of an ethnic group is controversial and complex, certainly with regards to speakers of Koman languages (Jedrej 2004). I also recognize that employing the idea of a “same language,” which means either a single language or a cluster of mutually intelligible dialects, is generally problematic, given that definitive criteria in establishing the difference between language and dialect beyond judgments of mutual intelligibility is highly complex (Good & Cysouw 2013). The Koman family is comprised five living languages: Gwama (ISO 693-3 code kmq), Opo (ISO 693-3 code lgn), Komo (ISO 693-3 code xom), Uduk (ISO 693-3 code udu), and Dana (no existing ISO code). The extinct language Gule (ISO 693-3 code gly), purported to also pertain to Koman in Greenberg’s (1963) classification, does not form part of this study (see §1.2.2). All of the living Koman languages have at least two varieties with the exception of Dana, a recently identified Koman language (Otero 2016). Aside from being virtually undocumented, Dana has not been recognized in the literature as a distinct language, but rather peripherally as a distant dialect within the Opo cluster (see §1.1.1.2). Lemi (2010) cites Dana as one of seven mutually intelligible dialects of Opo, five of which are spoken in South Sudan. Nonetheless, he provides no data on the dialects. Mellese (2017:3) states that Dana is the most distinct variety of Opo but recognizes that the degrees of mutual intelligibility among the Opo dialects remains to be investigated. My preliminary research suggests Dana is a first split from Proto-Dana- Opo (cf. §1.2.3). 5 Koman languages have been spoken for centuries if not millennia along what now constitutes the borderland of Ethiopia, Sudan and South Sudan (Fernández 2004). This particular borderland is demarcated by an escarpment of rock wall which falls from the Western Ethiopian highlands to the verdant lowlands of Ethiopia and further extends to the (South) Sudanese Nile valley. The forests of the Ethiopian and Sudanese lowlands are studded with mountain ranges, hills and dramatic inselbergs. Recent archaeological studies have discovered that the rock inselbergs were used as shelters as far back as the Middle Stone Age (González-Ruibal 2014, Fernández et al. 2007) and González-Ruibal (p.c.) has uncovered archeological remains of Koman communities that inhabited the area around Bambassi (Ethiopia) at least 2,000 years ago. This borderland has served as a refuge for Koman-speaking peoples and other minority groups fleeing oppression and enslavement from both sides of the border (James 1968, 1979, 1980; Triulzi 1982 inter alia). The history of Koman peoples, at least that of the past two centuries, has been fraught with struggle and strife. Koman speakers have been enslaved by Arab and Abyssinian forces and raided by the Nuer. Over centuries, this has led to the fragmenting of Koman communities as groups scattered into the mountains and valleys of the borderlands in a constant struggle for self-preservation. In describing the history of minority groups such as Koman in the Benishangul-Gumuz Regional State of western Ethiopia, the ethno-archaeologists González-Ruibal and Fernández-Martínez (2007:14–15) conclude, “[…] at the end of the day, it is all a matter of power. The traditional ethnic hierarchy of the region could be described as follows: historically the most oppressed and disempowered groups are those labelled Mao and Komo. Some groups related to this ethnic cluster were probably wiped out by slave raids in the 6 late 19th century and nowadays only their ethnic names and some dispersed and confusing data are preserved in the literature (Negaso Gidada 2001:61-91). The Mao and Komo were continually expelled from their lands and today they inhabit the most inhospitable and marginal frontier areas of Benishangul.” Koman’s history of being dominated and subjugated unfortunately continues in some part to the present. In Western Ethiopia, land grabbing is rampant, and many groups are being pushed out of ancestral territories (Meckelburg 2014). Ongoing war and strife for the last decade in South Sudan has led to forced migration of minority groups to refugee camps (UNHCR 2016). Many of the Koman groups in Sudan and South Sudan are currently found either in the midst of war or fleeing to refugee camps in Ethiopia, Kenya or Uganda. There are also small populations of Uduk refugees living in the Unites States since fleeing war in Sudan the mid 1990’s. A map of the areas in which Koman languages are spoken is in Figure 3. Note that “Tw’ampa” refers to the Uduk varieties, “Kwama” to the Gwama varieties and “Opoo” to the Opo varieties and what I suspect is also Dana. The Uduk varieties are spoken in the Blue Nile State of Sudan. The southernmost patch of Uduk in Figure 3 represents the Yabus Uduk variety, where they live alongside the Komo, and the Western Nilotic languages Jum Jum and Mabaan, among others. The river traversing from Sudan into Ethiopia in this area is the Yabus River. This map predates the creation of South Sudan in 2011. The border of South Sudan and Ethiopia is roughly at the latitude of Begi. 7 Figure 3 Map of the living Koman languages (from Killian 2015:6) The map in Figure 3, which Killian attributes to an unpublished manuscript by Roger Blench, appears to have been adapted from the map in James (1975:80) seen below in Figure 4. Note that in Figure 4, “Shita” refers to the Opo varieties and Dana. I suspect that the “Shita” group identified near the Daga River are speakers of Dana and possibly other Opo varieties (see §1.1.1.2). 8 Figure 4 Map of Koman language areas in James (1975:87) Multilingualism among speakers of Koman languages has been the norm historically and this continues to the present. James (1975:84-85) observes that Koman people have had long links with the Oromo and Amhara of Ethiopia as well as with the Arabs and Nilotes of Sudan. James notes that Koman people groups have retained their languages and their way of life despite their violent history of being the victims of the slave trade from both sides of the border. Nonetheless, she observes that Arabic and/or Oromo is spoken across the Koman hamlets (predominantly by men). James also 9 observes that Koman speakers further south are also fluent in Anuak (also known as Anywa).1 The following subsections discuss some issues in identifying Koman people groups and languages in the literature. 1.1.1 Issues in nomenclature of Koman groups The endonyms and exonyms for Koman peoples and the names for languages they speak also have a complicated history (e.g. Bryan 1945, James 1981). The literature on Koman, including historical, anthropological and linguistic is fraught with countless terms for individual communities as well as blanket terms for one or more language groups. This has led to an unfortunate history of mis-identifying Koman languages and Koman dialect groups. Some of the earliest accounts of speakers of Koman languages date back to the Dutch explorer Juan María Schuver’s travels in 1881-1882 along what then constituted the Ethiopia-Sudan borderlands (James et al. 1996). From his extensive travels, Schuver encountered many people groups with differing ethnonyms for speakers of Koman languages.2 At the end of the 19th century, according to Schuver, Koma or Goma was a blanket term for speakers of what are modernly Komo, Gwama, Dana and Opo. James et al. (1996:xc-xcii) provide an insightful summary of the wordlists that Schuver collected during that period mapped onto the terms they identified for Koman languages in 1996. Since 1996, some of the terms employed for Koman-speaking people 1 The Komo speakers residing in the Pokung village of the Gambela region often intermarry with the Anuak. I encountered many Komo who spoke Anuak there but Anuak people that spoke Komo were not as common, if practically non-existent. 2 James et al.’s (1996:xl) sketch map of Schuver’s travels indicates a round trip from Famaka, Sudan (approximately 11°18'00"N, 34°44'00"E) to as far south as Gobo, Ethiopia (approximately 8°56'11.27"N, 34°31'59.68"E). 10 groups have been refined. The following sections trace the histories of the nomenclature used for Koman peoples and Koman languages in the literature and update them with the current knowledge of Koman languages and language groups. 1.1.1.1 Gwama and Komo terms in the literature Some terms seen for modern-day speakers of Gwama in the literature are Gwama, Kwama, North Koma, South Koma, and Gogwama, among others.3 James (1980:67) states that she and M. Lionel Bender decided to employ the term Kwama for the modern-day Gwama speakers. Bender (1971:222) classifies modern-day Gwama into North Koma and South Koma, though Bender (1975:67) notes that this division may be unjustified. Note that “Koma” is employed by native speakers in the Sudanese side of the border while “Komo” is used in the Ethiopia. Currently, Gwama is classified into mutually intelligible Highland and Lowland Gwama varieties, which display some lexical and grammatical variation (Zelealem 2005, Joelle Goldberg 2015, Küspert 2015). The Highland Gwama speakers inhabit the area around Tongo, in the plateau above the escarpment, and speakers of the Lowland variety reside further west in the lowlands. Highland Gwama speakers can identify as Mao or Begi Mao, as they live in proximity with Omotic Mao speakers (M. Ahland 2012) or Sit Shwala ‘black people’. The Lowland speakers by contrast, identify either as Gwama or Komo though they recognize their language as entirely distinct from Komo (Küspert 2015, Meckelberg 2016). Incidentally, Schuver’s wordlist of Koma/Goma 3 Gogwama is most likely from Komo gʊ̀=gwàmá PL=gwama ‘Gwama people’. 11 collected in the village of Bosho in Ethiopia, at the end of the 19th century reproduced in James et al. (1996:329), corresponds to the Gwama data I have collected.4 Terms in the literature used to describe what constitute modern-day speakers of the Komo language are: Komo, Koma, Goma, Kwama, Gokwom, Kwom, Komo of Daga, Hayahaya, and Burun.5 The two terms that are widely used today are Koma, which is used in South Sudan and Sudan, while Komo is used in Ethiopia. In Ethiopia, Komo can be used to identify both the people and the language spoken by the Komo. But politically, Komo is employed for both Komo and Gwama speakers (Küspert 2015, Meckelburg 2016). Intermarriage between Komo and Gwama speakers is very common in Ethiopia though it is unclear whether the intermarriage is with both Highland and Lowland varieties of Gwama. There is evidence that at least in Ethiopia, there has been sustained contact between Komo and Gwama since that last century if not longer (Meckelberg 2016). Bender (1975) provides some insight on the history of the nomenclature used for modern-day Komo. Bender recognizes the problems of nomenclature and adopts the name Komo from Wendy James. He also recognizes that Komo and Koma are used by outsiders, especially the Oromo, for Komo as well as for speakers of other Koman languages. Bender (1975:63) observes that the term Central Koma employed in Bender (1971) is the same language described by Burns (1947), namely modern Komo. Further, Bender (1975) states that the term Komo of Daga is “inaccurate,” though he does not 4 James et al. (1996:xci) note that Bosho may have been a Komo-speaking village as it is was a Komo clan name at the time of their publication. Bosho also appears as a Komo clan name in my data. 5 Corfield (1938:128) notes that in Sudan (from the Dota to the Jokau rivers), the term Koma was being employed for at least two language groups at the time of his writing. 12 expand upon what he means by this.6 Bender (1975) observes that the Shita (i.e., Kigille Opo and possibly other Opo varieties) refer to the Komo as mɛdin. I assume this is Corfield’s (1938) Madiin, which data corresponds to my Komo data. Further, in my data, the Dana ethnonym for the Komo people is màdì̪n, which has no discernable etymology in Dana (or Komo) to my knowledge. 1.1.1.2 Dana and Opo terms in the literature James et al. (1996:334) observes, “while most Koman groups are ‘hill’ people, the Shyita are associated with the lowlands and with cattle pastoralism.”7 As noted earlier, Shyita is one of the blanket terms used to identify Opo (and most likely Dana) groups in the literature. In the literature, the terms Shita, Shyita, Ciita, Ansita, Po, Tʼapo, among others, are employed for what is now known to be Opo and/or Dana. Note that the Kigile Opo word for ‘person’ is ʃìtà, which undoubtedly corresponds to the terms Shita, Shyita, Ciita, Ansita found in the historical and anthropological literature. In Corfield’s (1938) exploration of what he referred to as “Koma” country in what now constitutes the Sudan/South Sudan border with Ethiopia, he comes across several Koma ethnic groups, which to him do not appear to speak the same language. Corfield collected wordlists, conducted interviews, and identified two Koma groups: the Madin 6 In my estimation, Komo of Daga may have either been a name for modern-day Dana or for the Komo speakers who inhabited Daga. If it is the latter, then the expression Komo of Daga can be taken as evidence that Komo speakers did in fact live in Daga and, as a result, were most likely in contact with Dana speakers. 7 An interesting point here is that an Opo ethnonym for the Komo, who have historically resided in mountainous territory, are compounds meaning ‘those who fly up mountains’. To illustrate, the Komo ethnonym in Pame Opo is bɪ̀=pāj zàw PL=fly rock – lit. ‘Those who fly (up and over) mountains’. Note the Opo use the same term for ‘rock/stone’ and ‘mountain’. 13 and the Ciita. The graphemes and employed by Corfield (1938) for linguistic data represents a voiceless (alveo)palatal fricative [ʃ].8 Corfield further subdivides Ciita into three dialect groups based on the proximity of the speech varieties: Kusgilo, Kigelle and the Buldiit. For Corfield, Kusgilo and Kigelle are closely related given the similarity he observed in speech. Further, the Kusgilo and Kigelle lived in proximity to one another around Daga Post (modern day Dajo, South Sudan). Corfield distinguishes Buldiit as a distinct dialect group, who lived further west from the Kigile and Kusgilo group citing greater variation (Corfield 1938:130). The Dana lexical data collected for this study correspond to Corfield’s Buldiit variant of Ciita. Anecdotally, Corfield (1938:129) also observes that, “the people of Buldiit replied that they were Dana, but this turned out to be the name of the area in which they lived.” In the late 1800’s, the Dutch explorer Schuver collected a wordlist of a language spoken by his slave which Schuver calls Gambiel or Kilai (James et al. 1996). Kilai refers to the area in which his slave lived prior to being bought. In 1996, Wendy James identifies Schuver’s Gambiel/Kilai language as Shiyta (James et. al 1996:xci).9 According to James, the term Gambiel is named after the Oromo word for Gambela, a regional state of southwest Ethiopia (ibid 1996:334). Nevertheless, Schuver’s wordlist of Gambiel/Kilai corresponds to my Dana data. 8 Corfield employs the English digraph to explain the phonetic representation of his use of and in his wordlists. 9 Whether Schuver’s Kilai is at all related to kɪ́ná(j), a term used by Komo, Gwama, Yabus Uduk and Dana speakers today as an ethnic term for Opo speakers, is uncertain at this point. 14 I find it noteworthy that both Schuver and Corfield found Dana to be a distinct speech community. Even though they were not trained linguists, their intuitions appear to be in the right direction. My Opo consultants, who are native Pame Opo speakers bilingual in Bilugu Opo, recognize that for them at least, Dana is not mutually intelligible with the Opo varieties they speak. Komo and Gwama speakers I interviewed recognize the Dana ethnicity and language as distinct from the Opo varieties. Komo and Gwama consultants distinguish the kɪ́ná ‘Opo’ from the dāná ‘Dana’. Further, the Yabus Uduk also distinguish between the cʰínaj ‘Opo’ and púr ‘Dana’. Returning to Schuver, he observes that “the self-name of speakers is sometimes distinguished as Ciita or Ansita; the people are termed Kina or Kena by Komo speakers, Pur by Uduk speakers, Kogo or Pau by the Oromo, some as Lango by the Anuak, others as Cai Buldiit by the Nuer (James et al. 1996:334)”. There is little doubt that Dana is closely related to the varieties of the modern-day Opo cluster. Many current linguists who have worked on Opo varieties have considered Dana as a variety of Opo, though few have actually collected Dana data. Lemi (2010), Mellese (2017) and Smolders (forthcoming) suggest, albeit loosely, that Dana and the Opo varieties form a dialect continuum. While this may turn out to be true, at least historically, it is yet to be definitively established. In discussing Dana’s history of nomenclature, we inevitably delved into Opo’s history but I now discuss Opo more specifically. In the earlier literature, Bender (1975) employed the term Shita and later adopted Opo (1983-2000) for what constitutes modern-day Opo. The most recent work on Opo, Smolders (forthcoming), identifies seven dialects, within which he includes Dana. The Bilugu and Modin varieties are 15 spoken in the Gambella Regional State, while Pame, Modin, Kigile, Pilakoy and Bikol are spoken in the Upper Nile State of South Sudan. By examining the autonyms employed by the distinct varieties, we see where some of the terms in the linguistic (and historical) literature came from. The Bilugu and Modin varieties employ the term pɔ̀ ‘people’ as an autonym, the Kigile and Pilakoy use ʃ̀ità ‘people’ and Pame employs zìtà. Thus, the terms Shita, Shyita, Ciita, Ansita likely came from researchers working with South Sudanese varieties, while Opo, Upo, Opoo, Opuoo derive from the Ethiopian varieties. In keeping with the linguistic literature, I follow Bender (1983) and Smolders and employ Opo as a general term for the Opo varieties though, for now, I treat Dana as a distinct language until proven otherwise. 1.1.1.3 Uduk terms in the literature The Uduk varieties have the least terminological confusion of any of the Koman groups. Further, Uduk is by far the most studied people and language of the Koman family. Wendy James has produced detailed anthropological and historical accounts of the Uduk in an extensive collection of work dating from the 1970’s until the time of this writing. Early wordlists date back to Evans Pritchard (1932) in his exploration of the inhabitants of the Blue and White Nile areas in 1926. All of the Uduk varieties are spoken in the Blue Nile State in Sudan. James (1979), Killian (2015) and my consultants identify three Uduk varieties: Chali Uduk, Bonya (or Bellila) Uduk, and Yabus Uduk. According to my consultants, the Chali and Bonya varieties are very close and easily mutually intelligible, while the Yabus variety is more distant (see also James 1979 and Killian 2015). The geographical distribution of the Uduk varieties in 1968, which holds to the present day, is seen in Figure 5. Note that 16 Central Uduk corresponds to Chali Uduk, which is in proximity to Bellila Uduk (or Bonya Uduk). Belatoma Uduk corresponds to modern-day Yabus Uduk, spoken along the Yabus river in Sudan. Figure 5 Distribution of Uduk communities in James (1968:18) 17 The Uduk cluster is the only Koman language that is not spoken in Ethiopia excluding the Uduk in Ethiopian refugee camps or other emigrant locales, particularly in the U.S. This terminological confusion for the majority of the Koman languages continues well into the 2010’s as noted by Küspert (2015), who attempts to unravel the Komo, Gwama, Kwama, Mao terms employed for several Koman and Omotic groups of western Ethiopia. While Küspert makes significant progress in identifying the language and people group names internally and externally, he observes that even in 2015, the picture is not as clear as one would like it to be. We turn now to a discussion of how issues such as the identification of people groups has hindered determining the population of Koman communities. 1.1.2 Population of Koman groups today Determining the population of Koman speakers has been historically challenging for a number of reasons including the nomenclature issue addressed in the preceding section. These continue to be challenges to the present day. A major issue has been the result of mis-identifying ethnic groups, lumping groups together, and/or not identifying a group altogether (James 1981). Moreover, these difficulties are coupled with the fact that many Koman communities inhabit rural areas, communicable only by footpaths, and often intermarry with other Koman groups as well as larger groups including the Oromo and Anuak. To make matters more complicated, Sudanese and South Sudanese population estimates of Koman groups other than the Uduk are not presently available. A widely used figure employed in recent editions of the Ethnologue (e.g. Lewis, Simons & Fennig 2015; Simons & Fennig 2018) is a population estimate of 60,000- 18 100,000 Koman speakers. This figure, which was originally taken from James (1975), includes the Gumuz whose population vastly outnumbers that of any Koman group (C. Ahland 2012) and whose linguistic genetic affiliation to Koman is still under debate (see §1.2).10,11 Bender’s (1997:211) population estimates of Koman groups taken from the 1988 version of the Ethnologue (Grimes 1988) and Bender’s own data seem closer to the current situation with respect to the Uduk (11,000), Opo (2,000-5,000) and Gwama (15,000).12 Bender (1997) estimates 4,500 Komo, which I assume to include the Ethiopian and Sudanese populations. Hudson’s (2004) survey of the 1994 Ethiopian census lists roughly 1,500 ethnic Komo and 140 ethnic Gwama. Another major issue that impacts determining the populations of Komo and Gwama speakers in Ethiopia is ethnic self-identification within the larger cultural context. In the Lowlands of Ethiopia, Komo and Lowland Gwama speakers self-identify as ethnically Komo, while speakers of Highland Gwama (tʼwa sit ʃwala) who inhabit the highlands often self-identify either via their clan name Kuro or Kirin, or as ethnically Mao (Küspert 2015:8). In the highlands of Western Ethiopia, Highland Gwama speakers are in contact with speakers of Omotic languages such as Northern Mao, Seze and Hozo (M. Ahland 2012, Küspert 2015). The map in Figure 6 provides a rough overview of the distribution of the “official” ethnic terms employed by the Ethiopian government in 10 The 2007 Ethiopian census cites 163,781 Ethnic Gumuz (FDCREPC 2008:44), which is higher than all of the living Koman groups combined. 11 If Koman and Gumuz are indeed genetically related, the split goes back to Pre-Proto-Koman (e.g. Bender 1983, 1997, 2008; Ehret 2001). This study of Koman does not include Gumuz, and any of my references to Koman languages do not include Gumuz unless stated explicitly. 12 Bender (1996) employs the term Kwama for modern-day Gwama. Note that in Bender (1996), Dana was either subsumed within Opo or not counted at all. 19 Küspert’s sociolinguistic survey of the Mao-Komo Special Woreda (district) of Western Ethiopia. Figure 6 Distribution of Mao and Komo ethnic terms in the Mao-Komo Special Woreda of Western Ethiopia (from Küspert 2015:9) This confusion in identifying Komo versus Gwama people is evident in the census data from Ethiopia. The 2007 Ethiopian census lists 7,773 ethnic Komo and 73 ethnic Qewama (sic), the latter of which I assume to be Gwama (FDCREPC 2008:44–45). Meckelburg (2016:190–191) compares the Ethiopian census data from 1994 and 2007 and notes that Gwama (listed as Kewama and Qewama in the two different census, respectively) is virtually absent in the census when, in fact, the Gwama significantly outnumber the Komo in Ethiopia.13 13 Meckelburg (2016:63) estimates the Gwama population at about 20,000 in Ethiopia. Note that Meckelburg identifies and groups together Lowland and Highland varieties of Gwama in his estimate. 20 In my estimate, the Koman languages vary significantly in terms of population. Uduk is the largest (approx. 20-25,000; Killian 2015), followed by Gwama (approx. 15,000; Siebert & Bryant 2007), Komo (approx. 2,000 in Ethiopia; Hudson 2004; Meckelberg 2016; and 2,000–5,000 in Sudan; Davies 1960, Theis 1995). Lemi (2010) cites 980 Opo in Ethiopia though Smolders (p.c) estimates around 2,000 total. The population of Dana speakers is unknown at present. 1.2 Koman classification This section discusses issues of the classification of Koman family’s internal structure as well as its purported affiliation to larger genetic groups. 1.2.1 Koman external classification The Koman family figured as one of six main branches in Greenberg’s (1963) initial proposal of Nilo-Saharan as an African super-family, presented in Figure 7. Greenberg based his classification on 48 pieces of grammatical evidence, plus mass comparison of the lexicons. Since his initial classification, the genetic affiliation and internal structure of the Koman family has been under scrutiny (e.g. Bender 1983, 1997; Ehret 2001). Figure 7 Greenberg’s (1963) Nilo-Saharan classification (adapted from C. Ahland 2012:22) 21 Bender’s (1971) external Koman classification within Nilo-Saharan (NS) is in line with Greenberg (1963), though Bender’s internal structure of Koman is distinct from Greenberg’s. Within his “Koman” family, Bender includes what he calls the “Koma languages”– Gumuz, Northern Mao and the Nara-Surma languages – though he acknowledges that “whether these are really coordinate branches of one grouping or whether they are related in some different way remains to be seen when more data is available” (Bender 1971:193). Bender’s “Koma” languages are split into two subgroups: Nuclear Koma contains Langa (a modern-day variety of Opo, most likely either Modin or Bilugu), Central Koma (modern-day Sudanese Komo), and Uduk.14 Bender’s Peripheral Koma subgroup contains North Koma and South Koma (what are most likely the modern-day Highland and Lowland Gwama varieties). Bender (1983) reconstructs what he calls “proto-orthodox-Koman” based on what he deems to be the five central languages of the family: Gwama, Komo, Uduk, Opo and Gule.15 Bender (1991, 1994) groups his orthodox-Koman and Gumuz as a single genetic unit (“Komuz”) within the “core” of Nilo-Saharan (cf. Figure 1). However, Bender (1996:63) recants his earlier proposal of a “Komuz” family stating, “current work shows them to be quite independent grammatically and lexically (except for lexical borrowings) and not standing in any special genetic relationship.” Koman maintains its position within the core of NS in Bender (1996, 2000). 14 I assume Bender’s Komo is the Sudanese variety. One indicator is that Bender’s (1971:273) Komo word for ‘road, path’ is kʰwa, which is identical to Burns’ (1947:6) word for ‘road, path’ in Sudanese Komo. The Ethiopian Komo consultants employed in this study also recognize the distinct pronunciations of ‘road, path’ in that the Sudanese Komo use kwà whereas the Ethiopian Komo use kʊ̀mà. 15 Bender used the term “Anej” for what is currently referred to as Gule. 22 Ehret (2001), by contrast with Bender, reconstructs Koman as one branch of an initial binary split from Proto-Nilo-Saharan (PNS), with a possible time-depth of around 10,000 BP. Ehret claims that a large piece of evidence for this split is the absence of Greenberg’s (1981) “moveable -k” in Koman, which is found in the rest of Nilo-Saharan. Another piece of evidence for Ehret was the multiplicity of consonants which he analyzed as reconstructing to PNS, for which he drew largely from Chali Uduk, which exhibits one of the most extensive consonant inventories in Koman. Ehret (2001) includes both Gule and Gumuz in higher level nodes within a single genetic unit he refers to as “Koman” (cf. Figure 2). The classification of Koman (in its many senses) within a larger genetic unit such as Nilo-Saharan is far from settled. Given the paucity of data, Dimmendaal (2014, 2018, to appear) considers Koman to be an independent family until further research comes to light. The following subsection further discusses the internal structure of Koman and discusses Gule and Gumuz. 1.2.2 Koman internal classification All of the relevant literature at least recognizes a genetic grouping of the languages employed in the reconstruction in this study: Gwama, Komo, Uduk and Opo. The Dana language, which has figured in wordlists collected by explorers, has most likely not formed part of any substantive linguistic work on Koman classification seen thus far. The two remaining languages and/or cluster of languages included in previous reconstructions, Gule and Gumuz, are briefly discussed below. I then turn to Koman- internal (i.e. sans Gule and Gumuz) classification. 23 Gule, an extinct language also known in the literature as Anej and Hamej, was spoken in Jebel Gule in the Blue Nile State of Sudan (James 1979:13). There is very little data collected on this language aside from wordlists in Marno (1874), Seligmann (1912), and Evans-Pritchard (1932). One main piece of evidence supporting an affiliation of Gule to Koman is a sound correspondence /ʃ/ in Koman to /f/ in Gule, though this occurs only in a handful of words. Seligmann (1912) gathered a bit of grammatical data which seems to indicate a correspondence in grammatical gender. Güldemann (2018:292) observes that while this is promising evidence, the scant data makes Gule almost impossible to classify. The issue of whether Gumuz, a dialect cluster newly referred to as Baga (C. Ahland p.c., Güldemann 2018), forms a genetic unit with Koman is yet to be resolved.16 M. Lionel Bender changed opinions on the status of “Komuz” (Koman + Gumuz) over his career. In his most recent work, Bender (2000) classifies them as distinct genetic units within Nilo-Saharan. Ehret (2001) includes Gumuz and Koman as a genetic unit, though Dimmendaal (2014) does not. Ahland (2010, 2013) attempts to link Koman and Gumuz via one sound correspondence in a handful of words between Gumuz and only Gwama. Güldemann (2018:298) observes that while this evidence may be promising, it does not justify positing a genetic link between Gumuz and Koman and he calls for a full-scale reconstruction of both proto-languages. I remain agnostic on the position of a Koman- Baga genetic link in this dissertation. While proposals about Koman’s affiliation to Nilo-Saharan has varied over the history of research, proposals about the internal structure of what I refer to as Koman 16 See Güldemann (2018:293–294) for discussion of the term Baga for what was formerly called Gumuz. 24 in the narrow sense (i.e. the living Koman languages employed in this study: Gwama, Komo, Uduk, Opo and Dana) has remained unchanged.17 The classifications in Bender (1971, 1983, 1991, 1996, 2000) and Ehret (2001) are identical with respect to the internal structure of Koman (without Dana). They concur that the first branch from Proto- Koman is Gwama followed by a split into a Komo-Uduk branch and an Opo branch. To illustrate, Bender’s (1983) internal classification of Koman, seen in Figure 8, is identical to Ehret’s (2001) in Figure 2.18 Proto-Koman Uduk Komo Opo Gwama Gule Figure 8 Bender’s (1983:286) internal classification of Koman More recent lexicostatistical methods produce similar results regarding internal Koman subgrouping. Starostin’s (2011-2016) online Global Lexicostatistical Database (OLD) is a compendium of annotated Swadesh wordlists of the world’s languages. Users can consult the website to produce trees generated from Starostin’s lexicostatistical analyses of the wordlists. An example of the Koman genetic tree based on 110 lexical 17 Dana has not been identified as a distinct language in any prior reconstructions of Koman. I don’t believe Bender collected any Dana data given the fact that the Dana consonant inventory is markedly distinct from the Opo varieties I have encountered, and Bender’s work does not reflect such complexity. 18 I have adjusted Bender’s language names in Figure 8 to reflect the current nomenclature. 25 items generated by the Global Lexicostatistical Database is in Figure 9.19 Note that Gule is included in this version of Koman. Also note that in Figure 9, Kwama refers to Lowland Gwama while Begi Mao is Highland Gwama. The numbers along the top of the figure represent glottochronological dating in millennia (0.00 = beginning of the 1st century A.D., 1.00 = 1000 A.D., and so on). Figure 9 Koman internal classification from the Global Lexicostatistical Database (Starostin 2011-2016) The cognacy percentages (i.e. percentages of pairwise retentions of cognates) employed to create the Starostin subgrouping is seen below the tree in Figure 9. Note that Gwama (Lowland Gwama) and Begi Mao (Highland Gwama) share 92% of the lexicon, while Gwama and Opo share only 22%. The main point here is not a discussion of lexicostatistical methods but rather that, over the body of research, proposals about the internal structure of Koman have remained fixed. One interesting point in Starostin’s (2011-2016) Koman tree is that it was calculated with assumptions about a 19 Starostin (2011-2016) employs the following sources for his classification: Gwama (Zelealem 2005), Opo (van Siflhout 2013), Komo (Krell 2011, Otero 2014, Otero et al. 2015), Uduk (Beam & Cridland 1970, Thelwall 1983, Killian 2015), Gule (Lejean 1865, Seligmann 1912) and (Bender 1983, Corfield 1938, Wedekind & Wedekind 2002) for general Koman. 26 conservative rate of lexical replacement: 5% loss every millennium (i.e. 5 items from a 100-word Swadesh list replaced every 1000 years). Calculation of lexical replacements is beyond the scope of this dissertation, but it is noteworthy to point out that according to these assumptions about lexical replacement, Starostin’s tree suggests Koman (or Proto- Koman) was spoken around 6,000 years ago, Proto-Gwama split roughly 5,000 years ago, followed by the binary split of what I refer to as Central Koman into the two remaining branches Komo-Uduk and Opo, roughly 4,500 before the time of this writing. The following subsection discusses the Koman internal structure that I propose in this study. I discuss how the new Koman data support previous internal classifications of Koman and I provide a more nuanced subgrouping with respect to the inclusion of Dana as well as individual language varieties. 1.2.3 Current Koman classification Determining the status of Koman with respect to larger genetic units is not the aim of this dissertation. In fact, in this study, I assume that, for the time being, Koman consists only of Gwama, Komo, Uduk, Dana and Opo and their varieties. Henceforth, I employ the term Koman to strictly mean the living Koman languages employed in this study. To help the reader throughout the rest of the dissertation, Figure 10 presages the results (conclusions – not presuppositions) of the study. It contains the Koman subgrouping of all of the languages and dialects argued for in this study. The internal structure I propose based on the reconstruction here turns out to be identical to prior classifications with the exception of Dana, which I propose was a split from Proto-Opo. The branch lengths in this tree are not significant nor are there any assumptions about time depth in this representation. In Figure 10, “P” represents “proto”. 27 P-Koman P-Central P-Komo-Uduk P-Dana-Opo P-Uduk P-Opo P-Gwama Chali Yabus Komo Bilugu Modin Pame Kigile Dana Highland Lowland Figure 10 Proposed Koman internal classification based on results of this study For ease, I employ the term “Proto-Central Koman” to describe the node which from which Proto-Komo Uduk and Proto-Dana-Opo diverged. I also recognize that the history of the Opo varieties may not be as distinct as represented in this tree. The Opo varieties may have evolved in a dialect continuum, which a tree model may not necessarily capture. That said, this subgrouping of the Opo dialects reflects shared similarities as well as shared innovations. Figure 11 contains a Koman tree generated with LingPy, a Python programming language suite designed for historical linguistics (List et al. 2018). This tree was constructed based on my current database and employing my cognacy judgments. LingPy employs similar lexicostatistical methods as those of the Global Lexicostatistical Database discussed in §1.2.2. Note that the internal structure of Koman generated from LingPy using my data is identical to prior proposed Koman subgroupings with respect to the branching of the major languages/language clusters (i.e. Gwama, Komo and Uduk). 28 Figure 11 Koman tree generated with Lingpy software distance measurements (List et al. 2018) In Figure 11, the edges (or nodes) are calculated with respect to cognate retention (i.e. maximum similarity) to create distances between pairs. From here, we can interpret that the Bilugu and Modin Opo varieties are the closest (edge.0) followed by the Kigile and Pame Opo varieties (edge.1). Taken together, the Opo varieties form a subgroup (edge.2), as does the Dana-Opo branch (edge.3). Next, we see the Uduk varieties forming a subgroup (edge.4) followed by a Komo-Uduk subgroup (edge.5). The Dana-Opo and Komo-Uduk branches form a subgroup, what I call Central Koman (edge.3). Lastly, the most distant subgroup consists of the Gwama varieties (edge.7). The percentages of shared cognates used in LingPy to output the tree in Figure 11 are seen in Table 1. 29 Table 1 Cognacy rates from the current database calculated with LingPy (List et al. 2018) Gwama Uduk Opo Komo Dana Hi Lo Yab Cha Bil Mod Pam Gwama Lo 97 Komo 45 47 Yab 36 36 65 Uduk Cha 35 37 66 91 Dana 37 38 63 51 52 Bil 36 38 65 49 51 79 Mod 36 37 64 49 50 79 98 Opo Pam 37 38 65 48 50 81 95 96 Kig 36 38 64 48 50 84 93 95 97 While this dissertation does not aim to make any claims about lexicostatistical methods in language classification, it is important to note that all of the proposed internal classifications of Koman (without Gule or Baga) are identical. While lexicostatistics is grounded in assumptions about cognatcy rates as a means of identifying relatedness, shared innovations in phonology and morphology are also classically considered to be fundamental in determining relatedness. Innovations in phonology and morphology also support this classification, as described in Chapter III and Chapter IV. I turn now to an outline of the database and the methods employed in this study. 1.3 Methods and the database In this dissertation, I follow the general tenets of the comparative method (Bloomfield 1973 [1933], Campbell 1998, Trask 1996, inter alia) in reconstructing the Koman family. I focus mainly on reconstructing proto-phonemes, tonemes/tone categories, and lexicon and morphology. It is understood that forms, and in some cases 30 the meanings that they represent, are subject to diachronic change. It is my imperative to uncover the archaic patterns that emerge as languages change over time. I collected the data employed in this study between 2011-2018 over many fieldtrips. Komo, Opo, Dana and Gwama data were collected in various locations in Ethiopia and the Uduk data were collected in Salt Lake City from Uduk refugees who fled Sudan in the 1990’s. A main aim was to obtain as much data from as many varieties of Koman languages as possible. Many of the specific varieties documented here have had very little prior academic treatment, while others were previously completely undocumented. I have also consulted all available published and unpublished works, as well as consulted in personal communication with other scholars working on Koman languages.20 The reconstruction of Proto-Koman and the subnodes in this study is based on the lexical database and grammatical elicitation that I have compiled by combining my data with all these other resources. My field data was collected phonetically and entered into a spreadsheet in a rough phonetic transcription. From there, phonological representations were refined over time across various elicitation sessions. This is to say that my ear became more finely adjusted to the nuances of the segmental and suprasegmental features of the various systems as I worked more with the speakers. My aim was to first capture the maximum systematic phonetic detail in each variety, in order to describe the major surface structures and the (morpho)phonological processes that underlie these patterns of surface realization. With this knowledge, I was able to build more informed 20 I must again acknowledge and thank my fellow Koman researchers Anne-Christie Hellenthal, Don Killian, Justin and Joelle Goldberg and Joshua Smolders for sharing their insights and data with me. 31 phonological representations of the segments and tone. It was these phonological representations that I employed in the reconstruction.21 Aside from lexical items, I elicited basic declarative sentences from many of the varieties. I examined the core nominal and verbal morphology, which I describe in §2.2. In order to remain as faithful to the data as possible, I adhere to the following principles when establishing cognate sets and reconstructing to particular nodes. First, in order to confidently reconstruct a given lexeme to Proto-Koman (PKMN), I take the stance that there must be a reflex in Gwama and at least in one language of the Central Koman branch. A high confidence level is given to cognates found in Gwama and in the both branches of Central Koman. One caveat to reconstructing to PKMN is if cognates are only found in Komo and Gwama. Given the fact that these two groups have been in contact in Ethiopia for what appears to be several generations if not hundreds of years, there is a high possibility of contact influence. As such, I proceed with caution with cognates that are unique just to Komo and Gwama. Second, in order to reconstruct to Proto-central Koman (PCTRL), I take the stance that there must be reflexes in both branches: Komo-Uduk and Dana-Opo. Likewise, cognates must be in Komo and at least one variety of Uduk in order to reconstruct to Proto-Komo-Uduk (PKOUD). The same principle is applied to reconstructing to Proto-Dana-Opo (PDAOP), i.e., there must be a cognate in Dana and at least one variety of Opo. While this method intends to tread a cautious path in reconstructing to PKMN, it does not account for possible PKMN losses in Gwama that are retained in the lower 21 The term “phonological representation” employed here specifically refers to an internally reconstructed form. I do not make any claims about cognitive phonological representation. 32 nodes of other branches. By a less cautious methodology, the number of actual PKMN cognates would be higher than what I posit here. I follow these methodological criteria for clarity and to be conservative, that is, quite certain that any conclusions I posit wil stand the test of time. After culling the database for cognates and excluding problematic data such as multi-word entries with no corresponding entry in another language or variety, I reconstruct 240 etyma to PKMN, 229 etyma to PCTRL (that are not also in PKMN) 66 to PKOUD (that are not also in PCTRL), and 89 to PDAOP (that are not also in PCTRL). Across these cognates, I employ the comparative method to reconstruct Koman phonology and lexica. An etymological wordlist is provided in Appendix B and the full dataset is in Appendix F. I also dedicate a part of this study to reconstructing some core Koman morphology, including pronominals and deictic directional verb morphology. The main aim is that this reconstruction will ultimately serve not only those interested in Koman languages and the histories of their peoples, but also those interested in higher-level genetic classification in East African language families. A secondary aim is to guide future synchronic research in individual Koman languages and language varieties. 33 CHAPTER II OVERVIEW OF KOMAN PHONOLOGY AND MORPHOLOGY 2 Typological Overview Even though Koman has figured in language classifications of Nilo-Saharan in the last several decades (e.g. Bender 1971, 1983, 1994 inter alia; Ehret 2001), the Koman languages have historically received little treatment in the literature. The main exception is the Chali variety of Uduk, which has a grammar sketch from the last century (Stevenson 1942), a detailed phonological sketch (Thelwall 1983), and a more recent and very comprehensive description of the phonology and morphosyntax (Killian 2015). The Yabus variety of Uduk is virtually undescribed, though there is a brief outline of the phonology and some texts in Killian (2015). Historical linguists, such as M. Lionel Bender, employed these and other short works coupled with some field notes to compile and reconstruct Proto-Koman (Bender 1983, 1994). While these efforts were groundbreaking in the field, the descriptions of individual Koman languages were still somewhat rudimentary. For instance, the only data on the Komo at the time was a short grammar sketch compiled by a missionary in Sudan (Burns 1947) and there was virtually nothing written on Gwama and Opo. This led to limited analyses which were employed in historical classifications. For example, Bender (1983) reconstructs Proto-Koman with a five-vowel contrastive system because he analyzed the Koman languages at the time with five-vowel systems. We now know that seven-vowel systems with Advanced Tongue Root (ATR) harmony are seen in almost all of the living Koman languages (e.g. Otero 2015, Justin Goldberg 2018, Smolders forthcoming) and I reconstruct a seven-vowel system to Proto-Koman (§3.3). 34 Further, all of the Koman languages exhibit contrastive tone. Tone reconstruction (and its impact on the synchronic and/or diachronic phonological systems) has not been treated in any historical classification of Koman. It wasn’t until late in the first decade of the 2000’s that comprehensive work on describing Koman languages began. In Ethiopia, this was largely due to literacy initiatives of minority languages sponsored by the Ethiopian national and regional governments. The Ethiopian government sponsored mother-tongue education projects which required field linguists to describe and codify previously undescribed and unwritten languages to be used in early childhood education. The aim was for children of minority languages to become literate in their mother tongue. The Gwama and Komo language projects in western Ethiopia began in early 2011 (Hellenthal & Kutsch-Lojenga 2011, Kutsch-Lojenga & Otero 2011a-b) and continue to the time of this writing. Since the turn of the century, more descriptions of Komo and Gwama by Ethiopian and international scholars has continued to enrich the field of Koman studies (Hellenthal 2005, 2018; Zelealem 2005; Teshome 2008; Lemi 2010; Kievet & Robertson 2012; van Silfhout 2013, Killian 2015, to appear; Otero 2015a-b, 2018, 2019; Mellese 2017; Justin Goldberg 2018; Joelle Goldberg 2018; Smolders forthcoming; among others). 2.1 Phonological comparanda The Koman languages all display rich phonological systems in both the segmental and suprasegmental domains. Koman phonemic consonant inventories typically contrast in three to five places of articulation across three to five manners. Koman languages generally exhibit a seven-vowel contrastive inventory with Advanced Tongue Root (ATR) contrast and harmony, as well as three to four level tones and 35 contour tones. The following sections examine the phonology of the Koman varieties employed in this study. I begin with overall patterns seen across Koman with regard to consonants, vowels and tone and then move on to individual phonological sketches. 2.1.1 Overview of Koman consonants Koman consonant inventories range from 22 to 34 contrasting phonemes. In Chali Uduk and Dana, which belong to the two branches of Central Koman, stops contrast in five places of articulation: bilabial, interdental, alveolar, palatal and velar, as well as in four manners of articulation: voiced, voiceless aspirated, voiceless unaspirated, and ejective. One noteworthy feature is glottalized pulmonic and non-pulmonic consonants. All of the Koman languages have ejective stops in at least the bilabial, alveolar and velar places of articulation /pʼ, tʼ, kʼ/ and the Uduk, Dana, and Opo varieties also display a palatal ejective stop /cʼ/ or an alveopalatal ejective affricate /tʃʼ/. Gwama, Komo and Dana also exhibit an ejective alveolar fricative /sʼ/ and the Yabus Uduk contrastive inventory contains a rare alveopalatal fricative ejective /ʃʼ/. Of crucial importance to the historical reconstruction of Koman are the allophonic processes among the consonant systems across the languages. Given the fact that some of the languages have extensive consonant inventories, examining their synchronic distributions and allophonic processes provides insight into the historical evolution. Koman languages tend to favor closed syllables at the root edge and CVC roots are very common, especially in verb roots.22 Complex onsets are limited to consonant plus glide sequences /Cw/ and /Cj/ though these largely only occur before /a/. Complex 22 I recognize that this is not a quantitative observation but rather an overall impression. 36 codas are typically nasals plus consonant /NC/. Overall, there is a tendency for word- final devoicing and hence loss of a phonological voicing contrast.23 Koman languages are by and large suffixing, and vocalic elements in the suffixes can provide an insight into the behavior of word-final consonants in an intervocalic environment across morpheme boundaries. There is considerable variation in the allophonic realizations of intervocalic consonants; lenition and voicing are common for stops and ejectives. Further, one peculiar tendency in Koman languages is the delayed release of either voiced stops or ejective consonants word-finally before a pause. Delayed release is realized by a full closure followed by a pause, and a less intense burst or release. This peculiar articulatory gesture is found across all of the branches. In order to capture the synchronic variation, the distributions of word-initial, intervocalic and word-final consonants are examined for the languages in this study for which data was collected. 2.1.2 Overview of Koman vowels and Advanced Tongue Root harmony All of the living Koman languages, with the exception of the Uduk cluster, exhibit a seven-vowel contrastive inventory /i, ɪ, ɛ, a, ɔ, ʊ, u/, with phonemic Advanced Tongue Root [ATR] contrast in the high vowels. The Uduk varieties only display a five-vowel phonemic inventory /i, ɛ, a, ɔ, u/ and no contrastive ATR feature. One distinct feature of the Koman languages, which was most likely inherited from Proto-Koman, is ATR vowel harmony (§3.3). Synchronically, all of the Koman languages with seven-vowel inventories exhibit distinct types of ATR vowel harmony. 23 Dimmendaal (forthcoming) observes that in Nilo-Saharan, word-final voicing contrasts are common in the eastern zones (Ethiopia, Sudan, South Sudan, Kenya and Tanzania) but not elsewhere in the phylum. 37 Casali (2008:497) observes that ATR harmony systems are often found in the Niger- Congo and Nilo-Saharan languages of sub-Saharan Africa.24 ATR vowel harmony is a type of feature assimilation in which vowels in certain contexts must agree on the value of the [±ATR] feature. The domain of harmony is typically the word and the assimilatory processes generally occur across morpheme boundaries involving roots (or stems), affixes and clitics. Traditionally, two types of ATR harmony processes are distinguished: dominant-recessive and stem-controlled or root-controlled (Kirchner 1993, Clements 2000, Baković 2000). In a dominant-recessive system, the feature that is considered ‘dominant’ (most frequently [+ATR]) spreads to the ‘recessive’ vowels of the opposite feature (typically [–ATR]) in roots/stems and/or affixes/clitics. In a stem- controlled system, neither feature is dominant, and affixes “harmonize” (or assimilate) to the ATR feature of the root or stem to which they are attached, regardless of its [ATR] specification. Casali’s (2003) typological survey of 110 languages found a direct correlation between the phonemic vowel inventory and the type of ATR harmony system displayed. The specific criterion is whether or not the phonemic ATR contrast is in the [+high] vowels. According to Casali (2008:520), languages with ATR contrast in the [+high] vowels, such as the Koman languages with contrastive seven-vowel inventories, overwhelmingly display [+ATR] dominant-recessive harmony. This is partially the case in Komo, which exhibits anticipatory (leftward-spreading) [+ATR] dominant harmony in one domain, but Komo also exhibits progressive (rightward- 24 While tongue root harmony systems are overwhelmingly found in Africa, they have also been described in languages beyond the African continent such as Karajá (Macro-Jê; Ribeiro 2000, 2002) and Even (Northern Tungisic; Aralova 2015). 38 spreading) [–ATR] dominant harmony in another domain (Otero 2015a, Olejarczuk et al. 2019). Further, the remaining Koman languages with contrastive ATR exhibit stem- controlled harmony involving only the [+high] vowels. 2.1.3 Overview of Koman tone In the suprasegmental domain, the Koman languages all display contrastive tone. A language with tone is defined here as “one in which the indication of pitch enters into the lexical realization of at least some morphemes” (Hyman 2001:1368). Koman languages all exhibit at least three distinct contrastive tone levels (L)ow, (M)id and (H)igh. The Opo cluster is unique in that it has innovated an extra-high (XH) tone, and thus has four contrastive level tones (Smolders 2017, forthcoming, p.c.). Koman languages also exhibit (R)ising and (F)alling contours in their contrastive tonal inventories. These contours can be analyzed as contrastive unitary tonemes. Vowel length is not contrastive in Koman languages and the TBU is a vowel (or a vocalic mora). All of the tones, including the level tones and contour tones, can occur on a single vowel nucleus. In some languages, the contour tones can appear to have a slightly longer duration than the level tones, though this could be a historical result from the abutting of two vowels of differing level tones. This process is also seen synchronically across morpheme boundaries. For instance, compare the data from Komo in (1). In (1a), the verb wʊ̀ ‘die’ occurs with a L tone and a phonetically “short” vowel as compared to (1b), in which it occurs with the H tone -ʊ́ DD1 morpheme and as a result, surfaces with a rising tone and a phonetically “longer” vowel. (1) Komo a. wʊ̀ [wʊ̀] b. wʊ̀-ʊ́ [wʊ̌ː] die die.DD1 39 Tone plays a prominent role in the reconstruction of Proto-Koman. The following subsections provide brief phonological sketches for each language cluster. I discuss the contrastive consonant, vowel and tone inventories and any significant morphophonological processes that impact the reconstruction. Some of the phonological sketches provide descriptions of heretofore undescribed Koman language varieties. 2.1.4 Gwama The Highland and Lowland varieties of Gwama employed in this study are spoken in the Benishangul-Gumuz Regional State of western Ethiopia. The varieties are mutually intelligible with some slight differences in the phonology and lexicon. Data and analysis in the following Gwama phonological sketch comes from fieldwork and from descriptions of the Lowland Gwama variety (cf. Hellenthall & Kutsch Lojenga 2011, Kievet & Robertson 2012, Goldberg et al. 2017). 2.1.4.1 Gwama contrastive consonant inventory The Gwama varieties exhibit the fewest consonants of all of the living Koman languages. The contrastive inventory for both Highland and Lowland varieties consists of 22 consonants, seen in Table 2. Plosives and fricatives contrast across three places of articulation (bilabial, alveolar and velar) and across three manners of articulation (voiceless, voiced and ejective). Note that Gwama is the only living Koman language that does not display contrastive implosives. 40 Table 2 Gwama contrastive consonant inventory Alveo- Bilabial Alveolar Palatal Velar Glottal palatal voiceless p t k voiced b d g ejective pʼ tʼ kʼ voiceless s ʃ h voiced z ejective sʼ nasal m n ŋ lateral l trill/flap r approximant w j Previous descriptions of the Gwama consonant inventory have differed slightly but not significantly. Bender (1983:264) proposes a 22-consonant phonemic inventory that lacks a velar nasal but includes a voiceless labiodental fricative, the latter of which is not contrastive; Bender most likely mistranscribed [ɸ] as [f]. The former is an allophone of the voiceless bilabial stop /p/ in intervocalic position, which has also been observed by Hellenthal (2005) (see 2.1.4.2 below). Zelealem’s (2005) unpublished sketch of Highland Gwama proposes two additional consonants to Bender’s (1983) inventory, a palatal nasal and a velar nasal.25 More recent in-depth studies of Gwama propose the same contrastive inventory seen in Table 2 (Hellenthal 2005, Hellenthall & Kutsch 25 Zelealem (2005) does not provide any phonological evidence for a phonemic voiceless labiodental fricative. 41 Lojenga 2011, Kievet & Robertson 2012, Amare 2013, Goldberg et al. 2017, Justin Goldberg 2018).26 2.1.4.2 Gwama consonant distribution and allophones Table 3 contains the distribution and allophones of the Gwama obstruents across bilabial, alveolar and velar places of articulations, contrasting in three manners of articulation: voiceless plosive, voiced plosive and ejective. The three-way manner contrast is maintained in word-initial position, though Goldberg (2018) observes that it is partially neutralized in intervocalic position as the ejectives /pʼ, tʼ, kʼ/ can be realized as plain voiced obstruents [b, d, g]. Further, he observes that the three way contrast is also neutralized word-finally as voiced obstruents do not occur in this position. Further, ejectives are optionally unreleased in word-final position, a very common trait of Koman ejective realization. Lastly, Gwama voiceless stops are phonetically aspirated in all positions and can lenite to fricatives intervocalically. Table 3 Distribution and allophones of Gwama obstruents #__ V__V __# #__ V__V __# /p/ pʰ p, pʰ, ɸ pʰ /s/ s s s /t/ tʰ t, tʰ tʰ /z/ z z – /k/ kʰ k, kʰ, x kʰ /sʼ/ sʼ sʼ, z sʼ /b/ b b, β – /pʼ/ pʼ pʼ, b pʼ, p̚ /d/ d d, ɾ – /tʼ/ tʼ tʼ, d tʼ, t̚ /g/ g g, ɣ – /kʼ/ kʼ kʼ, g kʼ, k ̚ /ʃ/ ʃ ʃ ʃ /h/ h h – 26 The palatal nasal is analyzed as an allophone of the alveolar and velar nasal before front vowels (Hellenthal 2005, Hellenthal & Kutsch Lojenga 2011, Goldberg et al. 2017). 42 Table 4 contains the distribution and allophones of the sonorants in word-initial, intervocalic and word-final positions. The bilabial and alveolar nasals can occur in all positions, while the velar nasal is restricted to intervocalic and final positions. The liquids can occur in all positions. Table 4 Distribution and allophones of Gwama sonorants #__ V__V __# #__ V__V __# /m/ m m m /l/ l l l /n/ n n n /r/ r ɾ r /ŋ/ – ŋ ŋ /w/ w w w /j/ j j j Consonant-glide combinations (Cw and Cj) are attested in Gwama though they are restricted to root-initial position. Cw sequences typically occur before /a/, though they can also occur before front vowels. In my database, I have 33 instances of Cw before /a/ and 14 instances before a front vowel. Cj sequences are less common (only 12 in my database) and they only occur before /a/. Goldberg et al. (2017) note that only /dj/, /ʃj/, and /zw/ do not occur. 2.1.4.3 Gwama vowels Both Highland and Lowland Gwama exhibit a contrastive seven-vowel inventory with ATR contrast in the high vowels. Vowel length is not contrastive in monosyllabic monomorphemic roots (Goldberg et. al 2017, Justin Goldberg 2018). The complete contrastive vowel inventory is given in Figure 12. 43 front central back i u [+ATR] [+high] ɪ ʊ [–ATR] ɛ ɔ [–ATR] [–high] a [–ATR] Figure 12 Gwama contrastive vowel inventory Examples of contrast in Gwama vowels are in (2). (2) /i/ sí ‘bone’ /ɪ/ sɪ ́ ‘fall’ /ɛ/ ʃɛ̀ ‘slaughter’ /a/ sā ‘flirt’ /ɔ/ sɔ́ ‘stab’ /ʊ/ sʊ́ ‘pierce’ /u/ sū ‘same (be)’ Gwama displays what Casali (2003, 2008), among others, refer to as root-controlled ATR harmony. In Gwama, root-controlled ATR harmony occurs in the high vowels at the level of the phonological word (Hellenthal & Kutsch Lojenga 2011, Goldberg et al. 2017). In this process, high vowels in affixes assimilate to the ATR value of the root. Gwama displays both prefixes and suffixes, all of which harmonize to the ATR value of the root to which they are attached. Goldberg et al. (2017) refer to this as [+ATR] dominant root-controlled harmony given the fact that only [+ATR] is spread, though it is important to note that in the Gwama system, affixes containing [+high] vowels are [–ATR]. To illustrate the Gwama ATR harmony system, the following data contain verb roots of contrasting ATR values inflected with the 3SG.M suffix /-nɪ/́, which contains a [+high, –ATR] vowel /ɪ/́. The roots in (3) contain [+high, –ATR] vowels /ɪ, ʊ/, and the 44 roots in (4) contain [–high, –ATR] vowels /ɛ, a, ɔ/. In all of these examples, the root and suffix vowels all surface as [–ATR]. (3) a. tɪ-̌nɪ~́tɪňdɪ ́ [tɪňɪt́ɪňdɪ]́ b. sʼʊ́p-nɪ-́sʼʊ́p [sʼʊ́pnɪśʼʊ́p] be.fat-3SG.M~RED be.cold-3SG.M~RED ‘He is fat.’ ‘It’s cold.’ (4) a. wǎ-nɪ-́wǎ [wǎnɪẃǎ] b. lɔ̀s-nɪ-́lɔ̀s [lɔ̀snɪĺɔ̀s] break-3SG.M~RED choose-3SG.M~RED ‘It is broken.’ ‘He chooses.’ c. ʃɛ̀-nɪ-́ʃɛ̀ [ʃɛnɪʃ́ɛ̀] slaughter-3SG.M ‘He slaughters.’ Compare this to (5), in which the roots contain a [+high, +ATR] vowel /i, u/. In (5), the [–ATR] suffix vowel in -nɪ ́3SG.M assimilates to the [+ATR] value of the [+high, +ATR] root vowel. (5) a. dwì-nɪ~́dwì [dwìnídwì] b. bǔʃ-nɪ-́bǔʃ [bǔʃníbǔʃ] buy-3SG.M~RED fart-3SG.M~RED ‘He buys.’ ‘He farts.’ Root controlled ATR harmony is also seen across morpheme boundaries on nouns occurring with the proclitic /ʊ̄=/ DEF.M. This clitic surfaces as [–ATR] when the root vowel is [–ATR], as in (6a); but if the root is [+ATR], the vowel of the clitic assimilates to [+ATR] and surfaces as [u], as in (6b). 45 (6) a. ʊ̄=kānā [ʊ̄kānā] b. ʊ̄=sízì [ūsízì] DEF.M=dog DEF.M=crocodile ʻthe dogʼ ʻthe crocodileʼ (adapted from Goldberg et al 2017:33) Hellenthal & Kutsch Lojenga (2011:2) and Goldberg et al. (2017:32) also describe what appears to be anticipatory (leftward-spreading) [+ATR] dominant harmony in which the [+high, +ATR] vowels /i, u/ cause the [–high, -ATR] vowels /ɛ, a, ɔ/ to surface as [+ATR] allophones [e, ə, o], respectively. Hellenthal & Kutsch Lojenga (2011:2) describe the allophones in this process as gradient realizations though they provide no evidence to support these claims. Goldberg et al. (2017:32) provide three examples of monomorphemic roots, reproduced here in (7). They argue that [+ATR] dominant harmony is anticipatory in that [+ATR] spreading occurs in (7a-b) but not in (7c) because in (7c), the [+high, +ATR] vowel precedes the [–high, –ATR] vowel. Note that (7b) is a borrowing and (7a) is also used in Komo, which does in fact exhibit [+ATR] dominant harmony (cf. §2.1.5.3, Otero 2015, Olejarczuk et al. 2019). I did not find any evidence for this [+ATR] dominant harmony in my field research in Gwama. (7) a. /sàlíjà/ [sə̀líjà] ʻcorn beerʼ b /làmúnù/ [lə̀múnù] ʻlemonʼ c. /bìdɔ́/ [bìdɔ́] ʻsheepʼ Previous Gwama researchers have noted that phonetic vowel nasalization can arise from the presence of a preceding or following nasal consonant, though they also recognize a small set of lexemes that contain nasal vowels without nasal consonants (Hellenthal 2005:9, Zelealem 2005:4, Hellenthal & Kutsch Lojenga 2011:7, Goldberg et al. 2017). Nasalized vowels, if indeed considered phonemic, are very rare and some may 46 have arisen from the historical presence of a nasal vowel which has subsequently been lost. Some examples of Gwama lexemes with nasal vowels are twā̃ ‘forehead’, kɛ̀̃ ʻsmellʼ, and zı̀̃ ‘green’. 2.1.4.4 Gwama Tone Both varieties of Gwama display contrastive tone. There are three level tones (L, M, H) as well as a rising contour tone (R). Goldberg et al. (2017) analyze the rising contour tone as phonetically LM.27 The three level tones as well as the rising tone can occur on a single vowel nucleus in a monomorphemic word as seen in the near minimal group in (8). Tone in Gwama is quite stable throughout the system; there are no significant tonal processes that occur across morpheme or word boundaries, nor is there downstepping or downdrifting at the level of the utterance. (8) H sí ‘bone’ M sīt ‘person’ L sìl ‘heavy (be)’ R sǐs ‘cane rat’ 2.1.5 Komo phonology Komo is spoken in villages along both sides of the border between Ethiopia and (South)Sudan. The majority of the ±1,500 Komo speakers in Ethiopia are found in remote villages of the Mao-Komo Special Woreda (district) of the Benishangul-Gumuz Regional State of western Ethiopia. Further, about 200 speakers reside in and around the village of Pokung in the Gambella Regional State of southwest Ethiopia. Komo 27 Hellenthal & Kutsch Lojenga (2011) observe that falling contour tones are marginal, at best. 47 speakers are also found in and around the village of Yabus, a town in the Blue Nile State of Sudan.28 There appears to be minimal dialectal variation between the Ethiopian and Sudanese varieties of Komo, though no definitive study has been conducted. Ethiopian Komo speakers all indicate that the Sudanese variety is mutually intelligible though they cite some minor lexical differences. The Komo employed in this study is solely the Ethiopian variety. The first data on Komo are wordlists collected in the Sudan in the late 1800’s by explorer Juan María Schuver (James et al. 1996) and Corfield (1938). The first linguistic analysis is by the missionary Burns (1947), who provides an unpublished grammatical sketch also containing didactic material of the Komo spoken in Yabus, Sudan.29 Tucker & Bryan’s (1966:356–362) Komo sketch is an aggregate of Burns (1947) and Corfield (1938). Bender (1983, 1989, 1994) examine aspects of Komo phonology within the larger Koman and Nilo-Saharan perspective, employing prior analyses by others as well as his own fieldwork. Teshome (2008) is a short phonological description of Ethiopian Komo. The Komo literacy project in the Benishangul-Gumuz region has produced several unpublished manuscripts on Ethiopian Komo phonology and lexicon (Kutsch Lojenga & Otero 2011a–b). More current works on Komo phonology and phonetics include Otero (2015a, 2018b) and Olejarczuk et al. (2019). 28 I was not able to travel to (South)Sudan given the ongoing armed conflict in the area. All of the information about Sudanese Komo is from Burns (1947) and from Komo and Yabus Uduk consultants. 29 Burns also translated the Gospel of John and the Gospel of Mark into Komo employing a working Komo orthography. Burns’ grammatical sketch is quite comprehensive. He gives an overview of the phonology and basic morphosyntax that is much in line with my data and analyses of Ethiopian Komo. 48 2.1.5.1 Komo consonants The Komo contrastive consonant inventory contains 23 phonemic consonants, seen in Table 5. There is a three-way contrast for plosives in the bilabial, alveolar and velar places of articulation and a three-way contrast in manner of articulation, which includes plain, ejective and implosive obstruents. This inventory is largely consistent with previous analyses of Komo, with the exception of the velar nasal and the glottal stop, which are marginally contrastive.30 Table 5 Ethiopian Komo contrastive consonant inventory† Alveo- Bilabial Alveolar Palatal Velar Glottal palatal voiceless p t k (ʔ) voiced b d g ejective pʼ tʼ kʼ implosive ɓ ɗ voiceless s ʃ h voiced z ejective sʼ nasal m n (ŋ) lateral l trill/flap r approximant w j † Parentheses indicate marginally contrastive consonants. Burns (1947) and Bender (1983) propose an identical consonant inventory to the current analysis with the exception of the velar nasal, which is not contrastive in their analyses. Similarly, Tesfaye (2015) includes the velar stop but lacks the velar nasal and 30 See Otero (2018b) for further details on the status of the velar nasal and glottal stop. 49 Teshome (2008) presents a 24-consonant inventory which lacks a bilabial implosive but includes a phonemic velar nasal, glottal stop and a voiceless bilabial fricative – the latter of which I consider to be an allophonic surface realization [β] of intervocalic /b/. 2.1.5.2 Komo consonant distribution and allophones In terms of distributional restrictions, the implosive series is restricted to word- initial position. All of the of the contrastive consonants can occur in word-initial, intervocalic and word-final position.31 Table 6 contains a complete distribution of the allophones of the stops and sibilants in word-initial, intervocalic and word-final positions. The voiceless stops /p, t, k/ can be realized with aspiration in all three positions [pʰ, tʰ, kʰ], and the bilabial and velar stops can lenite to voiceless fricatives intervocalically [ɸ, x]. The voiced bilabial and velar stops /b, g/ can lenite to voiced fricatives intervocalically [β, ɣ], and the voiced alveolar stop /d/ can be articulated as a flap intervocalically [ɾ]. All of the voiced stops are realized with a delayed release word-finally before a pause. The ejective stops are always articulated with glottal release word-initially [pʼ, tʼ, kʼ] and can be realized as voiced stops [b, d, g] intervocalically. In word-final position, the ejective stops are unreleased. The ejective sibilant /sʼ/ is realized with glottal release in initial and final positions and can either be realized with glottal release or as a voiced fricative intervocalically [z].32 31 The contrastive glottal stop only occurs in word-final position and the velar nasal only occurs in medial and final positions. 32 Otero (2018b) notes that word-final ejectives in monomorphemic noun roots behave distinctly from word-final ejectives in monomorphemic verb roots. Noun roots preserve the laryngeal feature while in verb roots, ejectives are optionally unreleased word-finally and neutralize with voiced non-ejective allophones intervocalically. 50 Table 6 Distribution and allophones of Komo stops and sibilants #__ V__V __# #__ V__V __# /p/ p, pʰ p, pʰ, ɸ pʰ /s/ s s s /t/ t, tʰ t, tʰ tʰ /z/ z z z /k/ k, kʰ k, kʰ, x kʰ /sʼ/ sʼ sʼ, z sʼ /b/ b b, β b̚ /pʼ/ pʼ pʼ, b pʼ, p̚ /d/ d d, ɾ d ̚ /tʼ/ tʼ tʼ, t, d tʼ, t̚ /g/ g g, ɣ g ̚ /kʼ/ kʼ kʼ, g kʼ, k ̚ 2.1.5.3 Komo vowels Komo exhibits a seven-vowel /i, ɪ, ɛ, a, ɔ, ʊ, u/ contrastive inventory distinguished by the feature of Advanced Tongue Root (ATR), seen in Figure 13. The phonetic allophones are enclosed in brackets. The Komo system exhibits phonemic ATR contrast in the [+high] vowels. Further, three [–high, +ATR] vowels [e, o, ə] occur on the surface as allophones of the three [–high, –ATR] vowels /ɛ, ɔ, a/, respectively, as a result of anticipatory [+ATR] harmony.33 Evidence for the ‘basic’ status of [–ATR] is the fact that the vowels [e, o, ə] do not occur in monosyllabic monomorphemic roots and only occur in a specific environment which contains a [+high, +ATR] vowel /i, u/. 33 Burns (1947) proposes a ten-vowel inventory including seven oral vowels /i, e, ɛ, a, ɔ, o, u/ and three nasal vowels /ĩ, ɛ̃, ã/. Nasal vowels, if they are indeed phonemic, are at best marginally phonemic and are always followed by a nasal consonant. Tesfaye (2015) proposes a nine vowel /i, e, ɛ, ɨ, ə, o, ʊ, u/ system though he does not provide any evidence for phonemic [–high, –ATR] vowels [e, ə, o], considered allophones in my analysis, nor do I find any evidence for a central vowel [ɨ]. 51 [+ATR] [–ATR] [+high] i u ɪ ʊ [–high] [e] [o] ɛ $ [ə] a Figure 13 Ethiopian Komo vowel system Komo exhibits two distinct types of ATR harmony. Anticipatory [+ATR] dominant harmony is seen in monomorphemic roots and across morpheme boundaries from suffix to root. This process spreads [+ATR] leftward from a [+high, +ATR] vowel /i, u/ to a [–high, –ATR] vowel /ɛ, a, ɔ/, causing the latter to surface as [e, ə, o]. The data in example (9) show three monomorphemic monosyllabic verbs containing each of the three [–high, –ATR] vowels /ɛ, a, ɔ/ in the root. Each verb occurs with a deictic directional (DD) suffix containing both [+high, +ATR] vowels /i, u/, followed by a person marking suffix. In every case, [–ATR] root vowels surface as [+ATR] allophones [e, ə, o]. Anticipatory [+ATR] harmony is further evidenced in (9a-c) by the fact that the suffix -à 1PL.IN, which is to the right of the [+ATR] suffix -úk DD2, does not assimilate to [+ATR] and surface as [ə]. (9) a. gɛm̄-úk-à [gēmúkà] d. gɛm̄-í-r [gēmír] weed-DD2-1PL.IN weed-DD∅-3SG.M ‘We weeded.’ ‘He weeds.’ b. bàb-úk-à [bə̀búkà] e. bàb-í-r [bə̄bīr] bury-DD2-1PL.IN bury-DD∅-3SG.M ‘We buried.’ ‘He buries’ 52 c. kɔ́p-úk-à [kópúkà] f. kɔ́p-í-r [kópír] bale-DD2-1PL.IN bale-DD∅-3SG.M ‘We baled.’ ‘He bales’ The same anticipatory [+ATR] dominant harmony seen across morpheme boundaries in (9) also occurs in monomorphemic words as in (10). The result of this harmony process is three non-contrastive allophones [e, ə, o] of /ɛ, a, ɔ/, respectively. (10) /ɛ/ bɛ̀zí [bèzí] ‘be thin’ mɛ̀tí [mètí] ‘chase’ /a/ pātí [pə̄tí] ‘liver’ gàbút [gə̀bút] ‘gruel’ /ɔ/ kɔ́sí [kósí] ‘bread (trad.)’ gɔ̀dùm [gòdùm] ‘sow’ The data in (11) show that the spreading of [+ATR] is strictly anticipatory, as an initial [+high, +ATR] vowel does not spread its [+ATR] feature rightward to a following [–high, –ATR] vowel. This particular vocalic environment remains disharmonic for the ATR feature. (11) /ɛ/ dúrɛ́ [dúrɛ]́ ‘louse’ /a/ dìbàl [dìbàl] ‘grasshopper’ /ɔ/ mílɔ̄ [mílɔ̄] ‘flute’ The second type of ATR harmony in Komo only involves the high vowels /i, ɪ, ʊ, u/. This process is progressive, spreading [–ATR] rightwards from a [+high, –ATR] /ɪ, ʊ/ root vowel to a [+high, +ATR] suffix vowel /i, u/, causing the latter to surface as [–ATR]. The data in (12) display monomorphemic monosyllabic verb roots occurring with the identical vocalic suffixes seen on the verbs which spread [+ATR] in (9). The 53 difference is that the verb roots in (12) contain a [+high, –ATR] vowel /i, ʊ/, which spreads [–ATR] rightward to the [+high, +ATR] suffix vowels. Note how in progressive [+ATR] dominant spreading, the same vowels that were the triggers of [+ATR] harmony in one domain are the targets of [–ATR] harmony in another domain. (12) a. ʃɪt́-úk-à [ʃɪt́ʊ́kà] d. ʃɪt́-í-r [ʃɪt́ɪŕ] whistle-DD2-1PL.IN whistle-DD∅-3SG.M ‘We whistled.’ ‘He whistles.’ b. bʊ̀d-úk-à [bʊ̀dʊ́kà] e. bʊ̀d-í-r [bʊ̄dɪr̄] unthatch-DD2-1PL.IN unthatch-DD∅-3SG.M ‘We unthatched.’ ‘He unthatches.’ Komo exhibits two fully productive ATR harmony processes (Otero 2015, 2018b). Recall that Casali’s (2003, 2008) typology of African ATR systems proposes that languages with ATR contrast in the high vowels tend to exhibit dominant-recessive ATR harmony, which is typically exemplified by [+ATR] affix vowels causing [–ATR] stem vowels to assimilate to [+ATR]. Komo does indeed exhibit anticipatory (leftward- spreading) [+ATR] dominant harmony, which conforms to Casali’s proposed typology. But Komo also exhibits progressive (rightward-spreading) [–ATR] dominant harmony. Taken together, these two harmony systems ultimately call into question the notion that only one ATR feature can be dominant in a given language. In a recent acoustic study of the Komo ATR system, Olejarczuk et al. (2019) provide phonetic evidence for the two distinct types of ATR harmony described in Otero (2015a). Olejarczuk et al. (2019) analyzed recording of verb paradigms inflected with distinct suffixes from twelve native speakers of Komo and examined the acoustic correlates of anticipatory and progressive ATR harmony. While the acoustic results are 54 not empirical proof of ATR harmony, the results confirm the transcriptions of the vowels in both harmony processes. Olejarczuk et al. (2019) found that while F1 was the most reliable acoustic correlate in establishing ATR contrasts in the Komo system, other cues such as voice-quality and vowel duration also play an important role. 2.1.5.4 Komo tone Komo displays three level tones – (L)ow, (M)id and (H)igh – as well as (R)ising and (F)alling contour tones.34 Minimal tone pairs are seen in (13). The level and contour tones can occur on a single vowel nucleus though contour tones can result from two vowel nuclei of differing level tones coming together at a morpheme boundary. (13) L bàm ‘move (as a herd)’ H bám 1SG.POSS M bā DEM.F R bǎ ‘father’ L ɓɪ̀ ‘mountain’ F ɓî ‘eye’ M dā Q F dâ ‘mother’ M jɪ ̄ ‘axe’ R jǐ ‘water’ Overall, tone is quite stable in Komo: there is no evidence of downstep or downdrift in prosodic words or even across longer sentences. The functional load of tone is moderate in the lexicon, more present in the verbal system where tonal suppletion in verb roots is a means of distinguishing nominal and verbal number.35 34 Burns (1947) claims there are four level tones in Sudanese Komo though I find no evidence of this in Ethiopian Komo. I also briefly worked with a Komo speaker from Sudan and did not encounter four level tones. The modern-day Opo varieties clearly display four level tones (Smolders 2017; also see §2.1.7 of this dissertation). 35 See Otero (2015b) for a description of the interaction between tone in the verb and participant indexing as a means of dual and/or paucal participant number marking. 55 Another aspect of Komo verbs that is of tangential relevance to the historical reconstruction are verb tone classes. There are four main tone classes for monosyllabic roots, which are established by comparing the tone of a non-finite bare root, such as what follows an inflected auxiliary, with that of a finite root inflected with Deictic Directional (DD) and person-indexing suffixes. The four root-level tone classes are H, M, L1 and L2. The H and M roots maintain their tone when inflected with additional morphology. The L tone roots split into two categories: one set retains a L tone and the other surfaces as M tone (set L1 and L2 respectively). The examples in (14) contain CVC verb roots of the four tone classes (H, M, L1, L2 ) inflected with /-í/ 2SG to form the imperative. Note that the M, H and L1 roots retain the root tone when inflected and the suffix tone remains H; while the in the L2 class, both the stem and suffix are realized as M. (14) class root + /-í/ 2SG gloss H ɓʊ́g bʊ́gɪ ́ ‘wait’ M tɔ̄n tɔ̄ní ‘pass’ L1 dùm dùmí ‘hit’ L2 kɛ̀d kɛd̄ī ‘push’ 2.1.6 Uduk phonology The Uduk dialect cluster of languages is spoken in the Blue Nile State of the Sudan. There are two close northern varieties, Chali Uduk and Bonya Uduk and a more divergent southern dialect, Yabus Uduk. James (1979), Thelwall (1983) and Killian (2015) as well as my consultants observe that the northern varieties are very close and there is “scarcely a problem of mutual understanding” (James 1979:263). By contrast, the southern dialect, which has been historically spoken along the Yabus river in Belatoma 56 and in the town of Yabus, is recognized as the most divergent dialect in the literature as well as by my consultants. James (1979:263) notes that the southern variety, what my consultants refer to as Yabus Uduk, is “markedly different” in the sound system and the lexicon. The degree of mutual intelligibility between the northern and southern varieties has yet to be empirically investigated. When asked informally about the differences between the two, Chali and Yabus Uduk speakers note significant differences in the sound systems, grammar and in the lexicon. With regards to mutual intelligibility, both Killian (2015) and I observe that Chali speakers find it “challenging” to understand Yabus Uduk speech. By contrast the Yabus consultants I worked with say they can understand the Chali variety. Whether or not this reported difference in intelligibility is system internal, due to diverging structural properties of the varieties, and/or whether it is a social matter of prestige as Chali is the dominant variety, remains to be investigated. Also note that the Uduk orthography, which has been used in schools and in Bible translations in Sudan since the 1960s as well as in refugee camps, is of the northern Chali variety. The first linguistic study of Uduk phonology is Stevenson’s (1942) unpublished grammar sketch of the Chali variety. Interestingly, Stevenson never actually conducted fieldwork but used notes collected by the missionaries Malcom and Enid Forsberg who began working with the Uduk for the Sudan Interior Mission (SIM) in Chali in 1941. Tucker & Bryan’s (1966) short survey of Uduk is based on Stevenson (1942). The first detailed study of Uduk phonology is Thelwall’s (1983) phonological sketch of Chali Uduk. Thelwall employed data collected from a native speaker in Khartoum from 1968- 1971. Thelwall (1983:323) notes that his phonological analysis does not differ 57 significantly from the Chali Uduk orthography created by the Sudan Interior Mission. This orthography was used to write Beam & Cridland’s (1970) Chali Uduk dictionary (first drafted as a manuscript in 1956). Chali Uduk phonology is also discussed in Killian (2015), which only differs significantly from the aforementioned studies in the phonemic consonant inventory, which is discussed below in (§2.1.6.1). The Yabus Uduk variety has received very little attention in the linguistic literature. The only prior work on the phonology is Killian’s (2015:305–306) note on the phonemic consonants and vowel inventories. Killian also includes several glossed texts in Yabus Uduk in an appendix. The following sections provide phonological sketches for the Chali and Yabus varieties based on my data and previous analyses. I focus on the synchronic phonemic inventories and lay out allophonic processes that are integral to the systems. 2.1.6.1 Chali Uduk contrastive consonant inventory The Chali Uduk variety has the most phonemic consonants of any living Koman language. Prior descriptions of the Chali Uduk consonant inventory differ significantly: Stevenson (1942) provides an inventory of 31 consonants, while Killian (2015) proposes an inventory of 55 contrastive consonants.36 My analysis of the Chali Uduk consonant inventory is consistent with Thelwall (1983) and the Uduk orthography employed in Beam & Cridland (1970 [1956]). I analyze 34 contrastive phonemic consonants, which are presented in Table 7. The plosive series robustly contrasts in five places of 36 It is not clear whether Stevenson’s inventory is a phonemic and/or a phonetic representation of the consonant system as he includes several sounds which are considered to be allophones of the current contrastive inventory. To his credit, Stevenson does recognize the shortcomings of his description and calls for a more in-depth investigation, given the fact that he hadn’t ever heard the Uduk language. 58 articulation: bilabial, interdental, alveolar, palatal and velar. The glottal stop is marginally contrastive in word-final (pre-pause) position. Further, the obstruents contrast in five manners of articulation: voiceless aspirated, voiceless unaspirated, voiced, ejective and implosive.37 Table 7 Chali Uduk contrastive consonant inventory† Inter- Alveo- Bilabial Alveolar Palatal Velar Glottal dental palatal unaspirated p t ̪ t c k (ʔ) aspirated pʰ t ̪h tʰ cʰ kʰ ejective pʼ t ̪̓ tʼ cʼ kʼ voiced b d̪ d ɟ g implosive ɓ ɗ fricative s ʃ h nasal m n ɲ ŋ lateral l trill/flap r approximant w j †Parentheses indicate a marginally contrastive phoneme. Killian’s (2015:20) contrastive consonant inventory of 55 consonants is presented in Table 8. The major difference between Killian’s proposal and the current analysis is that Killian adds a partial series of phonemic labialized consonants. 37 Chali Uduk, along with Dana, are the only Koman languages that contrast interdental and alveolar stops across four manners of articulation. 59 Table 8 Killian’s (2015:20) Chali Uduk contrastive consonant inventory† Post- Bilabial Dental Alveolar Palatal Velar Glottal alveolar unaspirated p t ̪ t c k ʔ labialized pʷ (t ̪̫ ) cʷ kʷ aspirated pʰ t ̪h tʰ cʰ kʰ labialized pʰʷ (tʰʷ) cʰʷ kʰʷ ejective pʼ t ̪̓ tʼ cʼ kʼ labialized tʼʷ cʼʷ kʼʷ voiced b d̪ d ɟ g labialized bʷ dʷ ɟʷ gʷ implosive ɓ ɗ labialized ɓʷ ɗʷ fricative s ʃ h labialized ʃʷ nasal m n ɲ ŋ labialized (mʷ) (ɲʷ) ŋʷ lateral l trill/flap r approximant w j † Parentheses indicate “rare or marginal phonemes” (Killian:2015:20). Killian proposes seventeen contrastive labialized consonants (including labialized voiceless unaspirated, labialized implosive and labialized velar nasal, among others) and four marginally contrastive labialized consonants. Killian’s analysis is problematic for several reasons. First, it increases the inventory of 34 contrastive consonants by 21. Second, it proposes a rather phonologically unbalanced system: of the 31 possible labialized consonants in his inventory, there are 10 gaps of unattested labialized consonants (i.e., /pʼʷ, t ̪h ʷ, t ̪̓ ʷ, d̪ʷ, tʷ, sʷ, nʷ, lʷ, rʷ, hʷ/ are unattested).38 Further, Killian 38 This is assuming that all of the 34 consonants, with the exception of /ʔ, j, w/, are “labializable”. 60 (2015:23–24) recognizes that the labialized consonants have a limited distribution. They can only occur before the vowel /a/ in word-initial position.39 The alternative proposed here is a consonant-glide sequence (Cw) rather than a series of labialized consonants. Further, there is no evidence to reconstruct a labialized series of consonants at any node within Koman. 2.1.6.2 Chali Uduk consonant distribution and allophones The distribution of Chali Uduk consonants is discussed in Thelwall (1983) and Killian (2015). Killian (2015) also discusses allophonic processes such as intervocalic lenition and word-final delayed release of pulmonic and non-pulmonic consonants. All of the contrastive Chali Uduk consonants can occur in word-initial position. The nasals, lateral, trill, fricatives /s, ʃ/ can occur in all positions and do not exhibit any significant allophonic variation. The glottal fricative /h/ only occurs word-initially and intervocalically. Table 9 contains the distributions and allophonic realizations of the voiceless aspirated and unaspirated plosives in word-initial, intervocalic and word-final positions. The voiceless bilabial, palatal and velar aspirated plosives can lenite to voiceless fricatives intervocalically as well as word-finally. The voiceless unaspirated series does not occur word-finally resulting in complete neutralization of contrast between voiceless aspirated and voiceless unaspirated plosives in this position. 39 Killian additionally notes that his proposed labialized consonants can also occur word-medially in compounds. 61 Table 9 Distribution and allophones of Chali Uduk voiceless plosives #__ V__V __# #__ V__V __# /pʰ/ pʰ pʰ, ɸ pʰ, ɸ /p/ p p – /t ̪h / t ̪h (t ̪h )† t ̪h /t/̪ t ̪ t ̪ – /tʰ/ tʰ tʰ tʰ /t/ t t – /cʰ/ cʰ cʰ, ç cʰ, ç /c/ c c – /kʰ/ kʰ kʰ, x, h kʰ, x, h /k/ k k – † Not attested in my data but attested in Thelwall (1983:332) The distribution and allophones of the voiced, ejective and implosive obstruents are in Table 10. Table 10 Distribution and allophones of Chali Uduk voiced, ejective and implosive obstruents #__ V__V __# #__ V__V __# /ɓ/ ɓ ɓ, b ɓ, ɓ ̚ /ɗ/ ɗ ɗ, d ɗ, ɗ ̚ /pʼ/ pʼ pʼ, b (pʼ)† /b/ b b, β – /t ̪̓ / t ̪̓ t ̪̓ t ̪̓ /d̪/ d̪ d̪ – /tʼ/ tʼ (tʼ)‡ – /d/ d d – /cʼ/ cʼ cʼ, ɟ cʼ, c ̚ /ɟ/ ɟ ɟ, ʒ – /kʼ/ kʼ kʼ kʼ, k,̚ ʔ /g/ g g, ɣ – † Unattested in my data and in Killian (2015); of restricted distribution in Thelwall (1983:332) ‡ Unattested in my data and of “restricted distribution” in Thelwall (1983:332) The implosives can be realized as implosives or as plain voiced stops intervocalically and they can also optionally be realized with a delayed glottal release word-finally. This word-final delayed glottal release also occurs in allophonic realizations of the 62 ejectives.40 The only voiced stops occurring in word-final position are the implosives; the plain voiced stops are restricted to initial and medial positions. One point of discrepancy across analyses, noted earlier, is Killian’s (2015) proposed (partial) series of labialized consonants, which Thelwall (1983) and I analyze as consonant-glide sequences. Killian’s (2015:24) arguments for proposing a labialized series of consonants are due to consonant-tone restrictions, depressor consonant effects, and syllable onset restrictions. Killian observes that one main restriction is that voiced consonants in syllable onsets cannot be followed by a M tone, but voiceless onsets can be followed by a M tone (see §2.1.6.6 for discussion of depressor consonants).41 Killian notes that when in syllable onset, the labiovelar glide /w/ patterns with the voiced obstruents with regard to tone restrictions but the tones following his “labialized” consonants pattern with respect to the onset (e.g. /g/ or /w/ cannot occur with a following M tone, but /k/ and /kʷ/ can occur with a M tone). He argues that if his proposed labialized consonants were a Cw sequence, then the tone rules should be governed by the glide that precedes the vowel not by the initial consonant (e.g. per Killian, /kw/ should not be followed by a M tone since presumably /w/ would count as “voiced”. If we treat Killian’s unitary “labialized” consonant as a Cw sequence instead, Killian’s analysis holds: voiced onsets cannot occur with a M tone and voiceless onsets can. 40 Delayed glottal release of ejectives involves a complete closure and delayed burst. This is a very common allophonic realization for word-final ejectives in the living Koman languages. 41 Beam & Cridland (1970) do have instances of w-initial words with M tones in their dictionary. None of the 22 w-initial words in my database occur with a M tone vowel immediately after /w/. 63 Appealing to economy, Killian (2015) argues that if one were to analyze his unitiary labialized consonants instead as a consonant + glide sequence, it would necessitate positing a syllable of the CGV type; his stance is that Uduk is a language without complex onsets. Further he notes that, “If one takes the approach that these are consonant sequences, the tone restrictions would result in an additional 30 or so rules for each consonant sequence, and justifications would have to be made for the vowel- consonant interactions (Killian 2015:24).” Unfortunately, Killian does not provide examples of any rules or explain in detail what he specifically means by “vowel- consonant interactions”. The current analysis and Thelwall (1983) do not posit a series of labialized consonants but rather treat a consonant-glide (Cw) syllable onset as a sequence of two consonants that can only occur before /a/. Within other Koman languages, Cw sequences restricted to occurring before the vowel /a/ are very common; they have been analyzed not as labialized consonants but rather as Consonant-glide sequences also in Gwama (Hellenthal & Kutsch Lojenga 2011) and Komo (Kutsch-Lojenga & Otero 2011, Otero 2018b). 2.1.6.3 Yabus Uduk consonant inventory The only prior work on Yabus Uduk is a brief note on the contrastive consonants and vowels in Killian (2015:305–306). The following analyses here are based on my own fieldwork and Killian (2015). The Yabus Uduk inventory consists of 34 contrastive consonants including the glottal stop, which is contrastive only in word final position (Table 11). There is a 4-way contrast in manner of articulation for plosives across bilabial, alveolar, palatal and velar 64 places of articulation. Yabus Uduk exhibits contrastive aspiration in voiceless stops but lacks the interdental series of stops that is present in the northern varieties. Table 11 Yabus Uduk contrastive consonant inventory Alveo- Bilabial Alveolar Palatal Velar Glottal palatal unaspirated p t c k ʔ aspirated pʰ tʰ cʰ kʰ ejective pʼ tʼ cʼ kʼ voiced b d ɟ g implosive ɓ ɗ voiceless s ʃ h voiced z ʒ ejective sʼ ʃʼ nasal m n ɲ ŋ lateral l trill/flap r approximant w j With respect to glottalic consonants, Yabus Uduk exhibits a full ejective series across bilabial, alveolar, palatal and velar places of articulation, as well as bilabial and alveolar implosives. Further, there are two fricative ejectives, an alveolar /sʼ/ and a rare palatal fricative ejective /ʃʼ/, the latter of which is articulated with clear frication before the glottal burst. Yabus Uduk also exhibits a voicing contrast among the alveolar fricatives /s, z/ as well as in the (alveo-)palatal fricatives /ʃ, ʒ/. By contrast to the proposed 34-consonant inventory in Table 11, Killian (2015:306) proposes a contrastive inventory of 50 consonants for Yabus Uduk, though he indicates that his inventory is based on a limited amount of research. What differentiates Killian’s inventory from the inventory proposed here are is the lack of a voiced palatal 65 fricative in Killian’s inventory plus his 17 additional labialized consonants /pʷ, pʰʷ, bʷ, ɓʷ, tʼʷ, dʷ, ɗʷ, cʷ, cʰʷ, cʼʷ, ɟʷ, ʃʷ, kʷ, kʰʷ, kʼʷ, gʷ, ŋʷ/. Killian does not provide any evidence for the partial series of labialized consonants nor does he mention whether they are restricted to occurring before the vowel /a/, as he mentions for the Chali variety. 2.1.6.4 Yabus Uduk consonant distribution and allophones This section outlines the distribution of the Yabus Uduk consonant system as well as the allophonic variation in word-initial, intervocalic and word-final positions within the phonological word. The distribution of voiceless aspirated and unaspirated plain stops is seen in Table 12. All of the voiceless aspirated stops, with the exception of the voiceless alveolar stop /tʰ/, can lenite to voiceless fricatives in intervocalic and word- final positions. The five-way contrast in manner of articulation for plosives seen in word-initial position is reduced to a three-way contrast in word-final position; voiceless unaspirated stops can occur in initial and intervocalic positions, but not word-finally. Table 12 Distribution and allophones of Yabus Uduk voiceless plosives #__ V__V __# #__ V__V __# /pʰ/ pʰ pʰ, ɸ pʰ, ɸ /p/ p p – /tʰ/ tʰ tʰ tʰ /t/ t t – /cʰ/ cʰ cʰ, ç cʰ /c/ c c – /kʰ/ kʰ kʰ, x kʰ, x /k/ k k – To illustrate the intervocalic lenition of the voiceless aspirated stops, the data in (15) contain word-final aspirated stops occurring with a vocalic suffix. All of the voiceless 66 aspirated stops lenite to fricatives intervocalically with the exception of the voiceless alveolar stop. (15) a. /tūpʰ/ áhā tūpʰ-á kʼwà [áhā tūɸá kʼwà] 1SG pierce-1SG gourd ‘I pierce a gourd.’ b. /bàtʰ/ áhā bàtʰ-á ʃùm [áhā bàtʰá ʃùm] 1SG hunt-1SG animal ‘I hunt animals.’ c. /dǔrīcʰ/ dǔrīcʰ-ām=pɛ́m [dǔrīçāmpɛḿ] young.people-ASS=1SG.POSS ‘my young people’ d. /ʃʼɔ̄kʰ/ áhā ʃʼɔ́kʰ-á mā [áhā ʃʼɔ́xá mā] 1SG stir-1SG porridge ‘I stir the porridge.’ Table 13 contains the distribution and allophones of the implosives, voiced plosives and the obstruent ejectives. Note that in Yabus Uduk, the implosives are restricted to word-initial position. The ejectives all occur in word-initial position, though they have distinct distributions medially and word-finally. The bilabial ejective is very infrequent in the database, only occurring in three words and only in initial position. The alveolar ejective only occurs in initial or medial position. By contrast, the palatal and velar ejectives occur in initial and final positions. Further, the palatal ejective is realized as a voiced fricative and the velar ejective can also be realized as a voiced stop intervocalically. The voiced plosives are realized as unreleased or with a delayed release 67 word-finally.42 The voiced palatal stop is unattested in word-final position in my database. Table 13 Distribution and allophones of Yabus Uduk implosives, ejectives and voiced plosives #__ V__V __# #__ V__V __# /ɓ/ ɓ – – /ɗ/ ɗ – – /pʼ/ pʼ –† –† /b/ b b b̚ /tʼ/ tʼ tʼ –† /d/ d d d ̚ /cʼ/ cʼ ɟ cʼ /ɟ/ ɟ ɟ – /kʼ/ kʼ kʼ, g kʼ /g/ g g g ̚ /sʼ/ sʼ sʼ, z sʼ /ʃʼ/ ʃʼ ʒ ʃ’ † Unattested in my data though possibly likely to occur. The data in (16) contain word-final alveolar, palatal and velar ejectives inflected immediately with a vowel-initial suffix. The ejectives are realized as voiced stops intervocalically across morpheme boundaries. (16) a. /ʃītʼ/ áhā ʃítʼ-á mɛ ̄ [áhā ʃídá mɛ]̄ 1SG tether-1SG goat ‘I tie up the goat.’ b. /tʰúcʼ/ tʰúcʼ-ām-pɛ́m [tʰúɟāmpɛḿ] work-ASS-1SG.POSS ‘my work’ c. /tʰūkʼ/ áhā tʰūkʼ-á īs [áhā tʰūgá īs] 1SG turn-1SG body ‘I turn around.’ 42 This seen in the numeral múzɛ̀d ‘five’, which can be realized as [múzɛ̀d ̚] or [múzɛ̀ʔ] in Yabus Uduk. Note that it is cognate with múdɛ̪̀ɗ in Chali Uduk. 68 Evidence for the intervocalic voicing of the fricative ejectives is seen in (17). Word final fricative ejectives are realized as voiced fricatives intervocalically across morpheme boundaries. (17) a. /kàsʼ/ kàsʼ-ām-pɛ́m [kàzāmpɛḿ] husband-ASS-1SG.POSS ‘my husband’ b. /bíʃʼ/ bíʃʼ-ām-pɛ́m [bíʒāmpɛḿ] fishhook-ASS-1SG.POSS ‘my fishhook’ The distribution and allophones of the fricatives and sonorants is seen in Table 14. The voiceless fricatives (plain and ejective) occur in all positions, with the exception of the glottal fricative, which does not occur word-finally. The alveolar ejective fricative can also be realized as a plain voiced fricative intervocalically, while the voiceless alveopalatal fricative ejective is realized as a voiced postalveolar fricative. Lastly, the nasals, lateral, trill and glides can occur in all positions and there is no significant allophonic variation. 69 Table 14 Distribution and allophones of Yabus Uduk plain fricatives and sonorants #__ V__V __# #__ V__V __# /s/ s s s /z/ z z – /ʃ/ ʃ ʃ ʃ /ʒ/ ʒ ʒ – /h/ h h – /m/ m m m /r/ r r r /n/ n n n /l/ l l l /ɲ/ ɲ ɲ ɲ /w/ w w w /ŋ/ ŋ ŋ ŋ /j/ j j j 2.1.6.5 Uduk contrastive vowels Stevenson (1944) proposes a 10-vowel system in Chali Uduk consisting of seven peripheral vowels /i, e, ɛ, a, ɔ, o, u/ and three central vowels /ɨ, ə, ɐ/. Excluding the central vowels, Stevenson’s inventory is representative of a current-day seven-vowel system with ATR contrast in the mid-vowels (cf. Casali 2008:503).43 By contrast, Thelwall (1983) proposes a five-vowel phonemic inventory /i, e, a, o, u/ and Killian (2015:32) proposes a distinct five-vowel system /i, ɛ, a, ɔ, u/.44 Taken together, Thelwall and Killian’s inventories combined yield Stevenson’s (1942) inventory minus the central vowels. Killian (2015:32) states that surface forms of /i ɛ a ɔ u/ in Chali Uduk extend into [i, ɪ, ɛ, e, a, ʌ, ɔ, o, ʊ, u].45 43 Recall that the other living Koman languages have a contrastive seven-vowel system /i, ɪ, ɛ, a, ɔ, ʊ, u/. This is the system I reconstruct to Proto-Koman. 44 Note that Killian employs the graphemes and to represent /ɛ/ and /ɔ/, respectively. 45 Killian employs the symbols [ɨ, ʉ] for what I transcribe as [ɪ, ʊ], the [+high, –ATR] vowels. 70 In Yabus Uduk, by contrast, Killian (2015:306) proposes a “tentative” seven-vowel system with ATR contrast as well as three [+ATR] allophones [e, ʌ, o] of /ɛ, a, ɔ/, respectively, though he provides no evidence to support these claims. I find no evidence for a seven-vowel contrastive inventory in Yabus Uduk, though phonetically the Uduk vowel space does seem to extend through the range of the seven vowel systems as discussed for Chali Uduk.46 In Figure 14, I propose a five-vowel contrastive inventory for the Chali and Yabus Uduk varieties. FRONT CENTRAL BACK HIGH i u MID ɛ ɔ LOW a Figure 14 Uduk contrastive vowel inventory The living Uduk varieties have lost phonemic ATR contrast in the high vowels and the reflexes of Proto-Koman (PKMN ) *ɪ and *ʊ in cognate lexemes correspond to synchronic Uduk /i, ɛ/ and /u, ɔ/ respectively. On the surface, reflexes of historical *ɪ can be realized phonetically as [i, ɪ, e] while reflexes of *ʊ could be realized as [u, ʊ, o]. See §3.3 for discussion of the PKMN vowel system. 2.1.6.6 Uduk tone Chali Uduk displays three contrastive level tones as well as contrastive rising and falling contour tones. The only prior studies of Uduk tone is of the Chali variety. Although Stevenson (1942) recognizes tone in the language, he does not investigate it 46 I recognize that a more detailed study of the Yabus Uduk vowels is warranted. 71 fully. Beam & Cridland’s (1970 [1956]) dictionary marks tone on all of the lexical entries. In their dictionary, level tones occur only on a single vowel nucleus while the contour tones, typically rising, are always transcribed with two sequential vowels of the same quality and differing tone heights. Killian’s (2014) tone transcriptions of Beam & Cridland’s lexical entries are very similar with one notable exception discussed further below.47 Killian (2015) is by far the most in-depth treatment of Chali Uduk tone. He investigates lexical and grammatical tone as well as the complex system of tonal aspectual marking in inflected verbs. Killian (2015:40) analyzes four of what he calls “basic” tonemes, L, M, H, LH and three “rare” tonemes, MH, HL and HM. He provides a near-minimal septuplet of these tonemes on single vowel nuclei in monosyllabic words as evidence though the HL and HM entries are Arabic borrowings.48 Further, in Killian’s (2015:45) list of tonemes in polysyllabic words, there are only two examples of HM and MH– both of which are found in names of species of birds. In sum, Killian proposed three contrastive level tonemes (L, M, H), two contrastive rising tonemes (LH, MH) and two contrastive falling tonemes (HL, HM). I did not find evidence for phonemic contrast for LH vs. MH or HL vs. HM as a unitary toneme on a single vowel nucleus. I find only one contrastive R(ising) tone and one contrastive F(alling) toneme. If such fine-grained contour tone distinctions are realized as Killian describes, I suspect 47 I am grateful to Don Killian for providing me with his updated version of the Beam & Cridland dictionary. 48 Killian (2015:40) states that MH, HL and HM “occur only rarely, but are possible in both mono- and disyllabic nominal roots” and “LM and ML contours do not appear in any lexeme roots”. 72 they are phonetic, not phonological.49 Killian (2015) does not mention tone in Yabus Uduk and I analyze this variety also with three contrastive level tones, one contrastive rising and one contrastive falling contour toneme. One of Killian (2015)’s major contributions with regard to the tone in Uduk has largely to do with its interaction with the consonant system. In Chali Uduk, Killian (2015:42) observes the following co-occurrence restrictions for (word-initial) consonants and the tone of the immediately following vowel in monosyllabic and polysyllabic words: (18) a. No syllable beginning with a voiceless onset occurs with a L tone. 1. Vowel-initial roots are included in this category, as they optionally begin with a glottal stop. 2. Implosives also are included in this category. b. No syllable beginning with a voiced onset occurs with a M tone. c. Voiced plosives can only occur with L or LH tones. d. Voiced sonorants can occur with L, LH, or H tones; the sonorant can be either a syllable onset with a following vowel, or it may be a as syllabic TBU. I find the same distributions in Chali Uduk as those observed by Killian, though I have only examined the distribution of the tones in word-initial syllables. In Beam & Cridland (1970), by contrast, all word-initial voiced consonants can occur with a following L, H or rising tone. Lexemes transcribed with contour tones in Beam and Cridland (1970) are rare and the rising contour tone is always transcribed as a MH across two adjacent identical vowels such as in dā̪ápʰán ‘beggar’ and dā̪átú̪ ‘evil eye’ (ibid 1970:56). All of the transcriptions of H tones after voiced stop in Beam and Cridland are rising contour tones in Killian (2014). One question is whether these H 49 By comparison, no other Koman language has two contrasting sets of rising or falling tones on monosyllabic monomorphemes (e.g. LH vs. MH, or HL vs. ML). 73 tones were realized as a level H tone at the time of Beam & Cridland’s fieldwork and have since become a rising contour tone, or whether they were realized as contour tones and just not perceived as such.50 This is the only discrepancy between Beam & Cridland’s (1970) and Killian’s (2014) transcriptions; all of the other observations put forth by Killian in (18) hold for Beam & Cridland’s work. Phonologically, it appears that the voiced plosives (the implosives are not included in this class) and the voiceless plosives are in complementary distribution with regard to the tone on a following vowel. Voiced stops can only occur with a L or LH tone, while voiceless stops can only occur with M or H tone. These observations are employed by Killian as evidence of a historical depressor consonant effect as well as for proposing a series of labialized consonants (see discussion above in §2.1.6.1). Killian (2015:43) argues that pre- or proto-Uduk only displayed two level tones (L and H). He hypothesizes that the voiced consonants (excluding implosives) functioned as depressor consonants and lowered the L tone to an extra L tone which, over time, phonologized into a new tone level, raising the total to the three level tones that he finds synchronically (L, M, H).51 Killian posits a subsequent sound change to explain the LH contour tone seen synchronically after voiced plosives and some sonorants. He argues that, following voiced plosives and sonorants, some H tones underwent L-tone insertion which created a LH contour. By this he suggests that the synchronic LH contour tones were 50 Impressionistically, I found it difficult to hear the rising contour tone following voiced stops at times, though when asked, the speakers would clearly whistle a rising tone. 51 It appears that Ehret (2001) predates Killian’s analysis of Uduk tone. Ehret (2001:60) proposes the following sound change in Uduk: *(C1)v̀(C) > (C1)v̀(C) where C1 = [+voice/-glottal]; > (C1)v̄(C) elsewhere. 74 historically all H tones. Killian’s (2015:43) example is reproduced in (19). He argues that the verb bǔkʰ ‘pull out’ was historically H tone *búk, which underwent L-tone insertion, which gave rise to the synchronic LH tone. While there is no evidence to prove this, it is interesting when examining the data from Beam & Cridland (1970), in which they transcribed H tones after voiced sonorants and plosives, which correspond to LH in Killian (2014) and in my data. (19) The Chali Uduk verb system is particularly revealing with respect to the tone distribution. Killian (2015:44) observes that Chali Uduk employs tonal alternations on the verb root to mark aspect – perfective (PFV) and imperfective (IPFV). He groups the Uduk verbs into five classes based on the behavior of the tone of the root when expressing either aspect. I have reproduced Killian’s (2015:44) tone classes in (20) and I also include the consonants that can occur as the word-initial syllable onset. Class PFV IPFV Consonants in word-initial syllable onset (20) 1 M H voiceless consonant, implosive 3 H M voiceless consonant, implosive 2 L LH voiced plosive, sonorant 4 LH L voiced plosive, sonorant 5 H L sonorant Notice that class 1 and class 3 only occur with a voiceless consonant or implosive syllable onset. Class 2 and class 4 comprise the voiced plosives and sonorants, and class 75 5 has only sonorants. If the LH contour were indeed a historical *H tone, the tone classes could be condensed into four classes, as seen in (21). Classes 4 and 5 could be collapsed into one. Notice that the distribution of the tone classes also patterns with the complementary distribution of the consonant in the syllable onset – Classes 1 and 2 have initial syllables with strictly voiceless consonants and implosives while classes 3 and class 4/5 have voiced consonants. Class PFV IPFV Consonants in word-initial syllable onset (21) 1 M H voiceless consonant, implosive 3 H M voiceless consonant, implosive 2 L *H voiced plosive, sonorant 4, 5 *H L voiced plosive, sonorant Killian’s (2015) hypothesis of a historical depressor consonant effect in a prior stage of Uduk actually extends further back in the history of the Koman languages. Killian’s intuitions are correct though in §3.1.2, I propose that voiced plosives functioned as depressor consonants not only at a Pre-Uduk stage, but more precisely prior to this, in a Proto-Koman stage. 2.1.7 Opo Opo is cluster of language varieties that are spoken along both sides of the border of southwest Ethiopia and South Sudan. The Bilugu and Modin varieties are spoken in Ethiopia, while the Pame, Kigile, Pilakoy and Bikol varieties are spoken in South Sudan. The degree of mutual intelligibility among these varieties has not been studied in detail though the Ethiopian varieties are mutually intelligible. Lastly, many researchers include Dana as a variety of Opo though there has not been any description of Dana 76 prior to this dissertation. I discuss Dana phonology in §2.1.8 and propose that Dana split from Proto-Opo Dana in §4.9. Opo phonology has received very little treatment in the linguistic literature until recently. Bender (1983) provides a very brief sketch, though he does not specify the variety (or varieties) in his study. Lemi (2010) and van Silfhout (2013) outline aspects of the phonology of Ethiopian varieties (presumably Bilugu and/or Modin Opo), and Mellese (2017) provides the most in-depth analysis of Bilugu Opo, though a full detailed study is needed. In the following subsections, I provide an overview of the consonant, vowels and tone of some of the Opo varieties.52 2.1.7.1 Opo consonants The Opo varieties exhibit 26-28 contrastive consonants. This inventory, seen in Table 15, includes the glottal stop and the palatal nasal, which are marginally contrastive. Bilabial, alveolar and velar obstruents contrast in four manners of articulation: voiceless aspirated, voiceless unaspirated, ejective and voiced. The alveopalatal affricates contrast in voiceless, ejective and voiced manners of articulation. Further, all varieties display bilabial and alveolar implosives. All of the varieties contain a voiceless alveolar fricative /s/ and, in my data, only the Pame variety displays a voiced counterpart /z/. The phonological status of the alveopalatal fricative /ʃ/ in the Kigile variety is unresolved. At present it is unknown whether [ʃ] is contrastive or in free variation with the voiceless alveolar fricative [s] in Kigile Opo.53 52 I am grateful to Joshua Smolders who hosted me and who I worked alongside during the Opo fieldwork. Much of the analysis presented here was reviewed and discussed together. 53 The Kigile Opo data was elicited from native Pame Opo speakers who are bilingual in the Bilugu varieties. They have knowledge of the Modin and Kigile varieties, but I was unable to conclude whether there is a phonemic contrast between /s/ and /ʃ/ in the Kigile variety. Smolders (forthcoming) states that 77 Table 15 Opo contrastive consonant inventories Alveo- Bilabial Alveolar Velar Glottal palatal unaspirated p t tʃ k (ʔ) aspirated pʰ tʰ kʰ ejective pʼ tʼ tʃʼ kʼ voiced b d dʒ g implosive ɓ ɗ voiceless s ʃ † h voiced z‡ nasal m n (ɲ) ŋ lateral l trill/flap r approximant w j ‡Only occurs in the Pame variety. †Only occurs in the Kigile variety. The inventory in Table 15 differs slightly from those proposed by previous researchers. Bender’s (1983:265) contrastive inventory of 22 consonants lacks an aspirated series of stops, implosives, and either velar or palatal nasals, though he includes a labiodental fricative.54 Lemi (2010:4), van Silfhout (2013) and Mellese (2017) propose somewhat similar inventories which include contrastive implosives, though they lack a contrastive aspirated series of stops.55 Mellese (2017:20–21) analyzes three contrastive dental stops /t,̪ t ̪̓ , d̪/ in addition to the alveolar series, though I find no in the Bilugu, Modin and Pame varieties, [ʃ] is an allophone of [s] before the [+high, +ATR] vowels /i, u/ but there is no phonological contrast between /ʃ/ and /s/ in these varieties. 54 The labiodental fricative, which is not phonemic, is most likely the allophone [ɸ] of /pʰ/. Bender notes that the implosives, which he transcribes in phonetic brackets, can only occur word-initially; likewise for the velar nasal, which can only occur word-finally. 55 van Silfhout (2013) recognizes [pʰ, kʰ], though she treats them as allophones of [p, k]. 78 evidence of an interdental (or dental) series in Opo.56 The contrastive inventory proposed here is more in line with Smolders (2017, forthcoming), who analyses a contrastive series of aspirated/unaspirated voiceless stops.57 In terms of the (alveo- )palatal obstruents, Bender (1983), van Silfhout (2013), Mellese (2017) and Smolders (forthcoming) analyze them as stops, while Lemi (2010), Smolders (2017) analyze them as alveopalatal affricates. These consonants /tʃ, tʃʼ, dʒ/ are articulated as affricates – i.e., a closure, followed by a burst with significant frication.58 The phonetic realization of these affricates differs significantly from the contrastive palatal stops in Dana and Chali Uduk. In these Dana and Chali Uduk, the palatal obstruent series is articulated further back in the palatal region, much more reminiscent of stops than of affricates. Further, in terms of the phonological system, I do not find any evidence of a contrastive aspirated alveopalatal stop in synchronic Opo.59 Lastly, only Smolders (forthcoming) proposes a phonemic voiceless labiovelar semivowel /ʍ/ (in the Bilugu and Modin varieties), though I analyze it as a consonant-glide sequence /hw/ as it only occurs before /a/ and patterns with other Cw sequences (see (23) below). 56 Note that Dana does indeed exhibit a contrastive interdental series of plosives (see §2.1.8) 57 Neither Smolders (p.c.) nor I have found any evidence for contrastive (or even phonetic) dental or interdental stops in Opo. Dana, the most closely related language/most divergent dialect does indeed exhibit a contrastive interdental series of stops. 58 Joshua Smolders (p.c.), a linguist who is currently working with Opo language speakers in Ethiopia, proposes that the (alveo-)palatal consonants /tʃ, tʃʼ, dʒ/ can be realized as stops or affricates depending upon the speaker. He notes that older speakers realize these consonants as palatal stops while the younger speakers realize them as alveo-palatal affricates. 59 Smolders (forthcoming) indicates that the status of the voiceless aspirated palatal stop /cʰ/ is uncertain, as it has only been observed to be contrastive at morpheme boundaries; he unfortunately does not provide examples. Further, Smolders (2017) does not analyze an aspirated alveopalatal affricate /tʃʰ/ in his Bilugu Opo wordlist. 79 2.1.7.2 Opo consonant distribution and allophones All of the Opo consonants can occur in word-initial position. Further, the fricatives, nasals, lateral, trill/flap and glides can occur in all positions. One exception is the voiceless glottal fricative, which can only occur in word-initial and intervocalic positions. The distribution and allophones in initial, intervocalic and final positions of the plosives, implosives and affricates is seen in Table 16. The aspirated stops can occur in all positions. The voiceless unaspirated stops, voiced stops and the bilabial and velar ejectives can be unreleased word-finally. The implosives do not occur in word-final position. Table 16 Distribution and allophones of Opo plosives, implosives and affricates #__ V__V __# #__ V__V __# /pʰ/ pʰ pʰ pʰ /p/ p p, b p, p̚ /tʰ/ tʰ tʰ tʰ /t/ t t, d t, t̚ /kʰ/ kʰ kʰ kʰ /k/ k k, g k, k ̚ /tʃ/ tʃ tʃ, dʒ tʃ /pʼ/ pʼ pʼ (b*) pʼ, p̚ /b/ b b b̚ /tʼ/ tʼ tʼ, (d*) tʼ /d/ d d d ̚ /kʼ/ kʼ kʼ, (g*) kʼ, k ̚ /g/ g g g ̚ /tʃʼ/ tʃʼ tʃʼ, dʒ tʃʼ /dʒ/ dʒ dʒ dʒ /ɓ/ ɓ b – /ɗ/ ɗ d – * Does not appear in my data, though it is likely to occur. Opo employs reduplication and partial reduplication of verb stems to make distinctions in number of participants and/or number of events (cf. Smolders forthcoming). The behavior of the obstruents in intervocalic position can be viewed in 80 monomorphemic words and across morpheme boundaries in (partially) reduplicated verb stems. The implosive stops are always realized as voiced stops intervocalically in verbal reduplication. The voiceless unaspirated, voiced and ejective stops and affricates can occur in intervocalic position in monomorphemic roots. Across morpheme boundaries in (partially) reduplicated verb stems, the voiceless unaspirated stops and can also surface as voiced stops. Some examples of reduplication in verb stems are seen in (22). (22) pàj ‘be.wide.SG’ pɪ ̄ ‘throw.SG’ /p/ pāpáj ‘be.wide.PL’ pɪb̄ɪ ́ ‘throw.PL’ tɛ́l ‘hammer.SG’ tɪ ̄ ‘tie.SG’ /t/ tɛt̄ɛl̄ ‘hammer.PL’ tɪ̀dɪ ́ ‘tie.PL’ ʃ tʃár ‘think.SG’ tʃɔ́t ‘pierce.SG’ /t / tʃātʃār ‘think.PL’ tʃɔ̄dʒɔ́t ‘pierce.PL’ kál ‘bypass.SG’ káw ‘be.strong.SG’ /k/ kākāl ‘bypass.PL’ kàgáw ‘be.strong.PL’ ʼ tʼʊ́ ‘be.fat.SG’ /t / tʼʊ̄tʼʊ̄ ‘be.fat.PL’ ʃʼ tʃʼɛ́m ‘be.sweet.SG’ tʃʼɛ ̄ ‘shave.SG’ /t / tʃʼɛ́tʃʼɛ́m ‘be.sweet.PL’ tʃʼɛ́dʒɛ́ ‘shave.PL’ ʼ kʼáj ‘be.good.SG’ /k / kʼākʼáj ‘be.good.PL’ In the first column, the initial consonant surfaces identically as when it is in a reduplicated stem. In the second column, the initial consonant surfaces as voiced in the reduplicated stem. I do not have any examples of reduplicated /pʼ/ nor do I have any 81 examples of intervocalic /t, kʼ/ being realized as voiced [d, g] in reduplicated stems or otherwise. Like the other Koman languages, Opo displays consonant+labiovelar glide (Cw) onsets word-initially (or syllable initially) before the vowel /a/. Some examples from Bilugu Opo are given in (23). (23) bwájɛ̀ ‘color’ gwàlɪ ́ ‘bean’ dwàr ‘hunt in group’ à=dʒwàlà ‘potato’ ɗwākʼ ‘earwax’ hwàn ‘four’ kʼwātʼ ‘tick’ ʔwáj ‘dog’ 2.1.7.3 Opo contrastive vowels In my data, the Opo varieties all exhibit a seven-vowel phonemic inventory with Advanced Tongue Root (ATR) contrast in the [+high] vowels. This analysis is consistent with Mellese (2017) and Smolders (2017, forthcoming, p.c.) though other scholars have proposed different systems.60 Bender (1983) proposes a five-vowel /i, e, a, o u/ system, and van Siflhout (2013:23) proposes a seven-vowel /i, ɪ, e, ɛ, a, ɔ, u/ system with ATR contrast that lacks the [+high, –ATR] vowel /ʊ/. Lemi (2010:9) proposes a somewhat unbalanced nine-vowel /i, e, ɨ, ə, a, o, ɔ, ʊ, u/ system with two front vowels, three central vowels, and four back vowels. 60 Smolders (forthcoming) notes that the mid vowels /ɛ, ɔ/ tend to be realized quite high (e.g. as [e, o]). He proposes that the Biligu, Pama and Modin varieties have a set of lower [–ATR] phonemes /ɛ,̞ ɔ̞/ which occur only in words borrowed from Nuer and Anuak. 82 front central back i u [+ATR] [+high] ɪ ʊ [–ATR] ɛ ɔ [–ATR] [–high] a [–ATR] Figure 15 Opo contrastive vowel inventory Some evidence for a seven-vowel contrastive system is seen below in (24). (24) /i/ sī ‘beer /ɪ/ sɪ ́ ‘rain’ /ɛ/ sɛ ̄ ‘tooth’ /a/ sá ‘eat (soft food)’ /ɔ/ sɔ́ ‘run.PL’ /ʊ/ sʊ̄ ‘buy.SG’ /u/ sūg ‘wake’ Opo displays root-controlled ATR harmony among the [+high] vowels. In this process, a [+high, –ATR] affix vowel /ɪ, ʊ/ will assimilate to [+ATR] [i, u] when a [+high, +ATR] vowel /i, u/ occurs in the stem to which it is attached. This process is seen when comparing verbs inflected with the Deictic Directional (DD) suffix /-(w)ʊ́/ DD1, whose vowel is [+high, –ATR], to those occurring with /-(j)á/ DD2, whose vowel /a/ does not undergo or trigger harmony. Consider the data from the Bilugu variety in (25), which contains verb stems with all seven of the contrastive vowels in the bare form as well as inflected with /-(w)ʊ́/ DD1 and /-(j)á/ DD2. Note that the vowel in the suffix /-(w)ʊ́/ DD1 becomes [ú] when the root vowel is [+ATR] /i, u/, as in the entries for ‘fart.SG’ and ‘exit.SG’. 83 (25) stem -(j)á DD2 -(w)ʊ́ DD1 meaning /i/ tīsí [tīsá] [tīsú] ‘fart.SG’ /ɪ/ gɪ̀l [gɪ̀lá] [gɪ̀lʊ́] ‘see.SG’ /ɛ/ kʼɛr̄ɛm̄ [kʼɛr̄ɛm̄á] [kʼɛr̄ɛm̄ʊ́] ‘burp.SG’ /a/ hátʃʼ [hátʃʼá] [hátʃʼʊ́] ‘chew.SG’ /ɔ/ kʼɔ́sɔ́ [kʼɔ́sá] [kʼɔ́sʊ́] ‘kill.SG’ /ʊ/ sʊ̄ [sʊ̄já] [sʊ̄wʊ́] ‘buy.SG’ /u/ pʰűtʼ [pʰűtʼá] [pʰűtʼú] ‘exit.SG’ Lastly, Mellese (2017) proposes three [–high, +ATR] allophones [e, ə, o] of /ɛ, a, ɔ/, respectively. He argues that Opo displays [+ATR] dominant harmony from suffix to stem, though neither myself nor the aforementioned Opo scholars claim similar findings. 2.1.7.4 Opo tone In the suprasegmental domain, one defining feature of the Opo cluster is the presence of four contrastive level tones. While all previous work on Opo recognized a tone system to varying degrees, the most in-depth analysis of tone is Smolders (2017, forthcoming, p.c.).61 Smolders analyzes low, mid, high and extra-high (L, M, H, XH) level tones as well as rising and falling contour tones. Smolders observes that the XH tone level only occurs on /i, a, u/, while all other tones can occur on any vowel. Evidence for the four level tone contrast on monosyllabic monomorphemes is seen in (26). (26) XH pıs̋ ‘disregard’ H pʰɪ ́ ‘drink.SG’ M pɪ ̄ ‘throw.SG’ L pɪ̀ ‘year’ 61 Bender (1983:265) notes “distinctive tone is probable”, Lemi (2010:9) cites “at least low and high tones”, van Silfhout (2013:36) states that Opo has “at least a high and a low tone” but she transcribes three phonetic tone levels in some lexemes, and Mellese (2017:44–45) analyzes three level tones. I agree with Smolders in that Opo displays four phonemic tone levels. 84 Grammatical tone plays a crucial part of the Opo verb system. Opo employs tone alternations in verb stems as one means to make distinctions in the number of participants in an event as well as in the number of events, or pluractionality (cf. Newman 1990). Smolders’ (forthcoming) study details the tonal alternations in verb stems along with other morphophonological means of expressing number in verbs. To illustrate, example (27) contains the singular (SG) and plural/pluractional (PL(U)) verb stem forms for monosyllabic verbs in Bilugu Opo found in my database. Tonal alternations functioning as a means to contrast number in verbs is a common strategy in the languages of the Central Koman branch. (27) SG PL(U) SG PL(U) gloss L H pɔ̀j pɔ́j ‘hide oneself’ L M jàl jāl ‘return’ L R kʰɛ̀m kʰɛm̌ ‘cut (with sickle)’ tʰɔ̀r tʰɔ̌r ‘hit’ M R kɔ̄r kɔ̌r ‘climb’ tīŋ tǐŋ ‘aim at’ M H pʰāj pʰáj ‘fly’ tʃɔ̄pʰ tʃɔ́pʰ ‘squat’ M XH pīt pıt̋ ‘fat, healthy (be)’ tūn tűn ‘raise, bring up’ XH M űpʰ ūpʰ ‘wash oneself’ 2.1.8 Dana In Corfield’s (1938) exploration into the so-called “Koma” people of the Ethio-Sudan borderlands, he comes across several ethnic groups that he called “Koma” that don’t appear to speak the same language. As such, he proposes two groups, the Madin and 85 the Ciita. The latter group he further subdivides into three dialect groups: Kusgilo and Kigelle, whose speech is almost identical, and the Buldiit, who live further west and whose speech shows greater variation. According to Corfield (1938:129) “The people of Buldiit replied that they were Dana, but this turned out to be the name of the area in which they lived.” Corfield’s Buldiit data correspond to my Dana data. The issue of whether Dana constitutes its own language (i.e. whether it is mutually unintelligible with any variety of Opo) is yet to be determined empirically. I have chosen to treat Dana as a distinct language based on the following criteria: The Dana recognize their language to be mutually unintelligible with the Opo varieties (and the Opo do as well). Other Koman groups such as the Komo and Gwama recognize the Dana people and their language (dàná) as being distinct from the Opo people and their language (kɪ́ná). Lastly, the Dana consonant inventory is vastly distinct from the Opo varieties and exhibits an interdental/alveolar contrast in plosives otherwise found only in the Chali variety of Uduk.62 2.1.8.1 Dana contrastive consonant inventory The Dana contrastive consonant inventory is seen in Table 17. 62 Anecdotally, I had a Dana speaker talk in Dana to an Opo speaker who is fluent in the Bilugu and Pame varieties of Opo. The Opo speaker informed me that he could not understand the Dana language and the Dana speaker confirmed this. 86 Table 17 Dana contrastive consonant inventory Inter- Alveo- Bilabial Alveolar Palatal Velar Glottal dental palatal aspirated pʰ t ̪h tʰ kʰ unaspirated p t ̪ t c k (ʔ) ejective pʼ t ̪̓ tʼ cʼ kʼ voiced b d̪ d ɟ g implosive ɓ ɗ voiceless s ʃ h voiced z ejective sʼ nasal m n ŋ lateral l trill/flap r approximant w j Dana displays 35 contrastive consonants, including the glottal stop, which is of marginal status. In the plosive series, Dana exhibits a 4-way contrast in manner of articulation across four places of articulation: bilabial, interdental, alveolar and velar, though a contrastive voiceless aspirated palatal stop /cʰ/ is lacking. Some evidence for consonant contrast in word-initial position is given in (28). 87 (28) C ontrast examples in word-initial position /pʰ, p, pʼ, b, ɓ/ /h, ʔ/ pʰád ‘fly (v)’ hâm ‘yawn’ pád̪á ‘cross (v)’ ʔjâ ‘go, walk’ pʼátʼá ‘white (be)’ bál ‘flirt’ ɓár ‘fall’ /t ̪h , t,̪ t ̪̓ , d̪/ /tʰ, t, tʼ, d, ɗ/ t ̪h ád̪ ‘hunger’ tʰápʰ ‘plaster (v)’ tá̪j ‘sun’ táj ‘break’ t ̪̓ád̪ ‘dull (be)’ tʼámà ‘maybe’ d̪ā ‘go, walk’ dā polar interrogative ɗáb ‘adhere’ /kʰ, k, kʼ, g/ /c, cʼ, g/ kʰàb ‘wind (n)’ kàm ‘find’ cà ‘dig’ kʼámá ‘eat (hard things)’ cʼàcʼ ‘chest’ gàm ‘toss’ ɟámɔ̀ ‘spear type’ /s, ʃ, sʼ, z/ /m, n, ŋ/ sám ‘baboon’ mámà ‘carry on back’ ʃàmʊ́ ‘mother’s brother’ nád̪à ‘cut, divide’ sʼámá ‘blood’ ŋwān ‘four’ zàlít ̪h ‘tendon’ hâm ‘yawn’ /l, r/ /j, w/ lájī ‘miscarry’ já ‘come’ rág ‘stir’ wàr ‘absent (be)’ 2.1.8.2 Dana consonant distribution and allophonic variation The distribution of allophones in initial, intervocalic and final positions of the (non- glottalic) plosive series is seen in Table 18. The five-way contrast in place of articulation across three manners of articulation is maintained in word-initial and intervocalic 88 positions, with the exception of [cʰ], which is an allophone of /c/ in intervocalic and word-final positions. In intervocalic position there are bilabial voiceless aspirated and voiced plosives, and the voiceless palatal stop can lenite to fricatives in fast speech.63 By contrast, in word-final position, voiced stops are realized as voiceless aspirated stops. Table 18 Distribution and allophones of Dana plosives #__ V__V __# #__ V__V __# /pʰ/ pʰ pʰ, ɸ pʰ /p/ p p pʰ /b/ b b, β pʰ /t ̪h / t ̪h t ̪h t ̪h /tʰ/ tʰ tʰ tʰ /t/̪ t ̪ t ̪ t ̪h /t/ t t tʰ /d̪/ d̪ d̪ t ̪h /d/ d d tʰ /kʰ/ kʰ kʰ kʰ /c/ c c cʰ /k/ k k kʰ /ɟ/ ɟ ɟ cʰ /g/ g g, ɣ kʰ Dana displays a productive process of verbal derivation; it is employed to distinguish nominal and/or verbal number. The singular form of a verb root is typically of a (C)VC(V) shape. The derived plural or pluractional form can involve a tone shift with an additional final vowel copied from the root (e.g. tɛ̪́l ‘pound.SG’ vs. tɛ̪̀lɛ́ ‘pound.PL) or partial reduplication (e.g. tɔ̀r ‘hit.SG’ vs. tɔ̀dɔ́r ‘hit.PL’). The data in (28) contain monomorphemic verb roots in the singular form (SG) with a (C)VC shape occurring with all of the plosives except for /t, k/. Each verb also occurs in the 63 In general, the palatal plosives are realized more as affricates – articulated with closure, burst and some frication. The range of frication is gradient, with the palatal ejective displaying the least amount of frication, and the voiceless aspirated palatal stop displaying the most frication. 89 plural/plurational (PL(U)) form, in which the final consonant of the singular form occurs in intervocalic position and the three-way contrast in manner holds. Interestingly, the voiceless unaspirated /p, t, k/ stops in word-final position in verbs are very infrequent in my data. (29) SG PL(U) /pʰ/ úpʰ [úpʰ] ùpʰú [ùpʰú] ‘wash oneself’ /p/ tʰáp [tʰápʰ] tʰàp [tʰàpʰ] ‘kick’ /b/ wàb [wàpʰ] wàbɪ ́ [wàbɪ]́ ‘fan (v)’ /tʰ/ kʰɔ́tʰ [kʰɔ́tʰ] kʰɔ̀tʰɔ́ [kʰɔ̀tʰɔ́] ‘weed, prepare earth’ /t/ – – /d/ búd [bútʰ] búdúd [búdútʰ] ‘tall (be)’ /t ̪h / kát ̪h [kát ̪h ] kàt ̪h ɪ ́ [kàt ̪h ɪ]́ ‘protect /t/̪ sɔ̀t ̪ [sɔ̀t ̪h ] sɔ̀tɔ̪́ [sɔ̀tɔ̪́] ‘remove quickly’ /d̪/ sɛ́d̪ [sɛt́ ̪h ] sɛ̀d̪ɛ́ [sɛ̀d̪ɛ]́ ‘deny, not take’ /c/ kàc [kàcʰ] kɔ̄cá [kɔ̄ɟá] ‘ascend’ /ɟ/ gáɟ [gácʰ] gàɟɪ ́ [gàɟɪ]́ ‘sow seeds’ /kʰ/ dâkʰ [dâkʰ] dàkʰɪ ́ [dàkʰɪ]́ ‘finish’ /k/ – – /g/ tʰág [tʰák ̚ʰ] tʰàgɪ ́ [tʰàgɪ]́ ‘transplant’ The remaining consonants display very few distributional restrictions and limited allophony. The implosives /ɓ, ɗ/ can occur in all three positions though they are phonetically unreleased word-finally. The fricatives, nasals, trill, lateral and glides can occur in all positions with the exception of /h/, which only occurs word-initially. Consonant-glide sequences are very limited in Dana. Consonant+labiovelar glide sequences (Cw) occur mostly after velars and before the vowel /a/, though here are few instances of Cw before /i/. There is only one consonant+glide sequence involving a palatal glide in my data, ʔjâ ‘go, walk.SG’. 90 2.1.8.3 Dana contrastive vowels Dana exhibits a seven-vowel contrastive inventory with an ATR contrast in the high vowels /i, ɪ, ɛ, a, ɔ, ʊ, u/, seen in Figure 16. The only co-occurrence restriction observed in disyllabic monomorphemic roots is that [+high] vowels of opposing ATR values cannot co-occur. All other combinations are attested in disyllabic words. Phonetically, the mid vowels /ɛ, ɔ/ have a wide range of articulation, often sounding higher [e, o], especially when the vowel occurs as a nucleus for high tone, though there is no evidence of any productive ATR harmony system creating [e, o] allophones of /ɛ, ɔ/ in Dana. FRONT CENTRAL BACK i u [+ATR] [+high] ɪ ʊ [–ATR] ɛ ɔ [–ATR] [–high] a [–ATR] Figure 16 Dana contrastive vowel inventory Dana does not exhibit extensive morphology in general, though it does seem to exhibit what appears to be root-controlled ATR harmony among the [+high] vowels. This is seen by comparing the surface forms of the vowels in monosyllabic verb stems when they occur with the Deictic Directional (DD) suffixes: /-ɪ/́ DD1 and /-á/ DD2. Dana verb stems are largely of the (C)VC(V) shape and the final vowel can either be replaced by these morphemes or it can undergo a tonal alternation (H to M). Consider the data in (30) which contains verb stems with each of the seven contrastive vowels occurring with the DD suffixes. Notice how all of the stems retain 91 their initial vowel when occurring with /-á/ DD2. The only exceptions are verbs containing /a/, which becomes /ʊ/ when inflected with DD2. The DD1 morpheme /-ɪ/́ assimilates to [+ATR] [i] when preceded by a [+high, +ATR] vowel /i, u/. This morpheme surfaces as [–ATR] [ɪ] when preceded by a [–ATR] vowel /ɪ, ɛ, ɔ, ʊ/. What is extremely peculiar about the Dana system is that the DD1 morpheme surfaces as [+ATR] [i] in roots containing /a/ as the first vowel.64 (30) stem -á DD2 -ɪ ́DD1 gloss /i/ ísá [ísā] [ísī] ‘ripe (be)’ /ɪ/ pɪd̪́á [pɪd̪́ā] [pɪd̪́ɪ]́ ‘be good’ /ɛ/ ɗɛ̀m [ɗɛm̄á] [ɗɛm̄ɪ]́ ‘cook, stew’ /a/ hásʼ [hʊ́sʼā] [hásʼī] ‘chew’ /ɔ/ tɔ́r [tɔ́rā] [tɔ́rɪ]̄ ‘slaughter’ /ʊ/ kʼʊ́sʼ [kʼʊ́sʼā] [kʼʊ́sʼɪ]̄ ‘dry (be)’ /u/ būd [būdá] [būdí] ‘tall (be)’ 2.1.8.4 Dana tone Dana displays three contrastive tone levels (L, M, H) as well as a falling (F) contour tone on a single vowel nucleus. Evidence for the tonal contrast is seen in (31). (31) H d̪ám ‘thing’ M d̪ā ‘go.SG’ L d̪à ‘what’ F d̪â ‘go.PL’ 64 An explanation for why this occurs is beyond the scope of this paper and warrants further investigation. My data points to the fact that this behavior is not idiosyncratic to this root. The DD1 morpheme surfaces as [+ATR] [-i] on all verbs containing /a/ as the first vowel in my data, even disyllabic roots that contain a final /ɪ/ such as t ̪̓àd̪ɪ ́‘blunt.SG’ is t ̪̓ad̪-í with DD1. 92 Tone plays an important role in the verb system. Dana employs tonal alternations on verb roots and stems to mark nominal number of participants as well as verbal number, or pluractionality (cf. Newman 1990). The general pattern for monosyllabic verb roots is seen in (32). Most common is the alternation between L and H. The M/H and F/LH alternations are less frequent in the database overall.65 (32) SG PL(U) SG PL(U) gloss L H bàkʼ bákʼ ‘hide oneself’ pʰàd̪ pʰád̪ ‘fly’ H L(H) t ̪h áp t ̪h àp ‘plaster (with mud)’ ɔ́g ɔ̀gɔ́ ‘avenge’ M H nāŋ náŋ ‘long, tall (be)’ kʰɔ̄r kʰɔ́r ‘arrogant (be)’ F L.H dâg dàgɪ ́ ‘finish’ kâ kàgá ‘refuse’ 2.2 Morphosyntactic comparanda While this dissertation does not aim to reconstruct Koman syntax, a brief overview of the basic syntactic patterns in the living Koman languages is provided before examining nominal and verbal morphology. Koman languages by and large exhibit AVP/SV word order in declarative main clauses.66 All of the languages with the exception of the Dana-Opo (DAOP) branch employ S/A argument indexing morphology on finite verbs to varying degrees. The robustness of verbal argument indexation in 65 I recognize that this is not an exhaustive study of all of the possible tonal alternations in Dana verbs. 66 Following Comrie (1989), I define A arguments as the most agent-like participants of transitive/ditransitive clauses, P arguments as the most patient-like participant of a transitive verb, and S as the single argument of an intransitive verb. Bender (1994) analyzes all of the Koman languages as having canonical AVP word order in main clause transitive constructions. 93 Gwama and Komo allows for more flexibility in word order. As such, PVA constructions are also possible in these languages. Further, in Gwama and Komo, up to three arguments can be indexed on the verb and a fully finite verb can constitute a clause as seen in (33) and (34). (33) Gwama (Lowland) a. ʊ̄hāj pāt-ná-b~pāt b. pāt-ná-b~pāt 3SG.M touch-3SG.M-3SG.F~RED touch-3SG.M-3SG.F~RED ‘He touches her.’ ‘He touches her.’ (34) Komo a. ʃɛ̀r-kú-r gùbí gá=hàpʼ b. ʃɛ̀r-kú-r-g-ápʼ-ɪ ̄ sweep-DD2-3SG.M house BEN=3SG.F sweep-DD2-APPL-3SG.F.-3N ‘He swept the house for her.’ ‘He swept it for her.’ Killian (2015) describes Chali Uduk as a V2 (verb second) language with a flexible word order with respect to A and P arguments in declarative main clause grammar. Consider the data in (35), in which AVP and PVA constructions are employed. Word order and argument indexation generally follow a nominative-accusative pattern throughout the Koman languages, though Killian (2015) analyzes Chali Uduk as having ergativity in the argument marking on the verb in a particular PVA construction, seen in (35b).67 67 I employ Killian’s (2015) glossing of nominal and verbal morphology in this dissertation. See §2.2.3.1.3 for an overview of Chali Uduk ergativity and Killian (2015) for a detailed description. 94 (35) Chali Uduk a. wàtí̪ d̪ìt ̪ gùɓ b. gùɓ d̪ìt-̪ā wàtí̪ man sweep.PFV house house sweep.PFV-CL1.ERG man ‘The man swept the house.’ ‘The man swept the house.’ (Killian 2015:70–71) The Koman languages all employ auxiliary verbs for negation. In Gwama and the Komo-Uduk (KOUD), the negative auxiliary precedes the lexical verb [AUX V (NP)] and is optionally followed by a direct object. Gwama and Komo index the S/A argument on the negative auxiliary and the lexical verb can optionally occur with a Deictic Directional suffix, as seen below in (36) and (37) respectively. (36) Gwama (Lowland) ʊ̄hāj dɛ̀b-ɛ̀ ʃɛ̀ njã 3SG.M NEG-3SG.M slaughter goat ‘He didn’t slaughter a goat.’ (37) Komo bāʃ-í-r tɛ́r-ʊ́ dɛ ̌ á=gùbí NEG-DD∅-3SG.M carry.SG-DD1 thing LOC=house ‘He didn’t carry the thing into the house.’ (Speaker is inside the house) One unique innovation in the Dana-Opo (DAOP) branch is seen in the order of the direct object and lexical verb in a transitive negative construction. In these languages, the direct object occurs between the negative auxiliary and the lexical verb [AUX NP V], as seen in (38). (38) Bilugu Opo ār=ɗʊ̄kʼʊ́ sɪ ̄ pʰɪ-́ʊ́ 3SG.M=NEG beer drink-DD1 95 ‘He didn’t drink beer (at another location)’. Interestingly, oblique phrases in negative constructions are treated the same as canonical direct objects in Dana and Opo. In (39a), the oblique constituent ádáɟʊ́ ‘in Dajo’ follows the lexical verb in the affirmative, while in (39b), it follows the negative auxiliary and precedes the lexical verb. (39) Dana a. hāpʼ pʰɛ̀ á=dáɟʊ́ b. hāpʼ kʼà á=dáɟʊ́ pʰɛ̀ 3SG.F live LOC=D. 3SG.F NEG LOC=D. live ‘She lived in Dajo.’ ‘She didn’t live in Dajo.’ With respect to other order features, Koman languages employ prepositions (as seen in (39)), possessive constructions follow a possessed–possessor order and demonstratives follow the noun.68 Thus, the family dominantly displays head-initial syntactic constructions. In §2.2.1 I give an overview of Koman independent pronouns, and in §2.2.2 I describe core nominal morphology in each language or language group. This includes the morphology employed in Koman nominal number/gender systems (cf. Smolders forthcoming; Killian 2015, to appear) as well as independent pronominals and demonstratives. In §2.2.3, I discuss core Koman verbal morphology, focusing on bound pronominal argument indexing and deictic directional morphology. 68 See §2.2.2 for descriptions of possessive constructions and demonstratives in the individual languages. 96 2.2.1 Koman independent pronouns Table 19 contains the independent pronominals of the living Koman languages. Language varieties within each cluster have been collapsed and variation in forms within a language group is indicated in the footnotes.69 Table 19 Koman independent personal pronominals Gwama Komo Uduk Dana Opo 1SG gà ākā áhā āgā āgā 2SG ɪk̄ àj ɛ́ āj āj 3SG.M ʊ̄hāj / ʊ̄hāl* hàr áɗi / hádī† hār ʊ̀tà‡ / wàr‡§ 3SG.F hāpʼ hàpʼ áɗi / hádī† hāpʼ ɓā‡ 3NH – hɪ̀n ~ hàn áɗi / hádī† hān nà‡ 1PL.IN mɪn̄ɪ̀ ānà ánā mīnā mìnà 1PL.EX mà āmʊ̀n ámān / âm† mānā mànà 2PL ʊ̄m ʊ̀m úm ʊ̄mā ʊ̄mā 3PL hʊ̄n hʊ̀n únī / húnī† hʊ̄n bìjà‡ * This form occurs in the Highland Gwama variety. † This form occurs in the Yabus Uduk variety. ‡ This form cannot occur independently (cf. Smolders forthcoming, §2.2.2.5.1) § This form only occurs in the Kigile variety. All of the languages distinguish singular and plural number across 1st/2nd/3rd persons as well as clusivity in the 1st person plural. Most of the languages express two to three genders in the third person singular with the exception of Uduk, which employs a single form. The Opo 3rd person forms in Table 19 cannot occur as standalone 69 In Table 19, “Gwama” comprises both Highland and Lowland varieties, “Uduk” comprises the Chali and Yabus varieties, and “Opo” comprises the Bilugu, Modin, Pame and Kigile varieties. 97 pronouns but they serve as base forms for demonstratives and relativizing enclitics (Smolders forthcoming and §2.2.2.5.1 of this dissertation). 2.2.2 Koman nominal morphology Nominal morphology occurs to varying degrees across the Koman languages. Many subclasses of noun words in Koman languages have ‘general number’ meaning that singular or plural is not inherent in the noun form (Corbett 2001). If a Koman language does express number morphologically on the noun, it is always realized by prefixes or proclitics. Further, Koman nominal number morphology is often conflated with class/gender, meaning that some of the languages exhibit portmanteau morphemes expressing both gender and number. Following Corbett (1991:1), who cites Hockett (1958:231), “genders are classes of nouns reflected in the behavior of words”. The criteria here is whether the gender of a given noun can result in agreement or indexation of particular morphosyntactic features on other elements of the grammar. The Koman languages that do exhibit productive nominal gender systems exhibit agreement in the argument indexing morphology in the verbal system and on demonstratives. Bender (1994:45–46) briefly analyzes and compares some the Koman nominal morphology (what he calls “noun formatives”).70 He observes that Koman languages distinguish masculine and feminine gender and that only Uduk also distinguishes what he calls “neuter” gender. Recent papers by Koman scholars have investigated the nominal number and gender systems in Chali Uduk (Killian 2015, to appear), Bilugu 70 Bender’s analysis here assumes a higher Komuz (Koman + Gumuz) genetic node situated within Nilo- Saharan. 98 Opo (Mellese 2017, Smolders forthcoming) and Gwama (Zelealem 2005, Kievet & Roberston 2012, Goldberg et al. 2017) to differing extents. This study draws largely from these works as well as from elicited fieldwork. There are two types of nominal number systems in the family, a “general number” (i.e. unspecified for number) for all nominals, and one that distinguishes number only for human nouns (cf. Killian 2015 and Smolders forthcoming). These systems will be detailed in subsections below. Morphological number encoding is also widely attested in some Koman languages though there are some idiosyncrasies. To illustrate, some examples of a singular/plural number opposition in basic (i.e. underived) human nominals realized through suppletion or via morphological means are seen in (40). Note that most of the languages have suppletive roots for the plural forms of ‘man’ and ‘woman’, while Gwama employs the plural feminine proclitic /ɪ=̄/ (in ‘woman’ vs. ‘women’) and the masculine plural proclitic /mā=/ (in ‘man’ vs. ‘men’). Note also that Komo and to a certain extent Dana employ a cognate masculine singular proclitic /jɛ= ~ jī=/ to mark the singular in ‘man’. (40) Gwama Komo Chali Uduk Dana Bilugu Opo ‘woman’ kɪḱjātà ɓāmít à=ɓɔ́m ɓāpʰā ɓāpʰà ‘women’ ɪ=̄kɪḱjātà ʊ̄p ūpʰ ɔ̄pʰ ɔ̄pʰɔ̄ ‘man’ kíkīzì jī=gwàz wàtí̪ jɛ=̄kàz ʊ̄=kàdʒ ‘men’ mā=kíkīzì gwàz gwàt ̪h kwàz bàdʒ As stated earlier, many of the Koman number proclitics are portmanteau morphemes that also express gender. Further, some researchers have analyzed the number/gender morphemes as functionally extending into other grammatical domains 99 such as definiteness and case marking.71 The number/gender systems for each of the language groups are discussed in the following subsections. 2.2.2.1 Gwama nominal morphology 2.2.2.1.1 Gwama independent and possessive pronouns Table 20 contains the independent and possessive pronominals in the Lowland variety of Gwama. Clusivity in 1PL is distinguished, as is masculine and feminine grammatical gender. Note that the plural possessive pronouns strongly resemble the independent forms while the singular forms vary in their resemblance to the independent pronominals. Table 20 Lowland Gwama independent and possessive pronouns Independent Possessive 1SG gà =nā 2SG ɪk̄ =kɛ́ 3SG.M ʊ̄hāj =dɛ́ 3SG.F hāpʼ =dápʼ 1PL.IN mɪn̄ɪ̀ =mɪn̄ɪ̀ 1PL.EX mà =má 2PL ʊ̄m =kʊ́m 3PL hʊ̄n =bʊ́n The Gwama possessive pronominal enclitics occur within a noun phrase. They connect to a noun with an intervening associative (ASS) enclitic =á and are within a 71 See Killian (2015) for Uduk gender/case marking morphology and Goldberg et al. (2017) for Gwama definite marking on nouns. 100 complex NP headed by that noun as schematized in (41). The complete possessive pronoun paradigm is seen in (42). (41) [N=á=POSS.PRO]NP (42) 1SG bǐt=á=nā ‘my bird’ 2SG bǐt=á=kɛ́ ‘your (sg.) bird’ 3SG.M bǐt=á=dɛ́ ‘his bird’ 3SG.F bǐt=á=dápʼ ‘her bird’ 1PL.IN bǐt=á=mɪn̄ɪ̀ ‘our (inclusive) bird’ 1PL.EX bǐt=á=má ‘our (exclusive) bird’ 2PL bǐt=á=kʊ́m ‘your (pl.) bird’ 3PL bǐt=á=bʊ́n ‘their bird’ While it appears that the associative morpheme is lexicalizing with the possessive enclitics, =á ASS is also fully productive in constructions that do not contain the possessive enclitics. The associative morpheme is employed in constructions to link two nouns as seen in (43a-b) or a noun and a modifying non-finite verb as seen in (43c). (43) a. ʊ̄=pwǎŋ=á=sɛ́nɛ hɔ̀-nɪ ́ zì sɪs̄ɛ ̄ kɪ̀ná DEF.M=road=ASS=one go-3SG.M resemble be.uphill like.that ‘A certain road goes up like that.’ (Goldberg et al. 2017:49) b. ʊ̄hāj hàj ɪ=́kʼwás=á=ɪḿɪ ́ 3SG.M COP.3SG.M LOC=back=ASS=cow ‘He is in front of the cow.’ (Goldberg et al. 2017:75) c. sānzā=á=hú~húnù dà-nɪ-̄à ʃám animal=ASS=REDUP~be.sick NEG-1SG-3N want ‘I don’t want a sick animal.’ (Goldberg et al. 2017:54) 101 2.2.2.1.2 Gwama deictic enclitics Table 21 contains the Lowland Gwama pronominal demonstratives, adapted from Goldberg et al. (2017:44). The pronominal enclitic demonstratives distinguish two genders (masculine and feminine) in the singular and three distances from the origo (proximal, medial and distal). Table 21 Lowland Gwama pronominal demonstrative enclitics (Goldberg et al. 2017:44) PROXIMAL MEDIAL DISTAL M.SG =ɛ̀ =tɛ̀ =nɪ ̌ F.SG =ɔ̀ =tɔ̀ =nǔ PL =nʊ̀n =tʊ̀n =nʊ̀n The demonstrative enclitics can occur in noun phrases headed by a lexical noun, such as in (44), or on another independent pronominal demonstrative root, such as mɪń DEM.RT in (45).72 Note that in (44b), the feminine copula co-occurs with the feminine demonstrative; taken together these morphemes express a biologically feminine bovine. (44) a. bɛ̀rmɛ́tʼá=tɛ̀ b. ɪḿ=tɔ̀ àpʼ=tǔ hat=DEM:MED.M bovine=DEM:MED.F COP.3SG.F=be.tall ‘that hat’ ‘That cow is tall.’ (Goldberg et al. 2017:44) (45) a. ū=sīt mɪń=ɛ̀=gɛ́lɛ̀ b. mā=sītkwì mɪń=nǔn DEF.M=person DEM.RT=DEM:MED.M=gɛ́lɛ̀ M.PL=farmer DEM.RT=PL.DIST ‘that man there ‘Those farmers’ (Goldberg et al. 2017:44) (Goldberg et al. 2017:46) 72 Goldberg et al. (2017) analyze the morpheme gɛ́lɛ̀ as having a discourse-related function that requires further investigation. 102 2.2.2.1.3 Gwama nominal gender/number morphology Gwama has been described as having a productive nominal classification system that distinguishes singular and plural number crosscut with a three-way gender distinction (masculine, feminine and neuter/non-human) in the singular and a two-way distinction (masculine, feminine) in the plural (Hellenthal 2005, Zelealem 2005, Goldberg et al. 2017). The full paradigm of Lowland Gwama gender/number proclitics, summarized from Goldberg at al. (2017), is in Table 22. Note that the feminine singular morpheme pʼā F.SG is no longer productive and only occurs in words that are synchronically monomorphemic. Thus, for singular nouns, feminine gender is “unmarked” in the sense that some nouns have lexicalized the historical feminine gender prefix while masculine gender is overtly coded on the noun. Table 22 Lowland Gwama nominal gender/number proclitics SG PL F (pʼà ~ pʼā)* ɪ=̄ M ʊ̄= mā= N à=† – *Lexicalized, no longer productive † Restricted distibution Goldberg et al. (2017) propose that the proclitic ʊ̄= M.DEF on animate nouns also marks definiteness. An example of a singular male human referent, marked with the masculine singular definite proclitic ʊ̄=, and triggering verbal masculine singular agreement for the A (marked by -nɪ)́ is seen in (46). (46) ʊ̄=wāl hɔ̄-ā kàm-nɪ ́ bʊ̄ M.DEF=boy go-SV find-3SG.M hole ‘The boy went and found a hole.’ (Goldberg et al. 2017:43) 103 Interestingly, the proclitic ʊ̄= M.DEF also appears on inanimate referents, such as kʼɔ́ŋ ‘bottle’ in (47), which is an excerpt from a text. At the second mention of this bottle, it is marked with ʊ̄= M.DEF. Note also that pʼàbɔ̌ŋgɔ́ ‘frog’, which occurs with the lexicalized feminine singular proclitic /pʼa/, also occurs with ʊ̄=. This suggests that the biological masculine gender information in ʊ̄= is being bleached and that the function of ʊ̄= is extending beyond marking of (biological) gender into a discourse function of marking definiteness of animates and inanimate referents. Another possibility is that the proclitic ʊ̄= M.DEF is expanding from marking “masculine” gender to marking “animate” or even some more abstract noun class. This requires further study. (47) ʊ̄=kʼɔ̀ŋ-ā zɛ̀-gɪ ́ ʊ̄=pʼàbɔ̌ŋgɔ́ M.DEF=bottle-REL sit-DD2 M.DEF=frog ‘… the bottle in which the frog sat’ (Goldberg et al. 2017:52) There are also nouns in Gwama that appear to have lexicalized the ʊ̄= M.DEF morpheme. These include kin terms and some animal names. Some examples are given in (48). (48) ʊ̄dɔ́k ‘weaver bird’ ʊ̄bàbá ‘father’ ʊ̄bár ‘heron’ ʊ̄màʃí ‘sister’s husband’ Goldberg et al. (2017) note that the singular neuter (N) proclitic /à=/ has a more limited distribution, only occurring in non-verbal predicate constructions and right- dislocated constructions in which the neuter referent occurs after the verb as a full NP, 104 and is indexed on the verb with the singular neuter object marking. Examples of this are in (49).73 (49) a. mɪn̄=kwànsʼ à=hàntʼà thing=write N=be.big ‘The book is big.’ (Goldberg et al. 2017:39) b. ∅-kǎʃ-à~kǎʃ à=tʼʊ̄tʼʊ̄mʊ̀ 2SG-close-3SG.N~REDUP N=door ‘Close it, the door!’ (Goldberg et al. 2017:40) The feminine singular morpheme /pʼā-/, which is cognate across Koman, is no longer productive in Gwama. This morpheme has been lexicalized in Gwama in at least certain kin terms and animals names. Examples of the lexicalized feminine /pʼā-/ morpheme are in (50). (50) pʼàbɔ̌ŋgɔ́ ‘frog’ pʼāmàʃ ‘wife’ pʼāpɛ̀ŋgɛ̀ ‘frog (sp.)’ With respect to plural marking on nominals, Goldberg et al. (2017) observe that there is a clear distinction in biological gender for human referents and less so for animal referents. Human masculine plural referents occur with mā= M.PL and feminine plural referents occur with ɪ=̄ F.PL, as seen in (51).74 Hellenthal (p.c) notes that the 73 Goldberg et al. (2017) do not comment on whether the singular neuter nominal proclitic also marks definiteness, though a cognate morpheme in Komo does appear to fulfill such a function (see Otero 2015c and §2.2.2.2.3 below). 74 Goldberg et al. (2017:76) note that the expression of definiteness is not as prominent with the plural morphemes as it is with ʊ̄= M.DEF. 105 Highland Gwama speakers are using mā= M.PL as a general plural marker for all genders. (51) sīt ‘person’ mā=sīt ‘people (M.PL)’ kàká ‘grandmother’ ɪ=̄kàká ‘grandmothers’ pʼwākām ‘sister’ ɪ=̄pʼwākām ‘sisters’ When the plural proclitics are employed on non-human animate referents such as animals, the feminine plural appears to mostly express feminine biological gender while the masculine morpheme does not necessarily express masculine biological gender. Some nouns can only occur with either the masculine or feminine proclitic such as in (52a), while other nouns can occur with either plural proclitic as in (52b). It is not clear in (52b) whether either plural proclitic expresses biological gender for the referent ‘frogs’. (52) a. wāŋā ‘chicken’ ɪ=̄wāŋā ‘hens’ b. pʼàbɔ̌ngɔ́ ‘frog’ mā=pʼàbɔ̌ngɔ́ ‘frogs’ ɪ=̄pʼàbɔ̌ngɔ́ 2.2.2.2 Komo nominal morphology Of the living Koman languages, Komo has the least nominal morphology. Biological gender in humans and other animates is indexed on the verb and in demonstratives, but nouns receive very little marking on the whole compared to the other Koman systems. 106 2.2.2.2.1 Komo independent and possessive pronouns Table 23 contains the Ethiopian Komo independent and possessive pronominals. Note that the third person forms distinguish three genders: masculine, feminine and neuter/non-human. Table 23 Ethiopian Komo independent and possessive pronouns Independent Possessive 1SG ākā bám 2SG àj bīní 3SG.M hàr bɪŕ 3SG.F hàpʼ bɪṕʼ 3N hɪ̀n ~ hàn bɪń 1PL.IN ānà bānà 1PL.EX āmʊ̀n bābʊ̀n 2PL ʊ̀m bʊ́m 3PL hʊ̀n bʊ́n Burns (1947:14) states that in Sudanese Komo all of the independent pronominals can optionally occur with what he calls a “pronominal distinguishing suffix” /-ná/ I have also observed the same in Ethiopian Komo though I have not been able to find the difference in meaning between a bare pronoun and one occurring with /-ná/ (e.g. 1SG ākā vs. ākāná). 107 In Komo, the possessive pronouns occur immediately after possessed nouns with no intervening or additional morphology.75 Examples of the possessive pronouns are seen in (53).76 (53) 1SG gùbí bám ‘my house’ 2SG gùbí bīní ‘your (sg.) house’ 3SG.M gùbí bɪŕ ‘his house’ 3SG.F gùbí bɪṕʼ ‘her house’ 1PL.IN gùbí bānà ‘our (ex.) house’ 1PL.EX gùbí bābʊ̀n ‘our (ex.) house’ 2PL gùbí bʊ́m ‘your (pl.) house’ 3PL gùbí bʊ́n ‘their house’ The possessive enclitics appear to be built on the morpheme bā which exists synchronically in Komo and functions as a demonstrative and/or as a possessive morpheme with lexical possessors, as seen in (54a-b), respectively. In Sudanese Komo, Burns (1947:20) transcribes alternate forms of bám 1SG.POSS as disyllabic [bāʔám] with a M.H tone melody or as monosyllabic [bǎm] with a rising tone. This suggests that bám 1SG.POSS could have been historically bimorphemic, where the initial segment was the possessive morpheme, attached to a 1SG pronominal element (i.e. bā=ám POSS=1SG > bám ‘my’). 75 The possessive pronouns are becoming clitics. I have chosen to transcribe them as independent morphemes following Burns (1947) and given the fact that the vowels in these morphemes do not undergo ATR harmony. 76 Burns’ original transcription of bám 1SG.POSS is [běm]. His grapheme corresponds to what I analyze as /ɪ/. Further, for some morphemes within a language, there can be allomorphs in which the vowel /a/ can vary with /ɪ/. One example is the locative proclitic in Komo, which can be realized as /á=/ or /ɪ=́/. 108 (54) a. gùbí=bā b. gùbí bā à=sʊ̀málɛ̀ house=DEM.PROX house POSS ID.SG=S. ‘this house’ ‘Sumale’s house’ 2.2.2.2.2 Komo pronominal demonstratives The Ethiopian Komo pronominal demonstratives are seen in Table 24. This paradigm distinguishes masculine and feminine genders across four distances from the origo: proximal, medial, distal and remote. The remote distance, which is employed when the referent is out of sight (and known to be very distant), is a compound of the medial demonstrative plus the morpheme wɪʃ́ɪt́ɪ̂n, which is of unknown origin or meaning.77 The demonstratives do not distinguish number and can be employed with singular or plural referents. The demonstrative =bā, briefly discussed in the preceding sections, indicates a proximal referent and can be used for all genders. Table 24 Komo pronominal demonstrative enclitics PROXIMAL MEDIAL DISTAL REMOTE M =bā, =nɪ ̌ =dɪ ̌ =dɪťɪ̂n =dɪ̀wɪʃ́ɪt́ɪ̂n F =bā, =nɔ̀ɪ ̄ =dʊ̀ɪ ̄ =dʊ̌tɪ̂n =dʊ̀wɪʃ́ɪt́ɪ̂n Some examples of Komo demonstrative enclitics are seen in (55). Note that ‘goat’ occurs with the proclitic /à=/ ID.SG, which is discussed below in §2.2.2.2.3. 77 This morpheme wɪʃ́ɪt́ɪ̂n appears to contain the verb ʃɪt́ʼ ‘be.far’ in which the alveolar ejective has weakened to a plain stop. Speakers often lengthen the final vowel emphatically. 109 (55) PROXIMAL MEDIAL DISTAL REMOTE gùbí ‘house’ gùbí=bā gùbí=dɪ ̌ gùbí=dɪťɪ̂n gùbí=dɪ̀wɪʃ́ɪt́ɪ̂n gùbí=nɪ ̌ mɛ́ ‘goat’ à=mɛ́=bā à=mɛ́=dʊ̀ɪ ̄ à=mɛ́=dʊ̌tɪ̂n à=mɛ́=dʊ̀wɪʃ́ɪt́ɪ̂n à=mɛ́=nɔ̀ɪ ̄ What appear to be reduced forms of the medial demonstratives can function as relativizing particles. The relativizers dɪ̀ REL.M and dʊ̀ REL.F are employed for singular masculine and feminine referents, respectively.78 The relativizers follow the head noun and agree in gender, as seen in (56). (56) a. à=sʊ̀málɛ, hàr dɪ̀ kʼɛw̄-kú-r dɛ=̌bā ID.SG=S. 3SG.M REL.M write-DD2-3SG.M thing=DEM.PROX ‘Sumale, he is the one who wrote this thing here.’ b. à=ɓāmít dʊ̀ tā-pʼ-g-ák jɛ̀n ID.SG=girl REL.F COP-3SG.F-APPL-1SG oil ‘I (really) desire this woman. (lit. This woman who (is) oil to me)’ 2.2.2.2.3 Komo nominal gender/number morphology Like the other Koman languages, Komo expresses general number for non-human referents, meaning that singular or plural is not specified at the root level, as seen in (57a). Some nouns denoting human referents are lexically specified for singular and plural number, as seen in (57b). 78 Note that the masculine medial demonstrative enclitic has a rising tone and the feminine demonstrative occurs with a final /ɪ/̄, differently from the relativizers. 110 (57) a. kʼáw ‘dog(s)’ gùbí ‘hut(s)’ kʼʊ́p ‘head(s)’ b. ɓāmít ‘girl’ jībā ‘person’ gībā ‘people’ Morphological number specification in Komo can occur by employing nominal enclitics and/or by argument indexation on the verb (§2.2.3.1.2). The only morphemes fulfilling any number and/or gender marking on nominals in Ethiopian Komo are presented in Table 25. Table 25 Ethiopian Komo nominal gender/number proclitics ID.SG à= M.SG.HUM jī= F.SG.HUM ɓā(bɪ)̄= ID.PL gʊ̀= F.SG(HUM) bā=† PL.HUM gī= † Very limited distribution and lexicalized. Can be employed with human and non-human animate referents. Komo has two productive derivational proclitics which derive human singular nominal referents: jī= M.SG.HUM and ɓā(bī)= F.SG.HUM. These morphemes can derive nominals from verbs, as in (58a-b) and (59a-b); or they can be employed to specify a singular member of an ethnic group, as seen in (58c) and (59c). (58) a. jī=kɔ́t b. jī=swǎ-r-à c. jī=kɔ̀mɔ̀ M.SG.HUM=farm(v.) M.SG.HUM=be.old-3SG.M-ADJZ M.SG.HUM=komo ‘male farmer’ ‘old man’ ‘male Komo person’ (59) a. ɓābɪ=̄kɔ́t b. ɓābɪ=̄swǎ-pʼ-à c. ɓābɪ=̄kɔ̀mɔ̀ F.SG.HUM=farm(v.) F.SG.HUM=be.old-3SG.F-ADJZ F.SG.HUM=komo ‘female farmer’ ‘old woman’ ‘female Komo person’ 111 The morpheme gī= PL.HUM can derive human referents from verbs, as in (60a), but it cannot be employed for plural ethnic group members, shown by (60b). For plural human ethnic group referents, the morpheme gʊ̀= PL is employed, as in (60c). (60) a. ʊ̄m tā-m gī=kɔ́t dā b. *gī=kɔ̀mɔ̀ 2PL COP-2PL PL.HUM=farm Q ‘Are you farmers?’ c. gʊ̀=kɔ̀mɔ̀ ID.PL=komo ‘Komo people’ The lexemes jībā ‘person’ and gībā ‘people’ could historically have been composed via a combination of jī= M.SG.HUM and gī= PL.HUM plus bā DEM.SG. Note that modernly, ɓā(bɪ)̄= F.SG.HUM can occur on both of these lexemes to derive singular and plural feminine human referents, as seen in (61). (61) a. ɓābɪ=̄jībā b. ɓābɪ=̄gībā F.SG.HUM=person F.SG.HUM=people ‘woman’ ‘women’ Synchronically, the only obligatory singular gender marking is à= ID.SG or bā= F.SG on male and female human proper names, as seen in (62). This is the only productive instance of the bā= F.SG morpheme. (62) a. à=sʊ̄málɛ b. bā=wálāpʼ ID.SG=S. F.SG=W. ‘Somale’ (man’s name) ‘Walapʼ (woman’s name) 112 The two feminine markers in Ethiopian Komo /bā=/ and /ɓā(bɪ)̄/ appear to be historically related given the similarity in phonological shape, though one form clearly contains a bilabial stop and the other a bilabial implosive. Further, both forms correspond to *ɓa in cognates denoting feminine referents across Koman.79 Interestingly, both morphemes have become lexicalized in a handful of Komo words. The bā= F.SG morpheme has become lexicalized in some kin terms and names for animals, as seen in (63); while /ɓa=/ only occurs lexicalized in some Komo kin terms, as seen in (64).80 (63) bākàká ‘female ego’s husband’s mother’s/father’s sister’ (older than ego) bāʔájí ‘female ego’s husband’s mother’s/father’s sister’ (younger than ego) bāɓáʃ ‘female ego’s husband's brother's wife’ bākʼʊ̄mâk ‘centipede’ bātʼʊ́rsá ‘woodpecker’ (64) ɓàkʊ̀màn ‘sister’ ɓâ ‘daughter’ ɓásʊ̄má ‘father’s sister’ In discourse, the morphemes à= ID.SG and gʊ̀= ID.PL are used for referents that are identifiable or known to the hearer.81 These clitics never appear on new referents introduced in discourse or on referents which the speaker believes that the hearer 79 See §5.2 for a reconstruction of PKMN /*ɓa/. 80 Note that the tones vary in the /ɓa/ morpheme: M, L and F. I do not have a synchronic or a historical explanation for this tonal variation at present. 81 See Otero 2015c for discussion. 113 cannot identify. Komo speakers employ these nominal clitics on core arguments to reactivate previously mentioned referents across a span of discourse and/or to disambiguate the unique identity of referents. An example from a text is seen in (65). In (65a), the first mention of kʼáw ‘dog(s)’ occurs independently, while in the second mention in (65b), it occurs with the gʊ̀= ID.PL proclitic. This proclitic signals to the hearer that these are the particular dogs that were mentioned prior. (65) a. mà gɪm̄ɪʃ́-∅=gɪ̀, ʊ̄l-ɪ-̄n kʼáw bʊ́n SUB get.up.PL-3PL=gɪ̀ call-DD∅-3PL dog 3PL.POSS ‘After they got up, they called their dogs.’ b. gʊ̀=kʼáw pá-ʊ́-n ʃāpʼ-ī-n sìn ID.PL=dog run.PL-DD1-3PL hit-DD∅-3PL tail ‘The dogs ran (towards them) wagging (their) tails.’ The number gender markers in Sudanese Komo described by Burns (1947:12) are presented in Table 26.82 The Sudanese Komo forms in Table 26 largely correspond to the Ethiopian Komo forms seen in Table 25 with two exceptions.83 Burns identifies opi- F.PL.HUM specifically as a feminine plural prefix and ke- PL.HUM as a plural prefix only for human referents. These plural prefixes stand in opposition to go- PL (what I identify 82 To Burns’ glosses, I have added HUM ‘human’ in the glosses of the forms that can only be employed for human referents. 83 I have retained Burn’s (1947) original transcriptions of these morphemes with one exception. Burns employs for the palatal glide which I transcribe with the IPA grapheme to maintain consistency throughout this dissertation. Note that Burn’s vowel transcriptions in which he employs and largely correspond to what I analyze as /ɪ/ and /ʊ/, respectively. 114 as gʊ̀= PL), which can be employed for human referents as well as non-human animate referents.84 Table 26 Sudanese Komo nominal gender/number proclitics (Burns 1947:12) F.SG ba- M.SG.HUM je- F.SG.HUM bapi- PL go- F.PL.HUM opi- PL.HUM ke- Burns’ (1947) prefix opi- F.HUM.PL can be analyzed as consisting of the noun ʊ̄p ‘women’ and the associative morpheme -ɪ ̄ASS. Komo productively employs an associative construction to link two noun roots into a compound noun, as seen in (66). (66) ʃùm-ɪ-̄lǎw animal-ASS-home ‘domestic animal’ 2.2.2.3 Uduk nominal morphology 2.2.2.3.1 Uduk independent and possessive pronouns The independent and possessive pronouns in Chali Uduk and Yabus Uduk are presented in Table 27. The Yabus Uduk 3rd person independent pronouns retain an initial voiceless glottal fricative, which has been lost in Chali Uduk.85 The plural 84 Burns does not have any examples of go- PL on non-animate referents. 85 See §5.1 for a reconstruction of the Koman pronominal system. 115 possessive pronouns display variation in the initial bilabial consonant: Chali has an initial voiced stop while Yabus has an initial voiceless unaspirated stop. Table 27 Chali Uduk and Yabus Uduk independent and possessive pronouns Independent Possessive Chali Yabus Chali Yabus 1SG áhā áhā pɛ́m pɛ́m 2SG ɛ́ ɛ́ pīní pīní 3SG áɗi hádī píɗī pádī 1PL.IN ánā ánā bǎnà pānà 1PL.EX ámān âm bâm pâm 2PL úm úm bǔm púm 3PL únī húnī bǔnī púnī The Uduk possessive pronouns follow the possessed head noun. The pronominal possessive morphemes are in the process of cliticization. One piece of evidence is that nouns ending in vowels insert what Killian (2015:119) calls an “associative nasal” consonant that agrees in place of articulation with the initial consonant of the possessive pronoun. This is seen in (67b) as compared to (67a); the head noun in (67a) ends in a consonant and does not receive the associative nasal. (67) a. mɛ̀ɗ pīní b. à=mùgú-m =pɛ́m hand 1SG.POSS CL2=friend-ASS=1SG.POSS ‘your hand’ ‘my friend’ (data adapted from Killian 2015:119) 116 2.2.2.3.2 Chali Uduk pronominal demonstratives By far the most extensive study of demonstratives in any Koman language is Killian’s (2015:149–166) work on Chali Uduk. The current study only touches upon what Killian (2015:149) refers to as adnominal demonstratives, or those “which are lexically nominal and may either head an NP or modify another noun.” Chali Uduk pronominal demonstratives are polymorphemic. Killian identifies three adnominal submorphemes that serve as bases (or bound roots that cannot occur independently) from which the demonstratives are constructed. Killian’s adnominal (pronominal) demonstrative submorphemes (bases) are seen in Table 28. The two singular forms differ semantically in proximity, while the plural form can be used for all distances. Table 28 Chali Uduk pronominal demonstrative bases (adapted from Killian 2015:150) DISTANCE SG PL Proximal já gwǎ Non-proximal ɟǎ gwǎ According to Killian (2015), these pronominal demonstrative bases combine with two other sets of morphemes to distinguish four distances from the origo: proximal, medial, remote and distal. Killian (2015:151) proposes two classes of complex demonstrative words, presented in Table 29. These demonstrative words are combinations of the pronominal demonstrative bases occurring with what Killian calls “submorphemes that mark distance”, followed by a final “configuration morpheme”. 117 Table 29 Chali Uduk pronominal demonstrative configurations (Adapted from Killian 2015:151–153) Base DEM Class 1 DISTANCE Example Gloss SG/PL affix Class 2 -n jánsān DEM.PROX.SG.1 -nh já / -ān gwǎnsān DEM.PROX.PL.1 PROXIMAL -ns gwǎ -ɛ ̄ jánsɛ ̄ DEM.PROX.SG.2 -h gwǎnsɛ ̄ DEM.PROX.PL.2 -s ɟǎɗān DEM.MED.SG.1 -ɗ ɟǎ -ān gwǎɗān DEM.MED.PL.1 MEDIAL -t gwǎ -ɛ ̄ ɟǎɗɛ ̄ DEM.MED.SG.2 -nt gwǎɗɛ ̄ DEM.MED.PL.2 ɟǎtāːn DEM.REM.SG.1 ɟǎ -t -ān ~ -āːn gwǎtāːn DEM.REM.PL.1 REMOTE gwǎ -tː -ɛ ̄~ -ɛː ɟǎtɛ̄ː DEM.REM.SG.2 gwǎtɛ̄ː DEM.REM.PL.2 ɟǎtːāːn DEM.DIST.SG.1 ɟǎ -t -ān gwǎtːāːn DEM.DIST.PL.1 DISTAL gwǎ -tː -ɛ̄ː ɟǎtːɛ̄ː DEM.DIST.SG.2 gwǎtːɛ̄ː DEM.DIST.PL.2 The labels in first column of Table 29 indicate the distance from the origo expressed by the entire demonstrative word. The second column contains the the SG/PL base forms presented in Table 28. These base forms can combine with the morphemes in the third column labeled “DEM affix”.86 The fourth column contains the “configuration” morphemes for the two configuration classes -ān Class 1 and -ɛ ̄Class 2. Lastly, examples of demonstrative words and glosses are in the final two columns. 86 The “DEM affix” column contains the forms in Killian’s (2015:151) Table 7.2 titled “Submorphemes marking distance distinctions”. 118 The complex demonstrative words are sensitive to distances from the speaker as well as from the addressee. Killian (2015:150) states that “Class 1” demonstratives are employed when the speaker and addressee are equidistant from the referent, while “Class 2” demonstratives are employed when the degrees of separation from the speaker, addressee and referent vary. Further, Killian (2015:154) analyzes a set of anaphoric adnominal demonstratives employed for movement away from a deictic center as well as for anaphoric reference in discourse. The basic forms are jín DEM.ANAPH.SG and gǔn DEM.ANAPH.PL. Killian (2015:95) also discusses two “definite generic nouns” which are “used to replace other nouns known or understood by context”. The forms are jì DEF.GEN.SG and kū DEF.GEN.PL. Killian states that they “resemble pronominal anaphoric demonstratives but are not lexically related.” 2.2.2.3.3 Uduk gender/number morphology Like the other Koman languages, Chali and Yabus Uduk also display general number for non-human referents, though for some nouns denoting human referents, number can be specified either by lexical suppletion or via the proclitic ī= PL. Some examples of lexical suppletion and morphological plural marking in Chali Uduk are in (68). (68) a. wàtí̪ ‘man’ kʼwàní ‘men’ b. kúm ‘mother’ ī=kúm ‘mothers’ One of the most distinguishing features of the Chali Uduk and Bonya Uduk varieties are complex systems of what Killian (2015) calls nominal “gender”. In these gender 119 systems, distinct classes of nouns trigger distinct morphological agreement patterns in other areas of the grammar. The Chali Uduk and Bonya Uduk systems are entirely different from the gender systems in all other Koman languages primarily given that the Chali and Bonya systems are not based in biological gender. I only provide a brief overview of the Chali Uduk gender system and invite the reader to consult Killian (2015, to appear) for details. Killian (2015:67) argues that all Chali Uduk nouns fall into two classes: Class 1 and Class 2. Further, he analyzes Chali Uduk gender agreement as also being conflated with case marking on nouns in transitive declarative clauses. Further, the class of a postverbal noun has an effect on whether the verb takes argument indexing morphology. In terms of morphological marking on the noun, class 1 nouns are always unmarked while class 2 nouns take a proclitic à= CL2 in isolation and/or as a clause- initial (preverbal) noun. Postverbal class 2 nouns can trigger two types of agreement patterns depending on whether they are A or P arguments. To illustrate, in (69)a, an AVP construction, the class 2 noun kʼá ‘dog’ is the preverbal A argument (marked with the à= CL2 proclitic) and ɟɛ̀ ‘elephant’ is a class 1 noun (unmarked) functioning as the postverbal P argument.87 In (69)b, a PVA construction, the postverbal A argument kʼá ‘dog’ occurs with the proclitic mā= CL2.ERG, which Killian (2015) analyzes as the ergative case marker for class 2 nouns. We see two distinct markers of class 2 nouns in these constructions. In (69)c, a AVP construction, class 2 kʼá ‘dog’ is the postverbal P argument and occurs with ā= CL2.ACC, 87 Class 1 nouns are unmarked in clause-initial (preverbal) position as well as in isolation. 120 which Killian analyzes as an accusative case marker for class 2 nouns.88 Note also that in this construction, the verb indexes the subject while the verb is unmarked in (69)a-b. (69) Bonya Uduk A V P P V A a. à=kʼá kʼɔ̄ʃ ɟɛ̀ b. ɟɛ̀ kʼɔ̄ʃ mā=kʼá CL2=dog kill elephant elephant kill CL2.ERG=dog ‘The dog kills an elephant.’ ‘The dog kills an elephant.’ A V P c. ɟɛ̀ kʼɔ̄ʃ -ɔ́ɗ=ā=kʼá elephant kill-3SG=CL2=dog ‘The elephant kills a dog.’ We have seen that the Chali Uduk gender system is sensitive to the class of noun serving as A or P argument as well s the position of the noun with respect to the verb. Gender agreement is seen on the noun and it also impacts whether the verb indexes the A argument. The Chali and Bonya Uduk gender systems have many other agreement targets which are well beyond the scope of this dissertation. Killian (2015) notes that the Yabus Uduk variety does not exhibit the gender system found in the Chali and Bonya varieties, which I also confirm. The unique gender system in Chali and Bonya Uduk appears to be a significant Proto-Chali-Bonya-Uduk innovation. 88 This morpheme is phonologically bound to the preceding verb. 121 2.2.2.4 Dana nominal morphology 2.2.2.4.1 Dana independent and possessive pronouns The independent and possessive pronouns in Dana are presented in Table 30. Dana, like Komo, displays three genders in pronouns: masculine, feminine and neuter/non- human. Unlike the Opo cluster, Dana has a set of independent third person pronouns, which are cognate across the family (§2.2.2.5). The possessive pronouns, with the exception of 1SG and 2SG, are cliticized forms of the independent pronouns. Table 30 Dana independent, possessive and bound pronominals Independent Possessive 1SG āgā =ɪńà 2SG āj =mīn 3SG.M hār =ɪr̄ 3SG.F hāpʼ =ɪp̄ʼ 3N hān =ɪn̄ 1PL.IN mīnā =mīnā 1PL.EX mānā =mánā 2PL ʊ̄mā =mʊ̄ 3PL hʊ̄n =mʊ̄n Examples of possessive pronominal enclitics in data from texts are in (70)-(71). Note that an epenthetic palatal glide [j] is inserted between the final vowel of gʊ́pʰà ‘basket’ and the possessive enclitic =ɪr̄ 3SG.M.POSS in (70). (70) jɛ̀=tɔ̪́n jʊ́ pún gʊ́pʰà=jɪr̄ nɪ̀=sīs M.SG=be.big come.SG search basket=3SG.M.POSS REL=be.lost.SG ‘The elder came to look for his basket that was lost.’ 122 (71) nā hān d̪ɔ̀ d̪ɪ̀ nāŋ ɛ̀s=ɪn̄=d̪à COP 3N thing REL do.SG body=3N.POSS=DEM.MED ‘That’s what happened.’ (Lit. ‘It is the thing that happened).’ 2.2.2.4.2 Dana pronominal demonstratives A preliminary analysis of Dana pronominal demonstrative enclitics is presented in Table 31. Dana does not distinguish number or gender in the demonstratives. Dana demonstrative enclitics express three distances from the origo and have two forms for the proximal.89 Table 31 Dana demonstrative enclitics PROXIMAL MEDIAL DISTAL =nɪ̀ =d̪à =dɪn̄tɪn̄ =īnà Some elicited examples of Dana pronominal demonstratives are seen in (72). An example from a text of the medial demonstrative =d̪à DEM.MED is above in (71) and a proximal demonstrative =nɪ̀ DEM.PROX is seen in (73). The second proximal demonstrative =īnà DEM.PROX is seen in (74).90 (72) nā búk=nɪ̀ b. nā jɛ̀=tɔ̪́n jɛ̀=nɛ̀ COP book=DEM.PROX COP M.SG=be.big M.SG=DEM.PROX ‘This is a book.’ ‘This is an old man.’ 89 This is by no means an exhaustive study of Dana pronominals. These forms were elicited in Komo with Komo speakers who spoke Dana. Data from a Dana text also corroborates the proximal and medial forms. 90 At present it is unclear what the difference is between the two proximal demonstratives. 123 (73) nā jɛ̀tɔ̪́n=nɪ̀ tʰījātʰ COP M.SG=be.big=DEM.PROX be.old.SG ‘This elder is old.’ (74) hʊ̄n ɓʊ́-wà-gár d̪ō=ɪn̄à á=mà kɛ̀ŋgɛ̀l=ɪr̄ 3PL put-DD2-APPL:3SG.M thing=DEM LOC=on bicycle=3SG.M.POSS ‘They put these things on his bicycle for him.’ It also appears that the proximal demonstrative =nɪ̀ DEM.PROX can be employed as a relativizer. Evidence from a text excerpt is in (75). Note that what appears to be the proximal demonstrative =nɪ̀ DEM.PROX employed as a relativizer occurs with a rising tone as compared to the pronominal demonstrative enclitic which has a low tone. (75) hān kɔ̀t jɛ̀tɔ̪́n ɗɛ̂ nɪ ̌ pʼɔ́t ̪h à máŋgà 3N have.SG M.SG=be.big one REL pick.PLU mango ‘There is an old man who is picking mangos.’ 2.2.2.4.3 Dana nominal number/gender morphology The Dana number/gender derivational proclitics are in Table 32. The proclitics distinguish masculine and feminine biological gender only in the singular. The plural proclitic can be employed for both genders. It appears that these proclitics are also employed derivationally to derive human referent nominals (though this requires further investigation). 124 Table 32 Dana nominal gender/number proclitics M.SG jɛ̀= F.SG jʊ̀= PL kɛ=̄ The Dana proclitics can derive nouns from verb roots, as seen in (76) with the verb root kʰɔ́t ̪h ‘farm’. (76) a. hār nā jɛ̀=kʰɔ́t ̪h ʃɔ̄ b. hāpʼ nā jʊ̀=kʰɔ́t ̪h ʃɔ̄ 3PL COP PL=farm grass 3SG.F COP F.SG=farm grass ‘He is a farmer.’ ‘She is a farmer.’ c. hʊ̄n nā kɛ=̄kʰɔ́t ̪h ʃɔ̄ 3PL COP PL=farm grass ‘They are farmers.’ Some Dana lexemes appear to contain a lexicalized prefixal element /à-/, a nominal formative which is found across Koman.91 Dana lexemes containing the lexicalized /à-/ element include animal and plant names, kin terms and inanimate objects. Some examples are seen in (77). 91 Cognate /à-/ prefixes are productive in Chali and Bonya Uduk (cf. §2.2.2.3.3) and are semi-productive in Komo (cf. §2.2.2.2.3). Further, the /à-/ morpheme is lexicalized in kin terms, and in plant and animal terms in all of the Koman languages with the exception of Gwama. 125 (77) àɓɔ̄ŋkʼɔ́ʔ ‘frog’ àmáɗɔ̄ ‘nephew/niece’ àkīʃ ‘antelope’ àbʊ́râŋ ‘cloth’ àdwákʰ ‘weaver bird’ àdɔ̀j ‘moon’ àkɪĺ ‘cattle egret’ In my database, there are also a few Dana lexemes that contain what appears to be a lexicalized feminine prefix /ɓā-/, which is also cognate across Koman. The Dana lexemes containing /ɓā-/ are mostly female kin terms and a few bird species, as seen in (78). (78) ɓāgʊ̂lːā ‘kite bird’ ɓāmʊ́ ‘sister’ ɓājírɔ́ŋ ‘bee-eater bird’ ɓāsɔ̂m ‘father’s sister’ ɓākɪl̄ɔ̄ʔ ‘fish eagle’ ɓākʊ̀m ‘mother’s sister’ 2.2.2.5 Opo nominal morphology 2.2.2.5.1 Opo independent and possessive pronouns The independent pronouns from the four Opo varieties employed in this study are presented in Table 33. The Opo independent pronouns distinguish three genders in the singular and clusivity in the 1st person plural.92 All of the pronouns are identical across the varieties with the exception of the 3rd person forms, which vary.93 Note that the 3rd 92 The 3N category subsumes non-human animate and inanimate referents. This category is 3NH (non- human) in Smolders (forthcoming). 93 Mellese (2017:55) also identifies muna 3PL, a distinct 3rd person plural independent pronoun that I have not encountered. 126 person forms in Table 33 are bound roots that cannot occur independently and must combine with distinct modifiers, such as demonstrative enclitics; these are discussed in below in §2.2.2.5.2.94 Table 33 Independent pronouns in four Opo varieties Bilugu Modin Pame Kigile 1SG āgā āgā āgā āgā 2SG āj āj āj āj 3SG.M ʊ̀tà† ʊ̀tà† wàr† ʊ̀† 3SG.F ɓā† ɓā† ɓā† ɓā† 3N nà† / hà† ní † nà† nà† 1PL.IN mìnà mìnà mìnà mìnà 1PL.EX mànà mànà mànà mànà 2PL ʊ̄mā ʊ̄mā ʊ̄mā ʊ̄mā 3PL bìjà† bìjà† bìjà† bìjà† † This form cannot occur independently. The Opo possessive pronominal enclitics are presented in Table 34. The possessive pronominals cliticize to the right of a noun. Note that the 1st and second plural possessive enclitics are cliticized versions of the independent pronominals with tonal modifications. The possessive enclitics all exhibit H or XH tones, while the independent pronouns have L or M tone. Note also that the 3rd person possessive forms are cognate with Koman independent pronouns outside of the Opo branch.95 94 Smolders (forthcoming) also shows how these pronominal bound morphemes combine with an associative suffix and a relativizing enclitic. 95 See §5.1 for reconstruction of Koman pronominals. 127 Table 34 Possessive pronominal enclitics in four Opo varieties Bilugu Modin Pame Kigile 1SG =(ɪ)́má =(ɪ)́má =(ɪ)́má =ɪḿá 2SG =mɪńɪ ́ =mɪńɪ ́ =mɪn̋ =mɪn̋ 3SG.M =ɪŕ =ɪŕ =ɪŕ =ɪŕ 3SG.F =ɪb́ =ɪb́ =ɪb́ =ɪb́ 3N =ɪń =ɪń =hɪń =jɪń 1PL.IN =mın̋á =mın̋á =mın̋á =mın̋á 1PL.EX =máná =máná =máná =máná 2PL =mwá =mwá =mwá –† 3PL =mʊ́n =mʊ́n =mʊ́n =mʊ́n † This form failed to be elicited. Some examples of possessed nouns are in (79). Note that the initial vowel of the possessive enclitic /ɪ/́ can either elide following a high vowel as in (79b) or it can coalesce to [ɛ] following /a/ as in (79c). (79) Bilugu Opo a. àdwa̋kʰ=ɪḿá b. àhű=má weaver=1SG.POSS fish.sp=1SG.POSS ‘my weaver (bird)’ ‘my fish (sp.)’ c. dʒùgà=ɪŕ [dʒùgɛŕ] name.SG=3SG.M.POSS ‘his name’ 2.2.2.5.2 Opo pronominal demonstratives Table 35 contains the Opo demonstrative enclitics, adapted from Smolders’ (forthcoming) Bilugu Opo data combined with some data from the three other varieties I 128 elicited for this study. Bilugu Opo distinguishes three distances from the origo: proximal, medial and distal.96 Neither gender nor number are distinguished in the demonstratives. Table 35 Bilugu Opo demonstrative enclitics PROXIMAL MEDIAL DISTAL Bilugu =ɪńɔ̄ =ɪńī =ɪńī tīrı ̋ =ínī Modin =ínɔ̄ =íntīn‡ – =ínī Pame =ínɔ̄ =ɪj́ɛ†̄ – Kigile – =ɪj́ɛ†̄ – – Unable to elcit this form. † This form is only employed to form 3SG.M, 3SG.F, and 3PL independent pronouns. ‡ This form is only employed to form 3SG.F and 3PL independent pronouns. Table 36 contains the 3rd person demonstrative pronouns, which are formed from the pronominal bound roots seen in Table 33 and the demonstrative enclitics seen in Table 35. Many of the forms have undergone vowel coalescence at morpheme boundaries. For instance, roots ending in /a/ followed by enclitics beginning with /ɪ/ will coalesce to [ɛ]. This is seen in Bilugu ɓɛ́nɔ̄ 3SG.F.PROX which can be decomposed into /ɓā + ɪńɔ̄/ as seen in Modin Opo ɓáɪńɔ̄ 3SG.F.PROX.97 Some of the varieties exhibit 96 It is very likely that the remaining Opo varieties can also distinguish three distances though this remains to be investigated. 97 See Smolders (forthcoming) for further details in Bilugu Opo. 129 two distinct enclitics employed for different genders or persons. For instance, in Modin Opo, the medial demonstrative enclitic =ínī MED.1 is employed for 3SG.M and for 3SG.N, while the medial demonstrative enclitic =íntīn MED.2 is employed for 3SG.F and 3SG.N.98 Table 36 Opo third person demonstrative pronouns PERSON variety PROXIMAL MEDIAL Bilugu ʊ̀tɛ́nɔ̄ ʊ̀tɛ́nī Modin ʊ̀tínɔ̄ ʊ̀tínī 3SG.M Pame wàrɛ́nɔ̄ wàrɛ́ɛ ̄ Kigile – ʊ̄ɪj́ɛ ̄ Bilugu ɓɛ́nɔ̄ ɓɛ́nī 3SG.F Modin ɓínɔ̄ ɓājɛ́ntīn Pame ɓáɪńɔ̄ ɓáɪj́ɛ ̄ nɛ́nɔ̄ nɛ́nī Bilugu hɛ́nɔ̄ hɛ́nī 3N Modin nínɔ̄ nínī Pame nàɪńɔ̄ nàɪɪ́ ̄ Bilugu bìjɛ́nɔ̄ bìjɛ́nɪ ̄ Pame bìjɛ́nɔ̄ bìjɛ́nɪ ̄ 3PL Modin bìínɔ̄ bìíntīn Kigile – bìíjɛ ̄ 2.2.2.5.3 Opo nominal number/gender morphology The Opo varieties have the most productive number/gender morphological system of any of the Koman languages. Smolders (forthcoming) is an in-depth study of number in Bilugu Opo, though he also discusses number in other Opo varieties. Smolders analyzes two distinct number encoding systems at the lexical level: human referents are specified for singular and in some cases plural number (and there are morphological 98 The MED.1 and MED.2 glosses are subsumed under MEDIAL in Table 36. 130 means to specify number), while non-human referents are by default general number. Smolders also describes a productive system of nominal gender/number morphology for human referents in Bilugu Opo, which is presented in Table 37.99 Note that there are distinct forms for masculine and feminine in the singular while the plural proclitic is employed for both genders.100 Table 37 Bilugu Opo human nominal gender/number proclitics (Smolders forthcoming) SG PL M ʊ̀= bɪ̀= F ɓā= ʊ̄-† † Only occurs in kin terms. Some examples of number/gender morphology employed to derive human referent nominals from verbs are in (80). Note that the plural number morpheme must occur with the plural form of the verb root tʃɛ̀gɛ̀ ‘be.big.PL’ and the singular proclitics must occur with the singular form of the verb. (80) a. tɔ́n ‘be big (SG.)’ ʊ̀=tɔ́n ‘male elder’ ɓā=tɔ́n ‘female elder’ 99 Smolders describes these morphemes as prefixes though they can occur at the phrase level. As such, I describe them as clitics. 100 Smolders (forthcoming:12) states that these morphemes are “clearly reduced forms of the third person pronoun bases” though he discounts the possibility that some of the bases could have been historically bi- morphemic. He also does not explain the synchronic distinction in ATR of the high front vowels in 3PL: the pronominal base bìjà contains [+ATR] /i/ while the proclitic bɪ̀= contains [–ATR] /ɪ/. 131 b. tʃɛ̀gɛ̀ ‘be big (PL.)’ bɪ̀=tʃɛ̀gɛ̀ ‘elders’ Outside of the system described above, the Opo kin system contains one productive plural morpheme ʊ̄- PL, which can be employed on masculine or feminine referents as in (81).101 Note that this morpheme only differs in tone from masculine singular morpheme ʊ̀= M.SG. (81) a. hàm ‘brother’ ʊ̄-hàm ‘brothers’ b. ɓām ‘sister’ ʊ̄-hàm ‘sisters’ This concludes the presentation of nominal and NP-related grammatical morphology. In the next section, I turn to Koman verb morphology. 2.2.3 Koman verb morphology Two salient morphological features of Koman verbs are argument indexing by way of bound pronominal suffixes and deictic directional verb suffixes. Another significant feature of the verb in the languages of the Central Koman branch (i.e. all languages except Gwama) is the expression of nominal and verbal number in the verb. “Nominal number” here specifically refers to participant number while “verbal number” refers to the number of events, or pluractionality (cf. Newman 1990).102 101 Whether this morpheme is cognate with Komo gʊ̀= PL remains to be determined. 102 The marking of number in the verb appears to have been a significant Proto-Central Koman innovation. The strategies observed to mark number in the verb are suppletion, partial reduplication of the verb stem, tonal alternation and combinations therein. Opo and Dana by far exhibit the most productive processes for marking number in the verb, followed by Komo. Uduk exhibits some suppletive roots that express participant number, and Uduk is the only Koman language that uses tonal alternations in the verb root to make aspectual distinctions (Killian 2015). 132 All of the Koman languages have paradigms of Deictic Directional (DD) verb morphology which, depending on the semantic profile of the root to which they are attached, can express a rage of functions including direction of motion, associated motion, and exchoative aspect (Otero 2018, accepted). The bound pronominal argument indexing morphology is discussed in §2.2.3.1.2 and the deictic directional systems are discussed in §2.2.3.2. 2.2.3.1 Koman argument indexing verb morphology Languages with robust bound pronominal argument indexing morphology, such as Komo and Gwama, obligatory index S/A arguments on finite verbs in declarative main clause grammar. The Uduk varieties exhibit argument indexing on the verb, though not as extensively as in Gwama and Komo. The Dana-Opo branch displays the least argument indexing of all of the Koman languages. However, this branch is currently innovating argument indexing verb morphology. The following subsections describe argument indexing morphology in the individual Koman languages. 2.2.3.1.1 Gwama argument indexing verb morphology The Gwama varieties exhibit robust argument indexing morphology on the verb. Table 38 contains the Lowland Gwama bound pronominal suffixes alongside the independent pronouns for comparison. In many instances, the bound forms appear to have cliticized from the independent pronouns coupled with phonological erosion. The Gwama bound pronominal paradigms follow a nominative-accusative pattern in declarative main clauses: Set I indexes S/A arguments and Set II indexes P arguments (J. 133 Goldberg 2018).103 On finite verbs in main clauses, Gwama must index S/A arguments and can optionally index P. Table 38 Lowland Gwama independent and bound pronominals Independent Set I (S/A) Set II (P) 1SG gà -nɪ ̄ -gà 2SG ɪk̄ -gɪ ́ -ɪ̀ 3SG.M ʊ̄hāj -nɪ ́ -ɛ̀ 3SG.F hāpʼ -á /-bɪ ́† -àpʼ 3N – – -à 1PL.IN mɪn̄ɪ̀ -nɪ̀ -nɪ̀ 1PL.EX mà -mɪ ̄ -mà 2PL ʊ̄m -mɪ ́ -ʊ̀m 3PL hʊ̄n -bɪ ́ -ʊ̀n † The variation in form depends on the preceding deictic dirtectional morpheme (cf. Joelle Goldberg 2018) We noted earlier (§2.2.2.1.3) that the biological gender of human referents is indexed on nouns, verbs and demonstratives in Lowland Gwama (Goldberg et al. 2017:39). Verbal indexing morphology also agrees in gender with human referents that function as core arguments, as shown in (82). Note that the verb indexes 3SG.M when the subject is male in (82a), and 3SG.F when the subject is female in (82b). (82) kíkīzì gʊ̀s-nɪ~́gʊ̀s b. kíkjātà gʊ̀s-á~gʊ̀s a. man run-3SG.M~RED woman run-3SG.F~RED ‘A man runs.’ ‘A woman runs.’ (Goldberg et al. 2017:39) 103 Joelle Goldberg (2018) labels the bound pronominal sets “Set I” and “Set II”. She observes idiosyncrasies with the person marking paradigm, specifically that negative auxiliaries index S arguments with P (Set II) morphemes. 134 Goldberg et al. (2017:40) also note that some non-human animate referents can trigger both masculine and feminine gender verb indexation in the singular, as seen in (83). Note that none of the referents in (82) and (83) carry nominal number morphology, and all are translated as indefinite. (83) a. kʼɪḱʼɪʃ́ kɔ̄-nɪ~́kɔ̄ b. kʼɪḱʼɪʃ́ gʊ̀s-á~gʊ̀s tortoise say-3SG.M~RED tortoise run-3SG.F~RED ‘A tortoise said...’ ‘A tortoise runs.’ (Goldberg et al. 2017:40) Given the richness of Gwama argument indexing morphology, a fully finite verb can constitute a clause. A full paradigm for the transitive verb kʊ́p ‘hit’ indexing a 3PL A argument with the Set I suffix -bɪ ́3PL and the full range of P arguments marked by the Set II suffixes is in (84). (84) a. kʊ́p-bɪ-́gà~kʊ́p f. kʊ́p-bɪ-́nɪ̀~kʊ́p hit-3PL-1SG~RED hit-3PL-1PL.IN~RED ‘They hit me.’ ‘They hit us.’ b. kʊ́p-bɪ-́ɪ̀~kʊ́p g. kʊ́p-bɪ-́mà~kʊ́p hit-3PL-2SG~RED hit-3PL-1PL.EX~RED ‘They hit you.’ ‘They hit us.’ c. kʊ́p-bɪ-́ɛ̀~kʊ́p h. kʊ́p-bɪ-́ʊ̀n~kʊ́p hit-3PL-3SG.M~RED hit-3PL-2PL~RED ‘They hit him.’ ‘They hit you all.’ d. kʊ́p-bá-àpʼ~kʊ́p i. kʊ́p-bɪ-́ʊ̀n~kʊ́p hit-3PL-3SG.F~RED hit-3PL-3PL~RED ‘They hit her.’ ‘They hit them.’ e. kʊ́p-bɪ-́à~kʊ́p hit-3PL-3N~RED ‘They hit it.’ 135 Gwama employs an applicative morpheme gá= APPL to mark BENEFACTIVE or RECIPIENT semantic roles. This morphemic form can occur as a benefactive proclitic on a noun phrase (i.e. as a preposition), as seen in (85a); or it can occur in the verb as an applicative, where it follows the S/A suffix as in (85b). Note that in (85b), the Gwama verb marks three participants (3PL, 3SG.F, and 3N),which is possible only when the -gá APPL morpheme occurs on an otherwise transitive verb. (85) a. á-bɪ ́ tɔ̄p tàm gá=hàpʼ INT-3PL harvest honey BEN=3SG.F ‘They will harvest honey for her.’ b. tàm, ʃám-bɪ-́gá-àpʼ-à~ʃám honey search-3PL-APPL-3SG.F-3N~RED ‘Honey, they searched for it for her.’ In affirmative imperative constructions, the addressee argument is indexed by a prefix if 2PL but is unmarked if 2SG. Examples with intransitive and transitive verb are in (86a-b) and (86c-d), respectively. Note that 2SG is unmarked (but shown here by a zero to clarify the contrast with 2PL). (86) a. ∅-sɪ~̄sɛl̄ b. mɪ-́gɪ̀~gʊ̀s 2SG-RED~climb 2PL-RED~run ‘Climb!’ ‘Run!’ c. ∅-kʊ̀ʃ-á~kʊ̀ʃ d. mɪ-́kʊ́ʃ-á~kʊ̀ʃ 2SG~kill-3SG.N~RED 2SG~kill-3SG.N.DD1~RED ‘Kill it!’ ‘Kill it (and then come)!’ (Goldberg et al. 2017:62) 136 2.2.3.1.2 Komo argument indexing verb morphology Of the Koman languages, Komo displays the richest system of morphological argument indexation on the verb.104 Table 23 contains the full paradigms of Ethiopian Komo bound pronominal argument suffixes alongside the independent pronouns for comparison.105 Table 39 Ethiopian Komo independent and bound pronominals Independent Set I (S/A)† Set II (A)‡ Set III (P) 1SG ākā -(n)á -g -āk 2SG àj -í -ɛ́ -ɛ ̄ 3SG.M hàr -r -r -ār 3SG.F hàpʼ -pʼ -pʼ -āpʼ 3N hɪ̀n ~ hàn -n -n -ɪ ̄~ ɪn̄ 1PL.IN ānà -(n)à -(n)àn -ānà 1PL.EX āmʊ̀n -(n)á ~ -(n)ám -(n)án -ām ~ -ān 2PL ʊ̀m -m -m -ʊ̄m 3PL hʊ̀n -n ∅ -ʊ̄n † This set is employed only when just one argument is indexed in the verb. ‡ This set is employed only when more than one argument is indexed in the verb. It is evident that almost all of the bound pronominals are reduced forms of the independent pronouns, which have become phonologically eroded over time. One notable exception is Set I -(n)á 1SG, which does not bear a resemblance to the corresponding synchronic independent pronoun. Also the Set II and Set III markers are 104 I find no difference between the verb morphology in Ethiopian Komo from the Sudanese variety described by Burns (1947). 105 The nasals (in parentheses) in the 1PL suffixes surface only after a vowel. 137 clearly phonologically reduced forms of the independent pronouns. There appears to be (at least synchronic) free variation between a bilabial and alveolar nasal in the Set I and Set II 1PL.EX suffixes as well as in the Set III 3N suffixes. Komo must index S/A arguments on all finite verbs in main clauses. If only one argument is indexed on the verb (either S or A), then the Set I markers are employed. If more than one argument is indexed in the verb, then the Set II markers are employed for the A argument and the Set III markers are employed for the P argument. Consider the 1st person singular bound pronominal indexing in (87). In (87a), 1SG is the A argument, realized by a preverbal independent pronoun as well as by the Set I bound pronominal verb marker -ná 1SG. In (87b), a Set II 1SG bound pronominal -g 1SG is employed as the verb also indexes a 3SG.M P argument. In (87c), the Set III P argument bound pronominal -āk is employed. (87) a. ākā tūs-ʊ́-ná hàr b. tūs-ʊ́-g-àr 1sg push-DD1-1SG 3SG.M push-DD1-1SG-3SG.M ‘I push him.’ ‘I push him.’ c. tūs-ʊ́-r-āk push-DD1-3SG.M-1SG ‘He pushes me.’ Komo can also index three participants on the verb. This is typically realized when the -gá APPL morpheme licenses a third argument, increasing the overall valence by one. The =gá morpheme can either occur as an prepositional proclitic on a noun or free pronoun outside of the verb, as in (88a); or it can function as an applicative -gá within the inflected verb, as in (88b). Inside the verb, -gá APPL occurs immediately after the S/A 138 bound pronominal suffix. In (89), -gá occurs on the intransitive verb ɪ ̄‘go.PL’ and increases the valence to two. (88) a. ākā pʼʊ́r-á ɔ̀ʃ gá=hàpʼ 1SG throw-1SG stone GOAL=3SG.F ‘I throw a stone to her.’ b. ākā pʼʊ́r-á-gá-pʼ-ɪ̀ 1SG throw-1SG-APPL-3SG.F-3N ‘I throw it to her.’ (89) a. ānà ɪ-̄nà-g-ár kʊ̀mà jángú 1PL.IN go.PL-1PL.IN-APPL-3SG.M to Y. ‘We’re going to Yangu for him.’ 2.2.3.1.3 Uduk argument indexing verb morphology Argument indexation on the verb in Uduk is by far the least transparent of the Koman languages. Chali and Bonya Uduk exhibit a unique morphological argument indexing system on the verb that is sensitive the position of independent NP arguments with respect to the verb, as well as the gender class of the postverbal NP argument (Killian 2015).106 Table 40 contains the independent pronouns and the argument indexing suffixes in Chali Uduk, adapted from Killian (2015). Killian analyzes two distinct types of verb agreement suffixes, which I have labeled for convenience as “Set 1” and “Set 2” in Table 40. Occurrence of the optional elements in the Set 1 suffixes depends on whether the 106 See Killian (2015) and §2.2.2.3.3 for Uduk nominal gender classes. 139 verb root to which they attach ends in a vowel or a consonant.107 According to Killian (2015), the Set 2 suffixes grammaticalized from the independent pronouns preceded by the gender Class 2 morpheme -ā CL2.ACC.108 Table 40 Chali Uduk independent pronouns and argument indexing verb morphology (adapted from Killian 2015) Independent SET 1 (S/A) SET 2 (ERG) 1SG áhā -(n)á -kāʔ 2SG ɛ ́ -(V)n -ɛ ̌ 3SG áɗi -(V)ɗ -ǎɗi 1PL.IN ánā -(n)à -ǎnā 1PL.EX ámān -(n)á -âm 2PL úm -(V)n -ǔm 3PL únī -(V)n -ǔnī In all of the Uduk varieties, core arguments must be realized as independent NPs or free pronouns and these arguments can be simultaneously indexed on the verb morphologically. The only exception is when postverbal A arguments are indexed on the verb employing the Set 2 suffixes, which is discussed below in this section. Killian (2015) observes that in Chali Uduk, argument indexing with the Set 1 suffixes only occurs in SV/AVP clauses.109 In SV clauses, the single argument is indexed on the verb with a Set 1 suffix. In AVP clauses, 1st person (singular and plural) are always indexed 107 For the Set 1 suffixes, the nasals in parentheses only occur after vowel-final roots and the (V) indicates that a vowel identical to the root vowel occurs after consonant-final roots. Further, the tone of all of the suffixes can vary, with the exception of the 1st person suffixes which always surface as the forms seen in Table 40. See Killian (2015) for details. 108 According to Killian, the rising contour tone on the initial vowel of the suffix is a result of the coalescence of a gender morpheme, which is M tone, with the initial vowel of the pronominal suffix, which is H tone. 109 Note that the Set 1 suffixes cannot be used pronominally. 140 on the verb while the indexing of all other persons depends upon the gender class of the postverbal argument. To illustrate, consider the data in (90) which contain AVP clauses with postverbal nouns of both gender classes. Examples (90a-g) contain Class 1 postverbal arguments; note that indexing on the verb only occurs for the first person – no other persons are indexed on the verb. By contrast, the postverbal arguments in (90h-n) are Class 2 nouns and now the preverbal argument is always indexed on the verb. While the examples in (90) do not appear to be semantically transitive clauses, this verb indexation pattern behaves identically to that found in more “prototypical” transitive clauses. See Killian (2015) for evidence. (90) Chali Uduk a. áhā gǔs-á pā h. áhā gǔs-á ā=wácʰā 1SG run.SG.IPFV-1SG home 1SG run.SG.IPFV-1SG CL2.ACC=fast ‘I am running home.’ ‘I am running fast.’ b. ɛ́ gǔs pā i. ɛ́ gǔs-ún ā=wácʰā 2SG run.SG.IPFV home 2SG run.SG.IPFV-2SG CL2.ACC=fast ‘I am running home.’ ‘I am running fast.’ c. áɗī gǔs pā j. áɗī gǔs -úɗ ā=wácʰā 3SG run.SG.IPFV home 3SG run.SG.IPFV-3SG CL2.ACC=fast ‘He/she is running home.’ ‘He/she is running fast.’ d. ánā sɔ́n-à pā k. ánā sɔ́n-à ā=wácʰā 1PL.IN run.PL.IPFV-IPL.IN home 1PL.IN run.PL.IPFV-IPL.IN CL2.ACC=fast ‘We are running home.’ ‘We are running fast.’ e. âm sɔ́n-á pā l. âm sɔ́n-á ā=wácʰā 1PL.EX run.PL.IPFV-IPL.EX home 1PL.EX run.PL.IPFV-IPL.EX CL2.ACC=fast ‘We are running home.’ ‘We are running fast.’ f. úm sɔ́ pā m. úm sɔ́-n ā=wácʰā 2PL run.PL.IPFV home 2PL run.PL.IPFV-2PL CL2.ACC=fast ‘You are running home.’ ‘You are running fast.’ 141 g. únī sɔ́ pā n. únī sɔ́-n ā=wácʰā 3PL run.PL.IPFV home 3PL run.PL.IPFV-3PL CL2.ACC=fast ‘They are running home.’ ‘They are running fast.’ According to Killian (2015:197–198), the Chali Uduk Set 2 suffixes only occur in PVA clauses in which solely the A argument is indexed on the verb.110 The data in (91) from the Bonya variety of Uduk support Killian’s claims about the distribution of the Set 2 suffixes. Note that in the Bonya variety, all of the argument indexing suffixes with the exception of 1SG co-occur with a formative -ī, which occurs immediately after the root and whose function or status is unknown at present.111 (91) Bonya Uduk a. ɟɛ̀ kʼɔ̄ʃ-kāʔ e. ɟɛ̀ kʼɔ̄ʃ-âm elephant kill.PFV-1SG elephant kill.PFV-IPL.EX ‘I killed the elephant.’ ‘We killed the elephant.’ b. ɟɛ̀ kʼɔ̄ʃ-ɛ ̌ f. ɟɛ̀ kʼɔ̄ʃ-ī=ūm elephant kill.PFV-2SG elephant kill.PFV-ī-2PL ‘You killed the elephant.’ ‘You killed the elephant.’ c. ɟɛ̀ kʼɔ̄ʃ mā=ɟí g. ɟɛ̀ kʼɔ̄ʃ-ī=ūn elephant kill.PFV CL2.ERG=3SG elephant kill.PFV-ī-3PL ‘He/she killed the elephant.’ ‘They killed the elephant.’ d. ɟɛ̀ kʼɔ̄ʃ-ī=ánā elephant kill.PFV-ī-IPL.IN ‘We killed the elephant.’ 110 Killian does not provide any evidence for this claim besides the Set 2 forms seen in Table 40. 111 As such, I have glossed this morpheme as -ī. 142 Table 41 contains the Yabus Uduk independent pronouns alongside the verb indexing suffixes. The Yabus Uduk variety exhibits the same Set 1 suffixes seen in the Chali and Bonya varieties though Yabus has a high front vowel in the second and third person suffixes after a consonant-final verb root, whereas Chali and Bonya insert a vowel copied from the root. Table 41 Yabus Uduk independent pronouns and argument indexing verb morphology Independent SET 1 (S/A) 1SG áhā -(n)á 2SG ɛ ́ -(i)n 3SG hádi -(i)d 1PL.IN ánā -(n)à 1PL.EX ámān -(n)á 2PL úm -(i)n 3PL únī -(i)n The Yabus variety does not exhibit a nominal gender system like the Chali and Bonya varieties. Recall that in transitive clauses, the gender of the postverbal argument has a bearing on whether or not the verb indexes the A argument in Chali and Bonya Uduk. The Yabus variety by contrast, always indexes the S/A on every verb. Elicited Yabus Uduk examples containing parallel data to the Chali data in (90) are presented in (92). (92) Yabus Uduk a. áhā gǔs-á pā h. áhā gǔs-á kāŋ=wácā 1SG run.SG.IPFV-1SG home 1SG run.SG.IPFV-1SG OBL=fast ‘I am running home.’ ‘I am running fast.’ 143 b. ɛ́ gǔs-ín pā i. ɛ́ gǔs-ín kāŋ=wácā 2SG run.SG.IPFV-2SG home 2SG run.SG.IPFV-2SG OBL=fast ‘I am running home.’ ‘I am running fast.’ c. áɗī gǔs -íɗ pā j. áɗī gǔs -íɗ kāŋ=wácā 3SG run.SG.IPFV-3SG home 3SG run.SG.IPFV-3SG OBL=fast ‘He/she is running home.’ ‘He/she is running fast.’ d. ánā sɔ́n-à pā k. ánā sɔ́n-à kāŋ=wácā 1PL.IN run.PL.IPFV-IPL.IN home 1PL.IN run.PL.IPFV-IPL.IN OBL=fast ‘We are running home.’ ‘We are running fast.’ e. âm sɔ́n-á pā l. âm sɔ́n-á kāŋ=wácā 1PL.EX run.PL.IPFV-IPL.IN home 1PL.EX run.PL.IPFV-IPL.IN OBL=fast ‘We are running home.’ ‘We are running fast.’ f. ūm sɔ́-n pā m. ūm sɔ́-n kāŋ=wácā 2PL run.PL.IPFV-2PL home 2PL run.PL.IPFV-2PL OBL=fast ‘You are running home.’ ‘You are running fast.’ g. únī sɔ́-n pā n. únī sɔ́-n kāŋ=wácā 3PL run.PL.IPFV-3PL home 3PL run.PL.IPFV-3PL OBL=fast ‘They are running home.’ ‘They are running fast.’ 2.2.3.1.4 Dana bound pronominal argument indexing morphology The Dana bound pronominal indexing enclitics, alongside the independent pronouns, are presented in Table 30. Dana does not obligatorily index S/A arguments on the verb, though optional P argument indexing is possible via the pronominal enclitics. The P-indexing enclitics are identical in form to the independent pronouns with the exception of 3SG and 3PL, in which the initial voiceless glottal fricative has been lost. 144 Table 42 Dana independent and bound pronominals Independent Bound (P) 1SG āgā =āga 2SG āj =āj 3SG.M hār =ār 3SG.F hāpʼ =āpʼ 3N hān =ān 1PL.IN mīnā =mīnā 1PL.EX mānā =mānā 2PL ʊ̄mā =ʊ̄mā 3PL hʊ̄n =ʊ̄n The Dana data in (93) contain declarative main clauses in which the P argument is indexed pronominally on the verb. (93) a. hāpʼ kʰɪẃ=āgā pì tì̪s e. hāpʼ kʰɪẃ=mānā pì tì̪s 3SG.F give.SG=1SG cow three 3SG.F give.SG=1PL.IN cow three ‘She gave me three cows.’ ‘She gave us three cows.’ b. hāpʼ kʰɪẃ=āj pì tì̪s f. hāpʼ kʰɪẃ=mīnā pì tì̪s 3SG.F give.SG=2SG cow three 3SG.F give.SG=1PL.EX cow three ‘She gave you three cows.’ ‘She gave us three cows.’ c. hāpʼ kʰɪẃ=ār pì tì̪s g. hāpʼ kʰɪẃ=ʊ̄mā pì tì̪s 3SG.F give.SG=1SG cow three 3SG.F give.SG=2PL cow three ‘She gave me three cows.’ ‘She gave you all three cows.’ d. hāpʼ kʰɪẃ=āpʼ pì tì̪s h. hāpʼ kʰɪẃ=ʊ̄n pì tì̪s 3SG.F give.SG=1SG cow three 3SG.F give.SG=3PL cow three ‘She gave me three cows.’ ‘She gave them three cows.’ Dana has a prepositional proclitic morpheme gá that can occur on NPs to mark a recipient or beneficiary participant, as in (94a). This morpheme can also occur within 145 the verb as an applicative, as in (94b). Note that the form -gápʼ APPL.3SG.F in (94b) is composed of -gá APPL plus the bound pronominal P argument marker -āpʼ 3SG.F. (94) a. hāpʼ ɗɛ̀d pì tìs gá=ísà 3SG.F send cow three BEN= I. ‘She sent Isa three cows.’ b. hāpʼ ɗɛ̀d-á-gápʼ pì tìs 3SG.F send-DD2-APPL.3SG.F cow three ‘She sent her three cows.’ 2.2.3.1.5 Opo bound pronominal argument indexing morphology The Bilugu Opo independent and bound pronominals are in Table 43. The bound S/A proclitics only occur when the arguments are realized pronominally and cannot co- occur with full NPs.112 The Bilugu Opo bound forms are synchronically cliticized independent pronouns with the exception of the third person forms. Recall that this distinction also occurs in the possessive pronominal enclitic paradigm (§2.2.2.5.1). 112 The Opo system is distinct from Komo and Gwama whose bound S/A pronominals are required on every finite verb regardless of whether the arguments also occur as independent NPs. 146 Table 43 Bilugu Opo independent and bound pronominals Independent Bound (S/A) Bound (P) 1SG āgā āgā =āga 2SG āj ī= =āj 3SG.M ʊ̀tà† ār= =ār 3SG.F ɓā† āb= =āb 3N nà† / hà† ān= =ān 1PL.IN mìnà mìnà =mìnà 1PL.EX mànà mànà =mànà 2PL ʊ̄mā ʊ̄mā =ʊ̄mā 3PL bìjà† ʊ̄n= =ʊ̄n † This form cannot occur independently The Opo pronominal S/A markers cliticize to the main lexical verb when the argument is pronominal, as seen in (95). If tense/aspect verb morphology occurs in the clause, such as the morpheme =á INT, the S/A markers cliticize to the tense/aspect morphology. This is seen in (96).113 (95) a. āb=sá mà b. āb=sá mà 3SG.F=eat.SG food 3SG.F=eat.SG food ‘She ate food.’ ‘He ate food.’ c. ʊ̄n=ùsà mà 3SG.M=eat.PL food ‘They ate food.’ (96) a. āb=á sá mà b. āb=á sá mà 3SG.F=INT eat.SG food 3SG.F=INT eat.SG food ‘She will eat food.’ ‘He will eat food.’ 113 The (future) Intentative grammatical category, which expresses future intent or desire, is realized morphologically in all Koman languages though not all forms are cognate. 147 c. ʊ̄n=á ùsà mà 3SG.M=INT eat.PL food ‘They will eat food.’ The morpheme gá in Bilugu Opo can be employed to mark a recipient or benefactor. It can occur on a noun phrase, as seen in (97a), or it can fuse with a bound pronominal outside of the verb, as in (97b). Note that the form gár BEN.3SG.M is composed of gá + ār 3SG.M, with which it which has coalesced. The gá morpheme can also occur in the verb as an applicative, as in (97c), and it can be followed by a bound pronominal enclitic in the verb as, in (97d). (97) a. āb=sʊ̄ pì tùsù gá=ʊ̀tɔ́n b. āb=sʊ̄ pì tùsù gár 3SG.F=buy.SG cow three BEN=man 3SG.F=buy.SG cow three BEN.3SG.M ‘She bought the man three cows.’ ‘She bought him three cows.’ c. āb=sʊ̄=gár pì tùsù d. āb=sʊ̄=gár=ān 3SG.F=buy.SG=APPL.3SG.M cow three 3SG.F=buy.SG=APPL.3SG.M=3N ‘She bought him three cows.’ ‘She bought them for him.’ 2.2.3.2 Koman Deictic Directional (DD) verb morphology Deictic Directional (DD) verb morphology is a core feature of the Koman verb system. The Koman DD systems consist of two to three suffixes that occur immediately on the verb stem (Otero 2017, under revision). These suffixes are presented in Table 44. Note that all of the languages have two DD morphemes with the exception of Komo, which has three.114 The parentheses in the Dana and Opo forms indicate epenthetic 114 The Komo DD∅ morpheme is devoid of any semantic material though it stands in morphological opposition to the other DD morphemes. It is treated as the unmarked form of the verb as compared to the 148 glides that surface when suffixed onto a vowel-final verb. While all of the DD suffixes may not be cognate in form, they display remarkably similar idiosyncratic behavior. Table 44 Koman Deictic Directional morphemes Gwama Komo Uduk Dana Opo Opo (Lowland) (Ethiopia) (Chali) (Bilugu) (Pame) DD1 -ɪ ́ -ʊ́ -í -(j)ɪ ́ -(w)ʊ́ -(j)ʊ́ DD2 -gɪ ́ -úk/-kú -kú/kí -(j)á -(j)á -(j)á DD∅ – -í / -á – – – – – indicates a bare or unmarked form of the verb The Koman DD morphology exhibit a range in functions when collocated with verb roots of distinct semantic profiles (cf. Otero 2018, accepted on Komo specifically). On verbs of motion, the DD suffixes express DIRECTION (or the trajectory) of motion relative to a deictic reference point. The Dana examples in (98) contain the verb pɔ́j ‘run.PL’ inflected with both the deictic directional morphemes. Observe that motion towards the 1st person is expressed by -ɪ ̄DD1 in (98a), and motion towards the second person is expressed by -wa DD2 in (98b).115 Compare these to the bare form (uninflected) of the verb in (98c) in which no direction is specified. other Koman languages which do not display a DD∅ morpheme a can employ a verb devoid of a DD morpheme in a clause. 115 The labels DD1/DD2 are employed mnemonically as they encode motion towards the 1st and 2nd persons, respectively. 149 (98) Dana a. hʊ̄n-ɪ ́ pɔ́j-ɪ ̄ b. hʊ̄n-ɪ ́ pɔ́-wà 3PL-PROG run.PL-DD1 3PL-PROG run.PL-DD2 ‘They are running (to me).’ ‘They are running (to you).’ c. hʊ̄n-ɪ ́ pɔ́j 3PL-PROG run.PL ‘They are running.’ (direction unspecified) Identical behavior is seen on the Lowland Gwama motion verb tìndì ‘roll’ in (99) and on the Komo verb pʊ́k ‘cross’ in (100). (99) Gwama (Lowland) a. mā=pīdìn tìndì-í-bá~tìndì b. mā=pīdìn tìndì-bí-gí~tìndì PL=stone roll-DD1-3PL~RED PL=stone roll-3PL-DD2~RED ‘The stones roll (towards me).’ ‘The stones roll (towards you).’ c. mā=pīdìn tìndì-bí~tìndì PL=stone roll-3PL~RED ‘The stones roll.’ (direction unspecified) (100) Komo a. ʊ́-r pʊ́k-ʊ́ sʼʊ́ b. ʊ́-r pʊ́k-kʊ́ sʼʊ́ INT-3SG.M cross-DD1 river INT-3SG.M cross-DD2 river ‘He will cross the river (towards ‘He will cross the river me).’ (towards you).’ c. ʊ́-r pʊ́k-∅ sʼʊ́ INT-3SG.M cross-DD∅ river ‘He will cross the river.’ (direction unspecified) 150 Creissels et al. (2008:148) observe that directional morphemes attached to the verb are common in Nilo-Saharan.116 A traditional view of verbal directionality posits a binary opposition between motion towards and motion away from a particular deictic center or point of reference. Hither, centripetal and ventive are terms commonly employed to describe motion toward a reference point, while thither, itive, andative and centrifugal describe motion away from a given reference point (e.g. Dimmendaal (2003) for Nilotic and Surmic, Payne 2013 for Maa, Kiessling (2007) for Datooga). The defaukt deictic reference point is most commonly the speaker. Thus, an Itive/Ventive contrast typically codes the direction of motion towards a speaker GOAL and away from a speaker SOURCE. However, the Koman languages show that the morphological expression of “towards a reference point” can have two distinct deictic goals: either the speaker or the addressee. Further, this is not in morphological opposition to ‘away’. The data in (98)-(100) above show that one set of Koman DD suffixes expresses motion towards the speaker, while the other set expresses motion towards the addressee. A second function of the Koman DD morphology is to express ASSOCIATED MOTION (AM), which is commonly defined as an event of translational motion coded in the verb by a morpheme other than the lexical main verb root (Koch 1984; Wilkins 1989, 1991). In Koman, associated motion is generally expressed by the DD suffixes on dynamic verbs whose lexical roots do not contain inherent motion (e.g. ‘kill’, ‘drink’, ‘build’ etc.).117 To illustrate, consider the Dana examples in (101), which employ the same DD 116 This also includes languages that express directional meanings by changes in root-vowel length, vowel quality, and/or tone (e.g. Dinka (W. Nilotic), as reported by Andersen 2012a, b). 117 The term “dynamic” here is employed to contrast with stative. Koman languages all have stative as well as dynamic verbs. The deictic directional morphemes which have distinct semantic effects with the two classes of verbs. 151 morphemes seen in (98). In (101a), the DD1 morpheme expresses that the subject will come to the speaker’s location after the event of the lexical verb root has been realized (in this case, the slaughtering of a goat). In (101b), the DD2 morpheme expresses that the subject will go to the hearer’s location after the event of the lexical verb root has been realized. Compare these to the unmarked form of the verb in (101c), in which no additional motion event is included in the meaning of the verb word. (101) Dana a. hār-ɪ ́ t ̪̓ ɔ̀ɗ-ɪ ̄ mɛ ̄ b. hār-ɪ ́ t ̪̓ ɔ̀ɗ-á mɛ ̄ 3SG.M-PROG slaughter.SG-DD1 goat 3SG.M-PROG slaughter.SG-DD2 goat ‘He is slaughtering a goat.’ ‘He is slaughtering a goat.’ (and then will come here) (and then will go to you) c. hār-ɪ ́ t ̪̓ ɔ̂ɗ mɛ ̄ 3SG.M-PROG slaughter.SG goat ‘He is slaughtering a goat.’ In the Bilugu Opo data in (102), we see the same behavior as in Dana – associated motion arises when the deictic directional morphemes occur with dynamic verb roots that do not inherently express motion. Further, the DD2 morphemes express AM toward the addressee. (102) Opo (Bilugu) a. ār=tʰáp-ʊ́ kù b. ār=tʰáp-á kù 3SG.M=plaster-DD1 hut 3SG.M=plaster-DD2 hut ‘He plastered the hut.’ ‘He plastered the hut.’ (then came here) (then left towards you) c. ār=tʰáp kù 3SG.M=plaster hut ‘He plastered the hut.’ 152 The coding of associated motion is an integral part of the Koman DD system, where AM is expressed even when the lexical verb is negated. To illustrate, consider the Komo data in (103). Notice that negation has scope over the lexical verb event but not over the AM component expressed by the DD morphemes. In Koman, DD2 expresses AM towards the addressee only in the Dana-Opo branch. In the remaining languages (Gwama, Komo, Uduk), DD2 expresses AM directed away from the point of reference. (103) Komo a. bāʃ-í-n tʼɔ̀r-ʊ́ mɛ́ b. bāʃ-í-n tʼɔ̀r-úk mɛ́ NEG-DD∅-3PL slaughter-DD1 goat NEG-DD∅-3PL slaughter-DD2 goat ‘They didn’t slaughter a goat.’ ‘They didn’t slaughter a goat.’ (at other location, then came (they left) here) c. bāʃ-í-n tʼɔ̀r mɛ́ NEG-DD∅-3PL slaughter goat ‘They didn’t slaughter a goat.’ The Koman languages provide counterevidence to an implicational hierarchy proposed for associated motion systems in Guillaume (2016), namely that the Koman systems only express AM ‘subsequent’ to the lexical root event, and do not express ‘concurrent’ nor ‘prior’ AM (cf. (101)-(103)). In some African languages, it appears that Direction (of motion) and Associated Motion are often expressed by the same coding material; this is observed primarily when taking the semantic profile of the verb into account (see Belkadi 2015, 2016 for an 153 overview).118 For instance, Mietzner (2012:171) argues that in certain Southern Nilotic languages deictic directional morphemes can also express an “altrilocal” or ‘other place’ function when occurring on a (potentially non-motion) lexical verb. In this case, the event denoted by the verb root can be construed as having occurred at a distinct location from that of the speech act or primary reference point. Similarly, Keissling (2007:128) describes “altrilocality” in Datooga (S. Nilotic), providing evidence that the ‘centrifugal’ extension on non-motion verbs indicates that the event denoted by the lexical verb root took place at a location that is not the deictic center. Alamin et al. (2012) use the term “alloying” to describe AM in Tima (Niger-Congo) in which the ventive marker codes deictic direction on motion verbs and subsequent associated motion (on non-motion verbs). In a typological study of 20 languages from the four African phyla, Belkadi (2016:46) observes that such morphology with lexical motion roots all trigger directional readings, while the same morphology with lexical verbs describing certain (non- motion) activities trigger associated motion readings. She also observes that the distinction between motion and non-motion lexical verbs and their subsequent behavior (i.e. deictic direction or AM) is not categorical but rather gradient, depending largely on the semantic verb class. Belkadi (2016:64) proposes a tentative ranking of some lexical verb classes and the likelihood of whether an AM or directional reading would occur, reproduced here in (104). She proposes that verb classes on the left 118 Prior attestations of what do appear to be AM systems in African languages are found scattered within descriptions of directional systems (Serzisko 1988, Reh 1996, Heath 2005, Bourdin 2006, Andersen 2012a, b, Payne 2013, Storch 2014, Kramer 2017 inter alia). 154 extreme are more likely to derive a directional reading while the classes towards the right extreme or more likely to derive an AM interpretation.119 (104) Path Motion > Motion > Causative motion> Perception> (Natural phenomena and bodily secretions?) > Activities not involving motion > States (adapted from Belkadi 2016:64) Komanists have also observed the overlapping of direction, associated motion and aspectual functions coded by the same morpheme(s) on different classes of verbs. Burns (1947) describes “aspectual suffixes which also can also code direction” in Sudanese Komo. More recently, Hellenthal (2018) describes directional/associated motion in the Lowland Gwama DD system, and Killian (2015) does likewise for Chali Uduk, and Otero (2018, accepted) for Ethiopian Komo. In sum, the Koman DD systems exhibit an overall tendency for coding of direction of motion on intransitive motion verbs and AM on dynamic (non-motion) verbs as seen above in (98)-(100) and (101)-(103), respectively. Across the Koman family, stative verbs are the least predictable and display a variety of nuanced meanings when combined with the DD morphology. In Gwama and Komo, DD1 on a stative verb retains the AM component directed towards the addressee as in (105a) and (106a), while DD2 expresses “exchoativity”, in which the subject has “left the state” or is “no longer in the state”. Examples of exchoativity in Gwama and Komo are seen in (105b) and (106b), respectively. 119 The only example of what appears to be a stative verb in Belkadi’s study is íŋg ‘stay’ in Kenga (Nilotic), which expresses an AM interpretation with ventive marking (2016:62 ex 36b). 155 (105) Gwama (Lowland) a. zɛ̀ tɪňd-ɪ-́ná~tɪňdɪ ́ b. zɛ̀ tɪň-nɪ-́gɪ~́tɪňdɪ ́ PROG be.fat-DD1-3SG.M~RED PROG be.fat-DD2-3SG.M~REDt ‘He was becoming fat.’ ‘He was becoming fat.’ (at another location, then came (He is no longer fat.) here) (106) Komo a. máŋgà pʼɛ̀l-ʊ́-n ɪ=́ȷá́ngú b. máŋgà pʼɛ̀l-kú-n ɪ=́ȷá́ngú mango be.red-DD1-3PL LOC=J. mango be.red-DD2-3PL LOC =J. ‘(The) mangos were red in Yangu.’ ‘(The) mangos were red in (then were brought here) Yangu.’ (They are no longer red.) Deictic directional morphology on stative verbs in the Dana-Opo branch functions a bit differently than in Komo and Gwama. The examples in (107) and (108) contain data from Dana and Bilugu Opo with the DD morphemes on the verb ‘be tall’. In both languages, the DD1 morpheme behaves identically to the DD1 category in Gwama and Komo, expressing AM towards the speaker, as seen in (107a) and (108a). The DD2 morpheme, by contrast, strongly expresses AM to the addressee in Dana, as seen in (107b); but it is not allowed with stative roots in Bilugu Opo, as seen in (108b).120 (107) Dana a. hār -â būd-í b. hār -â būd-á 3SG.M-INT be.tall.SG-DD1 3SG.M-INT be.tall.SG-DD2 ‘He will get tall.’ ‘He will get tall.’ (then come here) (then go to you) 120 The Opo consultants rejected the DD2 morpheme on almost all of the stative verbs I attempted to elicit. 156 c. hār -â būd 3SG.M-INT be.tall.SG ‘He will get tall.’ (108) Opo (Bilugu) a. ār=sɪḱʼ-ʊ́ b. *ār=sɪḱʼ-á 3SG.M=be.tall.SG-DD1 3SG.M=be.tall.SG-DD2 ‘He got tall.’ (ungrammatical) (then came here) c. ār=sɪḱʼ 3SG.M=be.tall.SG ‘He is tall.’ This concludes the overview of synchronic Koman morphology. The next chapter begins the reconstruction of Proto-Koman phonology. 157 CHAPTER III RECONSTRUCTION OF PROTO-KOMAN PHONOLOGY 3 Reconstruction of Proto-Koman Phonology This chapter presents a reconstruction of Proto-Koman (PKMN) phonology. I discuss the phonological innovations that took place in the segmental and suprasegmental domains, in light of the analyses proposed by Bender (1983). One notable issue is the role of tone in the reconstruction of PKMN phonology. In §3.1, I provide an overview of tone and discusses its diachronic relationship to reconstructing word-initial (syllable onset) stop consonants. I move on to PKMN consonants in §3.2. To conclude, I discuss PKMN vowels and ATR harmony in §3.3. In Chapter IV, I discuss the phonological innovations of each subnode. 3.1 Proto-Koman (PKMN) tone A major aim of this phonological reconstruction is to reconstruct PKMN segments. Yet in doing so, it has become apparent that tone plays a significant role in the development of Koman consonants. There is historical interaction in Koman between tone and consonants specifically seen in word-initial stops and the feature [±voice]. For instance, in a particular word-initial correspondence sets, the Komo-Uduk branch exhibit /b, d, g/ corresponding to /p, t, k/ (respectively) in Gwama and the Dana-Opo branch. Bender (1983:283) notes these correspondence sets and reconstructs PKMN voiced stops *b, *d, *g which are retained as voiced in the Komo-Uduk branch and “devoice” in the remaining branches, though he does not offer a motivation for the devoicing. I agree with Bender’s analysis and propose that the motivation for these changes (e.g. *b > p) in Gwama and Dana-Opo are directly related to tone. As such, a 158 discussion of Proto-Koman tone is imperative before discussing the reconstruction of segments. Additionally, a word-initial voicing contrast exists in all of the Koman languages though in the Uduk cluster, Killian (2015) observes the following synchronic consonant- tone restrictions for Chali Uduk: in syllable onsets, voiceless obstruents can only occur with M or H tones and voiced obstruents can only occur with L or LH (Rising) tones (§2.1.6.6). The fact that Uduk word-initial (syllable onset) voiced and voiceless obstruents are in complementary distribution with tones is of crucial importance to the reconstruction of PKMN tone as well as the reconstruction of word-initial stops. Bender (1983:285) claims insufficient documentation of tone in Koman languages at the time of his writing to be able to reconstruct tone in PKMN, though he does presume that “proto-Koman was probably tonal”. Given the historical consonant-tone interactions in Koman, I must discuss the historical evolution of tone and stops in tandem. In this dissertation, I focus specifically on word-initial (syllable onset) consonants and the tone of the following vowel nucleus. In §3.1.1, I establish tone correspondence sets and look for correlations in the distribution of these tone correspondence sets with respect to word-initial stop onsets. In §3.1.2, I sketch out a historical scenario for the evolution of Koman tone in tandem with word initial consonants. In §3.1.3 I briefly examine tone and non-stop onsets and lastly, in §3.1.4, I discuss the exceptions to the main tone patterns and their implications for the historical scenario of consonant-tone evolution. 159 3.1.1 PKMN tone categories based on synchronic tone distribution All of the living Koman languages exhibit contrastive tone with at least three level tones (Low, Mid, High) as well as (R)ising and (F)alling contour tones (§2.1). Tone in Koman languages is generally stable in noun roots, meaning that noun words generally never alter their tone. By contrast, in all of the languages in the Central Koman branch, tonal alternations or tonal suppletion in verb roots is employed to mark nominal/verbal number or aspect.121 The functional load of grammatical tone is by far more prevalent in verbs than in nouns. The fact that verb root tone can vary must be taken into account when reconstructing tone categories. I employ the term tone “categories” as a neutral term to refer to reconstructed pitch realization on a single vowel nucleus. Given the diachronic development of tone and stop onset in Koman we shall see that a reconstructed tone category can be phonemic or allophonic. Determining exactly at what stage a particular tone category became phonemic can be challenging at times. As such, I employ tone category as a more general term. Examining the tone of syllable nuclei following word-initial syllable onset stop consonants yields three major tone correspondence patterns. The three tone correspondence sets, labeled “A, B, C”, are seen in Table 45. In Set A and Set C, all languages exhibit synchronic H or L tone correspondences, respectively.122 In 121 Komo uses tonal suppletion in verb roots to mark number of the S/A argument. Dana and Opo employ tonal suppletion to mark nominal and verbal number. Uduk employs tonal alternations to mark aspect. It is unclear whether suppletive tone in verbs was an innovation of PCTRL Koman. 122 A R tone (phonetically LH) in Uduk will correspond to H across the remaining languages when the word-initial consonant is a voiced stop. Also, Opo innovated an XH tone by splitting *H > XH on [+high, +ATR] /i, u/ vowel nuclei and retaining H elsewhere. See §3.1.2 for discussion. 160 correspondence Set B, Gwama, Uduk and Opo display M tone while Dana and Komo have L tone.123 Table 45 Tone patterns in PKMN correspondence sets Set Gwama Komo Uduk Dana Opo A H H H H H B M L M L M C L L L L L To illustrate, the data in (109) contain cognates from each of the three tone correspondence sets. The full data employed in establishing the tone correspondence sets can be found in Appendices C1-3. (109) Tone Set Gwama Komo Uduk Dana Opo Meaning A tápʼ táb tʰáɓ t ̪h áp tʰáp ‘kick’ B pāj pàj pʰāj pʰàd̪ pʰāj ‘fly (v.)’ C tàm dàm d̪àm tâ̪m tàm ‘honey’ The tone distributions in Table 45 suggest reconstructing three tone categories to PKMN given the three tone correspondence sets (A, B, C). Yet, the examination of how the tone correspondence sets distribute relative to word-initial stops in PKMN cognates yields more nuanced correspondence patterns. Here, it is important to highlight that it is the distribution of word-initial reconstructed consonants and reconstructed tone categories that is being examined. When doing so, the data suggest that reconstructed tones are generally in complimentary distribution relative to voicing of the initial 123 Recall that Dana and Komo belong to distinct branches of Central Koman. 161 reconstructed consonant. The distribution tendencies between word-initial stop onsets in PKMN cognates and the tone patterns in Table 45 are presented in (110). (110) a . Tone correspondence Set A robustly occurs with reconstructed voiceless stop onsets and very marginally with reconstructed voiced stop onsets. b. Tone correspondence Set B only occurs with reconstructed voiceless stop onsets. c. Tone correspondence Set C only occurs with reconstructed voiced stop onsets. Examples of reflexes with word-initial stops in tone correspondence Sets A-B are in (111). Note that all reflexes occur with word-initial voiceless onsets, which crucially also reconstruct as voiceless consonants. (111) Tone Set Gwama Komo Uduk Dana Opo Meaning a. A tápʼ táb tʰáɓ t ̪h áp tʰáp ‘kick’ b. A kúsʼ kʊ́sʼ kʰúsʼ kʰʊ́sʼ kʰʊ́tʃʼʊ́ ‘dry’ f. B pāj pàj pʰāj pʰàd̪ pʰāj ‘fly (v.)’ g. B kāgā kàʔ kʰāʔ kʰàkʼà kʰākʼā ‘bitter’ Tone correspondence Set C, in which all languages exhibit synchronic L tone, only occurs with reconstructed voiced stops.124 Some examples of reflexes in PKMN tone correspondence Set C are in (112).125 Observe that not all word-initial consonants in the 124 PKMN tone correspondence Set C can also occur in reflexes with other voiced consonants though I limit the analysis to stops, given the fact that their evolution is directly integrated with tone. See §3.1.3 for discussion of PKMN tone and word-initial consonants that are not stops. 125 The voiced alveolar fricatives in the Gwama and Komo reflexes of ‘grind wet’ are the result of spirantization before high front vowels. 162 reflexes in (112) are synchronically voiced, though the onset consonant correspondences in (112) reconstruct to voiced stops. (112) PKMN Gwama Komo Uduk Dana Opo Meaning a. *gàm kàm gàm gàm kàm kàm ‘find, meet’ b. *d̪àm tàm dàm d̪àm tâ̪m tàm ‘honey’ c. *d̪ìm zìŋā zìm d̪ìm tì̪m tìm ‘grind wet’ One crucial pattern in reconstructing tone and word-initial stops is what is seen in (112): tone Set C reflexes have word-initial voiced stops in the Komo-Uduk branch that correspond with word-initial voiceless unaspirated stops in the Dana-Opo branch and in Gwama. This represents two independent innovations of PKMN *C[+voice, -aspirated] > *C[–voice, –aspirated] in Proto-Dana-Opo and in Proto-Gwama. I propose that this split was historically conditioned by tone, or that, more succinctly, word-initial stops and the tone of the following vowel evolved as a unit, or constellation so to speak. To examine this more in detail, we must look at the correspondence sets for a full stop series to see the how the tone correspondences pattern. Table 46 contains the correspondence sets for word-initial *pʰ, *p and *b.126 As mentioned in (110), PKMN word-initial consonant correspondence sets 1 and 2 only occur before tone correspondence sets A and B. 126 Note that Komo and Gwama independently merged all of the voiceless unaspirated and voiceless aspirated stops across all places of articulation (§4.1.1). 163 Table 46 PKMN bilabial stops correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 1 *pʰ p p p pʰ pʰ pʰ pʰ pʰ pʰ pʰ 2 *p p p p p p p p p p p 3a *b p p b b b p p p p p 3 *b b b b b b b b b b b The primary split *b > *p in Proto-Gwama and Proto-Dana-Opo is reflected in PKMN *b set 3a in Table 46. PKMN *b correspondence set 3a only occurs with tone correspondence Set C, in which all languages exhibit synchronic L tone.127 Some cognates from consonant correspondence set 3a (and tone set C) are seen in (113).128 (113) Tone Set Gwama Komo Uduk Dana Opo Meaning a. C pʊ̀m ɓú bwà pùmá pǔmá ‘pregnant’ b. C pɔ̀pʼ bɔ̀b – – pɔ̀j ‘skulk’ c. C pɪ̀t bìt bǐtʰ – – ‘toss’ e. C pàj ɓàjá bɛ̀ – pàj ‘wide’ Now that we have outlined the major tone correspondences and their distributions with word-initial stop onsets, a historical scenario for the evolution of tone in Koman is discussed in subsection §3.1.2. 127 This correspondence between voiceless unaspirated and voiced stops occurs across all of the stops and is not limited to the bilabial place of articulation. Further, the correspondence set is limited to word- initial position. 128 Note that there are exceptions such as Komo ɓú in (113a), which means ‘pregnant’ or ‘swell’ and may not be cognate for two reasons: word initial *b does not generally correspond to [ɓ] and Komo has a H tone when all others have L tone. Another exception is the Komo reflex in (113e), which contains an initial [ɓ] yet it does correspond in L tone. 164 3.1.2 A historical scenario for the evolution of PKMN tone The distributions of synchronic tone (i.e. tone correspondence sets A-C) coupled with the distributions of word-initial stops point to a historical scenario in which PKMN had three tone categories. Two of these tones (Set B and Set C) were in complementary distribution with respect to the voicing of the onset stop consonant as outlined in (110). We can consider the reconstructed to categories instantiated by tone Set B and Set C as allotones in PKMN. These allotones are discussed more in detail in this section, though first we turn to an earlier stage in Koman’s history. In order to explain the distributions of tone and initial stops in PKMN, I reconstruct two tone categories to an earlier stage in PKMN, that is, to a Pre-Proto-Koman stage. Pre-Proto-Koman (P-PKMN) had two contrastive level tone categories. These P-PKMN tone categories were realized phonetically by two level pitches of contrasting F0, labeled here as **H and **L for simplicity. I argue that these two P-PKMN tone categories evolved into the three synchronic level tones seen across the family predominantly via a split in **L following voiced stop word-initial onsets. I provide a brief sketch in (114) wherein P-PKMN **L > PKMN *L. PKMN *L has two allotones, one Low allotone which is realized phonetically [L] (which corresponds to Tone Set B) and another Low allotone which is realized phonetically as Extra-Low [XL] (which corresponds to tone Set C). These allotones of PKMN *L eventually phonologize into modern M and L tones, which I subsequently discuss in detail. 165 (114) P-PKMN **L ↓ PKMN *L ↙↘ (Set B) [L] [XL] (Set C) ↓ ↓ Modern M L Table 47 contains a possible diachronic evolution of Koman tone in the environment following word initial onset stops. For clarity, I employ the bilabial graphemes as a schematized representation for all stops across four points of articulation (bilabial, interdental, alveolar, velar).129 Given the fact that there is a direct correlation between the proto-tone category and voicing/aspiration of the proto-stop onset, I employ these graphemes followed by a tone symbol (i.e. L, H, etc.) to represent an initial consonant-tone “constellation”. For instance, in the first row of Table 47, **pʰ**H represents a constellation of any word-initial voiceless pre-proto-aspirated stop followed by pre-proto **H tone.130 This P-PKMN **pʰ**H constellation evolved into PKMN *P*H and into synchronic Gwama pH. I have also included a column of PKMN allotones to illustrate the allotonic realization of the reconstructed PKMN tone categories. 129 To clarify, in Table 47, *pʰ represents all reconstructed voiceless unaspirated stops, *p represents all reconstructed voiceless aspirated stops, and *b represents all reconstructed voiced stops. 130 Recall that **H tone represents the reconstructed tone category instantiated by the synchronic tone correspondence set A in which all languages display H tone. 166 Table 47 Evolution of Koman tone in PKMN cognates PKMN Pre Pre- Onset SET P-PKMN PKMN Uduk Dana Opo allotones Gwama Komo pʰH A **pʰ**H > *pʰ*H [pʰH] > pH pH pʰH pʰH *pʰ pʰXH B **pʰ**L > *pʰ*L [pʰL] > pM pL pʰM pʰL pʰM pH A **p**H > *p*H [pH] > pH pH pH pH *p pXH B **p**L > *p*L [pL] > pM pL pM pL pM A **b**H > *b*H [bH] > bH bH bR bH bH *b C **b**L > *b*L [bXL] > pL bL bL pL pL Table 47 shows that in P-PKMN, all initial stops had two contrastive tone categories, **H and **L. Over time the initial voiced stops acted as depressor consonants and lowered the F0 of the following vowel, creating an extra-low register in PKMN ( **b**L > *b*L [bXL]). During the PKMN stage, three *L allotones coexist: one allotone with a lower F0 following voiced stops (*b*L [bXL]) and two allotones with a higher F0 following voiceless onsets (*pʰ*L [pʰL] and *p*L [pL]). When the *b*L allotone eventually phonologizes, three tone levels become contrastive: *H > H is realized with a higher pitch than the M tone that arises from the *L > M following voiceless onsets. This historical scenario yields three contrastive level tones in Gwama, Uduk and Opo.131 Further, Proto-Gwama and Proto-Dana-Opo split PKMN voiced *b*L and merge with voiceless unaspirated /pL/. Proto-Uduk (PUD) does not undergo the devoicing seen in Gwama and Dana-Opo. In the pitch register following voiced onsets, PKMN *H tone becomes synchronic Rising tone in PUD: *b*H > bR. 131 See §2.1.6.6 for discussion of Uduk tone. 167 One more tone split must be described. The final column of the **pʰ**H row in Table 47 contains a Proto-Opo innovation of an extra-high tone (XH) by splitting POP *pʰ*H. This split *pʰ*H > pʰXH only occurred in nuclei that contained a [+high, +ATR] vowel /i, u/. Some examples illustrating this split are in (115) and the full set of correspondences are provided in Appendix C2. (115) Gwama Komo Uduk Dana Opo Meaning a. ɪś íʃ ís ísá ıs̋á ‘ripen’ b. pīs píʃ píʃ píʃ pıs̋ ‘disregard’ c. tʊ̌kʊ̀ túk tú̪kʰ tú̪kʰ tűkʰ ‘acacia’ d. wǔt wǔt útʰ – hűtʰ ‘ostrich’ This historical tone scenario accounts for much of the data and yields the contrastive level tones seen in the modern languages with the exception of Komo and Dana. Recall that while *L > M following voiceless onsets (i.e PKMN *pʰ*L > pʰM and PKMN *p*L > pM) in the rest of the family, this did not occur in Komo and Dana. In Komo and Dana, the correspondences yield PKMN *L > L following all stop onsets (cf. Table 47). At present, I cannot account for how Komo and Dana developed a third level tone seen modernly in these languages and I leave this question for further research. 3.1.3 PKMN tone and non-stop consonants We have seen that Koman tone and word-initial stop onsets exhibit a strong correlation diachronically. Non-stop word-initial consonants in terms of voicing and tone yields similar patterns to those described for the PKMN stops. Voiced onsets, including nasals, glides and the lateral generally occur with a tone set C 168 correspondence set, which correlates with the voiced stop onsets (116). There is no evidence for devoicing of sonorants. (116) Gwama Komo Uduk Dana Opo Meaning a. màm màmá màm màmá màmā ‘carry on back’ b. màʃ màʃ màʃ màʃ màs ‘wife, marry’ d. àlàpɛ ́ lɛ̀pɛ ́ lɛ̀pʰɛ ́ lɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ ‘claves’ e. nàbɔ̀ŋà nàbɔ̀ŋà – nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ ‘pelican (sp.)’ f. wàsà wàʃākʼ wàsá – – ‘hail, ice’ g. wàsʼ wàsʼ wàʃʼ wàcʼà wàtʃʼà ‘fish’ h. wàpʼ wàpʼ wàɓ – – ‘boar’ i jɪ̀ jɛ̀k jɛ̀kʰ – – ‘sow seeds’ 3.1.4 PKMN tone residue Every historical reconstruction contains some residue and this reconstruction of PKMN tone is no exception. This section addresses some key residual issues in the reconstruction of PKMN tone categories. The major tone correspondences were discussed in §3.1.1 and their relationship to word-initial stop consonants as well as a historical scenario for the evolution of P-PKMN tone categories were discussed in §3.1.2. These analyses were based on the three major tone correspondence sets which exhibit consistent patterning (Table 47). Further, the independent innovations in Proto-Gwama and Proto-Dana-Opo of devoicing initial voiced stops and merging with voiceless unaspirated stops after the tone split creating the [XL] allotone of PKMN *L strengthens the likelihood of word-initial voiced stops acting as depressor consonants; this eventually resulted in the phonologization of three level tones. However, this doesn’t account for all of the data. 169 In terms of PKMN tone, some residual issues with word-initial stops are: i. Cognates in tone Set C in which both or either Gwama and/or Dana-Opo do not exhibit devoicing, ii. residual tone “correspondences” with very few members, iii. cognate sets which do not display reconstructable tone patterning. I recognize that these issues cannot be solved here and require further investigation. Nevertheless, I elaborate on these issues in order to provide a foundation for future efforts at reconstructing PKMN tone. There are a handful of cognates in which a word-initial voiced stop onset precedes L tone. This occurs in Gwama and the Dana-Opo branch, which contradicts the analysis presented in §3.1.2 that these languages devoiced word-initial voiced stops in tone set C contexts (*L [XL] for convenience). One issue is the reliability of the cognates that display this exceptional pattern as the cognates are usually plant, animal, or place names, which could have been borrowed. Some examples are seen in (117). (117) Gwama Komo Uduk Dana Opo Gloss a. bàbá bǎ àbàbá – àbá ‘father’ b. ūdùmɪ̀ dùmɛ̀ dùmáj dùmàj ‘tree (sp.)’ c. gʊ̀s gùʃ gùs – – ‘run, flow’ d. gjǎnpá – ɟàpʰá – – ‘throwing club’ While the Dana-Opo branch always exhibits a voiceless unaspirated stop with tone correspondence set C (*L [XL]), with the exception of the cognates in (117), the Gwama cluster displays more erratic behavior. Gwama does indeed exhibit historical devoicing of voiced stops before the *L [XL] allotone but there are cases in which what appear to be cognates do not exhibit this change. These reflexes are seen in (118). 170 (118) Gwama Komo Uduk Dana Opo Gloss a. bùʃùl bùʃ bùʃ – pùsà ‘belly’ b. bǔsʼ – bùt ̪̓ – pʰǔtʃʼ ‘choke’ c. dàlísʼ dɪ̀l dìl – tɪ̀lɪ ́ ‘stomp’ d. gɪ̀m – ɟìm – – ‘stuff, pack in’ e. gʊ̀s gùʃ gùs – – ‘run, flow’ f. gjǎnpá – ɟàpʰá – – ‘throwing club’ Recall that following voiceless stops, Dana and Komo retain PKMN *L > L tone while all other languages shift PKMN *L > M tone. Evidence for this tone pattern requires reflexes in Dana and Komo in order to determine whether they exhibit a synchronic L tone that corresponds to M tone in the other languages. There are four cognates, which all contain reflexes with M tone, seen in (119).132 While I can offer no explanation, I do not believe four reflexes constitutes enough evidence to posit an additional proto-tone category. Note also that Dana unexpectedly exhibits H tone in ‘fear’. (119) Gwama Komo Uduk Dana Opo Gloss a. *tʼwānkʼ bāɗāgí ɗākʰ ɗāgí ɗāgı ̋ ‘scorpion’ b. tʼwā tʼā tʼwā t ̪̓ āʔá tʼā ‘mouth’ c. pʼɪk̄ʼɪń pʼīn tʼīpʼīɲ pʼɪn̄ā pʼɪn̄ā ‘ash’ d. kwāgà kɔ̄g kʰɔ̄kʼ kʰɔ́k kʰɔ̄gɔ́ ‘fear’ Lastly, there are cognates that do not show any reconstructable patterning in tone. Some patterns may reflect erratic behavior in one intermediate proto-language, as in (120), in which no general tone correspondence can be found. 132 Note that Dana exhibits M tone in (119d) and the full Gwama reflex in (119a) is a reduplicated form tʼwānkʼɪt̄ʼwānkʼ ‘scorpion’ 171 (120) Gwama Komo Uduk Dana Opo Gloss M H H H H a. kɪl̄ à+kɪĺ cɛɲ́ akɪĺ kɪĺ ‘cattle egret’ H H M H H b. síʔ ʃúmákʼ sīmāʔ ʃʊ́j sʊ́j ‘bone’ Overall, there appears to be more inconsistency in synchronic tone correspondences with word-initial non-stop consonants than there is with word-initial stop consonants as seen in (121).133 No robust tone pattern can be found for the former. (121) Gwama Komo Uduk Dana Opo Gloss M M H M L a. sʼɛ ̄ sʼɛʔ̄ ʃʼɛ ́ kʼɛ ̄ tʃʼɛ̀ ‘tooth’ H H M L H b. sɔ́m ʃɔ́m sām sɔ̀m sɔ̄m ‘warm oneself’ M L M L L c. ʃʊ̃ʃ̄ ʃʊ̀nʃ ʃūʃ ʃʊ̀ʃ sʊ̀sʊ̀ ‘nose’ Overall the data suggest a strong correspondence between word-initial stop onsets and historical tone. This is seen not only by the fact that historical tone categories can be clearly reconstructed following word-initial stop onsets but also by the subsequent independent innovations of stop devoicing robustly in PDAOP and marginally in PGW. While the historical scenario proposed here is not without exceptions, and there is 133 I realize that this is an impressionistic observation that requires further research. 172 residue that cannot be accounted for, the notion of a consonant-tone constellation is crucial to reconstructing consonants and tone in Koman. 3.1.5 Observations on tonogenesis in Pre-Koman Voicing in word-initial stops among other obstruents has been well documented as interacting with the pitch realization of the following vowel. This can ultimately result in historical contrastive pitch, or tone, as Hombert (1978:78) observes: “The development of contrastive tones on vowels due to the loss of a voicing distinction on obstruents in prevocalic position is probably the most well documented type of tonogenesis. When such a development occurs, a relatively lower pitch register develops on vowels following the previously voiced series, and a relatively higher pitch is found after the previously voiceless or voiceless aspirated series. This process can lead to a multiplication by two of the number of tones. If the language is atonal, it will have two tones after this development; an already existing two-tone system can be transformed into a four-tone system, and so on.” In a study of tonogenesis in Chadic, Wolff (1987:199) observes that “the voicing distinction of syllable-initial obstruents plays the all-decisive role in assigning pitch realizations to following vowels”. Further, while “[…] the voicing distinction is still phonologically operative in these languages, there is no indication whatsoever that the contrast between voiced and voiceless obstruents, for instance, is being given up.” This observation appears to parallel at least one stage in the history of Pre-Koman tone in which there is a strong correlation between the voicing of the word-initial (syllable onset) stop and the tone of the following vowel.134 While this dissertation does not aim to reconstruct tonogenesis in Pre-Koman, the historical scenario for the evolution of 134 At present, there does not appear to be a relationship between the final consonant of a (CVC) root and the tone on the preceding vowel, though I recognize this requires further investigation. 173 tone and the interaction with stop onsets discussed in §3.1.2 does harken to Wolff’s observations in Chadic. I leave this issue for future research. We now turn to focus more specifically on the reconstruction of PKMN consonants. 3.2 Proto-Koman (PKMN) consonants The proposed Proto-Koman consonant inventory is in Table 48. A total of 29 consonants can be reconstructed not including four marginal consonants. One noteworthy observation is the reconstruction of an interdental stop series, with the exception of *t ̪h for which there is marginal evidence. Bender (1983) did not reconstruct an interdental series of stops and attributed the interdental series seen moderbly in Chali Uduk as an Uduk innovation. Bender did not have data from Dana, which exhibits interdental stops that correspond to Chali Uduk interdental stops. Another important result is the reconstruction of a series of voiceless aspirated stops, where Bender (1983) only reconstructed *kʰ. Note, however, that *cʰ cannot be reconstructed. Other notable features of the PKMN consonant inventory include ejective stops in all five places of articulation as well as a voiceless alveolar fricative/affricate ejective *sʼ. 174 Table 48 Proto-Koman (PKMN) consonant inventory (Parentheses indicate marginally reconstructable phonemes) *pʰ (*t ̪h ) (*tʰ) *kʰ *p *t ̪ *t *c *k *b *d̪ *d *ɟ *g *pʼ *t ̪̓ *tʼ *cʼ *kʼ *ɓ *ɗ *T̪ *s *ʃ *h *D̪ *sʼ *m *n (*ɲ) (*ŋ) *l *r *w *j The history and development of word-initial PKMN consonants has undergone a number of noteworthy innovations, some of which were a function of the tone in the syllable nucleus, as described in §3.1. In the following subsections I examine each proto- phoneme individually and discuss the pertinent PKMN innovations. I provide correspondence sets for each proto-phoneme, according to the schema seen in Table 49. The first column indicates a “correspondence set number” that is unique to a particular correspondence set.135 This is followed by the reconstructed proto-phoneme. Under each language or language cluster, I indicate the varieties such as “Lowland” (Lo), “Yabus” (Yab), etc.136 135 Sets 1-5 comprise the bilabial stops, sets 6-9 are interdentals, sets 10-15 are alveolars, sets 16-18 are palatals, sets 19-22 are velars, sets 23-25 are the voiceless fricatives, sets 26-27 are the lateral and trill, sets 28-31 are nasals, and sets 32-33 are glides. 136 The language varieties are as follows: Lo= Lowland Gwama, Hi= Highland Gwama, Yab= Yabus Uduk, Cha= Chali Uduk, Bil= Bilugu Opo, Mod= Modin Opo, Pam= Pame Opo, Kig= Kigile Opo. 175 Table 49 Schematic for PKMN correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig n *C C C C C C C C C C C After discussing a correspondence set, I provide a short set of data to illustrate the correspondence. In the illustrative data sets, I do not indicate the specific language variety for issues of space (e.g. Lowland Gwama), but the appendices provide all of the data used to establish every correspondence set. Thus, if the reader wishes to identify the particular variety of a particular lexical reflex, this information can be found in the appendices. Further, the reader can find there the data employed in reconstructing every consonant in this study. 3.2.1 PKMN bilabial obstruents The proto-Koman bilabial obstruent series contrasts across five manners of articulation: voiceless aspirated, voiceless unaspirated, voiced, ejective and implosive. The following five bilabial obstruents can be confidently reconstructed to PKMN *pʰ, *p, *b, *pʼ, *ɓ. Uduk, Dana and Opo are the most conservative languages, retaining all of the PKMN bilabial obstruents. Komo and Gwama exhibit parallel mergers of the voiceless aspirated and voiceless unaspirated bilabial stops: *pʰ + *p > p.137 Further, Gwama merged the bilabial implosive with the bilabial ejective *ɓ > *pʼ. 137 To be clear the grapheme

employed as the result of the merger does not represent a voiceless unaspirated consonant rather the merger of voiceless unaspirated and voiceless aspirated. Gwama and Komo do not display contrastive aspiration. 176 3.2.1.1 PKMN *pʰ A voiceless aspirated bilabial stop *pʰ can be reconstructed to PKMN in word initial/medial and word-final positions. The PKMN *pʰ correspondence set 1 is seen in Table 50. Reflexes of PKMN *pʰ are retained as [pʰ] in all of the languages except for Komo and Gwama, which independently merged PKMN *pʰ > *p.138 Table 50 PKMN *pʰ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 1 *pʰ p p p pʰ pʰ pʰ pʰ pʰ pʰ pʰ Some cognates containing reflexes of PKMN *pʰ are seen in (122) and the full set of cognates containing reflexes of *pʰ can be seen in Appendix D1. (122) PKMN Gwama Komo Uduk Dana Opo Meaning *pʰàd̪ pāl pàj pʰɛ ̄ pʰàd̪ pʰāj ‘fly (v.)’ *pʰuj pɪ ̄ pɪ̀ pʰí pʰùj pʰű ‘blow (with mouth)’ *úpʰ ʊ́p úp úpʰ úpʰ űpʰ ‘bathe’ 3.2.1.2 PKMN *p A voiceless unaspirated stop *p can be confidently reconstructed to PKMN in correspondence set 2 seen in Table 51. Uduk, Dana and Opo retain [p] reflexes of *p while Komo and Gwama merge *p > *pʰ.139 138 Whether or not these mergers in Komo and Gwama were also the result of close contact is undeterminable at present. 139 Gwama and Komo exhibit independent mergers of all *Cʰ > C stop consonants and as a result, are the only Koman languages that do not exhibit synchronic contrastive aspiration. 177 Table 51 PKMN *p correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 2 *p p p p p p p p p p p Some word initial reflexes of *p are in (123). As alluded to in §3.1, historical PKMN tone plays a factor in the development of word-initial *p. All word-initial reflexes of *p for which tone can be reconstructed exhibit PKMN tone correspondence Set A or Set B.140 Recall that in tone set A, all languages have synchronic H tone, as seen in (123a- b).141 In tone set B, all word-initial reflexes of PKMN *p in Dana and Komo have L tone corresponding to M elsewhere, as seen in (123c).142 It is important to mention here that there are no reflexes of word-initial PKMN *p that reconstruct to PKMN tone correspondence set C, in which all languages exhibit synchronic L tone. The full set of *p reflexes is provided in Appendix D.2. (123) PKMN Gwama Komo Uduk Dana Opo Meaning a. *pít ̪h pɪt́ pít – pít ̪h – ‘vagina’ b. *piʃ pīs píʃ píʃ píʃ pıs̋ ‘disregard’ c. *pàD pāt pàt pār – – ‘crawl, touch’ 140 See §3.1 for discussion of PKMN tone. 141 Also recall that the XH tone in Opo in (123b) is the result of a POP innovation which split *H > XH tone on [+high, +ATR] vowel nuclei /i, u/, and > H elsewhere 142 I cannot explain the H tone in Dana in (123c) though the rest of the languages exhibit the PKMN tone Set B correspondence which reconstructs to the *L tone category in PKMN. 178 Gwama and Komo share many cognates with an initial /p/ that do not have cognates in the other languages. As such, determining whether the initial voiceless bilabial stops in these lexemes are reflexes of PKMN *pʰ versus *p is not possible given the fact that a reflex in a language that retains contrastive aspiration is the indicator of the correspondence set to which the reflex belongs. Another possibility is that these lexemes were innovations in either Komo or Gwama that were then borrowed through contact, though this remains to be investigated. 3.2.1.3 PKMN *b A voiced bilabial stop *b can be reconstructed to PKMN in the two correspondence sets seen in Table 52. While the existence of two correspondence sets suggest reconstructing two proto-phonemes, the conditioning for these two PKMN *b correspondence sets in word-initial position is the reconstructed tone category of the following vowel nucleus. Table 52 PKMN *b correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 3 *b b b b b b b b b b b 3a *b p p b b b p p p p p The two *b correspondence sets appear to be in complementary distribution with respect to the proto-tone categories with which they occur. To illustrate, word-initial reflexes in *b correspondence set 3 can occur before PKMN tone correspondence set A, 179 in which all languages exhibit H tone except for Uduk which exhibits R tone, as seen in (124).143 (124) PKMN Gwama Komo Uduk Dana Opo Meaning *bángwà báŋgà bánkɔ̀ bǎŋgɔ̀ – – ‘cannabis’ *bár bár bár – bár bár ‘heron’ By contrast, reflexes in the PKmn *b correspondence set 3a, in which voiced bilabial stops in Komo and Uduk correspond to voiceless unaspirated stops in Gwama and the DAOP branch, only occur with PKMN tone correspondence set C, in which all languages exhibit synchronic L tone. Recall that PGW and PDAOP split and merged word-initial *b > p in tone set C and retained /b/ elsewhere. Some cognates containing reflexes of PKMN *b are in (125); the full set of reflexes is provided in Appendix D3a. (125) PKMN Gwama Komo Uduk Dana Opo Meaning *bit⁽ʰ⁾ pɪ̀t bìt bìt – – ‘toss’ *bwaŋ(a) pwǎŋ – bwàj – – ‘path, towards’ *bɔ̀ɓ pɔ̀pʼ bɔ̀b – – pɔ̀j ‘skulk’ 3.2.1.4 PKMN *ɓ All of the living Koman languages with the exception of the Gwama varieties exhibit bilabial and alveolar implosives. A bilabial implosive *ɓ can be confidently reconstructed to PKMN. All of the languages have retained a bilabial implosive reflex /ɓ/ of PKMN *ɓ with the exception of Gwama, which merged PKMN *ɓ with *pʼ. 143 Exceptions to this observation are discussed in the section on PKMN tone residue in §3.1.4. 180 Table 53 PKMN *ɓ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 4 *ɓ pʼ pʼ ɓ ɓ ɓ ɓ ɓ ɓ ɓ ɓ Some cognates containing reflexes of word-initial PKMN *ɓ are seen in (126) and the full set of correspondences is provided in Appendix D4. (126) PKMN Gwama Komo Uduk Dana Opo Meaning *ɓId̪a pʼɪl̄ ɓà ɓā ɓìd̪à ɓījā ‘neck’ *ɓít ̪̓ pʼɪ ́ ɓísʼ ɓít ̪̓ – – ‘strong’ *bɔŋkʼɔ́ (pʼà-)bɔ̌ngɔ́ (bā-)ɓɔ̌nkʼɔ́ – (à-)ɓɔ̄nkʼɔ́ (à-)ɓɔ̄nkʼɔ́ ‘frog’ Synchronically, implosives are restricted to word-initial position in most of the Koman languages, with the exception of Chali and Bonya Uduk, which exhibit implosives in all positions. Exceptional intervocalic implosives outside of Uduk can be attributed to historical implosives in word-initial position which became intervocalic due to the lexicalization of a gender prefix. This is seen in some of the reflexes of ‘frog’ in (126). The reflexes of ‘frog’ contain synchronic intervocalic bilabial implosives in Komo, Dana and Opo but historically the root was *ɓɔnkʼɔ, to which gender/number prefixes lexicalized. Note that in the Gwama lexeme *(pʼà-)bɔ̌ngɔ́, the *ɓ reflex is realized as /b/ intervocalically.144 144 See §2.1.4.2 for a description of intervocalic voicing of ejectives in Gwama and see §2.2 for a description of Koman number/gender morphology. 181 3.2.1.5 PKMN *pʼ A bilabial voiceless ejective *pʼ can be reconstructed to PKmn, based on correspondence set 5, seen in Table 54. PKMN *pʼ is retained in word-initial position as /pʼ/ in all of the languages. In word-final position, /pʼ/ is retained in all languages except the Uduk cluster. In word-final position, reflexes of *pʼ are often realized as unreleased or even delayed release, often resulting in full closure that lasts several seconds.145 Table 54 PKMN *pʼ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 5 *pʼ pʼ pʼ pʼ pʼ pʼ pʼ pʼ pʼ pʼ pʼ Selected reflexes of *pʼ are seen in (127) and the full set of reflexes are provided in Appendix D5.146 (127) PKMN Gwama Komo Uduk Dana Opo Concept *(tʼi)pʼIkʼIɲ(a) pʼɪk̄ʼɪń pʼīn tʼīpʼīɲ pʼɪn̄ā pʼɪn̄ā ‘ash’ *pʼɔt ̪̓(a) pʊ́t pʼɔ̀tʼ – pʼɔ̀t ̪h à pʼɔ̄tʼā ‘pick’ *wɔpʼ ~ hɔpʼ hɔ̌bɔ̀sʼ wɔ́pʼ kʰɔ̀bɔ̀sʼ hɔ̀pʼ hɔ̄pʼ ‘sip liquid’ The realization of the reflexes of *pʼ in intervocalic position is more unpredictable– some languages exhibit intervocalic /pʼ/ while others exhibit /b/ reflexes. Voicing of 145 It is common in the Koman languages for glottalized consonants to be unreleased in word-final position, especially in careful elicited speech. This has also been described in Chali Uduk (Killian 2015), Lowland Gwama (Goldberg 2018), and Komo (Otero 2018b). 146 My Uduk consultants provided tʼīpʼīɲ ‘ash’ while Beam & Cridland (1970) and Killian (p.c.) have pʼīɲ. I employ parentheses to indicate the discrepancy. The source of the initial segment tʼī in my transcription remains unknown. 182 ejective stops in intervocalic position appears to be an ongoing phonetic process in some of the Koman languages. 3.2.2 PKMN interdental stops The only Koman languages that display a synchronic contrastive interdental series of stops are in two main branches of Central Koman: Dana from the DAOP branch and the Chali and Bonya Uduk varieties of the KOUD branch.147 Synchronically, these stops are clearly interdental – articulated with the tongue between the teeth during closure followed by a release which may have some minor frication. Koman may have developed the interdental stops via contact with Western Nilotic. Phonemic interdental stops contrasting with alveolar stops are attested in all three branches of Western Nilotic, some of which are spoken in areas contiguous with Koman speakers.148 Given the fact that only two Koman languages display contrastive interdental stops, a particular challenge for reconstructing interdental stops to PKMN is finding cognates in Chali Uduk and Dana words that contain interdental reflexes. While such apparent cognates do occur, one must also weigh the validity of a proposed PKMN cognate against possible borrowing or contact. All of this taken into account, there does seem to be enough evidence for the following PKMN interdental stops: *t,̪ t ̪̓ and d̪. Evidence for reconstructing a voiceless aspirated interdental stop *t ̪h to PKMN is less convincing. 147 I was unable to precisely determine whether Bonya Uduk exhibits interdental stops as my time with a Bonya consultant was limited. The Bonya and Chali varieties are very close and I expect that Bonya does in fact display contrastive interdental stops, though I have opted to not include Bonya in this study due to unreliable data. 148 Contrastive interdental stops are attested in the Southern Burun branch in Mabaan (Andersen 1999a) and Jumjum (Andersen 2004), in the Northern Burun branch in Kurmuk (Andersen 2007) and Mayak (Andersen 1999b), in the Dinka-Nuer branch in Agar Dinka (Andersen 1987a) and in Nuer (Huffman 1929), as well as in the Luo branch in Päri (Andersen 1988). Andersen (1987b) also describes contrastive interdental stops in Lulubo, a Central Sudanic language spoken in South Sudan. 183 3.2.2.1 PKMN *t ̪h There is scant evidence for reconstructing *t ̪h to PKMN. Given that Chali Uduk and Dana are the only languages exhibiting interdental stops, a voiceless aspirated interdental stop reflex /t ̪h / must occur in at least one of the languages to even consider reconstructing *t ̪h . If *t ̪h were to be reconstructed, the correspondence set is seen in Table 55. The Uduk and Opo varieties merge *t ̪h with tʰ, while Gwama and Komo independently eventually merge all of the voiceless (aspirated and unaspirated) interdentals and alveolar stops *t ̪h > tʰ > t > t ̪> t.149 Table 55 PKMN *t ̪h correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 6 *t ̪h t t t tʰ tʰ t ̪h tʰ tʰ tʰ tʰ In my database, there are no cognates that contain interdental [t ̪h ] in both Chali Uduk and Dana. The only PKMN cognates with word-initial /t ̪h / reflexes in Dana are seen in (128). Of the two that have word-initial /t ̪h /, only ‘kick’ contains a reflex in Chali Uduk. Note that in ‘kick’, Chali Uduk displays a voiceless aspirated alveolar word-initial reflex /tʰ/. Given the paucity of data, I tentatively reconstruct *t ̪h to PKMN. (128) PKMN Gwama Komo Uduk Dana Opo Meaning *t ̪h u tū – – t ̪h úwà tʰűjhá ‘spit (v.)’ *t ̪h áɓ tápʼ táb tʰáɓ t ̪h áp tʰáp ‘kick’ *pít ̪h pɪt́ pít – pít ̪h – ‘vagina’ 149 See §4 for a more detailed discussion of the evolution of consonants in each Koman subnode. 184 3.2.2.2 PKMN *t ̪ A voiceless interdental stop can be reconstructed to PKMN based on the correspondence set in Table 56. In *t ̪set 7, Dana and Chali Uduk retain /t/̪ reflexes while all other languages merge reflexes of *t ̪with reflexes of *t. Table 56 PKMN *t ̪correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 7 *t ̪ t t t t t ̪ t ̪ t t t t The most solid evidence for reconstructing a single voiceless unaspirated interdental stop *t ̪is if both Chali Uduk and Dana exhibit voiceless interdental stop /t/̪ reflexes. Further evidence to support PKMN *t ̪is if the remaining languages exhibit alveolar stops. The two *t ̪cognates that have reflexes with interdental stops in both Chali Uduk and Dana and also have corresponding voiceless alveolar stops elsewhere in the family are seen in (129). It is crucial to note that none of the *t ̪reflexes can be reconstructed to PKMN tone set C, which only occurred following voiced stops (§3.1). (129) PKMN Gwama Komo Uduk Dana Opo Meaning *tú̪k⁽ʰ⁾(u) tʊ̌kʊ̀ túk tú̪kʰ tú̪kʰ tűkʰ ‘acacia’ *(ɔ)t⁽̪ʰ⁾ɔɗ twɛj́ tɔ́ tɔ̪́ɗ ɔ̀t ̪h ɔ́ ɔ́tɔ́ ‘grind (second grind)’ There are two cognates with [t]̪ reflexes in Dana corresponding to /t/ elsewhere except Gwama, which exhibits /s/, as seen in (130). There does not appear to be any conditioning outside of possible spirantization before a high front vowel in (130)a, but that does not account for (130)b. As such, these Gwama reflexes are treated as 185 idiosyncratic *t ̪and *s mergers. All of the data employed in reconstructing *t ̪can be found in Appendix D7. (130) PKMN Gwama Komo Uduk Dana Opo Meaning ‘roughen a. *tɪ̪t́ ̪ sɪt́ tɪt́ títʰ tɪ̪t́ ̪h tɪt́ɪ ́ stone (for grinding)’ b. *tU̪b(a) sʊ́ – tūpʰ tú̪bá tʰūbá ‘pierce’ 3.2.2.3 PKMN *d̪ A voiced interdental stop *d̪ can be reconstructed to PKMN based on the correspondence sets seen in Table 57. PKMN *d̪ correspondence set 8 can be reconstructed to initial, medial and final positions, while set 8a only occurs in word- initial/syllable onset position. For word-initial reflexes, these two *d̪ correspondence sets are in complimentary distribution with respect to the historical tone of the following vowel nucleus, which is described below. Table 57 PKMN *d̪ correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 8 *d̪ d d d d d̪ d̪ d d d d 8a *d̪ t t d d d̪ t ̪ t t t t In PKMN *d̪ set 8, Chali Uduk and Dana retain /d̪/ while mergers of *d̪ with /d/ occur elsewhere. Some reflexes of *d̪ in correspondence set 8 in initial and final positions are seen in (131); the full set of reflexes are provided in Appendix D8. In word-initial position, set 8 is limited to cognates which do not reconstruct to PKMN tone set C, in 186 which all languages exhibit L tone synchronically (§3.1.1). In medial/final position there is evidence for overall weakening [d̪ > d > r > l > j > ∅] though there does not appear to be a consistent pattern. Note that word-final devoicing and/or elision is common across Koman. (131) PKMN Gwama Komo Uduk Dana Opo Meaning *d̪ɛrkʼɛsʼ dɛr̄gɛśʼ – d̪ɛ̀rɛ̀s d̪ɛŕkʼɛś – ‘slip (v.)’ *sud̪(i) ʃʊ́l ʃùʔí sū sùd̪ swī ‘beer’ *gUd̪Um kʊ̀rʊ́m gùdúm – gùd̪ùm kùdùmà ‘pig’ The reflexes in PKMN *d̪ set 8a only occur with PKMN tone set C, in which all languages have L tone (§3.1.2). Historically, word-initial stops tone lowered the F0 on the following vowel in this PKMN *L correspondence set. Subsequent to this, PGW and PDAOP independently devoiced the word-initial voiced stops occurring with this lowered tone and merged them with the voiceless unaspirated stops. This pattern occurs across all of the voiced stop series. Some reflexes of *d̪ from set 8a are seen in (132) and the full set is provided in Appendix D8a. Note that Gwama, Komo and Yabus Uduk spirantize reflexes of *d̪ > z before high front vowels, as seen in ‘strain’. (132) PKMN Gwama Komo Uduk Uduk Dana Opo Meaning (Yabus) (Chali) *d̪àm tàm dàm dàm d̪àm tâ̪m tàm ‘honey’ *d̪ìm zìŋā zìm zìm d̪ìm tì̪m tìm ‘strain’ 187 3.2.2.4 PKMN *t ̪̓ An interdental ejective *t ̪̓ can likely be reconstructed to PKMN based on the correspondence sets in Table 58. PKMN *t ̪̓ set 9 occurs only in word-initial/syllable onset position, while set 9a occurs only in word-final position. Table 58 PKMN *t ̪̓ correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 9 *t ̪̓ tʼ tʼ tʼ tʼ tʼ t ̪̓ tʼ tʼ tʼ tʼ 9a *t ̪̓ tʼ tʼ tʼ d ɗ t ̪̓ tʼ tʼ tʼ tʼ Given the limited data available for *t ̪̓ reflexes in word-initial (syllable onset) position, it tentatively appears that Dana retains an interdental ejective reflex /t ̪̓/ while elsewhere in the family *t ̪̓ > tʼ. Note that according to these data, it appears that in Chali Uduk, *t ̪forms part of a chain shift wherein *t ̪̓ > tʼ and *sʼ > t ̪̓ .150 The word-initial or syllable onset reflexes of *t ̪̓ are in (133) and the full set of data is in Appendix D9. (133) PKMN Gwama Komo Uduk Dana Opo Meaning *t ̪̓wa tʼwā tʼwā tʼā t ̪̓ āʔá tʼā ‘mouth’ *t ̪̓wI tʼwɪ ́ – – t ̪̓wī – ‘enter’ *hat ̪̓is hǎtʼìʃ -- -- hát ̪̓ìs ha̋tʼīs ‘sneeze’ *t ̪̓ ɛn sʼɪn̄ sʼɛń tʼɛń t ̪̓ ɛń tʼɛń ‘alone, abstain’ 150 See §3.2.3.6 for discussion of PKMN *sʼ. 188 In word-final position, PUD shifts *t ̪> ɗ which leads to a merger in Yabus Uduk of reflexes of *ɗ with those of *d. Some examples of cognates with word-final *t ̪̓ reflexes are seen in (134) and the full set of data is provided in Appendix D9a. (134) PKMN Gwama Komo Uduk Dana Opo Meaning *sɪ̀t ̪̓ ʃɪt̄ʼ ʃɪ̀tʼ sīɗ sɪ̀t ̪̓ sɪt̄ʼ ‘far (be)’ *mɛ̀t ̪̓ mɪ̀tʼ – mɛ̀ɗ mɛ̀t ̪̓ mɪ̀tʼɪ ́ ‘hand’ *kʼwànt ̪̓ kʼwāntʼɪ ́̓kwāntʼ kʼwàtʼ kʼwāɗ kʼwàt ̪̓ kʼwātʼ ‘tick’ 3.2.3 PKMN alveolar stops and alveolar affricate ejective A full series of alveolar stops can be reconstructed in word-initial position to PKMN with the exception of *tʰ, which can be marginally reconstructed in word-final position only. There is robust evidence for reconstructing an alveolar affricate/fricative ejective *sʼ. 3.2.3.1 PKMN *tʰ An aspirated alveolar stop *tʰ cannot be reconstructed to PKMN in word-initial position, though what appear to be reflexes of PKMN *t can tentatively be reconstructed in word final position, as seen in correspondence set 10 in Table 59. This correspondence set does not instill much confidence in PKMN *tʰ for two important reasons. First, word-final devoicing (and subsequent aspiration) is common among all of the languages and the reflexes of word-final *tʰ could in fact be reflexes of *t, *t,̪ *t ̪h or *d.151 Lastly, *tʰ correspondence set 10 does not contain any cognates in Dana. Recall 151 Note that phonological contrast between voiced and voiceless stops across the languages varies and for some languages, adding a vowel-initial suffix to a root containing a word-final aspirated stop can 189 that in PKMN *t ̪h , Dana retains the interdental aspirated stop while Chali Uduk merges *t ̪h > tʰ. With these facts taken into account, reconstructing *t to PKMN remains tentative at best.152 Table 59 PKMN *tʰ word-final correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 10 *tʰ t t t tʰ tʰ – tʰ tʰ tʰ tʰ Some cognates containing reflexes of PKMN *tʰ in word-final position are seen in (135) and the full set of cognates is provided in Appendix D10. (135) PKMN Gwama Komo Uduk Opo Meaning *bit⁽ʰ⁾ pɪ̀t bìt bǐtʰ – ‘toss’ *wutʰ wǔt wǔt ʔútʰ hűtʰ ‘ostrich’ *kUtʰ gʊ̄t kùt kūtʰ – ‘short (be)’ 3.2.3.2 PKMN *t A voiceless unaspirated alveolar stop *t can be reconstructed to PKMN via the correspondence set seen in Table 60. In *t correspondence set 11, Komo and Gwama independently merge reflexes of PKMN *t with reflexes of PKMN *tʰ and /t/ is retained elsewhere in the family (§3.2.3.1). surface as a voiced stop. I have done my best to keep track of this variation, but I do recognize my possible shortcomings in transcriptions. 152 Note also that in the PCTRL *tʰ correspondence set 10, there are a few cognates in Dana but not in Chali Uduk. I can only confidently reconstruct *t to the PUD and POP subnodes. See §4 for discussions of PCTRL, PUD and POP phonology. 190 Table 60 PKMN *t correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 11 *t t t t t t t t t t t Some cognates containing reflexes of PKMN *t in word initial position are seen in (136); the full set of reflexes is seen in Appendix D11. Note that the tone on the vowel following all initial *t reflexes cannot be reconstructed to PKMN tone set C, as this tone correspondence set only occurred with historical word-initial voiced stops (§3.1). (136) Gloss PKMN Gwama Komo Uduk Dana Opo ‘tall (be)’ *tUr tǔ tʊ́l túr – – ‘shake (sth.)’ *tEŋ(g)(E) tɪḡɪ ̄ – – tɛŋ́ tɪŋ́há 3.2.3.3 PKMN *d There is solid evidence for reconstructing a voiced alveolar stop *d to PKMN, based on the two correspondence sets in Table 61. In *d set 12 all languages retain a /d/ reflex of *d, while in set 12a, Gwama and the Dana-Opo branch exhibit /t/ corresponding to /d/ in the Komo-Uduk branch. These two correspondence sets are in complementary distribution word-initially with respect to the historical tone on the following vowel. Table 61 PKMN *d correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 12 *d d d d d d d d d d d 12a *d t t d d d t t t t t 191 Word-initial reflexes in *d set 12 all occur with synchronic H tone or reconstructed PKMN tone set A. Some examples are seen in (137); the full set of cognates with reflexes for this correspondence set are seen in Appendix D12. (137) PKMN Gwama Komo Uduk Dana Opo Meaning *dʊ́ɟɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ – dʊ́ɟɛ̀ ‘pipe’ *dwakʰ dwâk dɔ̂k – dwákʰ dwa̋kʰ ‘bird_weaver’ Word-initial reflexes in *d set 12a all occur with synchronic L tone which can be reconstructed to PKMN tone set C. All of the word-initial voiced stops that reconstruct with tone set C became devoiced and merged with the voiceless unaspirated stops. Some examples of *d reflexes from *d correspondence set 12a are seen in (138); the full set is provided in Appendix 12a. (138) PKMN Gwama Komo Uduk Dana Opo Meaning *tOt tʊ̌t dɔ̀t dɔ̌tʰ – tɔ̄tɔ́ ‘ask’ *dùcʼá tùsʼ dʊ̀sʼ – tùcʼá – ‘urine_1, urinate_1’ *djalIsʼ dàlísʼ dɪ̀l dìl – tɪ̀lɪ ́ ‘stomp’ 3.2.3.4 PKMN *ɗ An alveolar implosive *ɗ can be reconstructed to PKMN based on the correspondence set in Table 62. In word-initial position, all of the languages retain /ɗ/ reflexes of *ɗ and Gwama merges *ɗ with tʼ.153 153 Anecdotally, in working with Gwama speakers that are fluent in Komo in a Komo literacy project, the Gwama speakers would often write a Komo alveolar implosive with an alveolar ejective grapheme. In general, I observed that Gwama speakers who are fluent in Komo found it difficult to perceive the implosive [ɗ] and [ɓ] sounds in Komo, often mistaking them for [tʼ] and [pʼ], respectively. 192 Table 62 PKMN *ɗ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 13 *ɗ tʼ tʼ ɗ ɗ ɗ ɗ ɗ ɗ ɗ ɗ Cognates with word-initial [ɗ] reflexes of *ɗ are seen in (139) and the full data set is provided in Appendix D13. (139) PKMN Gwama Komo Uduk Dana Opo Meaning *ɗ(w)ankʼI tʼwānkʼ ɗāgí ɗwākʰ ɗāgí ɗāgı ̋ ‘scorpion’ *ɗar(a) tʼájà ɗàr ɗār ɗɛ̀d ɗɛr̄ ‘send someone’ In medial and final positions there is considerable variation in the realization in PKMN *ɗ reflexes.154 There does not seem to be a strong enough pattern to posit mergers at any node or language in particular, but rather an overall tendency for deglottalization followed by general phonetic weakening. Some examples are provided in (140). A word-final /ɗ/ in Chali Uduk usually corresponds to /t, d, r, l/ in the rest of the family and there can be weakening to /j/. Note that there are no cognates that reconstruct to PKMN in which both Chali Uduk and Dana retain /ɗ/ reflexes of *ɗ, but there are many cognates that do at the PCTRL node. Further, Chali Uduk merges *t ̪with /ɗ/ in word-final position (see §3.2.2.4). 154 Recall that only Chali Uduk and Dana retain /ɗ/ in initial non-initial positions, while the rest of the family is restricted to /ɗ/ in word-initial position only. See §2.1 for synchronic phonological descriptions. 193 (140) PKMN Gwama Komo Uduk Dana Opo Meaning *wVɗ wɛť wàl wáɗ wâl – ‘become (angry)’ *burbuɗ bùrbūt – bǔɗ būrkʼùs būrkʼùs ‘dust’ *kʰáɗ(a) kálá kár kʰǎɗ kʰátá kʰátá ‘open’ One example of a full range of reflexes of medial/final *ɗ is seen in the cognate 3SG.M independent pronouns in (141). There appears to be a diachronic pathway of weakening *ɗ > d > r > l > j. (141) PKMN Gwama Gwama Komo Uduk Uduk Dana Gloss (Lo) (Hi) (Yabus) (Chali) *haɗ(i) hāl hāj hàr hádī áɗī hār 3SG.M 3.2.3.5 PKMN *tʼ A voiceless alveolar ejective can be marginally reconstructed to PKMN. There are only two cognates with possible *tʼ reflexes in initial position and four in medial/final position. Given the fact that Dana and Chali Uduk exhibit phonological contrast between /tʼ/ and /t ̪̓/ (Dana also exhibits contrastive /sʼ/), reflexes in one or both of these languages are crucial in determining the validity of reconstructing *tʼ to PKMN. A proposed correspondence set for *t is seen in Table 63. Table 63 PKMN *tʼ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 14 *tʼ tʼ tʼ tʼ tʼ tʼ tʼ tʼ tʼ tʼ tʼ 194 Some cognates with reflexes of *tʼ are seen in (142); and the full set of relevant cognates is in Appendix D14. Note that there is only one cognate with reflexes in all of the languages, ‘salt (made from ash)’.155 (142) P KMN Gwama Komo Uduk Dana Opo Meaning *tʼaʃ tʼāʃ tʼàʃ tʼāʃ tʼàʃ tʼās ‘salt (made from ash)’ *tʼwaŋ(k)a ~ ‘cut (split in half tʼāŋà – tʼwákʰ – – t ̪̓waŋ(k)a lengthwise)’ 3.2.3.6 PKMN *sʼ There is robust evidence for reconstructing an alveolar voiceless fricative/affricate ejective *sʼ to PKMN. The correspondence set for *sʼ is seen in Table 64. Gwama, Komo, Dana and Yabus Uduk retain a /sʼ/ reflex of *sʼ while Opo merges *sʼ with tʃʼ. Chali Uduk merges *sʼ with /t ̪̓/ as part of a chain shift involving *t ̪̓ and *tʼ (§4.6). Table 64 PKMN *sʼ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 15 *sʼ sʼ sʼ sʼ sʼ t ̪̓ sʼ tʃʼ tʃʼ tʃʼ tʃʼ Some cognates with reflexes of *s in initial and final positions are seen in (143); the full set of correspondences is in Appendix D15. 155 The gloss ‘salt (made from ash)’ is a substance traditionally made from the remains of burning a specific tree or plant. This results in a substance with which the Koman people season their food. 195 (143) Uduk Uduk PKMN Gwama Komo Dana Opo Meaning (Yabus) (Chali) *sʼÍk sʼɪ ́ sʼík sʼíʔ t ̪̓íkʰ – tʃʼıg̋ı ̋ ‘rat’ *sʼa sʼà sʼā sʼā t ̪̓ ā sʼʊ́wà tʃʼǎ ‘light (fire)’ *kʰʊ́sʼ kúsʼ kʊ́sʼ kʼúsʼ kʰút ̪̓ kʊ́sʼ kʰʊ́tʃʼʊ́ ‘dry (be)’ 3.2.4 Palatal consonants Only three palatal consonants can be reconstructed to PKMN. However, the evidence is not as robust as one would expect given the abundance of palatal consonants in both branches of Central Koman. Even though the Uduk varieties and Dana each exhibit a contrastive voiceless aspirated palatal stop /cʰ/, there is no evidence to reconstruct *cʰ to PKMN. Further, /cʰ/ appears to have been an independent innovation in PUD and Dana.156 3.2.4.1 PKMN *c A voiceless unaspirated palatal stop *c can tentatively be reconstructed to PKMN, though there are few cognates overall. Reconstructing *c seems plausible given that /c/ occurs in the two main branches of Central Koman: in the Uduk cluster of the KOUD branch as well as in all of the languages of the DAOP branch. The correspondence set for PKMN *c is seen in Table 65. Table 65 PKMN *c correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 16 *c s/ʃ s/ʃ s c c c tʃ tʃ tʃ tʃ 156 See §4.5 for PUD phonology and §4.9 for historical Dana phonology. 196 Some cognates that contain reflexes of PKMN *c are in (144); the full set of relevant cognates is in Appendix D16. While most of the languages retain /c/ reflexes, Komo merges *c with /ʃ/ and Gwama merges *c with /ʃ/ and *c with /s/. I find no conditioning for the Gwama mergers at present. (144) PKMN Gwama Komo Uduk Dana Opo Meaning *cwálá swálá sá cwá cáɟáʔ tʃá ‘tree’ *càkʰO sàkʊ́ sàkʊ́ – – tʃàkʰɔ́ ‘grandfather’ *cà ʃà – – cà tʃà ‘dig’ 3.2.4.2 PKMN *ɟ There is some evidence for reconstructing a voiced palatal stop *ɟ to PKMN based on the correspondence set seen in Table 66. The Dana-Opo branch retains a voiced (alveo-)palatal stop while Komo and Gwama independently shift *ɟ > z, and Yabus Uduk shifts *ɟ > ʒ. Note there are no PKMN *ɟ correspondences with reflexes in Chali Uduk. Bender (1983:284) notes the unexplained residue of /z/, which appears especially initially in Komo often corresponding to /c/ or /ɟ/ in other languages. Table 66 PKMN *ɟ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 17 *ɟ z z z ʒ – ɟ dʒ dʒ dʒ dʒ Some cognates with *ɟ reflexes are in (145) and the full set of relevant cognates is in Appendix D17. 197 (145) PKMN Gwama Komo Uduk Dana Opo Meaning *ɟàrú zɛ̀rú zɛ̀rú ʒàrú ɟɛ̀rú dʒɛ̀rű ‘stork’ *ɟana sjànà zɛ̀nā – – dʒɛ̀ná ‘sorghum’ *ɟaŋg(ɔ)aj zǎgʊ́ zǎgʊ́ ʒwǎŋgì ɟàŋwɛ̀j dʒāŋɔ́ ‘Nuer’ 3.2.4.3 A note on PKMN *c, *ɟ and tone One issue in the reconstructed plain palatal stop series is the fact that the historical tone correspondences seen with other voiced stops do not appear to hold for the palatal stops. All other voiced stop reflexes exhibit word-initial (syllable-onset) correspondences between voiced stops in the Komo-Uduk branch and voiceless unaspirated in the Dana-Opo branch and in Gwama. This occurs only when the tone of a particular etymon reconstructs to PKMN tone set C. As discussed in §3.1. Historically, in tone set C, the “low” tone following a voiced stop became phonetically lower than the “low” tone following a voiceless stop onset. This distinction in pitch eventually phonologized, which gave rise to contrastive three level tones. Synchronically, all reflexes of PKMN tone set C exhibit L tone. The issues with the reconstructed palatal stops *c and *ɟ are twofold. First, voiceless *c reflexes occur in correspondence sets which all exhibit synchronic L tone and in which we would expect a voiced onset in Komo such as (146a). Second, there are L tone reflexes in Dana and Opo that do not exhibit voiceless onsets, such as those seen in (145) and repeated here in (146b-d). 198 (146) PKMN Gwama Komo Uduk Dana Opo Meaning a. *càkʰO sàkʊ́ sàkʊ́ – – tʃàkʰɔ́ ‘grandfather’ b. *ɟàrú zɛ̀rú zɛ̀rú ʒàrú ɟɛ̀rú dʒɛ̀rű ‘stork’ c. *ɟana sjànà zɛ̀nā – – dʒɛ̀ná ‘sorghum’ d. *ɟaŋg(ɔ)aj zǎgʊ́ zǎgʊ́ ʒwǎŋgì ɟàŋwɛ̀j dʒāŋɔ́ ‘Nuer’ Note also that, unlike the bilabial, interdental and alveolar plain stops, the palatal stops can correspond to fricatives outside of the Dana-Opo branch. Perhaps the palatal stop series were not stops but rather affricates as seen in modern-day Opo. As such, the voiced affricate *dʒ did not form part of the PKMN proto-tone set C. This would explain the synchronic tone distribution of *c and *ɟ reflexes. Further, the correspondence between proto-affricates and fricatives seems articulatorily more likely. Further, we see that *cʼ ejective reflexes also correspond with fricative and affricate ejectives. 3.2.4.4 PKMN *cʼ There is ample evidence for reconstructing a voiceless palatal ejective *c’ to PKMN. The correspondence set for *cʼ, seen in Table 67, contains some unique innovations. PKMN *cʼ in is retained as /cʼ/ in Dana and the Uduk branch and as /tʃʼ/ in Opo. Yabus Uduk subsequently merges *cʼ with /ʃʼ/.157 Both Komo and Gwama merge *cʼ with /sʼ/ independently. 157 All cognates containing a /cʼ/ in Yabus Uduk are reflexes of *kʼ that palatalized before front vowels. See §3.2.5 for discussion. 199 Table 67 PKMN *cʼ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 18 *cʼ sʼ sʼ sʼ ʃʼ cʼ cʼ tʃʼ tʃʼ tʃʼ tʃʼ Examples of cognates with word-initial and word-final reflexes of *cʼ are seen below in (147).158 The full set of cognates containing reflexes of *cʼ can be found in Appendix D18. (147) PKMN Gwama Komo Uduk Dana Opo Meaning *cʼɛ sʼɛ ̄ sʼɛʔ̄ cʼɛ ́ kʼɛ ̄ tʃʼɛ̀ ‘ear’ *cʼɔ(tʼɔ)l sʼɔ́tʼɔ́ – cʼɔ̄cʼɔ̄lɔ̄cʼ cʼɔ̀ʔ tʃʼɔ̄ ‘drip’ *Bɪncʼ bɪn̄sʼ bɪńsʼ bǐcʼ bɪćʼ ɓītʃʼ ‘fishhook’ *wàcʼà wàsʼ wàsʼ wàcʼ wàcʼà wàtʃʼà ‘fish’ 3.2.5 PKMN velar stops Four velar stop consonants can be reconstructed to PKMN: *kʰ, *k, *g, *kʼ. One notable innovation in Proto-Uduk is a primary split via the palatalization of PKMN velar stops before front vowels. All synchronic Uduk lexemes that contain a velar stop followed by a front vowel are reflexes of erstwhile velar stops. Interestingly, there are no reflexes of *kʰ in Uduk which split and merged with *cʰ before front vowels that can be reconstructed to PKMN. Recall that *cʰ cannot be reconstructed to PKMN. The following subsections outline the history and development of the PKMN velar series of consonants. 158 I cannot account for the initial /kʼ/ in Dana in ‘ear’. 200 3.2.5.1 PKMN *kʰ A voiceless aspirated velar stop *kʰ can be confidently reconstructed to PKMN in word-initial, medial and word-final positions. The correspondence set for PKMN *kʰ is seen in Table 68. Note that all of the languages retain aspirated /kʰ/ reflexes while Komo and Gwama independently merge *kʰ with reflexes of *k. Table 68 PKMN *kʰ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 19 *kʰ k k k kʰ kʰ kʰ kʰ kʰ kʰ kʰ Some cognates containing reflexes of PKMN *kʰ are seen below in (148). Note that there are no cognates containing word-initial *kʰ reflexes that form part of PKMN tone correspondence set C (§3.1). If tone can be reconstructed in cognates with word-initial *kʰ reflexes, they will only correspond to PKmn tone set A or B. The full set of cognates with *kʰ reflexes is provided in Appendix D19. (148) PKMN Gwama Komo Uduk Dana Opo Meaning *kʰakʼa kāgā kàʔ kʰāʔ kʰàkʼà kʰākʼāʔ ‘bitter (be)’ *kʰʊ́sʼ kúsʼ kʊ́sʼ kʼúsʼ kʊ́sʼ kʰʊ́tʃʼʊ́ ‘dry (be)’ *kʰáɗ(a) kálá kár kʰǎɗ kʰátá kʰátá ‘open’ 3.2.5.2 PKMN *k A voiceless unaspirated velar stop *k can be confidently reconstructed to PKMN in word-initial, medial and word-final positions. There are two correspondence sets for PKMN *k in Table 69. In *k correspondence set 20, all languages retain voiceless 201 unaspirated /k/ reflexes while Komo and Gwama independently merge reflexes of *k and *kʰ into one synchronic velar stop /k/, which is synchronically unspecified for aspiration. Table 69 PKMN *k correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 20 *k k k k k k k k k k k 20a *k k k k c c k k k k k Cognates containing reflexes of *k from correspondence set 20 are seen in (149) and the full set of relevant cognates is provided in Appendix D20. (149) PKMN Gwama Komo Uduk Dana Opo Meaning *kum(bi) kùmbì kúm kūm kúmā kűmá ‘cover (v.)’ *kO(j) kù kò kɔ̄ kɔ̀j kwɛ ̄ ‘cry’ PKMN *k correspondence Set 20a reflects a Proto-Uduk innovation wherein all *k refl,exes palatalized before front vowels causing a split and merger of *k > *c as seen in (150) and in Appendix D20a.159 Synchronically, there are no lexemes in Uduk which contain a velar stop followed by a front vowel, indicating a complete merger. (150) PKMN Gwama Komo Uduk Dana Opo Meaning *kEɲ kɪl̄ kɪĺ cɛɲ́ kɪĺ kɪĺ ‘cattle egret’ *k⁽ʰ⁾ɪńáj kɪńá kɪńá cʰínáj kɪńáj – ‘Opo’ 159 I cannot account for why the Uduk reflex of *k is aspirated [cʰ] in ‘Opo’ instead of the expected [c]. This particular reflex is from the Yabus Uduk variety. 202 3.2.5.3 PKMN *g A voiced velar stop *g can be reconstructed to PKMN in word initial, medial and final positions. Table 70 contains the correspondence sets that reconstruct to *g. Word initially and word-medially, *g is retained in all of the languages as /g/ as seen in *g correspondence set 21. PKMN *g correspondence set 21a only occurs in word-initial position and the correspondence between voiced /g/ and voiceless unaspirated /k/ represents individual innovations in the PGW and PDAOP branches wherein they merged *g with *k. The conditioning factor for these two *g correspondence sets in word-initial position is the historical tone of the following vowel nucleus. Table 70 PKMN *g correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 21 *g g g g g g g g g g g 21a *g k k g g g k k k k k Voiced velar stop reflexes of *g are retained as /g/ in word-initial or syllable onset position when either the historical tone correspondence is PKMN tone set A (in which all languages exhibit H tone) as in (151a), or when the tone cannot be reconstructed but crucially no reflexes exhibit L tone as in (151b). See Appendix D21 for the full set of *g reflexes. 203 (151) PKMN Gwama Komo Uduk Dana Opo Meaning ‘bird (yellow- a. *gʊlɪla gʊ́lːā bāgʊ́lɪ̀lā – ɓāgʊ̂lːā ɓāgʊ́lʊ́l billed kite or black kite)’ b. *gaʃa gáʃà – gāʃáʔ gâʃ gātʃ ‘belt, sash’ Word-finally, *g is often devoiced to /kʰ/ or elided altogether as in (152). There is an overall tendency for word-final devoicing in Koman. (152) PKMN Gwama Komo Uduk Dana Opo Meaning *kʰOG kɔ́k – kʰúʔ kʰɛǵ kɛ̂k ‘giraffe’ *hag(a) hã́ʔ hág háʔ – hágá ‘have sex’ PKMN *g correspondence set 21a is by far the most robust, but is limited to word- initial position. In this set voiced /g/ reflexes in the KOUD branch correspond to voiceless unaspirated /k/ elsewhere. This correspondence set is conditioned by the historical tone of the following vowel nucleus. PKMN *g set 21a only occurs with tone correspondence set C, in which all languages modernly exhibit L tone (§3.1). Some examples of *g correspondence set 8a are in (153) and the full set of relevant cognates is in Appendix D8a. (153) PKMN Gwama Komo Uduk Dana Opo Meaning *gàm kàm gàm gàm kàm kàm ‘find, meet’ *gUd̪Um kʊ̀rʊ́m gùdúm – kùd̪ùm kùdùmà ‘pig’ 3.2.5.4 PKMN *kʼ A voiceless ejective *kʼ can be reconstructed to PKMN in word-initial, medial and final positions. The correspondence set for PKMN *kʼ is in Table 71. 204 Table 71 PKMN *kʼ correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 22 *kʼ kʼ kʼ kʼ kʼ kʼ kʼ kʼ kʼ kʼ kʼ All languages retain /kʼ/ in initial position as seen in (154). (154) PKMN Gwama Komo Uduk Dana Opo Gloss *kʼama kʼā kʼá kʼá kʼámá kʼámá ‘eat (hard food)’ *kʼʊ́s k’úʃ kʼʊ́ʃ kʼús kʼʊ́s kʼʊ́sʊ́ ‘throat’ *kʼOʃ kʼʊ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́sɔ́ ‘kill’ In word final position, there is evidence for *kʼ weakening to a glottal stop, or eliding altogether as seen in (155). Intervocalically, reflexes of *k can often voice to /g/ or elide completely. This is seen in ‘bitter (be)’ in (155), in which Dana and Opo preserve the intervocalic /kʼ/ reflex of *k, Gwama has voiced it to /g/ and Komo and Uduk have elided the segment at the word edge. There does not seem to be a consistent pattern in any of the languages to suggest a robust historical merger of *k and *g in intervocalic position. All of the cognates with reflexes of PKMN *kʼ are provided in Appendix D22. (155) PKMN Gwama Komo Uduk Dana Opo Meaning *pʰákʼá pák pá pʰā pʰákʼá pʰákʼ ‘hoof, shoe’ *sʊkʼa swījā sʊ̄ sú sʊ́kʼà sʊ̄kʼá ‘two’ *kʰakʼa kāgā kà kʰā kʰàkʼà kʰākʼā ‘bitter (be)’ 3.2.6 PKMN fricatives Three voiceless fricatives *s, *ʃ and *h can be confidently reconstructed to PKMN. Most of the languages exhibit retentions of each proto-phoneme. A notable exception 205 involves a chain shift in Proto-Komo whereby *s merged with /ʃ/ and *c merged with /s/. Proto-Gwama also displays primary splits in word-initial *s and *ʃ reflexes conditioned by the quality of the following vowel. The following subsections discuss these innovations in detail. 3.2.6.1 PKMN *s The correspondence sets that reconstruct to PKMN *s are seen in Table 72. Despite two *s correspondence sets, there does not appear to be evidence to justify two distinct proto-phonemes. Some of the innovations can be explained either diachronically or language-internally at the modern stage. Table 72 PKMN *s correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 23 *s ʃ ʃ ʃ s s s s s s s 23a *s s s ʃ s s s s s s s It appears that Komo innovated a chain shift of *s > ʃ and then *c > s word-initially (§4.4.1). Gwama splits *s and merges with ʃ in set 23 and retains /s/ in set 23a. The only conditioning for Gwama *s > ʃ word-initially appears to be the following vowel. Word- initially, Gwama *s > ʃ occurs only before the [+high, –ATR] vowels /ɪ, ʊ/ and PGW retains /s/ before [–high, –ATR] vowels as seen in (156). 206 (156) SET PKMN Gwama Komo Uduk Dana Opo Meaning 23 *sud̪(i) ʃʊ́l ʃùʔí sū sùd̪ swī ‘beer’ 23 *sɪ̀t ̪̓ ʃɪt̄ʼ ʃɪ̀tʼ sīɗ sɪ̀t ̪̓ sɪt̄ʼ ‘far (be)’ 23 *sʊm ʃʊ̄ʃʊ́m ʃʊ̌m súm sʊ̄mʊ́ sʊ̄mʊ́ ‘python’ 23a *s(w)am sɔ́m ʃɔ́m sām sɔ̀m sɔ̄m ‘warm oneself’ 23a *sɔ́p sɔ́ – sɔ́ – sɔ́p ‘stab’ 23a *sɛl sɛl̄ – sɛ ̄ – – ‘climb’ While the word-initial *s reflexes in Gwama appear to follow a pattern, the word- final *s reflexes are more difficult to explain given the fact that Gwama exhibits both /s/ and /ʃ/ reflexes of *s. I cannot account for the word-final split and merger of *s in Gwama. Some examples of both reflexes occurring before /ʊ, ɪ/ are seen in (157). See Appendix D23 and D23a for all of the cognate sets employed to reconstruct PKMN *s. (157) PKmn Gwama Komo Uduk Dana Opo Meaning *(j)Es jɪs̄ ɪʃ̄ īs ɛ̀s ɛs̄ ‘body’ *jEsI ɪʃ́ɪ̀ jɛ̀ʃ jɛ̀s sìʔ ‘slippery’ *gÙs gʊ̀s gùʃ gùs – – ‘run (SG)_1, flow, bleed’ *tU̪s dʊ̄ʃ tūʃ tūs – – ‘cotton, thread’ Lastly there are two cognates which do not follow either pattern, which are seen in (158). In these correspondences, every language retains /s/. I cannot offer an explanation for these cognates and attribute this to idiosyncratic behavior. (158) PKMN Gwama Komo Uduk Dana Opo Gloss *sʊkʼa swījā sʊ̄ sú sʊ́kʼà sʊ̄kʼá ‘two’ ‘tree_sp. (mahogany, *k⁽ʰ⁾ɪs kɪs̄ kɪs̄ cɛś kɪś kʰɪś Trichilia emetica)’ 207 3.2.6.2 PKMN *ʃ A voiceless (alveo)palatal fricative *ʃ can be reconstructed to PKMN based on the two correspondence sets seen in Table 73. The KOUD branch and Dana from the DAOP branch retain a [ʃ] reflex, while Gwama splits and merges *ʃ with s. This conditioned split in Gwama is described below. In the Opo branch, Bilugu, Modin and Pame Opo varieties merge *ʃ with *s. It appears that in some instances, Kigile Opo retains /ʃ/ though this has yet to be definitively determined.160 Table 73 PKMN *ʃ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 24 *ʃ ʃ ʃ ʃ ʃ ʃ ʃ s s s s/ʃ 24a *ʃ s s ʃ ʃ ʃ ʃ s s s s/ʃ Correspondence set 24 is by far the most robust of the two PKMN *ʃ correspondence sets with about twenty cognates in word-initial position. Some examples are seen in (159) and the full set of cognates can be found in Appendix D24. Note that in *ʃ set 24, all of the initial /ʃ/ reflexes in Gwama precede a front vowel, either a [+high, –ATR] vowel /ɪ, ʊ/ or a [–high, –ATR] vowel /ɛ/.161 160 In my database, the Kigile Opo reflexes are mostly [ʃ] in correspondence sets 24 and 24a. In the twelve cognates which contain a word-initial Kigile Opo reflex from sets 24 and 24a, eight occur with [ʃ] and four occur with [s]. This is why the [s/ʃ] symbol occurs in the Kigile Opo slot. My consultants were not first-language Kigile Opo speakers and were often not sure whether a particular lexeme contained [s] or [ʃ]. 161 There appears to be one exception in Gwama, *ʃī ‘tooth’, which contains [+high, +ATR] /i/ following the initial /ʃ/. The rest of the family has /ɛ/ reflexes in ‘tooth’. In Komo, there is a remnant of a plural suffix /-í/ which appears in ʃɛ̀ʔí ‘teeth’, which is synchronically pronounced [ʃɛ̀ʔí ~ ʃɛ̀í]. A plausible scenario for Gwama could have been *ʃɛʔi > ʃi via the loss of the internal glottal stop and the coalescence of the vowel. 208 (159) PKMN Gwama Komo Uduk Dana Opo Meaning *ʃʊnʃ ʃʊ̃ʃ ʃʊ̀nʃ ʃùʃ ʃʊ̀ʃ ʃʊ̀ʃ ‘nose’ *ʃOkʼ(VN) ʃʊ̄gʊ̀n ʃùwɛ̀n ʃɔ̄kʼɔ̄m ʃùkʼnáj ʃūkʼɛń ‘louse’ *ʃʊ(n)kʼ ʃʊ̄nkʼ ʃʊ̄ ʃúʔ ʃùkʼ ʃūkʼ ‘tendon, vein’ *ʃE ʃī ʃɛ̀ ʃɛ ̄ ʃɛ ̄ ʃɛ ̄ ‘tooth’ Cognate set 24a reflects a Gwama innovation in which a primary split *ʃ > s occurred word-initially. Some examples are provided in (160). The only conditioning factor here in Gwama is that word-initial /s/ reflexes of *ʃ occur either before a [+high, +ATR] vowel /i, u/ or a [–high, –ATR] vowel /ɔ/. This is in complementary distribution with the retention of *ʃ as /ʃ/ before /ɪ, ʊ/. (160) PKMN Gwama Komo Uduk Dana Opo (Kigile) Meaning *ʃum(a) sūm ʃùm ʃūm ʃùmà ʃūm ‘meat, animal’ *ʃUImakʼ síʔ ʃúmákʼ sīmāʔ ʃʊ́j sʊ́j ‘bone’ *ʃukʼ(i) sūgì ʃùg ʃūkʼ ʃùg ʃùg ‘wake (trs.)’ *ʃɔnkʼ sɔ̄nkʼ ʃɔ̀g ʃɔ̀kʼ ʃɔ̀g – ‘foot or leg’ It is tempting to try generalizing the Gwama split of *ʃ based solely in the [ATR] quality of the following vowel. The /ʃ/ retention only occurs before [–ATR] vowels and the *ʃ > s split occurs before [+ATR] vowels /i, u/. The problem is the Gwama cognate sɔ̄nkʼ ‘foot, leg’ in which initial /s/ occurs before [–ATR] /ɔ/. However, there are cases in which *wa > /ɔ/ across the family. If this /wa/ sequence were historically a diphthong *ua which occurred with an initial [+ATR] /u/, then the initial /s/ in Gwama would follow the generalization: word-initial *ʃ > s before [+ATR] and /ʃ/ is retained elsewhere (i.e. before [–ATR] vowels). This distribution also fits the Gwama split *s > ʃ, which only occurs before [–ATR] /ɪ, ʊ/ as seen above in §3.2.6.1. 209 3.2.6.3 PKMN *h A voiceless glottal fricative can be confidently reconstructed to PKMN in word- initial position via the correspondence set in Table 74. All branches retain /h/ reflexes. Table 74 PKMN *h correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 25 *h h h h h h h h h h h Some examples of reflexes of *h are in (161) and the full set of reflexes is provided in Appendix D25. (161) PKMN Gwama Komo Uduk Dana Opo Meaning *hag(a) hã́ʔ hág háʔ hágá ‘have sex’ *hasʼ hāsʼɪ̀ hásʼ hát ̪̓ hásʼ hátʃʼ ‘trample, ruminate’ *hʊn(i) hʊ̄n hʊ̀n húnī hʊ̄n – 3SG.M 3.2.7 PKMN *r and *l An alveolar trill *r and an alveolar liquid *l can be reconstructed to PKMN via the correspondence sets seen in Table 75. Table 75 PKMN *r and *l correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 26 *r r r r r r r r r r r 27 *l l l l l l l l l l l 210 Some examples of *r and *l reflexes are seen in (162) and the full sets are seen in Appendices D26 and D27, respectively. (162) PKMN Gwama Komo Uduk Dana Opo Meaning *rOk⁽ʰ⁾Oɲ rɔ́kɔ̀n rɔ́kɔ̀n rúgùɲ rɔ́kʰɔ̀n rʊ́kʊ̄n ‘corner’ *lili līlí lìlí līl līlí ‘sink (v.)’ 3.2.8 PKMN nasal consonants A bilabial nasal *m and an alveolar nasal *n can confidently be reconstructed to PKMN. There is marginal evidence for *ɲ and *ŋ. Bender (1983) notes that *ɲ and *ŋ were later developments. 3.2.8.1 PKMN *m and *n The bilabial *m and alveolar *n nasals easily reconstruct to PKMN. All of the languages retain /m/ and /n/ reflexes in initial, medial and final positions. The correspondence sets for *m and *n are in Table 76. Table 76 PKMN *m and *n correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 28 *m m m m m m m m m m m 29 *n n n n n n n n n n n Some examples of *m and *n reflexes are in (163) and the full sets are provided in Appendices D28-29. 211 (163) PKMN Gwama Komo Uduk Dana Opo Meaning *mVsʼ mɪ̀sʼ mʊ̀sʼ mút ̪̓ ɛ ̄ mútʼ mìsʼà ‘shut, close eyes’ *nɛḱʰ nɛḱ à+nɛḱ nɛḱʰ nɪḱʰ nɪḱʰ ‘bird (sp.)’ 3.2.8.2 PKMN *ɲ Only the Uduk varieties exhibit a contrastive palatal nasal synchronically. There is some evidence for reconstructing a palatal nasal to PKMN, though only in word-final or word-medial position. The correspondence set for *ɲ is seen in Table 77. Only Uduk retains /ɲ/ while all others merge *ɲ > n. Table 77 PKMN *ɲ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 30 *ɲ n n n ɲ ɲ n n n n n The only PKMN cognates containing a /ɲ/ in Uduk are seen in (164). Note that /l/ corresponds to /ɲ/ in ‘cattle egret’ and that Uduk initial /c/ < *k before front vowels (§3.2.5.2). At best, *ɲ is a marginal proto-phoneme. (164) PKMN Gwama Komo Uduk Dana Opo Meaning *rOk⁽ʰ⁾Oɲ rɔ́kɔ̀n rɔ́kɔ̀n rúgùɲ rɔ́kʰɔ̀n rʊ́kʊ̄n ‘corner’ *(tʼi)pʼIkʼIɲ(a) pʼɪk̄ʼɪń pʼīn tʼīpʼīɲ pʼɪn̄ā pʼɪn̄ā ‘ash’ *kʰàɲ ‘light (the kɪn̄ – kʰāɲ – – way)’ *kEɲ kɪl̄ kɪĺ cɛɲ́ kɪĺ kɪĺ ‘cattle egret’ 212 3.2.8.3 PKMN *ŋ There is some marginal evidence in favor of reconstructing a velar nasal *ŋ to PKMN. The correspondence set in Table 78 shows retentions of [ŋ] in Uduk and in the Dana-Opo branch, a merger *ŋ with g in Komo and a partial merger of *ŋ with *g in Proto-Gwama (PGW). Table 78 PKMN *ŋ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 31 *ŋ ŋ/g ŋ/g g ŋ ŋ ŋ ŋ ŋ ŋ ŋ The cognates showing reflexes of a proposed PKMN *ŋ are limited and always involve a voiced velar stop. Some examples of such cognates with potential *ŋ reflexes are in (165); the complete set of relevant cognates can be found in Appendix D31. (165) PKMN Gwama Komo Uduk Dana Opo Meaning *waŋa wāŋā wàgá ŋwá – – ‘chicken’ *gàŋ(a) kɛ̃ ̄ gàg – kàŋà kàŋà ‘smell (v.)’ *ɟaŋg(ɔ)aj zǎgʊ́ zǎgʊ́ ʒwǎŋgì ɟàŋwɛ̀j dʒāŋɔ́ ‘Nuer’ 3.2.9 PKMN glides 3.2.9.1 PKMN *w A labiovelar glide *w can be confidently reconstructed to PKMN in word-initial position only. The correspondence set for PKMN *w is in Table 79. All languages retain /w/ in word-initial position. 213 Table 79 PKMN *w correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 32 *w w w w w w w w w w w Some examples of cognates containing reflexes of PKMN *w are in (166) and the full set is seen in Appendix D31. (166) PKMN Gwama Komo Uduk Dana Opo Meaning *wàcʼà wàsʼ wàsʼ wàcʼ wàcʼà wàtʃʼà ‘fish’ *was(ikʼ) wàʃí wàʃíkʼ – wás wās ‘boil, bubble up’ *wasakʼ wàsà wàʃākʼ wàsáʔ – – ‘hail, ice’ 3.2.9.2 PKMN *j A voiced palatal glide *j can be tentatively reconstructed to PKMN in word-initial position based on the correspondence set in Table 80. If *j is to be reconstructed, it would be retained in all branches but and merged with *dʒ in Proto-Opo (§4.10 ). Table 80 PKMN *j correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 33 *j j j j j j j dʒ dʒ z ʃ There are very few cognates that contain reflexes of *j. Some are in (167) and the full set is in Appendix D33. If this analysis is accurate, the data here suggest fortition of PKMN *j to POP *dʒ in word-initial position. Another analysis is that this correspondence set represents *ɟ, which independently weakened to /j/ outside of Opo. 214 (167) PKMN Gwama Komo Uduk Dana Opo Meaning *jiɗE ìjáʔ jǐ jìɗɛʔ́ jìʔí dʒì ‘water’ *jEkʰ jɪ̀ jɛ̀k jɛ̀kʰ – – ‘sow seeds’ *jàgàl jàgɪ̀ – – – dʒīkʼāj ‘sweat’ 3.2.10 PKMN consonant residue There are some minor correspondences for which a proto-phoneme cannot be definitively reconstructed. These and other issues are outlined in the following subsections. 3.2.10.1 PKMN interdental stop residue: *T̪ and *D̪ There are two residual correspondences involving interdental stops, seen in Table 81. Whether or not these proposed correspondences genuinely reflect additional (interdental) proto-phonemes is yet to be determined. I employ *T̪ and *D̪ as temporary placeholders to represent these correspondences. Table 81 PKMN *T̪ and D̪ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig 34 *T̪ ʃ ʃ s/z s t ̪ s tʃ tʃ tʃ tʃ 35 *D̪ s/z s/z z – – d̪ dʒ dʒ z s/ʃ In the *T̪ reflexes, Chali Uduk /t/̪ corresponds to Dana /s/, Opo /tʃ/, Gwama /ʃ/ and Komo /z/.162 The only two potential cognates are in (168), but it is extremely difficult to 162 Note that the Komo reflex occurs with L tone which could suggest either a historical voiced consonant or that Komo irregularly voiced this particular consonant in this particular word. 215 reconstruct the proto-phoneme as all of the initial consonant reflexes appear to reconstruct to distinct correspondence sets. Perhaps this *T̪ correspondence set actually reconstructs to *cʰ, for which we have no correspondence set or perhaps the *T̪ and *t ̪ correspondence sets reconstruct to a single PKMN *t ̪proto-phoneme for which there was an unconditioned split. Nevertheless, taken together, the *t ̪and *T̪ correspondences at least support the likelihood of a PKMN voiceless interdental proto-phoneme. (168) PKMN Gwama Komo Uduk Dana Opo Meaning a. *T̪ʊ(n)G ʃʊ̌nkʼ zʊ̀g tū̪kʰ sʊ̄kʰ tʃʊ̄gʊ́ ‘guinea fowl’ b. *T̪Ol(a) ʃʊ̀ – tū̪l – tʃɔ́já ‘descend’ The two cognates with PKMN *D̪ reflexes are seen in (169). Observe that /d̪/ in Dana corresponds to /dʒ/ in Bilugu and Pame Opo, /z/ in Modin Opo and Gwama, and /ʃ/ in Kigile Opo before L tone in ‘eye’. Similar correspondences are seen in ‘person’ with the exception of Gwama which exhibits /s/, though with M tone in this descendent morpheme. (169) Opo Opo Opo Opo PKMN Gwama Komo Uduk Dana Meaning (Bil) (Pam) (Mod) (Kig) *D̪E zì – ɛ ̄ d̪ɛ̀ dʒɛ̀ dʒɛ̀ zɛ̀ ʃɛ̀ ‘eye’ *D̪ita̪ sīt – d̪ítà̪ – – zìtà ʃìtà ‘person’ The elements in these two proposed correspondence sets are erratic, though show some patterning of fricatives and/or affricates with interdental stops. Whether this motivates reconstructing one or two additional proto-phonemes is yet to be determined. 216 3.2.10.2 A note on PKMN *z There is only one cognate which may begin to suggest reconstructing a voiced alveolar fricative *z to PKMN, seen in (170). Given the fact that there are no other potential cognates with this set of correspondences, reconstructing *z to PKMN is questionable. Bender (1983) proposes that /z/ surfaces in the modern languages due to independent innovations, though this is puzzling given that Gwama, Komo, Yabus Uduk, Dana and Modin Opo all exhibit a contrastive /z/ synchronically. (170) PKMN Gwama Komo Uduk Dana Opo Meaning *zjaɗa(j) zɛt́ zjàntʼá – zāt ̪h ɛ ̄ dʒɛ̀rà ‘chili pepper’ 3.3 Proto-Koman (PKMN) vowels All of the modern Koman languages with the exception of the Uduk cluster exhibit a contrastive seven-vowel inventory /i, ɪ, ɛ, a, ɔ, ʊ, u/, with phonemic Advanced Tongue Root (ATR) contrast in high vowels (Hellenthal & Kutsch Lojenga 2011, Killian 2015, Smolders forthcoming, Otero 2018b, inter alia). Further, all of the languages with seven- vowel inventories exhibit active ATR harmony (cf. §2.1.2). Prior reconstructions of the Proto-Koman vowel system propose five contrastive vowels *i, *e, *a, *o, *u (cf. Bender 1983, Ehret 2001). Ehret (2001:43) sustains that Proto- Nilo-Saharan (PNS) exhibited seven short vowels (*i, *e, *ɛ, *a, *ɔ, *o, *u) and their long counterparts, which simplified into five vowels in Koman.163 These significant contributions to Proto-Koman reconstruction were carried out without detailed study of 163 Ehret (2001) assumes that Koman descended from Proto-Nilo-Saharan, an issue that is still under debate (cf. Dimmendaal 2011 and Güldemann 2018 for opposing views). 217 Koman vowel systems. In his reconstruction of PKMN phonology, Bender (1983) analyzes five vowels in each of the modern Koman languages. Ehret’s (2001) work on PKMN drew largely from Chali Uduk, which only exhibits five contrastive vowels. Neither scholar was informed of the active ATR harmony systems found in all of the Koman languages save the Uduk cluster. This led to reconstructions which do not stand up in the light of current analyses (Hellenthal & Kutsch Lojenga 2011; Goldberg et al. 2017; Otero 2015, 2018b, Olejarczuk et al. 2019, Smolders forthcoming), and in some cases these faulty analyses impacted their higher-level reconstructions, as is the case with Ehret’s (2001) analysis of how the PKMN vowel system evolved from Proto-Nilo- Saharan. My aim here is not to criticize these earlier reconstruction attempts given the data at hand at the time, but rather to highlight how important it is to have detailed knowledge of synchronic systems in order to conduct a valid historical reconstruction. In light of Koman language documentation and description, all of the synchronic and diachronic evidence points to a PKMN seven vowel system with ATR contrast in the high vowels *i, *ɪ, *ɛ, *a, *ɔ, *ʊ, *u. There is no evidence for a synchronic nor historical vowel length contrast in Koman. The correspondence sets are by and large stable, especially in the [–high] vowels, where all languages generally exhibit retentions. In the [+high] vowels, by contrast, there are identical reflexes in correspondence sets which allow us to reconstruct *i, *ɪ, *ʊ and *u. But, there are correspondence sets in which Gwama has a high vowel of opposing ATR value to that found in Central Koman languages (i.e.. Gwama /ɪ/ corresponding to /i/ in all of Central Koman and vice versa). These issues are discussed in the following subsections. All of the cognate sets employed in the reconstruction of PKMN vowels are in Appendix E. 218 3.3.1 PKMN high vowels PKMN had two sets of contrasting front and back high vowels distinguished by the ATR feature, giving a total of four proto-high vowels: *i, *ɪ, *u and *ʊ. Reflexes of these vowels are seen in the synchronic phonemic inventories of all Koman languages with the exception of the Uduk cluster, which merged *i with *ɪ and *u with *ʊ.164 The correspondence sets for high vowels suggest splits in either PGW or PCTRL as, for instance, /u/ can correspond with /u/ in one set but with /ʊ/ in another. The same holds for /i, ɪ/. There do not appear to be any conditioning factors for these splits though examining tone and reflexes of *s and *ʃ give some insight into possible directionality. The reconstruction of PKMN high vowels are discussed in the following subsections. 3.3.1.1 PKMN *i and *ɪ The three correspondence sets for PKMN high front vowels are in Table 82. The data suggest reconstructing only two proto high front vowels distinguished by the feature [ATR]: *i and *ɪ. In set V1, all languages retain [+high, +ATR] /i/ reflexes and in set V2, all languages retain [+high, –ATR] /ɪ/ reflexes. PUD merges *i with *ɪ into one contrastive vowel /i/. 164 In the synchronic Uduk varieties, both Killian (2015) and I recognize that /i, u/ can be realized phonetically as [i, e, ɪ] and [u, o, ʊ] respectively. The different phonetic vowels in each set appear to be in free variation in Uduk. 219 Table 82 PKMN *i and *ɪ correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig V1 *i i i i i i i i i i i V2 *ɪ1 ɪ ɪ ɪ i i ɪ ɪ ɪ ɪ ɪ V2a *ɪ2 ɪ ɪ i i i i i i i i Reconstruction of sets V1 and V2 is straightforward as two proto-phonemes *i and *ɪ. The question is what to reconstruct for correspondence set V2a, in which /ɪ/ in Gwama corresponds to /i/ in Central Koman. There are two correlates that might provide an indicator of the vowel quality at a certain stages of evolution: (i) Whether or not the Opo varieties exhibit a XH tone reflex. (ii) The Gwama reflexes of *s and *ʃ. Recall that a POP innovation in tone was a split *H > XH on [+high, +ATR] vowel nuclei (§3.1.2). Thus, if modern Opo exhibits XH tone on a high vowel, it must have been *i or *u in POP or in a previous stage. Second, recall the Gwama word-initial innovations: *s > ʃ before [+high, –ATR] vowels /ɪ, ʊ/ (§3.2.6.1) and *ʃ > s before [+high, +ATR] vowels /i, u/ (§3.2.6.2). Analyzing the ATR value of the vowels following word- initial reflexes of *s and *ʃ may also provide a clue to the ATR value of the proto-vowel. While these diagnostics will not solve the puzzle entirely, they at least provide insights into vowel quality at certain nodes. I discuss the tone evidence first. Example (171) contains cognates from each of the three correspondences sets in Table 82. Each of these sets reconstruct to PKMN tone set A, which gave rise to synchronic H tone across the family. Note that in set V1, Opo exhibits XH tone and the vowel reconstructs to *i. By contrast, in set V2, Opo exhibits 220 H tone on reconstructed *ɪ. In set V2a, in which Gwama exhibits /ɪ/ and Proto-Central Koman exhibits /*i/, Opo exhibits XH tone. This suggest that the vowel was at least /*i/ in PCTRL though it does not solve the issue of directionality (i.e. whether PGW retained *ɪ and PCTRL split *ɪ > *i, or vice versa). (171) Set Gwama Komo Uduk Dana Opo Meaning V1 hírɔ́kʼ (bā-)jírɔ́kʼ – (ɓā-)jírɔ́ŋ (ɓā-)jır̋ɔ̄kʼ ‘bird (bee-eater)’ V2 sɪt́ tɪt́ títʰ tɪ̪t́ ̪h tɪt́ɪ ́ ‘roughen stone’ V2a ɪś íʃ ís ísá ıs̋á ‘ripen’ I now turn to evidence related to *s and *ʃ. The Gwama reflexes of *s and *ʃ in (172a) show the Gwama innovation of *s > ʃ before /ɪ/. This vowel correspondence is set V2, which reconstructs to *ɪ with /ɪ/ reflexes across the family (and *ɪ > i in Uduk). In (172b), Gwama retains PKMN *ʃ as /ʃ/ before /ɪ/, though this is vowel correspondence set V2a in which Central Koman exhibits /i/. The question of whether (172b) reflects PKMN *ɪ, in which Gwama retains *ʃ before /ɪ/ with a subsequent change of PKMN *ɪ > PCTRL *i or whether (172b) reflects PKMN *i with Gwama innovating PKMN *i > ɪ cannot be definitively determined. There is a strong likelihood that (172b) was /i/ in PCTRL as all languages of this branch exhibit /i/ and Opo innovates *H > XH tone, the latter of which only occurs on /i/ and /u/. (172) Set PKMN Gwama Komo Uduk Dana Opo Meaning a. V2 *sɪ̀t ̪̓ ʃɪt̄ʼ ʃɪ̀tʼ sīɗ sɪ̀t ̪̓ sɪt̄ʼ ‘far (be)’ b. V2a *ʃɪntʼ ~ ʃintʼ ʃɪn̄tʼ ʃīn ʃīn ʃìnà sın̋á ‘blow nose’ 221 3.3.1.2 PKMN *u and *ʊ Synchronically the Koman languages with seven-vowel inventories only display two contrastive high back vowels of /u, ʊ/. Nevertheless, there are four correspondence sets that relate to high back vowels, presented in Table 83 and Table 84. In two of the sets, all of the languages retain /u/ and /ʊ/, which can easily reconstruct to *u and *ʊ, respectively. Though in the remain two correspondence sets, Gwama /u/ corresponds to /ʊ/in Central Koman and Gwama /ʊ/ corresponds to /u/ in Central Koman. I argue that these latter correspondence sets do not reconstruct to two additional vowels but rather reflect innovations in either Gwama or Central Koman: the issue here is to determine whether Central Koman or Gwama split *u and *ʊ. We first consider the correspondence sets in Table 83. Though there are two correspondence sets here, there is evidence that *u1 and *u2 reconstruct to a single proto-phoneme. Assuming that PKMN had a seven vowel inventory, at present I am unable to determine whether Gwama preserves the PKMN vowel system and the changes happened in PCTRL or if the changes in vowels occurred in PGW and the PKMN vowels are retained in PCTRL. Table 83 PKMN *u correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig V3 *u1 u u u u u u u u u u V3a *u2 ʊ ʊ u u u u u u u u In PKMN *u1 set V3, all languages retain /u/ while in set V3a, /ʊ/ in Gwama corresponds to /u/ in Central Koman. Deciding whether set V3a reflects an *u > ʊ 222 innovation in Central Koman or an *ʊ > u innovation in Gwama is challenging but there is one piece of evidence that may shed light on these correspondences, which involves ATR vowel quality and tone. Recall the POP innovation of splitting *H > XH on [+high, +ATR] vowel nuclei (§3.1 f). In PKMN *u1 set V3, Opo exhibits synchronic XH tone when the PDAOP tone is *H. To illustrate, consider the data in (173) in which all languages exhibit /u/ reflexes. Here, Dana exhibits H tone and Opo exhibits XH tone. In these reflexes it is clear that POP split *H > XH on a /u/ nucleus. Further evidence is that there are no reflexes in the *u1 correspondence set in which the POP tone category is *H and Opo does not exhibit XH tone. To be clear, this means that there are no reflexes in correspondence set V3 in which at least Dana exhibits H tone and Opo does not exhibit XH tone. This suggests that the tone was at least H at the DAOP node if not further back in Koman history. The fact that all langauges retain /u/ reflexes of *u plus the tonal split *H > XH in Opo provide solid evidence to reconstruct PKMN *u for set V3 *u1. (173) Set Gwama Komo Uduk Dana Opo Meaning V3 wǔt wǔt útʰ – hűtʰ ‘ostrich’ V3 tū – – t ̪h úwà tʰűjhá ‘spit (v.)’ V3 jàhú jàhú – àhút ̪h àhű ‘fish (sp.) V3 kùmbì kúm kūm kúmā kűmá ‘cover (v.)’ When examining *u2 correspondence set V3a, in which /ʊ/ in Gwama corresponds to /u/ in all other languages, we see identical patterning with regard to tone. Also in this set, if the proto-tone category is PKMN *H or POP *H, Opo always exhibits XH tone as seen in (174). This fact favors an analysis in which PGW split and merged *u with *ʊ in syllable nuclei which were of PKMN *H or corresponded to POP *H tone. 223 (174) *u Set Gwama Komo Uduk Dana Opo Meaning V3a ʊ́p úp úpʰ úpʰ űpʰ ‘bathe’ V3a tʊ̌kʊ̀ túk à-tú̪kʰ tú̪kʰ tűkʰ ‘acacia’ V3a kʊ̌ʃ ūs – húʃ űʃ ‘smoke out’ While there is a correlation between XH tone in Opo and a proposed *u reflex in which PKMN *u > PGW *ʊ, this only accounts for the reflexes in tone set A correspondence sets (i.e. sets that reconstruct to PKMN *H). This is to say that there are many reflexes in both of the PKMN *u correspondence sets that exhibit tones other than H. While PKMN *u1 (Set V3) can be confidently reconstructed to *u, there is no apparent conditioning factor for either an analysis of *u > ʊ or *ʊ > u innovations for *u2 (Set V3a). Consider the data in (175) which contain reflexes from both PKMN *u correspondence sets. The reflexes here all exhibit L tone which evolved from the PKMN *L tone category. In these data, there does not appear to be a conditioning for a change such as PKMN *u > Gwama /ʊ/, though this was most likely the case given the evidence presented for cognates that reconstruct to PKMN *H tone and that there are two correspondence sets which reconstruct to PKMN *Ʊ. All of the data for PKMN *u correspondence sets V3 and V3a are seen in Appendix EV3 and EV3a respectively. (175) *u Set Gwama Komo Uduk Dana Opo Meaning a. V3 bùʃùl bùʃ bùʃ – pùsà ‘belly’ b. V3a kʊ̀rʊ́m gùdúm kùd̪ùm kùdùmà ‘pig’ There are two additional high back vowel correspondence sets, which I propose are reflexes of PKMN *ʊ. These are seen in Table 84. Note that the distribution of reflexes patterns identically to the correspondences for *u in Table 83, though with opposing 224 ATR values: in *ʊ set V4 set all languages retain /ʊ/ reflexes and in set V4a /ʊ/ in Central Koman corresponds to /u/ in Gwama. Table 84 PKMN *ʊ correspondence set Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig V4 *ʊ1 ʊ ʊ ʊ ʊ ʊ ʊ ʊ ʊ ʊ ʊ V4a *ʊ2 u u ʊ ʊ ʊ ʊ ʊ ʊ ʊ ʊ Interestingly, there are no reflexes in either of these *ʊ correspondence sets in which Opo exhibits XH tone. This is good evidence that there are reflexes in which Opo exhibits H tone and these reflexes are all reconstruct to PKMN *H tone. In (176a), all languages exhibit /ʊ/ reflexes and H tone. In (176b), /ʊ/ in Central Koman corresponds to /u/ in Gwama and all languages exhibit H tone. If this were *u, then we would expect XH tone in Opo but that is not the case. This is a strong indicator that set V4a reflects a Gwama split *ʊ > u given the fact that Opo split *H > XH only on /i, u/ vowel nuclei. Again, this does not account for the data in which H tone is not present. (176) Set Gwama Komo Uduk Dana Opo Meaning a. V4 ʔʊ́p kʼʊ́p kʼúpʰ kʼʊ́pʰ kʼʊ́p ‘head’ b. V4a kúsʼ kʊ́sʼ kʰúsʼ kʰʊ́sʼ kʰʊ́tʃʼʊ́ ‘dry (be)’ The *u and *ʊ correspondences (SetV3a and Set V4a) taken together strongly suggest two developments in Gwama: *u > ʊ and *ʊ > u, respectively, yet there does not appear to be any segmental conditioning, at least synchronically. The only indicator of directionality of these splits is when Opo exhibits synchronic XH tone. Given the fact 225 that Opo split *H on [+high, +ATR] vowel nuclei /i, u/, vowel correspondence sets in which Opo exhibit XH provide evidence for *u (and *i). This combined with a reconstructed *H tone category provides even stronger evidence for reconstructing *u in Set V3a. 3.3.2 PKMN [–high] vowels *ɛ, *ɔ and *a Three [–high] vowels can be reconstructed to PKMN, *ɛ, *ɔ and *a through three correspondence sets seen in Table 85. These proto-vowels are generally retained in all languages. Table 85 PKMN *ɛ, *ɔ and *a correspondence sets Gwama Komo Uduk Dana Opo Set PKMN Lo Hi Yab Cha Bil Mod Pam Kig V5 *ɛ ɛ ɛ ɛ ɛ ɛ ɛ ɛ ɛ ɛ ɛ V6 *ɔ ɔ ɔ ɔ ɔ ɔ ɔ ɔ ɔ ɔ ɔ V7 *a a a a a a a a a a a Some cognates demonstrating each [–high] proto-vowel are seen in (177) and the full data is provided in Appendices E5-7. (177) *Vowel Gwama Komo Uduk Dana Opo Meaning *ɛ sʼɛ ̄ sʼɛʔ̄ ʃʼɛ ́ kʼɛ ̄ tʃʼɛ̀ ‘ear’ *ɛ zɛ̀rú zɛ̀rú ʒàrú ɟɛ̀rú dʒɛ̀rű ‘stork (maribou)’ *ɔ gɔ̀kʼɔ́ʃ gɔ̀nkʼí gɔ̀kʰ kɔ̀gɔ̀ kɔ̀gɔ̀ ‘skin or bark’ *ɔ sɔ̄ŋkʼ ʃɔ̀g ʃɔ̄kʼ ʃɔ̀g – ‘foot’ *a kàm gàm gàm kàm kàm ‘find, meet’ *a pāj pàj pʰāj pʰàd̪ pʰāj ‘fly (v.)’ 226 While the correspondences for *ɛ and *ɔ seem regular, there are cognates in which some reflexes exhibit distinct vowels so establishing a proto-phoneme proves more challenging. Some examples are seen in (178). In these probable cognates there are correspondences between /i ~ ɪ ~ ɛ/ and /ɔ ~ ʊ/ though there do not appear to be any patterns to this variability that would warrant reconstructing a distinct proto-phoneme. I attribute these alternations to the general nature of variability in vowels over time. (178) Gwama Komo Uduk Dana Opo Meaning jɪs̄ ɪʃ̄ īs ɛ̀s ɛs̄ ‘body’ mɪ̀tʼ mɪt́ʼ mɛ̀ɗ mɛ̀t ̪̓ mɪ̀tʼɪ ́ ‘hand’ sʼɪ ̄ sʼɛ̀ cʼɛ ̄ sʼɛ̀d̪ʰ tʃʼɛ ̄ ‘shave’ ʃʊ́ʃʊ̄ʃʊ̄ ʃɔ̀ʔí ʃɔ̄ ʃɔ̄ʔɔ́ sɔ̀ ‘grass’ 3.3.3 Consonant-glide sequences and diphthongs The only scholar to propose a series of contrastive labialized consonants is Killian (2015) for Chali Uduk. The only complex onset that is exhibited by all of the living Koman languages is Cwa, in which an initial Consonant is followed by what can be analyzed as a labiovelar glide or a high back vowel. When it is analyzed as a glide the /w/ segment crucially does not carry tone. Occasionally, the vowel following the Cw sequence will have a contour tone. This suggest that at least historically, some synchronic Cw sequences developed from a diphthong, i.e. a high back vowel followed by *a (e.g. *ua or *ʊa). There are some cognates that contain /wa/ in Gwama and Uduk but /a/ in Komo and Dana-Opo as seen in (179). These cognates are all nouns and the Cw sequence may 227 be explained through metathesis of an initial gender prefix *ʊ̄- and the initial consonant of the noun root, such that *ʊ̄-Ca > Cʊa > Cwa. (179) Gwama Komo Uduk Dana Opo Meaning swálá sá cwá cáɟáʔ tʃá ‘tree’ twã ̄ tàg – tàg bī+tʰāg ‘forehead’ tʼwā tʼā tʼwā t ̪̓ āʔá tʼā ‘mouth’ pwǎŋà – bwàj – – ‘path, towards’ There are also correspondences in noun cognates that do not exhibit the pattern of /wa/ in Gwama and Uduk but /a/ in Komo and Dana-Opo in (179), as seen in (180). (180) Gwama Komo Uduk Dana Opo Meaning kʼwāntʼɪ ́̓kwāntʼ kʼwàtʼ kʼwāɗ kʼwàt ̪̓ kʼwātʼ ‘tick’ kwám kàm kām āmʊ́ hàm ‘brother’ ʊ̄+dɔ́k à+dɔ̂k – à+dwákʰ à+dwa̋kʰ ‘weaver bird’ While this does not rule out the possibility of a PKMN set of labialized consonants, the data do not show robust correspondences supporting such a hypothesis either. This parallels Bender’s (1983:285) assessment that “no convincing case for labialized consonants is apparent, so I treat CʷV as Cui, Cue, Cuo, Cwa (the last is by far the most frequent).” This concludes the discussion of PKMN phonology. In the following chapter, I discuss the phonology of each subnode descending from PKMN to the modern languages. 228 CHAPTER IV PHONOLOGICAL RECONSTRUCTION OF KOMAN SUBNODES 4 Phonological reconstruction of Koman subnodes The following subsections reconstruct the core phonology of the Koman subnodes. I discuss the consonants, vowels and tone at each node down to the synchronic languages and/or languages varieties. Within each subsection, I lay out the innovations that characterize each node. I provide a schematized evolution of all consonants from PKMN to the particular node under scrutiny. I begin with the evolution of modern-day Gwama and then turn to Proto-Central Koman. From there I discuss the Komo-Uduk branch: Komo, Proto-Uduk and its modern varieties, Chali Uduk and Yabus Uduk. I then turn to the Dana-Opo branch and examine Proto-Dana-Opo followed by Dana, Proto- Opo and the modern Opo varieties. 4.1 Gwama Proto-Gwama (PGW) is one initial split from Proto-Koman. In the following subsections I treat Gwama as a single language given the proximity of the varieties. Thus, when referring to Gwama in the following subsections, I specifically mean both Highland and Lowland varieties. Gwama shows the most reductions in the consonant system of any Koman language. Several innovations that characterize this branch are the loss of PKMN implosives as well as interdental and palatal obstruents. Gwama also exhibits independent devoicing of word-initial voiced stops before PKMN *L. Another unique set of innovations are PKMN *s > ʃ /_ V[+high, –ATR] and PKMN *ʃ > s /_ V[+high, +ATR]. In terms of the morphology, Gwama innovates some of the participant indexing of core 229 arguments on verbs as well as an elaborate system of partial reduplication of the verb stem. The phonological innovations are discussed in the following subsections. 4.1.1 Gwama consonants Synchronically Gwama has the smallest consonant inventory of any modern Koman language, with 21 phonemes. The synchronic Gwama consonant inventory is presented in Table 2 and reproduced here in (181). What characterizes modern Gwama is the array of independent mergers which occurred at a Proto-Gwama (PGW) stage. These involve a loss of the PKMN three-way phonological contrast in VOT in stops. (181) p t k (ʔ) b d g pʼ tʼ kʼ s z ʃ h sʼ n m l (ŋ) r w j Synchronically, Gwama voiceless stops are phonetically aspirated at least in syllable onset and coda positions. This is not an indication of a historical aspiration as reflexes of both unaspirated and aspirated stops are aspirated synchronically in these positions. Another significant series of PGW innovations were the independent mergers of the interdental and alveolar obstruent series. This includes plain and ejective stops. Gwama is also the only Koman language that does not display implosives synchronically, as a result of mergers with ejectives. The complete set of PGW innovations and the reflexes in modern Gwama are seen in Table 86 and some notable details are discussed below. 230 Table 86 Gwama phonological innovations SET PKMN PGW Gwama Observations 1 *pʰ > *p > p Merger 2 *p > *p > p Merger 3 *b > *b > b Retention: elsewhere (see 3a) 3a *b > *p > p Split: only initial in tone set C 4 *ɓ > *pʼ > pʼ Merger (also [b] medial) 5 *pʼ > *pʼ > pʼ Retention: initial and final, [pʼ and b] medial 6 *t ̪h > *t > t Merger 7 *t ̪ > *t > t Merger 8 *d̪ > *d > d Merger: not in tone set C 8a *d̪ > *t > t Merger: only initial before tone set C 9 *t ̪̓ > *tʼ > tʼ Merger 10 (*tʰ) > *t > t Merger: only word-final 11 *t > *t > t Merger 12 *d > *d > d Retention: not in tone set C 12a *d > *t > t Split: only initial in tone set C 13 *ɗ > *tʼ > tʼ Merger 14 *tʼ > *tʼ > tʼ Retention 15 *sʼ > *sʼ > sʼ Retention 16 *c > *s > s Merger: no conditioning for either [s] or [ʃ] (cf.23a) 16 *c > *ʃ > ʃ Merger: no conditioning for either [s] or [ʃ] (cf. 23) 17 *ɟ > *z > z Shift 18 *cʼ > *sʼ > sʼ Merger 19 *kʰ > *k > k Merger 20 *k > *k > k Merger 21 *g > *g > g Retention: not before tone set C 21a *g > *k > k Merger: only initial before tone set C 22 *kʼ > *kʼ > kʼ Retention 23 *s > *ʃ > ʃ Split: initial /_V [+high, –ATR] [ɪ, ʊ] 23a *s > *s > s Retention: elsewhere (see 23) 24 *ʃ > *ʃ > ʃ Retention: initial /_V [ɪ, ɛ, ʊ] 24a *ʃ > *s > s Split: initial /_V [i, ɔ, u] 25 *h > *h > h Retention: only initial 26 *r > *r > r Retention 27 *l > *l > l Retention 28 *m > *m > m Retention 29 *n > *n > n Retention 30 (*ɲ) > *n > n Merger 31 (*ŋ) > *ŋ > ŋ Retention: only medial 31 (*ŋ) > *g > g Merger: only medial, no conditioning 32 *w > *w > w Retention 33 *j > *j > j Retention 34 (*T̪) > *ʃ > ʃ Merger: /_ V[+high, –ATR] [ʊ] and R or L tone 35 (*D̪) > *s > s Merger: /_ V[+high, +ATR] [i] and M or H tone 35 (*D̪) > *z > z /_ V[+high, +ATR] [i] and L tone 231 The evolution of the PKMN palatal stops have an interesting history in PGW, one which suggests a chain shift involving the alveolar and alveopalatal fricatives. PGW underwent two primary splits in PKMN *s and *ʃ. In syllable onset position, these splits were conditioned by the vowel of the syllable nucleus. In Gwama, PKMN *s > PGW *ʃ before [+high, –ATR] [ɪ, ʊ] and [s] was retained elsewhere in this position while PKMN *ʃ > PGW *s before [+high, +ATR] [i, u] and [ɔ]. These splits allowed for the palatal stops to merge as PKMN *c > PGW *s/*ʃ though no conditioning is apparent for either the [s] or [ʃ] reflex. With regard to the voiced palatal stop, PGW innovates *ɟ > z and the palatal ejective merges with the alveolar fricative/affricate ejective PKMN *cʼ > PGW *sʼ. This series of innovations led to the loss of a PKMN palatal stop series in PGW. The questionable PKMN proto-phonemes (*T̪, *D̪), only found in onsets, also correspond to PGW fricatives. These correspondences, while thin, nonetheless exhibit similar outcomes to the PKMN palatals. The PGW reflexes appear to be conditioned by the ATR value of the following vowel and also tone in some cases. For instance, PKMN *T̪ > PGW *ʃ before [–ATR] [ʊ] parallels *s > *ʃ before [–ATR] [ɪ, ʊ] and *D̪ > *s before [+ATR] [i] parallels *ʃ > *s before [i, u]. 4.1.2 Gwama vowels and tone PGW exhibited a seven-vowel contrastive inventory /i, ɪ, ɛ, a, ɔ, ʊ, u/ which exhibited some unique innovations, specifically with regard to the high vowels /i, ɪ, ʊ, u/. Recall that retentions or splits of PKMN *s and *ʃ in onset position were conditioned largely by the ATR value of the following vowel with [s] correlating with [+high, +ATR] and [ʃ] correlating with [+high, –ATR]. 232 The Gwama varieties synchronically exhibit stem-controlled ATR harmony only in the high vowels. An identical system is seen in the Dana-Opo branch and this would either suggest a retention of this harmony system in both branches from PKMN, or two independent innovations. Given that Komo is the only language exhibiting ATR harmony in the Komo-Uduk branch, it seems likely that the stem-controlled ATR system was inherited in PGW.165 Gwama most likely inherited a PKMN tone system that contained three level reconstructed categories. These three tone categories evolved into the three level tones seen modernly in the Gwama varieties (see §3.1 for discussion). The development of three level tones from an erstwhile tone system containing two level tones was specifically due to the interaction with voiced stop onsets. These voiced stop onsets functioned as depressor consonants and lowered the F0 of the following vowel nucleus. and ultimately gave rise to three level tones. A subsequent and independent PGW innovation was to devoice some stop onsets and merge them with voiceless stop onsets. This can be seen throughout the Gwama lexicon though there are exceptions to this pattern, which are discussed in §4.1.1. 4.2 Proto-Central Koman (PCTRL) Proto-Central Koman (PCTRL) consists of all of the living Koman languages with the exception of the Gwama branch. In order to reconstruct cognates to PCTRL, there must at least be a reflex in the Komo-Uduk branch that corresponds to a reflex in the Dana- Opo branch. 165 Komo exhibits two distinct ATR harmony systems that are not stem-controlled but rather [+ATR] and [–ATR] dominant systems (Otero 2015, 2018b). 233 Central Koman phonology does not differ that much from PKMN, given that Dana and Chali Uduk retain most, if not all of the PKMN consonant and vowel system, though it could be argued that several morphological innovations gave rise to Central Koman. One strong feature of the verb system in Central Koman is nominal and verbal number marking in the verb, which is seen robustly in the Dana-Opo branch (Smolders forthcoming) and in Komo of the Komo-Uduk branch (Otero 2018b). The Uduk varieties retain some segmentally suppletive verbs (Stevenson 1942, Killian 2015). 4.2.1 PCTRL consonants PCTRL largely retained the consonant system inherited from PKMN with few exceptions. The PCTRL consonant inventory is provided in Table 87. Table 87 Proto-Central Koman (PCTRL) consonant inventory *pʰ (*t ̪h ) *tʰ *kʰ *p *t ̪ *t *c *k *b *d̪ *d *ɟ *g *pʼ *t ̪̓ *tʼ *cʼ *kʼ *ɓ *ɗ *s *ʃ *h (*T̪) (*D̪) *sʼ *m *n *ŋ *l *ɲ *r *w *j The PCTRL consonant inventory strongly resembles that of PKMN, with retentions of all of the PKMN consonants. With regard to sound correspondences, PCTRL reflexes 234 exhibit the same sound correspondences as seen in PKMN (Chapter III). Given the strict criteria for reconstruction used here, I also recognize that some PCTRL cognates may well reconstruct to PKMN, having suffered a loss in Gwama. Table 88 provides a schematized version of PCTRL retentions and innovations. Table 88 Proto-Central Koman (PCtrl) phonological innovations SET PKMN PCTRL Observations 1 *pʰ > *pʰ Retention 2 *p > *p Retention 3 *b > *b Retention 4 *ɓ > *ɓ Retention 5 *pʼ > *pʼ Retention 6 *t ̪h > *t ̪h Retention 7 *t ̪ > *t ̪ Retention 8 *d̪ > *d̪ Retention 9 *t ̪̓ > *t ̪̓ Retention 10 (*tʰ) > (*tʰ) Retention (weak evidence overall) 11 *t > *t Retention 12 *d > *d Retention 13 *ɗ > *ɗ Retention 14 *tʼ > *tʼ Retention 15 *sʼ > *sʼ Retention 16 *c > *c Retention (weak PCTRL evidence) 17 *ɟ > *ɟ Retention (weak PCTRL evidence) 18 *cʼ > *cʼ Retention 19 *kʰ > *kʰ Retention 20 *k > *k Retention 21 *g > *g Retention 22 *kʼ > *kʼ Retention 23 *s > *s Retention 24 *ʃ > *ʃ Retention 25 *h > *h Retention 26 *r > *r Retention 27 *l > *l Retention 28 *m > *m Retention 29 *n > *n Retention 30 (*ɲ) > (*ɲ) Retention: only final 31 (*ŋ) > (*ŋ) Retention: only medial/final 32 *w > *w Retention 33 *j > *j Retention 34 (*T̪) > (*T̪) Retention 35 (*D̪) > (*D̪) Retention 235 The bilabial series is robustly attested across stops, nasals, and glides. There is good evidence for the interdental stops *t,̪ *d̪ and *t ̪̓ . Evidence for a voiceless aspirated interdental stop *t ̪h is still weak, given the fact that the best evidence for this proto- phoneme relies in cognates with a voiceless aspirated interdental reflex in Chali Uduk corresponding to a voiceless aspirated interdental reflex in Dana. The only reflexes of *t ̪h are seen in (170). Note that the only solid correspondence is in (170a), and possibly (170b), though Chali Uduk exhibits an unaspirated interdental. Lastly, in (170c-d), Chali exhibits [tʰ] corresponding to Dana [t ̪h ]. Evidence is weak overall for *t ̯h . (182) PCTRL Komo Chali Uduk Dana Opo Meaning a. *t ̪h at ̪̓ – t ̪h āt ̪̓ t ̪h ājd̪á – ‘mediate’ b. *t ̪h ipʰ – tī̪pʰ t ̪h ì – ‘raise (a child)’ c. *t ̪h ùd̪ tùd tʰūr t ̪h ùd̪ – ‘dry out’ d. *t ̪h (w)ád̪ twáj tʰɛ ́ t ̪h ád̪ tʰáj ‘hunger’ The palatal series of plain stops *c and *ɟ have few reflexes in PCTRL, though confidence in the existence of these stops is bolstered by identical correspondences in PKMN cognates (see *ɟ correspondence sets 17). Stops in word-final position are generally diachronically unstable. Some languages retain reflexes of word-final stops which correspond to glides or a total loss of the segment in other languages. For instance, in (170d), word-final /d̪/ is retained in Dana which corresponds to /j/ in Komo and Opo. Note that Chali Uduk has lost the final stop but its effect remains in the vowel as word-final /aj/ sequences can coalesce to /ɛ/. 236 4.2.2 A note on *T̪ and *D̪ in PCTRL In PCTRL, there is a bit more evidence for what in PKMN are marginal proto- phonemes: *T̪ and *D̪. These proposed proto-phonemes gain credence when examining PCTRL reflexes, though deciding whether to reconstruct a proto-phoneme and if so, what to reconstruct is challenging. The PCTRL *T̪ and *D̪ correspondence sets are in Table 89. Table 89 PCTRL *T̪ and D̪ correspondence set Komo Uduk Dana Opo Set PCTRL Yab Cha Bil Mod Pam Kig 34 *T̪ s/z s t ̪ s tʃ tʃ tʃ tʃ 35 *D̪ z/j j j/w d̪ dʒ dʒ z s/ʃ In the *T̪ correspondence set, Chali Uduk exhibits /t/̪ which corresponds to Dana /s/ and Opo /tʃ/. Komo alternates between /s/ and /z/, which cannot be explained except possibly by tone. The PCTRL *T̪ reflexes are seen in (183). Note that ‘resemble’ constitutes the only robust set. Dana /h/ could be explained as weakening from *s. I cannot explain the Komo /z/ correspondence in ‘nosebleed’ other than the fact that it is not H tone. The ‘arrow’ correspondence is problematic given that the initial correspondences fit the pattern, but the vowels do not correspond in Dana-Opo. Further it is important to mention that the Komo cognate sá is also ‘tree’, though this cognate forms part of a *c correspondence.166 I include this Komo word in parentheses to indicate it may not be cognate with the other ‘arrow’ forms. 166 *c reflexes are /s/ in Komo, /tʃ/ in Opo, and /c/ in Uduk and Dana. 237 (183) Komo Uduk Uduk Dana Opo Meaning (Yabus) (Chali) (Bil) a. sál sál tá̪l hálà tʃál ‘resemble’ b. – – tī̪t ̪̓ – tʃītʃʼá ‘urinate’ c. zīnà – – – àtʃwìnà ‘nosebleed’ d. (sá) sá à+tá̪ sɛt̄ ̪h tʃɛ ̄ ‘arrow’ The *D̪ reflex is even more puzzling. The reflexes are seen in (184). In the DAOP branch, Dana /d̪/ corresponds to /dʒ, dʒ, z, s/ʃ/ in the Opo varieties, with the exception of /d/ in Bilugu Opo in ‘penis’ and /dʒ/ in Pame Opo in ‘string (v.)’. This aside, there does seem to be a regular correspondence in DAOP. Turning to the KOUD branch, in Komo, the reflexes of *D̪ are /z/, /j/ or a loss altogether as in ‘string (v.)’, while the Uduk reflexes are /w/ and /j/. The /w/ could be the result of coalescence between an initial glide and a back vowel though this is speculation. (184) Komo Uduk Dana Opo Opo Opo Opo Meaning (Bil) (Mod) (Pam) (Kig) a. zǐt – – dʒǐtʰ dʒǐtʰ zǐtʰ ʃǐtʰ ‘shade’ b. zāgà – – dʒùgà – – ʃùgà ‘name’ c. jìʃ jīs d̪ɪ̀s dʊ̄s dʒʊ̄s zʊ̄s sʊ̄s ‘penis’ d. jà jà d̪ā dʒà dʒà zà ʃà ‘go.SG’ e. ɔ̀m wɔ̌m d̪ɔ̀m dʒɔ̀m dʒɔ̀m dʒɔ̀m ʃɔ̀m ‘string (v.)’ An important note here must be made for the deictic ‘go’ and ‘come’ verbs, which appear to be a Central Koman innovation. The segmental histories of these verbs are very difficult to tease apart for a number of reasons. First, there are both singular and plural forms for each verb as seen in (185). Second, some of the ‘come’ verbs can be constructed from ‘go’ roots plus a deictic directional morpheme that indicates motion 238 towards the speaker. Many of these verb forms have fused a historical root and a directional morpheme. (185) Komo Uduk Dana Opo Opo Opo Opo Meaning (Chali) (Bil) (Mod) (Pam) (Kig) a. jà jà d̪ā dʒà dʒà zà ʃà ‘go.SG’ b. ɪ̀ ī d̪â ʔjá ʔjá ʔjá ʔjá ‘go.PL’ c. – jà-í jʊ́ dʒʊ̄ dʒʊ̄ zʊ̄ ʃʊ̄ ‘come.SG’ d. ɪ̀-ʊ́ ī-ú já dʒā dʒā-ɪ ́ zā-jʊ́ ʃā-jʊ́ ‘come.PL.’ With specific regard to initial consonants, what stands out in the ‘go.SG’ reflexes is a *D̪ correspondence. Thus, ‘go.SG’ (or one form of a historical ‘go’ or ‘move’ verb) could conceivably reconstruct to *D̪a ‘go, move’. It appears that Proto-Opo (POP) constructs the ‘come’ forms from *D̪a plus a deictic directional *ʊ or *ʊɪ.167 The /d̪/ corresponds to /j/ in Komo-Uduk in the singular ‘come’ forms. Whether *D̪a is cognate with the plural ‘come’ verbs in Komo-Uduk is unclear, but likely. The plural ‘come’ forms may have arisen from a plural ‘go’ root *ʔj(a) seen in Dana. Note that the initial segment in this form is one of the few true glottal stop onsets in Koman. Komo has a distinct (and cognate) form ʔjá ‘go.SG’ which only occurs in a particular deictic directional paradigm (Otero 2015b, 2018b). I cannot offer a definitive reconstruction of Koman ‘go/come’ verb roots at present. Nevertheless, these verbs appear to be very old given the idiosyncrasy in the reflexes that may well reconstruct to *D̪. Given the fact that these *T̪ and *D̪ irregular correspondences hold to varying degrees from PKMN down to lower nodes, one possible solution is to posit contrasting 167 Whether to reconstruct *ʊ, *ɪ or a form such as *ʊɪ for deictic directional 1 (DD1) is unclear at present. See §2.2.3.2 for description of deictic directional morphology in Koman. 239 PKMN (or PCTRL) interdental fricatives *θ and *ð. The voicing is almost the same across the reflexes of the interdentals as is found in the corresponding reflexes of the proto fricatives and affricates. This is speculation at this point, yet by examining PCTRL and PKMN correspondences, the data suggest possibly reconstructing two additional proto- phonemes. 4.2.3 PCTRL vowels and tone The PCTRL vowel system reconstructs to a seven-vowel system /i, ɪ, ɛ, a, ɔ, ʊ, u/ with ATR contrast in the high vowels. There is evidence to suggest that PCTRL exhibited some kind of ATR vowel harmony system given that Komo and the Dana-Opo branch exhibit ATR harmony systems even though the systems are distinct. If PCTRL exhibited an ATR harmony system, it was most likely a stem-controlled system involving only the high vowels. This is due to the fact that there is phonemic ATR contrast only in the high vowels (i.e. there are no [+ATR] counterparts to /e, a, ɔ/), coupled with the observation that both Gwama and the Dana-Opo branch exhibit stem-controlled ATR harmony in the high vowels. Hypothesizing the retention of a stem-controlled ATR harmony system in Gwama and Dana-Opo requires fewer changes than positing that these two systems were independent yet identical innovations. This hypothesis also allows for the Komo innovation of a distinct harmony system from the inherited stem- controlled system in the other sub-branches. Therefore it is highly likely that PCTRL or even PKMN was characterized by a stem-controlled ATR harmony system involving only the high vowels /i, ɪ, ʊ, u/. Tone figures prominently in the history of Koman phonology. It’s historical relationship to word-initial (syllable onset) stops necessitates examining historical tone 240 and the onset consonants in tandem. Section 3.1 provided a historical scenario for the development of tone in PKMN beginning with a system which exhibited two contrasting level tones. From there, voiced stops acted as depressor consonants, lowering the pitch of the following vowel, which ultimately led to the phonologization of an additional third level tone. This accounted for much of the data with the exception how Dana and Komo acquired synchronic M tone. Recall that the correspondence for tone set B has M tone in all languages except for L tone in Komo and Dana (see Table 45 in §3.1.1). Further, there are no PKMN cognates with plain stop onsets in which all languages, including Komo and Dana, exhibit M tone. Examining the tone distributions of PCTRL cognates provides a slightly different picture than that we can deduce from PKMN cognates, namely one additional robust tone correspondence set. The PCTRL tone correspondence sets mapped on to the distribution of stop onsets is seen in Table 90. Note that the historical tone classes based on the distribution in PKMN cognates are still robustly attested in PCTRL (i.e. tone sets A-C) and that set C is still in complementary distribution with set B with regard to voicing of stop onsets. Tone set D, in which all languages have corresponding M tone, is robust in PCTRL. This accounts for the M tones seen synchronically in Komo and Dana, but I cannot offer an explanation as to how this arose at present. 241 Table 90 Tone categories in PCTRL correspondence sets and the distribution with stop onsets Set Komo Uduk Dana Opo stop onsets A H H H H voiceless aspirated/unaspirated B L M L M voiceless aspirated/unaspirated C L L L L voiced D M M M M voiceless aspirated/unaspirated 4.3 Proto-Komo-Uduk (PKOUD) PKOUD is one of two main branches which split from PCTRL Koman. There is virtually no change in the consonant, vowel and tone systems inherited from PCTRL. This is primarily due to the fact that Chali Uduk is extremely conservative, retaining virtually all of the PKMN consonants. One notable phonological innovation is the merger of *T̪ > *t.̪ PUD also retains voiced stops before PKMN *L tone as opposed to PDAOP which devoices and merges voiced stops with voiceless unaspirated stops word- initially before *L. This is also seen partially in Gwama. In terms of morphological innovations, PKOUD loses the masculine singular nominal proclitic/prefix *ʊ̀, which is retained in Gwama and Dana-Opo (§5.2.1). Another morphological innovation that characterizes this branch is the development of PKOUD *-ki/ku DD2 deictic directional morpheme (§5.4). 4.3.1 PKOUD consonants The PKOUD consonant inventory is seen in Table 91. The PKOUD consonant system exhibits retentions across the board from PCTRL with the exception of two consonants: *T̪ and *D̪. 242 Table 91 Proto-Komo-Uduk (PKOUD) consonant inventory *pʰ (*t ̪h ) *tʰ *kʰ *p *t ̪ *t *c *k *b *d̪ *d *ɟ *g *pʼ *t ̪̓ *tʼ *cʼ *kʼ *ɓ *ɗ (*T̪) *s *ʃ *h (*D̪) *sʼ *m *n *ɲ *ŋ *l *r *w *j While there were no significant innovations in the consonant system in PKOUD save innovation of the marginal proto-phonemes *T̪ and *D̪. I provide a list of retentions in Table 92. Note that the fact that a particular proto-phoneme not having cognates that reconstruct to this node does not imply that the phoneme was not in the proto-system as there are lexemes in higher nodes that exhibit all of the reflexes in this table. Table 92 Proto-Komo-Uduk (PKoUd) phonological innovations SET PKMN PCTRL PKOUD Observations 1 *pʰ > *pʰ > *pʰ Retention 2 *p > *p > *p Retention 3 *b > *b > *b Retention 4 *ɓ > *ɓ > *ɓ Retention 5 *pʼ > *pʼ > *pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h Retention (no cognates in PKOUD) 7 *t ̪ > *t ̪ > *t ̪ Retention (no cognates in PKOUD) 8 *d̪ > *d̪ > *d̪ Retention 9 *t ̪̓ > *t ̪̓ > *t ̪̓ Retention (no cognates in PKOUD) 10 (*tʰ) > (*tʰ) > (*tʰ) Retention (no cognates in PKOUD) 11 *t > *t > *t Retention 12 *d > *d > *d Retention 13 *ɗ > *ɗ > *ɗ Retention 14 *tʼ > *tʼ > *tʼ Retention (no cognates in PKOUD) 243 Table 92 Proto-Komo-Uduk (PKoUd) phonological innovations SET PKMN PCTRL PKOUD Observations 15 *sʼ > *sʼ > *sʼ Retention 16 *c > *c > *c Retention 17 *ɟ > *ɟ > *ɟ Retention 18 *cʼ > *cʼ > *cʼ Retention (no cognates in PKOUD) 19 *kʰ > *kʰ > *kʰ Retention 20 *k > *k > *k Retention 21 *g > *g > *g Retention 22 *kʼ > *kʼ > *kʼ Retention 23 *s > *s > *s Retention 24 *ʃ > *ʃ > *ʃ Retention 25 *h > *h > *h Retention (no cognates in PKOUD) 26 *r > *r > *r Retention (no cognates in PKOUD) 27 *l > *l > *l Retention (no cognates in PKOUD) 28 *m > *m > *m Retention 29 *n > *n > *n Retention (no cognates in PKOUD) 30 (*ɲ) > (*ɲ) > (*ɲ) Retention: only final (no cognates in PKOUD) 31 (*ŋ) > (*ŋ) > (*ŋ) Retention: only medial/final (no cognates in PKOUD) 32 *w > *w > *w Retention 33 *j > *j > *j Retention 34 (*T̪) > (*T̪) > (*T̪) Retention 35 (*D̪) > (*D̪) > (*D̪) Retention (no cognates in PKOUD) The *T̪ and *D̪ proposed proto-phonemes were discussed in PCTRL cognates in §4.2.2. The only PKOUD cognate with reflexes of *T̪ is in (186) and it follows the correspondences seen in higher nodes. (186) Komo Uduk (Yabus) Uduk (Chali) Meaning sìn sīn tī̪n ‘tail’ PKOUD most likely retained *D̪ realized as some type of fricative or affricate given that it is realized as a fricative in Komo /s, z/ but (possibly) a glide /j, w/ in Uduk. Thus, positing PCTRL*D̪ > PKOUD *j/*w does not seem likely given that a change such as PKOUD *j/*w > Komo /s, z/ is unattested. Rather, retention of *D̪ in PKOUD and the 244 subsequent *D̪ > /s, z/ seems more likely. Given the paucity of reflexes, this is speculation at this point. 4.3.2 PKoUd vowels and tone PKOUD exhibited a seven-vowel system /i, ɪ, ɛ, a, ɔ, ʊ, u/ with ATR contrast in the high vowels. If stem-controlled ATR harmony was present in PCTRL (inherited from PKMN), then it was also present PKOUD. Given the fact that modern Uduk has collapsed the seven-vowel system into five vowels with no ATR contrast, but Komo retains the PKMN seven-vowel system with ATR contrast in the high vowel, I reconstruct the ATR value of the high vowels exhibited in Komo to PKOUD. At the PKOUD stage in Koman’s history, two level tones were in complementary distribution with voiced and voiceless stop onsets, which was inherited from PCTRL (c.f. Table 45). This complementary distribution between tone and onsets is retained in the modern Uduk varieties.168 Further, this distribution holds across all Uduk reflexes in cognates reconstructed to PKOUD, PCTRL and PKMN. 4.4 Komo Komo underwent significant innovations from PKOUD specifically with regard to the consonant system such as the loss of interdental and palatal stops. Further, it developed a unique ATR harmony system unseen in any of the other Koman languages (§2.1.5.3). In the tone system, Komo extended the three level tone system to extend to the full range of stop onsets and onsets: synchronic Komo does not have any consonant-tone restrictions. In terms of morphological innovations, Komo innovated 168 See §2.1.6.6 for discussion of synchronic Uduk consonant-tone restrictions and §3.1 for discussion of PKMN tone. 245 “exchoativity” in the deictic directional system in which a particular morpheme on stative verb roots expresses a state that no longer holds (§2.2.3.2). Komo also innovated an extensive system of participant indexing on the verb (§2.2.3.1.2). In the following subsections. I treat the major Komo phonological innovations. 4.4.1 Komo consonants Significant Komo innovations, largely by way of mergers, gave rise to the consonant inventory seen in Table 5 and reproduced here in (187). Note that the glottal stop and velar nasal are marginally contrastive. (187) p t k (ʔ) b d g pʼ tʼ kʼ ɓ ɗ s ʃ h z sʼ m n (ŋ) l r w j One overarching innovation in Komo was the complete merger of aspirated and unaspirated voiceless stops. This merger took place across three places of articulation: bilabial, interdental and alveolar. Further, the interdental and alveolar stops merged completely, drastically reducing the number of phonemic consonants. A schematized version of the Komo innovations is provided in Table 93. Table 93 Komo phonological innovations SET PKMN PCTRL PKOUD Komo Observations 1 *pʰ > *pʰ > *pʰ > p Merger 2 *p > *p > *p > p Merger 3 *b > *b > *b > b Retention 246 Table 93 Komo phonological innovations SET PKMN PCTRL PKOUD Komo Observations 4 *ɓ > *ɓ > *ɓ > ɓ Retention: initial 4 *ɓ > *ɓ > *ɓ > pʼ Merger: medial, final 5 *pʼ > *pʼ > *pʼ > pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h > t Merger 7 *t ̪ > *t ̪ > *t ̪ > t Merger 8 *d̪ > *d̪ > *d̪ > d Merger 8 *d̪ > *d̪ > *d̪ > z Shift: spirantization before [i] 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > tʼ Merger 10 (*tʰ) > (*tʰ) > (*tʰ) > t Merger 11 *t > *t > *t > t Merger 12 *d > *d > *d > d Retention 13 *ɗ > *ɗ > *ɗ > ɗ Retention 13 *ɗ > *ɗ > *ɗ > r Merger: medial, final 14 *tʼ > *tʼ > *tʼ > tʼ Retention 15 *sʼ > *sʼ > *sʼ > sʼ Retention 16 *c > *c > *c > s Shift 17 *ɟ > *ɟ > *ɟ > z Shift 18 *cʼ > *cʼ > *cʼ > sʼ Merger 19 *kʰ > *kʰ > *kʰ > k Merger 20 *k > *k > *k > k Merger 21 *g > *g > *g > g Retention 22 *kʼ > *kʼ > *kʼ > kʼ Retention 23 *s > *s > *s > ʃ Merger 24 *ʃ > *ʃ > *ʃ > ʃ Retention 25 *h > *h > *h > h Retention 26 *r > *r > *r > r Retention 27 *l > *l > *l > l Retention 28 *m > *m > *m > m Retention 29 *n > *n > *n > n Retention 30 (*ɲ) > (*ɲ) > (*ɲ) > n Merger: only final 31 (*ŋ) > (*ŋ) > (*ŋ) > g Merger: only medial 32 *w > *w > *w > w Retention 33 *j > *j > *j > j Retention 34 (*T̪) > (*T̪) > (*T̪) > s Merger 35 (*D̪) > (*D̪) > (*D̪) > z Shift There appears be a chain shift in Komo beginning with the merger of the alveolar and alveopalatal fricatives *s > ʃ, which allowed for the subsequent shift *c > s. These shift must have occurred in this order or else an initial merger of *c > s would not justify only some /s/-initial words to have split to /ʃ/. 247 Another Komo innovation is the now-contrastive phoneme /z/. While this phoneme exists synchronically in three branches, I cannot reconstruct it to a proto-phoneme at any stage given the lack of solid correspondences. Komo /z/ could have arisen via the shift of *ɟ > z as it parallels the *c > ʃ shift in voicing. This is one possible source. Another possibility for the origin of /z/ in Komo is via the spirantization of *d̪ > z before high front vowels and L tone (PKMN *d̪ìm ‘strain’ > zìm) and/or the weak *D̪ correspondence of *D̪ > z. Note also that Komo merges *T̪ > s. 4.4.2 Komo vowels and tone The synchronic seven-vowel system /i, ɪ, ɛ, a, ɔ, ʊ, u/ in modern-day Komo was inherited all the way from PKMN. If an ATR harmony system existed in PKMN further down the nodes, it was most likely a stem-controlled system in which [+high] affix vowels harmonize to the ATR of a [+high] stem vowel. This is due to the fact that this ATR harmony system exists modernly in Gwama and in the Dana-Opo branch. Even if ATR harmony were not reconstructed to PKMN or any of the nodes below it, one very notable innovation in Komo is the unique bi-directional ATR harmony system seen modernly (§2.1.5.3 also Otero 2015, Olejarczuk et al. 2019). Komo inherited a tone system with three level tones and crucially with distributional restrictions regarding the voicing of word-initial (syllable onset) stops. Throughout its history, Komo lost the consonant-tone restrictions. Synchronic Komo onsets are not sensitive to the voicing of the initial consonant, be it a stop or otherwise. As such, all consonants can occur with all three level tones in synchronic Komo. The trajectory of how Komo gained the tone system relative to onsets requires further investigation. 248 4.5 Proto-Uduk (PUD) phonology Uduk consists of the Northern varieties, Chali and Bonya and the Southern variety Yabus Uduk. The Chali and Bonya varieties are much closer in structure and the Yabus variety is the most divergent. This study focuses only on the Chali and Yabus varieties and a cognate must have reflexes in both Chali and Yabus to be reconstructed to PUD. A notable PUD innovation is the collapse of the PKMN seven-vowel inventory into a five- vowel system along with a loss of ATR contrast and ATR harmony which reconstructs to PKMN. I discuss the major PUD phonological innovations in the following subsections. 4.5.1 PUD consonants The Proto-Uduk PUD reconstructed consonant inventory is in Table 94. Table 94 Proto-Uduk (PUD) consonant inventory *pʰ (*t ̪h ) *tʰ *kʰ *p *t ̪ *t *c *k *b *d̪ *d *ɟ *g *pʼ *t ̪̓ *tʼ *cʼ *kʼ *ɓ *ɗ *s *ʃ *h (*T̪) (*D̪) *sʼ *m *n *ŋ *l *ɲ *r *w *j PUD retained much of the PCTRL consonant inventory with a few notable exceptions. The PUD retentions and innovations are schematized in Table 95 and discussed below. 249 Table 95 Proto-Uduk (PUd) phonological innovations SET PKMN PCTRL PKOUD PUD Observations 1 *pʰ > *pʰ > *pʰ > *pʰ Retention 2 *p > *p > *p > *p Retention 3 *b > *b > *b > *b Retention 4 *ɓ > *ɓ > *ɓ > *ɓ Retention 5 *pʼ > *pʼ > *pʼ > *pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h > *t ̪h Retention 7 *t ̪ > *t ̪ > *t ̪ > *t ̪ Retention 8 *d̪ > *d̪ > *d̪ > *d̪ Retention 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > *t ̪̓ Retention: only initial 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > *ɗ Split: only final 10 (*tʰ) > (*tʰ) > (*tʰ) > (*tʰ) Retention 11 *t > *t > *t > *t Retention 12 *d > *d > *d > *d Retention 13 *ɗ > *ɗ > *ɗ > *ɗ Retention 14 *tʼ > *tʼ > *tʼ > *tʼ Retention 15 *sʼ > *sʼ > *sʼ > *sʼ Retention 16 *c > *c > *c > *c Retention 17 *ɟ > *ɟ > *ɟ > *ɟ Retention 18 *cʼ > *cʼ > *cʼ > *cʼ Retention 19 *kʰ > *kʰ > *kʰ > *kʰ Retention 20 *k > *k > *k > *k Retention 21 *g > *g > *g > *g Retention 22 *kʼ > *kʼ > *kʼ > *kʼ Retention 23 *s > *s > *s > *s Retention 24 *ʃ > *ʃ > *ʃ > *ʃ Retention 25 *h > *h > *h > *h Retention 26 *r > *r > *r > *r Retention 27 *l > *l > *l > *l Retention 28 *m > *m > *m > *m Retention 29 *n > *n > *n > *n Retention 30 (*ɲ) > (*ɲ) > *ɲ > *ɲ Retention + Extension 31 (*ŋ) > (*ŋ) > (*ŋ) > *ŋ Retention: only medial/final 32 *w > *w > *w > *w Retention 33 *j > *j > *j > *j Retention 34 (*T̪) > (*T̪) > (*T̪) > (*T̪) Retention 35 (*D̪) > (*D̪) > (*D̪) > (*D̪) Retention One significant innovation in the modern Uduk varieties is the palatalization of all velar stops (including the ejective) before front vowels. This innovation, or series of innovations, split the velar series and merged those that were palatalized with the existing palatal stops. These changes have spread through the lexicon of Chali and 250 Yabus Uduk in that there are no lexemes containing a velar onset followed by a front vowel.169 Thus it is tempting, and economical, to posit that these changes occurred once in PUD. This seems likely were it not for the fact that Yabus Uduk split PUD *ɟ > ʒ and PUD *cʼ > ʃʼ. Crucially, these Yabus Uduk splits occurred in palatal reflexes that were not historically velars. Thus, two independent innovations of the split and merger of velars in onsets with front vowels had to occur independently in Chali and Yabus Uduk, despite this not being the most elegant solution. PUD extends the palatal nasal to initial position. Recall that there is some evidence for *ɲ in cognates higher up in the family, though the distribution is restricted to medial and final position. The velar nasal is completely phonologized in PUD though it only occurs in syllable coda position. In the interdental series, PUD splits and merges *t ̪̓ > *ɗ in coda position. This merger progresses subsequently in Yabus Uduk as *ɗ > *d. 4.5.2 PUD vowels and tone One important innovation that distinguishes the Uduk varieties from the rest of the living Koman languages is the merger of the seven-vowel system into a five-vowel system in PUD. This merger occurred in the high vowels where both *i and *ɪ merged into one vowel /i/, and the back vowels *u and *ʊ merged into /u/. Previous scholars have recognized the phonetic variation in Chali Uduk vowels (Stevenson 1942, Killian 169 There are several exceptions in Chali Uduk, but they are all grammatical morphemes with the shape [ki] and distinct tones. These morphemes range in function from complementation, direction of motion to marking oblique phrases (see Killian 2015). I am grateful to Don Killian for pointing out these exceptions to me. 251 2015) though there is no evidence for anything other than a five-vowel contrastive system. The PUD tone system retains the system inherited from PCTRL (§4.2.3). This tone system had distributional restrictions with regard to voicing in onsets and this remains completely phonologized in the Uduk varieties today. 4.6 Chali Uduk Chali Uduk has figured prominently in reconstructions of PKMN given its extensive consonant system (Bender 1983, Ehret 2001). Chali Uduk retains most of the PKMN consonant system with some important innovations: the palatalization of velar stops before front vowels, *sʼ > t ̪̓ and *t ̪̓ > tʼ, among others. One important Chali Uduk innovation outside of phonology is the development of a nominal gender/class system. This gender/class system is outlined and discussed in Killian (2015) and summarized in §2.2.2.3.3 of this manuscript. 4.6.1 Chali Uduk consonants Chali Uduk is one of the most conservative Koman languages, retaining almost every consonant inherited from PKMN. The synchronic Chali Uduk consonant system is presented in Table 7 and reproduced in (188). (188) pʰ t ̪h tʰ cʰ kʰ p t ̪ t c k (ʔ) b d̪ d ɟ g pʼ t ̪̓ tʼ cʼ kʼ ɓ ɗ s ʃ h m n ɲ ŋ l r w j 252 The Chali Uduk retentions and innovations are schematized in Table 96. Table 96 Chali Uduk phonological innovations SET PKMN PCTRL PKOUD PUD Chali Observations Uduk 1 *pʰ > *pʰ > *pʰ > *pʰ > pʰ Retention 2 *p > *p > *p > *p > p Retention 3 *b > *b > *b > *b > b Retention 4 *ɓ > *ɓ > *ɓ > *ɓ > ɓ Retention 5 *pʼ > *pʼ > *pʼ > *pʼ > pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h > *t ̪h > t ̪h Retention 7 *t ̪ > *t ̪ > *t ̪ > *t ̪ > t ̪ Retention 8 *d̪ > *d̪ > *d̪ > *d̪ > d̪ Retention 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > *t ̪̓ > tʼ Merger: only onsets 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > *ɗ > ɗ Retention: only codas 10 (*tʰ) > (*tʰ) > (*tʰ) > (*tʰ) > tʰ Retention 11 *t > *t > *t > *t > t Retention 12 *d > *d > *d > *d > d Retention 13 *ɗ > *ɗ > *ɗ > *ɗ > ɗ Retention 14 *tʼ > *tʼ > *tʼ > *tʼ > tʼ Retention 15 *sʼ > *sʼ > *sʼ > *sʼ > t ̪̓ Merger 16 *c > *c > *c > *c > c Retention 17 *ɟ > *ɟ > *ɟ > *ɟ > ɟ Retention 18 *cʼ > *cʼ > *cʼ > *cʼ > cʼ Retention 19 *kʰ > *kʰ > *kʰ > *kʰ > kʰ Retention 19 *kʰ > *kʰ > *cʰ > *kʰ > cʰ Split: before front vowels 20 *k > *k > *k > *k > k Retention 20 *k > *k > *k > *k > c Split: before front vowels 21 *g > *g > *g > *g > g Retention 21 *g > *g > *g > *g > ɟ Split: before front vowels 22 *kʼ > *kʼ > *kʼ > *kʼ > kʼ Retention 22 *kʼ > *kʼ > *kʼ > *kʼ > cʼ Split: before front vowels 23 *s > *s > *s > *s > s Retention 24 *ʃ > *ʃ > *ʃ > *ʃ > ʃ Retention 25 *h > *h > *h > *h > h Retention 26 *r > *r > *r > *r > r Retention 27 *l > *l > *l > *l > l Retention 28 *m > *m > *m > *m > m Retention 29 *n > *n > *n > *n > n Retention 30 (*ɲ) > (*ɲ) > *ɲ > *ɲ > ɲ Retention 31 (*ŋ) > (*ŋ) > (*ŋ) > *ŋ > ŋ Retention: medial/final 253 Table 96 Chali Uduk phonological innovations SET PKMN PCTRL PKOUD PUD Chali Observations Uduk 32 *w > *w > *w > *w > w Retention 33 *j > *j > *j > *j > j Retention 34 (*T̪) > (*T̪) > (*T̪) > (*T̪) > t ̪ Merger 35 (*D̪) > (*D̪) > (*D̪) > (*D̪) > j/h Merger One significant innovation in Chali Uduk is what appears to be a chain shift in two PUD ejectives: *t ̪̓ and *sʼ. This presumably began with *t ̪̓ > tʼ followed by *sʼ > t ̪̓ . This is supported by the fact that in my database at least, there are no cognates in which both Chali Uduk and Dana both exhibit interdental ejective reflexes /t ̪̓/. Another innovation in Chali Uduk, which is paralleled in the Yabus variety, is the split of the velar stop series before front vowels. The split of *kʰ > cʰ led to the creation of a new phoneme which provided symmetry across the stop series completing the four-way contrast seen modernly: voiceless aspirated, voiceless unaspirated, voiced and ejective. 4.6.2 Chali Uduk vowels and tone The Chali Uduk vowel system retains a five-vowel contrastive inventory /i, ɛ, a, ɔ, u/ with no ATR contrast inherited from the mergers in the high vowels in PUD. The Chali tone system remain identical to those inherited from PUD, with significant synchronic consonant-tone restrictions that are the result of the historical evolution of Koman. 4.7 Yabus Uduk While Chali Uduk exhibits strong retentions in the consonant system, Yabus Uduk is marked by several significant mergers and several innovations in the consonant system that are unique to Koman. Some significant Yabus Uduk innovations are the loss 254 of the interdental series, *ɟ > ʒ, *cʼ > ʃʼ and the spirantization of *d̪ > z in some environments. 4.7.1 Yabus Uduk consonants The Yabus Uduk consonant inventory is presented in Table 11 and reproduced in (189). Note the loss of the interdental series from PUD. (189) pʰ tʰ cʰ kʰ p t c k (ʔ) b d ɟ g pʼ tʼ cʼ kʼ ɓ ɗ s ʃ h z ʒ sʼ ʃʼ m n ɲ ŋ l r w j Yabus Uduk underwent a series of mergers as well as innovated several phonemes. The Yabus Uduk innovations and retentions are presented in Table 97. The major innovations are discussed below. Table 97 Yabus Uduk phonological innovations SET PKMN PCTRL PKOUD PUD Yabus Observations Uduk 1 *pʰ > *pʰ > *pʰ > *pʰ > pʰ Retention 2 *p > *p > *p > *p > p Retention 3 *b > *b > *b > *b > b Retention 4 *ɓ > *ɓ > *ɓ > *ɓ > ɓ Retention 5 *pʼ > *pʼ > *pʼ > *pʼ > pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h > *t ̪h > tʰ Merger 7 *t ̪ > *t ̪ > *t ̪ > *t ̪ > t Merger 8 *d̪ > *d̪ > *d̪ > *d̪ > d Merger 255 Table 97 Yabus Uduk phonological innovations SET PKMN PCTRL PKOUD PUD Yabus Observations Uduk 8 *d̪ > *d̪ > *d̪ > *d̪ > z Split: spirantization before front vowels 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > *t ̪̓ > tʼ Merger 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > *ɗ > d Merger: only codas 10 (*tʰ) > (*tʰ) > (*tʰ) > (*tʰ) > tʰ Retention 11 *t > *t > *t > *t > t Retention 12 *d > *d > *d > *d > d Retention 13 *ɗ > *ɗ > *ɗ > *ɗ > ɗ Retention: only onsets 13 *ɗ > *ɗ > *ɗ > *ɗ > d Merger: only codas 14 *tʼ > *tʼ > *tʼ > *tʼ > tʼ Retention 15 *sʼ > *sʼ > *sʼ > *sʼ > sʼ Retention 16 *c > *c > *c > *c > c Retention 17 *ɟ > *ɟ > *ɟ > *ɟ > ɟ Retention 17 *ɟ > *ɟ > *ɟ > *ɟ > ʒ Split: unconditioned 18 *cʼ > *cʼ > *cʼ > *cʼ > ʃʼ Split 19 *kʰ > *kʰ > *kʰ > *kʰ > kʰ Retention 19 *kʰ > *kʰ > *cʰ > *kʰ > cʰ Split: before front vowels 20 *k > *k > *k > *k > k Retention 20 *k > *k > *k > *k > c Split: before front vowels 21 *g > *g > *g > *g > g Retention 21 *g > *g > *g > *ɟ > ɟ Split: before front vowels 22 *kʼ > *kʼ > *kʼ > *kʼ > kʼ Retention 22 *kʼ > *kʼ > *kʼ > *kʼ > cʼ Split: before front vowels 23 *s > *s > *s > *s > s Retention 24 *ʃ > *ʃ > *ʃ > *ʃ > ʃ Retention 25 *h > *h > *h > *h > h Retention 26 *r > *r > *r > *r > r Retention 27 *l > *l > *l > *l > l Retention 28 *m > *m > *m > *m > m Retention 29 *n > *n > *n > *n > n Retention 30 (*ɲ) > (*ɲ) > *ɲ > *ɲ > ɲ Retention 31 (*ŋ) > (*ŋ) > (*ŋ) > *ŋ > ŋ Retention: medial/final 32 *w > *w > *w > *w > w Retention 33 *j > *j > *j > *j > j Retention 34 (*T̪) > (*T̪) > (*T̪) > (*T̪) > s Retention 35 (*D̪) > (*D̪) > (*D̪) > (*D̪) > j/h Merger The most notable changes in Yabus Uduk involved the merger of the interdental series of plain stops with the alveolar series. In cognates reconstructed to nodes higher than PUD, Yabus Uduk generally exhibits /d/ reflexes of *d̪. There are some cases of /z/ 256 reflexes in higher nodes but these can be explained by spirantization before front vowels (PKMN *d̪ìm ‘strain’ > Yabus Uduk zìm). It appears that an ongoing innovation in Yabus Uduk has begun the creation of phonemic *d̪ > /z/ seen before non-front vowels (PUD *d̪àn ‘elder, big’ > Yabus Uduk zàn). Though this does not explain anomalies such as PUD *d̪ìsʼ ‘sweep’ > Yabus Uduk dìsʼ, in which we would expect Yabus Uduk zìsʼ ‘sweep’ Yabus Uduk innovated two other phonemes, a voiced palatal fricative /ʒ/ and a voiceless palatal fricative ejective /ʃʼ/. I cannot find any conditioning for the split PUD *ɟ > Yabus Uduk /ʒ/. Nevertheless, /ʒ/ is rare in Yabus Uduk overall. Perhaps /ʒ/ is a recent innovation that is making its way through the Yabus Uduk lexicon. The palatal fricative ejective appears to form part of a chain shift wherein PUD *cʼ > Yabus Uduk /ʃʼ/ and PUD *kʼ > Yabus Uduk /cʼ/. There are no instances of Yabus Uduk /ʃʼ/ reflexes that correspond to PUD *kʼ or even PUD *cʼ which was the result of PCTRL *kʼ > PUD *c’. This shift PUD *cʼ > Yabus Uduk /ʃʼ/ is a unique sound change that characterizes Yabus Uduk speech. The questionable proto-phonemes *T̪ and *D̪ have distinct outcomes in Yabus Uduk. Yabus Uduk merges PUD *T̪ > s, though this is attested in a few reflexes only. The only possible cognates which contain a reflex of *D̪ in Yabus Uduk correspond to /j/ and /w/ and are weak overall. 4.7.2 Yabus Uduk vowels and tone The synchronic Yabus Uduk vowel system is a five-vowel contrastive inventory /i, ɛ, a, ɔ, u/ with no ATR contrast. This system was inherited from the merger of the PCTRL seven-vowel system in PUD. The Yabus tone system is retained from PUD, with 257 significant synchronic consonant-tone restrictions that are the result of the historical evolution of Koman. 4.8 Proto-Dana-Opo (PDAOP) phonology The Dana-Opo (DAOP) branch constitutes what was most likely a dialect chain historically. Dana is by far the most divergent dialect and is no longer mutually intelligible with the Opo varieties examined here. While the degree of mutual unintelligibility has yet to be determined definitively, Dana does exhibit a very distinct sound system from the Opo varieties, notably in the consonant and tone systems. Some of the innovations that distinguish this branch are the devoicing of word-initial voiced stops and subsequent merger with voiceless unaspirated stops before *L tone (§3.1) and the innovation of the deictic directional morpheme *-á DD2, which strongly profiles the addressee (§5.4). 4.8.1 PDAOP consonants The Proto-Dana-Opo (PDAOP) consonant system is largely retained from PCTRL Koman. That is, Dana exhibits the most conservative contrastive consonant inventory in Koman. The PDAOP reconstructed consonant inventory is in Table 98. 258 Table 98 Proto-Dana-Opo (PDAOP) consonant inventory *pʰ (*t ̪h ) *tʰ *kʰ *p *t ̪ *t *c *k *b *d̪ *d *ɟ *g *pʼ *t ̪̓ *tʼ *cʼ *kʼ *ɓ *ɗ *s *ʃ *h (*T̪) (*D̪) *sʼ *m *n *ɲ *ŋ *l *r *w *j Table 99 outlines the innovations in consonants in PDAOP. The major innovations, including the impact of tone on word-initial (syllable onset) consonants, is discussed below. Table 99 Proto-Dana-Opo (PDaOp) phonological innovations SET PKMN PCTRL PDAOP Observations 1 *pʰ > *pʰ > *pʰ Retention 2 *p > *p > *p Retention 3 *b > *b > *b Retention 4 *ɓ > *ɓ > *ɓ Retention: only initial 5 *pʼ > *pʼ > *pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h Retention 7 *t ̪ > *t ̪ > *t ̪ Retention 8 *d̪ > *d̪ > *d̪ Retention: elsewhere (see 8a) 8a *d̪ > *d̪ > *t ̪ Split: Initial, before Tone set C 9 *t ̪̓ > *t ̪̓ > *t ̪̓ Retention 10 (*tʰ) > (*tʰ) > *tʰ Retention 11 *t > *t > *t Retention 12 *d > *d > *d Retention: elsewhere (see 12a) 12a *d > *d > *t Split: Initial, before Tone set C 13 *ɗ > *ɗ > *ɗ Retention 14 *tʼ > *tʼ > *tʼ Retention 15 *sʼ > *sʼ > *sʼ Retention 16 *c > *c > *c Retention 17 *ɟ > *ɟ > *ɟ Retention: elsewhere (see 17a) 17a *ɟ > *ɟ > *c Split: Initial, before Tone set C 18 *cʼ > *cʼ > *cʼ Retention 259 Table 99 Proto-Dana-Opo (PDaOp) phonological innovations SET PKMN PCTRL PDAOP Observations 19 *kʰ > *kʰ > *kʰ Retention 20 *k > *k > *k Retention 21 *g > *g > *g Retention: elsewhere (see 21a) 21a *g > *g > *k Split: Initial, before Tone set C 22 *kʼ > *kʼ > *kʼ Retention 23 *s > *s > *s Retention 24 *ʃ > *ʃ > *ʃ Retention 25 *h > *h > *h Retention 26 *r > *r > *r Retention 27 *l > *l > *l Retention 28 *m > *m > *m Retention 29 *n > *n > *n Retention 30 (*ɲ) > (*ɲ) > *n Merger 31 (*ŋ) > (*ŋ) > *ŋ Retention and extension 32 *w > *w > *w Retention 33 *j > *j > *j Retention 34 (*T̪) > (*T̪) > (*T̪) Retention 35 (*D̪) > (*D̪) > (*D̪) Retention The most significant innovation in PDAOP involves word-initial (onset) stop consonants and historical tone. Recall that PKMN exhibited three major correspondence patterns with respect to tone, from which two tone categories can be reconstructed for an even earlier stage (see §3.1). These tone categories were most likely level tones of opposing F0, realized as L and H tones, which over the course of Koman’s history, gave rise to three level tones. The innovation of a third level tone was directly linked to voiced stop onset consonants (§3.1). Voiced stops functioned as depressor consonants and lowered the F0 realized on the vowel nucleus, which subsequently phonologized into three level tones (L, M, H synchronically). There is evidence that at the PDAOP stage, the level tones were in complementary distribution with respect to stop onsets: L tone with voiced stops, M and L with voiceless onsets. A PDAOP innovation was then to devoice almost all voiced stops before L tone, ultimately merging with voiceless unaspirated stops. This 260 conditioned split is seen throughout the Dana-Opo lexicon and is robustly attested in the bilabial, interdental, alveolar and velar stops but absent in the palatal stops. Nevertheless, there are exceptions to this sound change of devoicing, which are discussed in §3.1.2. The implosives *ɓ and *ɗ are retained as implosives in word-initial (onset) position. In word-final, or syllable coda position, there is weakening of *ɓ > p/pʰ though there does not appear to be enough evidence for a consistent pattern or for a strong diachronic signal. The alveolar implosive by contrast, is maintained in all positions in PDAOP. Lastly, the velar nasal is extended in use to all positions from an earlier medial/final restricted distribution in PCTRL. 4.8.2 PDAOP vowels and tone PDAOP retained the seven-vowel system /i, ɪ, ɛ, a, ɔ, ʊ, u/ with contrastive ATR in the high vowels, seen modernly in both Dana and the Opo varieties. Both Dana and Opo exhibit stem-controlled ATR harmony which occurs only with the high vowels (§2.1.8.3 and §2.1.7.3). In this system, high affix vowels harmonize to the ATR feature of the high vowel in the stem. This ATR system can be confidently reconstructed to PDAOP. If PCTRL Koman exhibited an ATR harmony system it would have most likely been the stem-controlled harmony seen in the Dana-Opo branch and in modern day Gwama. This seems probable given that all of the Koman languages that have contrastive seven- vowel systems exhibit some type of ATR harmony. Another possibility would be that Gwama, Komo and Dana-Opo each innovated ATR harmony. 261 I discussed the role of tone and voiced stop onsets in PDAOP in §4.8.1. The general patterns correlate historical tone correspondence set C (in which all languages synchronically exhibit L tone) with word-initial voiced stops. There is strong evidence for subsequent devoicing of voiced stops and merging with voiceless unaspirated stops in PDAOP. Note, that crucially, there is absolutely no evidence for the devoicing and subsequent merger with voiceless aspirated stops in the DAOP branch, or elsewhere in Koman for that matter. Nevertheless, there are correspondences of word-initial voiced stops that occur with L tone in DAOP, which either did not devoice for some reason that I cannot account for, or which may have entered the lexicon via lateral transmission. These are issues to be dealt with in further research. Notwithstanding, the diachronic signal of devoicing and merger outlined above is strong. 4.9 Dana Dana has not figured in any previous work on Koman reconstruction (e.g. Bender 1983, Ehret 2001). Without data from Dana, this reconstruction would have been significantly different. Dana forms a key piece to the reconstruction of PKMN given its conservative consonant system. From a historical perspective, it is the retention of interdental consonants in Dana which correspond to interdental consonants in Chali Uduk that provides evidence for a PKMN interdental series of consonants. 4.9.1 Dana consonants With regard to the contrastive consonant inventory, Dana is the most conservative of all Koman languages. The Dana consonant inventory is seen in Table 17 and reproduced in (190). 262 (190) pʰ t ̪h tʰ cʰ kʰ p t ̪ t c k (ʔ) b d̪ d ɟ g pʼ t ̪̓ tʼ cʼ kʼ ɓ ɗ s ʃ h z m sʼ ɲ ŋ n l w r j Dana retains all of the consonants seen in PKMN. The Dana retentions and innovations are outlined in Table 100 and discussed below. Table 100 Dana phonological innovations SET PKMN PCTRL PDAOP Dana Observations 1 *pʰ > *pʰ > *pʰ > pʰ Retention 2 *p > *p > *p > p Retention 3 *b > *b > *b > b Retention 4 *ɓ > *ɓ > *ɓ > ɓ Retention: only initial 5 *pʼ > *pʼ > *pʼ > pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h > t ̪h Retention 7 *t ̪ > *t ̪ > *t ̪ > t ̪ Retention 8 *d̪ > *d̪ > *d̪ > d̪ Retention 8a *d̪ > *d̪ > *t ̪ > t ̪ Retention 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > t ̪̓ Retention 10 (*tʰ) > (*tʰ) > *tʰ > tʰ Retention 11 *t > *t > *t > t Retention 12 *d > *d > *d > d Retention 12a *d > *d > *t > t Retention 13 *ɗ > *ɗ > *ɗ > ɗ Retention 14 *tʼ > *tʼ > *tʼ > tʼ Retention 15 *sʼ > *sʼ > *sʼ > sʼ Retention 16 *c > *c > *c > c Retention 17 *ɟ > *ɟ > *ɟ > ɟ Retention 17a *ɟ > *ɟ > *c > c Retention 263 Table 100 Dana phonological innovations SET PKMN PCTRL PDAOP Dana Observations 18 *cʼ > *cʼ > *cʼ > cʼ Retention 19 *kʰ > *kʰ > *kʰ > kʰ Retention 20 *k > *k > *k > k Retention 21 *g > *g > *g > g Retention 21a *g > *g > *k > k Retention 22 *kʼ > *kʼ > *kʼ > kʼ Retention 23 *s > *s > *s > s Retention 24 *ʃ > *ʃ > *ʃ > ʃ Retention 25 *h > *h > *h > h Retention 26 *r > *r > *r > r Retention 27 *l > *l > *l > l Retention 28 *m > *m > *m > m Retention 29 *n > *n > *n > n Retention 30 (*ɲ) > (*ɲ) > *n > n Merger 31 (*ŋ) > (*ŋ) > *ŋ > ŋ Retention 32 *w > *w > *w > *w Retention 33 *j > *j > *j > *j Retention 34 (*T̪) > (*T̪) > (*T̪) > s Merger 35 (*D̪) > (*D̪) > (*D̪) > d̪ Merger The bilabial implosive is retained only word-initially and the velar nasal is retained only medially/finally. The only significant difference between the PDAOP consonant inventory and the Dana inventory is the outcome of the *T̪ and *D̪ proto-phonemes. Recall that evidence for this set is weak overall. Dana merges PDAOP *T̪ > s and PDAOP *D̪ > d̪. 4.9.2 Dana vowels and tone Aside from retaining the PKMN consonant inventory, Dana also retains the PKMN seven vowel inventory with ATR contrast in the high vowels /i, ɪ, ɛ, a, ɔ, ʊ, u/. Dana synchronically exhibits stem-controlled ATR harmony, in which only high vowels in affixes assimilate to the ATR value of the stem to which they attach (§2.1.8.3). This ATR harmony system was retained from PDAOP and most likely occurred in PKMN. 264 Dana synchronically exhibits three contrastive level tones as well as rising and falling contour tones (§2.1.8.4). This tone system was inherited from as far back as least as PCTRL or PKMN. One independent innovation in Dana was to extend the distribution of all three level tones to all consonant onsets. In Dana, there are no consonant-tone restrictions synchronically. 4.10 Proto-Opo (POP) The following sections outline the development of Proto-Opo (POP) from Proto- Dana-Opo. POP underwent significant innovations in the consonant system as well as in the tone system, with a most notable *H > XH tone innovation. POP lost the interdental series which is retained in Dana. A morphological innovation in POP is the creation of new 3rd person independent pronouns from demonstrative bases plus gender prefixes (§2.2.2.5). 4.10.1 POP consonants The Proto-Opo consonant inventory is presented in Table 101. Table 101 Proto-Opo (POP) consonant inventory *pʰ *tʰ *kʰ *p *t *tʃ *k *b *d *dʒ *g *pʼ *tʼ *tʃʼ *kʼ *ɓ *ɗ *s *ʃ *h (*D̪) *sʼ *m *n *ɲ *ŋ *l *r *w *j 265 POP exhibits mostly retentions of the PDAOP consonant system with some exceptions. The POP innovations in the consonant system are schematized in Table 102. The POP retentions and innovations are discussed below. Table 102 Proto-Opo (POp) phonological innovations SET PKMN PCTRL PDAOP POP Observations 1 *pʰ > *pʰ > *pʰ > *pʰ Retention 2 *p > *p > *p > *p Retention 3 *b > *b > *b > *b Retention 4 *ɓ > *ɓ > *ɓ > *ɓ Retention: only initial 5 *pʼ > *pʼ > *pʼ > *pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h > *tʰ Retention 7 *t ̪ > *t ̪ > *t ̪ > *t Retention 8 *d̪ > *d̪ > *d̪ > *d Merger 8a *d̪ > *d̪ > *t ̪ > *t Merger 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > *tʼ Merger 10 (*tʰ) > (*tʰ) > *tʰ > *tʰ Retention 11 *t > *t > *t > *t Retention 12 *d > *d > *d > *d Retention 12a *d > *d > *t > *t Retention 13 *ɗ > *ɗ > *ɗ > *ɗ Retention 14 *tʼ > *tʼ > *tʼ > *tʼ Retention 15 *sʼ > *sʼ > *sʼ > *tʃʼ Merger 16 *c > *c > *c > *tʃ Retention (shift in articulation) 17 *ɟ > *ɟ > *ɟ > *dʒ Retention (shift in articulation) 17a *ɟ > *ɟ > *c > *tʃ Retention (shift in articulation) 18 *cʼ > *cʼ > *cʼ > *tʃʼ Retention (shift in articulation) 19 *kʰ > *kʰ > *kʰ > *kʰ Retention 20 *k > *k > *k > *k Retention 21 *g > *g > *g > *g Retention: elsewhere (see 21a) 21a *g > *g > *k > *k Split: Tone set C 22 *kʼ > *kʼ > *kʼ > *kʼ Retention 23 *s > *s > *s > *s Retention 24 *ʃ > *ʃ > *ʃ > *s Merger 25 *h > *h > *h > *h Retention 26 *r > *r > *r > *r Retention 27 *l > *l > *l > *l Retention 28 *m > *m > *m > *m Retention 29 *n > *n > *n > *n Retention 30 (*ɲ) > (*ɲ) > *n > *n Retention 31 (*ŋ) > (*ŋ) > *ŋ > *ŋ Retention 32 *w > *w > *w > *w Retention 33 *j > *j > *j > *dʒ Merger: initial 266 Table 102 Proto-Opo (POp) phonological innovations SET PKMN PCTRL PDAOP POP Observations 33 *j > *j > *j > *j Retention: medial/final 34 (*T̪) > (*T̪) > (*T̪) > *tʃ Merger 35 (*D̪) > (*D̪) > (*D̪) > (*D̪) Retention The most significant innovations in POP consonants is the merger of all interdental stops with alveolar stops. This was a complete merger, one which distinguishes the modern day Opo varieties from Dana. The palatal series of stops, at least in some varieties or for some speakers, came to be articulated more in the alveopalatal place of articulation, which is seen in some modern day Opo varieties. Whether this shift in articulation occurred in POP or whether it was a more recent change has yet to be determined. Another significant POP innovation was the merger of the alveolar ejective with the alveopalatal ejective PDAOP *sʼ > POP *tʃʼ. The merger of the alveopalatal fricative with the alveolar fricative PDAOP *ʃ > POP *s, which definitively occurred in the Bilugu, Modin and Pame varieties, may not have occurred in the Kigile variety. Kigile Opo appears to exhibit a contrast between /s/ and /ʃ/ though it may be the case that [s] and [ʃ] are in free variation though I find the latter unlikely given the number of correspondences of Kigile /s/ with Dana /ʃ/.170 Thus, there are two possible outcomes for POP *ʃ: either PDAOP *ʃ > POP *s and Kigile later innovated /ʃ/ or *ʃ is retained in PDAOP and the *ʃ > s merger occurs deeper into Opo’s history. This question can be answered with further research into the Opo varieties examined here as well as the 170 My Opo consultants were not native Kigile speakers. The dialects of Opo that they spoke underwent a complete merger of *ʃ > *s. At times, the speakers hesitated between [s] and [ʃ] for particular lexemes. See consonant *ʃ correspondence sets 24 in the appendices. 267 varieties for which data has not been collected. Notwithstanding, I tentatively analyze a retention of *ʃ in POP. The palatal glide PDAOP *j merges with POP *dʒ in initial position and /j/ is retained in medial and final positions. 4.10.2 POP vowels and tone The seven-vowel system with ATR contrast in the high vowels is retained in POP. Given that Dana and the Opo varieties both exhibit stem-controlled ATR harmony involving the high vowels, this harmony system was also inherited from PDAOP, if not from an earlier stage in Koman’s history (§4.2.3). A significant innovation that distinguishes Opo from Dana and the rest of Koman lies in the contrastive tone system. All of the living Koman languages exhibit three contrastive level tones, with the exception of the Opo varieties, which exhibit four level tones (Smolders forthcoming). POP innovated a fourth level tone by splitting PDAOP *H tone to H and an extra-high (XH) tone. This conditioned split of PDAOP *H > POP *XH that occurred robustly on [+high, +ATR] /i, u/ vowels. Crucially, this split occurred in cognates whose reflexes reconstruct to PKMN tone set A. This reconstructed tone category is defined by synchronic reflexes exhibiting H tone which was only preceded by voiceless onsets. To be explicit, there are no cognates for which tone set A can be reconstructed on a reconstructed [+high, +ATR] vowel (*i, *u) in which an Opo reflex does not exhibit XH tone. There are XH reflexes in Opo for which a tone cannot be reconstructed but they will always involve [i, u] and the Dana reflex will always be H tone. This suggests that in these correspondences, at least a PDAOP tone can be reconstructed to *H. 268 Synchronic Opo varieties also exhibit XH tone on /a/. There is scant evidence for this XH tone being the reflex of a split that occurred at the POP stage given there are only two clear cognates, presented in (191). If this POp *XH before /a/ were a true split from PDAOP *H, there would be no conditioning for the split. As such, I argue that the POP split of a *H that at least reconstructs to PDAOP if not earlier only occurred on *i and *u. Thus, it appears that this split in a PDAOP *H tone in POP was conditioned by both the onset and also by the height and ATR quality of the vowel. (191) Komo Chali Uduk Dana Opo Meaning a. kárúm – kʰáɗúm kʰa̋rűm ‘roof’ b. – àhǎɗkī háɗìkʼ ha̋rīkʼ ‘hiccough’ 4.11 Opo This study employs data from four Opo varieties: Bilugu, Pame, Modin and more peripherally, Kigile Opo. The lexical and grammatical data were collected from native Pame Opo speakers who were bilingual in the Bilugu variety. Further, they had knowledge of the Modin variety, as it is very close to the Bilugu variety and some knowledge of the Kigile variety. Given the proximity in sound systems in the Opo varieties, I discuss the retentions as a whole and discuss the independent innovations individually. 4.11.1 Opo consonants The Opo consonant inventory is presented in Table 15 and reproduced in (192). Note that the voiceless alveopalatal fricative /ʃ/ only occurs in the Kigile variety and the 269 voiced alveolar fricative only occurs in the Pame variety. All other consonants occur in each variety. (192) pʰ tʰ kʰ p t tʃ k b d dʒ g pʼ tʼ tʃʼ kʼ ɓ ɗ s (ʃ) h (z) m n ɲ ŋ l r w j The Opo retentions and innovations are schematized in Table 103. The significant innovations are discussed below. Table 103 Opo phonological innovations SET PKMN PCTRL PDAOP POP Opo Observations 1 *pʰ > *pʰ > *pʰ > *pʰ > pʰ Retention 2 *p > *p > *p > *p > p Retention 3 *b > *b > *b > *b > b Retention 4 *ɓ > *ɓ > *ɓ > *ɓ > ɓ Retention: only initial 5 *pʼ > *pʼ > *pʼ > *pʼ > pʼ Retention 6 *t ̪h > *t ̪h > *t ̪h > *tʰ > tʰ Retention 7 *t ̪ > *t ̪ > *t ̪ > *t > t Retention 8 *d̪ > *d̪ > *d̪ > *d > d Retention 8a *d̪ > *d̪ > *t ̪ > *t > t Retention 9 *t ̪̓ > *t ̪̓ > *t ̪̓ > *tʼ > tʼ Retention 10 (*tʰ) > (*tʰ) > *tʰ > *tʰ > tʰ Retention 11 *t > *t > *t > *t > t Retention 12 *d > *d > *d > *d > d Retention 12a *d > *d > *t > *t > t Retention 13 *ɗ > *ɗ > *ɗ > *ɗ > ɗ Retention 14 *tʼ > *tʼ > *tʼ > *tʼ > tʼ Retention 15 *sʼ > *sʼ > *sʼ > *tʃʼ > tʃʼ Retention 16 *c > *c > *c > *tʃ > tʃ Retention 270 Table 103 Opo phonological innovations SET PKMN PCTRL PDAOP POP Opo Observations 17 *ɟ > *ɟ > *ɟ > *dʒ > dʒ Retention 17a *ɟ > *ɟ > *c > *tʃ > tʃ Retention 18 *cʼ > *cʼ > *cʼ > *tʃʼ > tʃʼ Retention 19 *kʰ > *kʰ > *kʰ > *kʰ > kʰ Retention 20 *k > *k > *k > *k > k Retention 21 *g > *g > *g > *g > g Retention 21a *g > *g > *k > *k > k Retention 22 *kʼ > *kʼ > *kʼ > *kʼ > kʼ Retention 23 *s > *s > *s > *s > s Retention 24 *ʃ > *ʃ > *ʃ > *ʃ > ʃ Retention: Kigile 24 *ʃ > *ʃ > *ʃ > *ʃ > s Merger: Bilugu, Pame and Modin 25 *h > *h > *h > *h > h Retention 26 *r > *r > *r > *r > r Retention 27 *l > *l > *l > *l > l Retention 28 *m > *m > *m > *m > m Retention 29 *n > *n > *n > *n > n Retention 30 (*ɲ) > (*ɲ) > *n > *n > n Retention 31 (*ŋ) > (*ŋ) > *ŋ > *ŋ > ŋ Retention 32 *w > *w > *w > *w > w Retention 33 *j > *j > *j > *dʒ > dʒ Retention: initial 33 *j > *j > *j > *j > j Retention: medial/final 34 (*T̪) > (*T̪) > (*T̪) > *tʃ > tʃ Retention 35 (*D̪) > (*D̪) > (*D̪) > (*D̪) > dʒ Merger: Bilugu and Modin 35 (*D̪) > (*D̪) > (*D̪) > (*D̪) > z Shift: Pame 35 (*D̪) > (*D̪) > (*D̪) > (*D̪) > ʃ/s Merger: Kigile The modern Opo varieties exhibit mostly retentions from significant innovations in POP. Some notable differences in the varieties are the merger of *ʃ > s in the Bilugu, Pama and Modin Opo varieties. Further research will be needed to determine if these three varieties reconstruct to a node. Kigile Opo appears to have retained POP *ʃ: this is discussed in §4.10.1. Lastly, the marginal phoneme *D̪ has interesting reflexes in the Opo varieties. Bilugu and Modin Opo merge *D̪ > dʒ, Pame Opo shifts *D̪ > z and Kigile Opo merges *D̪ > ʃ/s. Whether it is significant or not that *T̪ and *D̪ were fricatives or affricates of some kind (if they were indeed PKMN phonemes) and whether they correspond to the only 271 languages that display alveopalatal affricates as opposed to stops, remains to be determined, though it is interesting that *T̪ and *D̪ do not correspond to palatal stops in either Dana or Chali Uduk. 4.11.2 Opo vowels and tone The Opo varieties retained the seven-vowel system /i, ɪ, ɛ, a, ɔ, ʊ, u/ with contrastive ATR in the high vowels from POP. It appears that all of the Opo varieties studied here exhibit stem-controlled ATR harmony in the high vowels. In this system, high affix vowels harmonize to the ATR feature of the high vowel in the stem to which they are attached. The modern Opo varieties studied here retain the four level tones which resulted from a POP innovation (§4.10.2). I now turn to a reconstruction of some PKMN pronominal and deictic morphology. 272 CHAPTER V RECONSTRUCTION OF PROTO-KOMAN PRONOMINAL AND DEICTIC MORPHOLOGY 5 Reconstruction of Proto-Koman pronominal and deictic morphology While the main aim of this dissertation is to reconstruct PKMN phonology, some attention to morphosyntax is presented to provide a more holistic picture of innovations at different stages of the Koman family. The only attempt at reconstructing any PKMN morphosyntax is Bender (1994), though his analysis included Gule, Gumuz and Shabo.171 Further, it was framed within the assumption that Koman (plus the other languages mentioned) formed part of Nilo-Saharan. Thus, at times, Bender’s analyses are obscured by efforts to link certain proto-forms outside of nuclear Koman to larger proposed genetic units. Proto-Koman most likely displayed SV/AVP word-order, with a possible tendency towards verb-second (V2) word order. PKMN was most likely head-marking, which is seen in the nominal domain through reconstructed number/gender morphology, as well as in the verbal domain where deictic directional suffixes can be reconstructed. Many languages also index core S/A/P arguments on the verb. Much of this morphology likely came from the cliticization of independent pronouns which phonologically eroded as they developed into obligatory argument indexing verb morphology. This chapter is not an exhaustive reconstruction of Koman morphology. I focus on reconstructing morphological forms (in some cases segments) to certain grammatical 171 To be clear, Bender (1994) was not entirely convinced of Shabo’s genetic affiliation. 273 categories. For instance, while many languages exhibit argument indexing verb morphology which may have originated in independent pronouns, I do not reconstruct the scenario for how this came to be. Rather, I discuss the cognacy of a particular bound morphological reflex to its source and leave the historical scenario of free pronouns developing into bound indexation for future work. I begin with a reconstruction of PKMN pronominal elements (§5.1). Then I turn to nominal gender/number morphology (§5.2). I briefly discuss a reconstruction of deictic demonstrative elements in light of Bender’s (1994) findings (§5.3). Lastly, I discuss a reconstruction of a PKMN deictic directional system on the verb (§5.4). 5.1 Reconstruction of Koman pronominal elements A system of independent pronominals can be reconstructed to PKMN with few deviations from the system proposed by Bender (1994). He reconstructs a three-way gender distinction in 3SG as well as clusivity in 1PL. Bender (1994:37) proposes that both gender and clusivity are not original to Proto-Nilo-Saharan and both were Proto-Komuz (Koman + Gumuz) innovations. I reconstruct a similar system though I present an alternative analysis for the origin of clusivity seen in the modern Koman languages. The Koman languages display some cognate independent pronouns that can be reconstructed to PKMN. The reconstructable independent pronouns appear to be monomorphemic, at least synchronically. Other pronominal forms can also be reconstructed to an earlier stage in Koman’s history and in many cases, appear to have been polymorphemic. Further, many languages have subsequently cliticized or affixed erstwhile independent pronouns into argument indexing morphology on verbs and possessive pronominal clitics. 274 Table 19 contains the synchronic Koman independent pronouns that can probably be reconstructed to PKMN. The starred forms indicate a proto-form reconstructed from the dialectal variation within a language group (e.g. *háɗ(i) in the “Uduk” column of Table 1 is reconstructed for Proto-Uduk/Pre-Uduk). The question mark in Gwama 2SG indicates the uncertainty of this form being cognate with the other 2SG forms. The plural first person forms have a more complex history which is discussed in detail below. Table 104 PKMN reconstructable independent personal pronouns PKMN Gwama Komo Uduk Dana Opo 1SG *aGa gà ākā áhā āgā āgā 2SG *aj ~ ?*aɪk ?(ɪk̄) àj ɛ́ āj āj 3SG.M *haɗ(i) *hāl hàr *háɗi hār – 3SG.F *haɓ hāpʼ hàpʼ – hāpʼ – 3NH *hɪ̀n ~ hàn – hɪ̀n ~ hàn – hān – 1PL.IN (see §5.1.2) mɪn̄ɪ̀ ānà ánā mīnā mìnà 1PL.EX (see §5.1.2) mà āmʊ̀n ámān mānā mànà 2PL *ʊm(a) ʊ̄m ʊ̀m úm ʊ̄mā ʊ̄mā 3PL *hʊ̄n(i) hʊ̄n hʊ̀n *húnī hʊ̄n – – indicates the lack of a cognate form Table 105 contains additional pronominal forms which are reconstructable. These forms may have been pronominal elements in a Pre-PKMN stage although *ma ~ *am can only be reconstructed to Proto-Central Koman given the lack of a cognate in Gwama. 275 Table 105 Additional reconstructable pronominal elements 1SG *na, *ma ~ *am 2SG *(m)ini The following subsections discuss the reconstruction of the independent pronouns in Table 19 as well as the morphemes presented in Table 105. 5.1.1 PKMN 1SG reconstructed forms I discuss three reconstructable 1SG morphemes in the following subsections. 5.1.1.1 PKMN 1sg *aGa An independent 1SG pronoun *aGa can be reconstructed to PKMN. Table 106 contains all of the reflexes of *aGa found in Koman languages. “Morpheme type” in the last column indicates the type of morpheme the reflexes instantiate modernly. Table 106 Reflexes of PKMN 1SG *aGa PKMN Gwama Komo Uduk Dana Opo Morpheme type 1SG *aGa gà ākā áhā āgā āgā 1SG independent pronouns -g – -kāʔ =āgā =āgā 1SG bound S/A/P VERB? affixes -āk The medial consonant appears to have been historically voiced given voiced reflexes in two distinct branches: Gwama and Dana-Opo. It is unclear whether *aGa become /aka/ at the Proto-Komo-Uduk (PKOUD) stage given the voiced bound reflex in Komo /-g/. Alternatively, Komo /-g/ 1SG always occurs with a following suffix which is vowel- initial. Thus, the possibility that Komo /-g/ derived from an independent form /aka/ and then subsequently revoiced intervocalically is also possible. If this were the case, a 276 subsequent PKOUD innovation would be the devoicing of *aGa > [aka]. This may have also been likely. Note that Bender (1994:37) reconstructs PKMN *aka 1SG. 5.1.1.2 PKMN 1SG *na A second 1SG pronominal form *na can tentatively be reconstructed to PKMN based on the reflexes in Table 107. Reflexes of *na are found in bound subject indexing verb morphology as well as in possessive pronouns. Bender (1994) does not reconstruct this morpheme though he acknowledges its presence. Table 107 Reflexes of PKMN 1SG *na PKMN Gwama Komo Uduk Dana Opo Morpheme type -nɪ ̄ -(n)á -(n)á – – 1SG bound S/A affixes =nā – – =(ɪ)́nā – 1SG possessive pronouns 1SG *na – – – mīnā mìnà 1PL.EX independent pronoun – ānà ánā mānā mànà 1PL.IN independent pronoun I cannot account for the vowel variation /ɪ ~ a/ seen across 1SG bound S/A affixes including in the Gwama forms. This vowel variation occurs sporadically elsewhere in Koman, often within a language (e.g. the Komo independent 3N pronoun hàn ~ hɪ̀n). The Gwama bound S/A form -nɪ ̄may or may not be cognate with the other forms in Table 107, but the possessive form is surely cognate. I discuss the cognacy and the structure of the 1PL forms in §5.1.2. 277 5.1.1.3 PCTRL 1SG *ma ~ *am One other 1SG pronominal element *ma ~ *am can be reconstructed to Proto-Central Koman (PCTRL) via the reflexes in Table 108. Proto-Komo-Uduk (PKOUD) employs *ma ~ *am to innovate a 1PL.EX pronoun. This innovation is discussed in §5.1.2. Table 108 Reflexes of PCTRL 1SG *má ~ *ám PCTRL Komo Uduk Dana Opo Morpheme type (b-)ám (p-)ɛ́m – (ɪ)́má 1SG possessive pronouns 1SG *má ~ *ám ām(ʊ̀n) ám(ān) – – 1PL.EX independent pronoun Whether *ma ~ *am reconstructs to PKMN and has subsequently been lost in Gwama has yet to be determined. Given the strict criteria employed here for reconstructing to nodes, I only reconstruct *ma ~ *am to PCTRL due to the lack of a reflex in Gwama. I assume metathesis of segments in either the Opo cluster or the Komo-Uduk branch but cannot account for the directionality at present. The vowel /ɛ/ in Uduk is problematic though Killian (2015) suggests it is phonetically [ɪ], which would give another example of an /a ~ ɪ/ vowel variation. Nonetheless, I assume the Uduk reflex is cognate. 5.1.2 1PL reconstructed forms A clusivity contrast in first person plural is found in every Koman language. Hence, it is tempting to reconstruct an inclusive/exclusive distinction to PKMN though when examining the 1PL independent pronominals, but the picture is not as clear as one would expect. The synchronic Koman 1PL independent pronouns are in Table 109. 278 Table 109 Koman 1PL independent pronouns Gwama Komo Uduk Dana Opo 1PL.IN mɪn̄ɪ̀ ānà ánā mīnā mìnà 1PL.EX mà āmʊ̀n ámān mānā mànà The historical scenario tentatively proposed here presupposes a stage earlier than PKMN, one in which an inclusive/exclusive distinction had yet to evolve. Pre-Proto- Koman exhibited a general 1PL pronoun **mana, which could have been bi-morphemic, composed of **ma PL + **na 1SG. A reflex of **ma is retained in Gwama, seen synchronically as a plural marker on nouns /mā=/ PL (cf §5.2.2). Once clusivity enters the Koman languages, possibly via contact with other groups and calquing, some branches employ a reflex of *mana 1PL for the 1PL.EX form, and innovate a 1PL.IN form, and others employ a reflex of *mana 1PL for the 1PL.IN form, and innovate a 1PL.EX form. To illustrate, Proto-Komo-Uduk (PKOUD) employs *mana > ana (with a loss of initial /m/) as the 1PL.IN independent pronoun. A later innovation is PKOUD *amʊn 1PL.EX composed from *am 1SG + *ʊn 3PL, with a composite meaning of ‘me and them (i.e. not you)’. The reflexes of *mana and *amʊn in the Komo-Uduk branch are in Table 110. The independent pronouns subsequently cliticized/affixed as argument indexing suffixes on verbs and over time become phonologically reduced. Another PKOUD innovation is possessive pronouns, which were most likely formed from a formative *ba plus an independent pronoun (e.g PKOUD *ba + *ana > Komo bānà 1PL.IN). This formative *ba could have been a demonstrative with a possible reflex /bā/ DEM.SG in Komo, which is found synchronically. 279 Table 110 Reflexes of 1PL in Komo-Uduk (KOUD) PKMN PKOUD Komo Uduk Morpheme type 1PL *mana *ana ānà ánā 1PL.IN independent pronoun -(n)à -(n)àn -(n)à 1PL.IN bound S/A/P verb affixes -ānà bānà bǎnà 1PL.IN possessive pronoun 1PL.EX *amʊn āmʊ̀n ámān 1PL.EX independent pronoun -(n)á(m) -(n)án -(n)á 1PL.EX bound S/A/P verb affixes -ām ~ -ān bābʊ̀n bâm 1PL.EX possessive pronoun While the PKOUD innovations seem plausible and provide a nice innovation to a single branch, the following scenario requires independent innovations in Proto- Gwama (PGW) and Proto-Dana-Opo (PDAOP). It is proposed that in these two branches, reflexes of *mana 1PL are employed for 1PL.EX: PGW *mana > /mà/ 1PL.EX and in PDAOP *mana > /mana/ 1PL.EX. The subsequent independent innovations of *minina 1PL.IN come about via *mini 2SG + *na 1SG with a meaning akin to ‘you and me’. In PDAOP, *minina > *mina via haplology and in PGW *minina > *mɪnɪ via apocope of the final syllable. Reflexes of *mana and *minina in Gwama, Dana and Opo are seen in Table 111. Note that Dana and Opo do not display robust argument indexing on the verb. Rather, once-independent pronouns are beginning to encliticize. Such pronominal enclitics are also employed as possessive pronouns in the Dana-Opo branch. Presumably, this was a PDAOP innovation. Gwama, by contrast, exhibits robust S/A argument indexing on the verb, which has been significantly phonologically reduced. 280 Table 111 Reflexes of 1PL in Gwama, Dana and Opo Gwama Dana Opo Morpheme type 1PL.IN *minina mɪn̄ɪ̀ mīnā mìnà 1PL.IN independent pronoun -nɪ̀ =mīnā =mìnà 1PL.IN bound S/A/P verb affixes =mɪn̄ɪ̀ =mīnā =mın̋á 1PL.IN possessive pronoun 1PL *mana mà mānā mànà 1PL.EX independent pronoun -mɪ ̄ =mānā =mànà 1PL.EX bound S/A/P verb affixes -mà =má =mánā =máná 1PL.EX possessive pronoun 5.1.3 PKMN 2SG reconstructed forms The two forms that reconstruct to 2SG pronominal elements are discussed below. 5.1.3.1 PCTRL *aj and questionable PKMN 2SG ?*ai(k) A 2SG independent pronoun *aj can be confidently reconstructed to PCTRL via the independent and bound reflexes in Table 112. Further, the sound change *aj > /ɛ, ɪ/ seen in the bound forms also seems likely as this is also seen in roots (e.g. PKOUD *pʰaj > Chali Uduk pʰɛ́ ‘fly (v.)’). Table 112 Reflexes of PCTRL 2SG *aj PCTRL Komo Uduk Dana Opo Morpheme type àj ɛ́ āj āj 2SG independent pronoun -í 2SG *aj ɪ=̄ -ɛ́ -ɛ ̌ =āj 2SG bound S/A/P verb affixes =āj -ɛ ̄ 281 An issue here is whether *aj can be reconstructed to PKMN. The problem is due to the dubiously cognate reflexes in Gwama seen in Table 113. Note that almost all Gwama reflexes contain a velar stop, which is absent in Central Koman. If one were to reconstruct a PKMN pronoun that also reflected the Gwama data, it may have been of the shape ?*aɪk. I employ a question mark here to indicate that this is a more questionable reconstruction. Table 113 Reflexes of a questionable PKMN 2SG ?*aɪk PCTRL Gwama Komo Uduk Dana Opo Morpheme type ɪk̄ àj ɛ́ āj āj 2SG independent pronoun -gɪ ́ -í ɪ=̄ 2SG ?*aɪk (-ɪ̀)* -ɛ́ -ɛ ̌ =āj 2SG bound S/A/P verb affixes =āj -ɛ ̄ =kɛ́ – – – – 2SG possessive pronoun * This form may be a reflex of *mini (see §5.1.3.2) The synchronic Gwama independent 2SG pronoun may have simplified the diphthong *aɪk > /ɪk̄/. Metathesis and voicing could account for the Gwama bound verb affix *aɪk > ɪk̄ > kɪ > gɪ, though this is speculation. Further, Gwama is the only branch that employs a possible *aɪk reflex in a possessive pronoun, whose form may have arisen via metathesis: *aɪk > kaɪ > kaj > kɛ. In sum, while we can clearly reconstruct PCTRL *aj 2SG, attempting to reconstruct a 2SG morpheme to PKMN by including the Gwama reflexes is problematic. This would entail a loss of the velar stop in PCTRL, which is plausible, though not robustly attested elsewhere. Nevertheless, Gwama seems markedly different here from Central Koman and PKMN ?*aɪk does not inspire confidence as a reconstructed form. Further, there is 282 another 2SG pronominal element which can be reconstructed to PKMN, discussed in the following subsection. Note that Bender (1994) does not reconstruct a 2SG independent pronoun for the reasons presented here. 5.1.3.2 PKMN 2sg *mini A 2SG morpheme *mini can be reconstructed to PKMN via the reflexes in Table 114. Reflexes of this morpheme occur primarily in possessive pronouns and in some composite 1PL.EX independent pronouns as described in §5.1.2. It is likely that the Gwama bound verb affix /-ɪ̀/ is a reflex of *mini given that the vowel and tone are identical to the final vowel of the independent pronoun /mɪn̄ɪ̀/. Table 114 Reflexes of PKMN 2SG *mini PKMN Gwama Komo Uduk Dana Opo Morpheme type =mɪn̄ɪ̀ bīní pīní =mīn =mɪńɪ ́ 2SG possessive pronoun 2SG *mini -ɪ̀ – – – – 2SG bound S/A/P verb affixes mɪn̄ɪ̀ – – mīnā mìnà 1PL.EX Independent pronoun 5.1.4 PKMN 2PL *ʊm(a) One 2PL pronoun *ʊm(a) can be confidently reconstructed to PKMN via the reflexes in Table 115. It is unclear whether the final vowel reconstructs to PKMN or whether it was a later innovation in Proto-Dana-Opo (PDAOP).172 Note that a PDAOP innovation is metathesis of PKMN*ʊma > PDAOP *mʊa, which is retained as /mwá/ in Opo and truncated to /mʊ̄/ in Dana. 172 Bender (1994) reconstructs PKMN *um. 283 Table 115 Reflexes of PKMN 2PL *ʊm(a) PKMN Gwama Komo Uduk Dana Opo Morpheme type ʊ̄m ʊ̀m úm ʊ̄mā ʊ̄mā 2PL independent pronoun -mɪ ́ -m 2PL *ʊm(a) -ǔm =ʊ̄mā =ʊ̄mā 2PL bound S/A/P verb affixes -ʊ̀m -ʊ̄m =kʊ́m bʊ́m bǔm =mʊ̄ =mwá 2PL possessive pronoun I cannot account for the initial velar stop in Gwama /=kʊ́m/ 2PL.POSS, though it appears to pattern with the velar stop seen in the Gwama 2SG forms discussed in §5.1.3. It seems likely that these velar stops are related though a definitive explanation requires further investigation. 5.1.5 PKMN 3SG reconstructed forms Synchronic Koman languages display either two or three genders in their synchronic third person pronominal system: masculine, feminine and possibly neuter/non-human (§2.2.2). Some of the languages retain this three-way distinction, such as Komo and Dana, while others have lost it altogether and only exhibit one 3SG pronoun, such as the Uduk varieties. A note must be made on the 3rd person pronominal forms in general. All of the third person independent pronouns can be reconstructed with an initial voiceless glottal fricative *h- or a voiceless glottal fricative and a vowel *hV- (e.g. Komo hàr 3SG.M, hàpʼ 3SG.F, hɪ̀n/hàn 3N). While this suggests either *h or *hV as some sort of formative base to which gendered morphology may have attached, I cannot find a source for this proposed formative base. Thus, I reconstruct the 3rd person pronouns together with the voiceless glottal fricative for simplicity even though I recognize they may have had a 284 more complex history. The following sections examine the 3SG pronominal forms and discuss possible sources for their evolution. 5.1.5.1 PKMN 3N *han ~ *hɪn A 3SG pronoun *han ~ *hɪn can be reconstructed to PKMN via the reflexes in Table 116. I cannot account for the variation in vowel quality but again we see /a ~ ɪ/ alternations within and across languages in the 3N reflexes. Note that all languages retain reflexes of *han ~ *hɪn with the exception of the Uduk cluster, which has presumably lost any vestiges of this proto-pronoun. Gwama only retains a P argument indexing suffix /-à/ in limited constructions (cf. §2.2.3.1.1). Table 116 Reflexes of PKMN 3N *han ~ hɪn PKMN Gwama Komo Uduk Dana Opo Morpheme type – hà† 3N independent hɪ̀n/hàn – hān nà† pronoun 3N *ʊm(a) -à -n ān= 3N bound S/A/P verb – =ān -ɪ ̄~ ɪn̄ =ān affixes – bɪń – =ɪn̄ =hɪń 3N possessive pronoun †This form cannot occur independently and requires demonstrative enclitcs to derive the independent pronoun (cf. §2.2.2.5.1). One notable Proto-Opo innovation is a distinct set of third person independent pronouns (see §2.2.2.5 for discussion). Opo employs bound formatives, or “bases” (such as 3N /hà/ and /nà/ in Table 116) in combination with demonstrative elements to form innovated 3N pronouns. Further, we see grammaticalized bound reflexes of *han in the Opo verb affixes, which suggests that POP cliticized /han/ on the verb while innovating a new independent set of 3N pronouns with the 3N formatives /hà/ and /nà/. 285 Bender (1994:37) reconstructs *hen 3N to PKMN and assumes it was only retained in Komo.173 Note that the data from Dana, which Bender presumably did not have, reinforces the validity of an independent PKMN pronoun *hɪn ~ *han 3N, given that independent pronoun reflexes are seen in two branches: PKOUD and PDAOP. 5.1.5.2 PKmn 3sg.f *haɓ A 3SG.F pronoun *haɓ can be reconstructed to PKMN through the reflexes in Table 117. Note that reflexes are retained in all branches but Uduk, which lost the gender distinction in pronouns and only retains a reflex of the PKMN 3SG.M pronoun *haɗ(i), which is employed for all 3SG genders (cf. §2.2.2.3 for discussion). Table 117 Reflexes of PKMN 3SG.F *haɓ PKMN Gwama Komo Uduk Dana Opo Morpheme type hāpʼ hàpʼ – hāpʼ ɓā† 3SG.F independent pronoun -á /-bɪ ́* -pʼ āb= 3SG.F *haɓ – =āpʼ 3SG.F bound S/A/P verb affixes -àpʼ -āpʼ =āb =dápʼ bɪṕʼ – =ɪp̄ʼ =ɪb́ 3SG.F possessive pronoun †This form cannot occur independently (cf. §2.2.2.5.1). Proto-Opo innovated a 3SG.F pronoun employing a formative base *ɓa combined with demonstrative elements (see §2.2.2.5.1 for description). It seems very likely that this Opo formative base /ɓa/ is the same morpheme employed to derive feminine human nominals in modern-day Opo /ɓā=/ F.SG (cf. §5.2.1). This reinforces the validity 173 Bender’s (1994) “/e/” vowel most likely represents the same vowel that others have written as /ɪ/. Note that Bender only reconstructed five vowels /i, e, a, o, u/ to PKMN, though I reconstruct a seven-vowel /i, ɪ, ɛ, a, ɔ, ʊ, u/ system. 286 of at least some of the gender morphology in the PKMN pronominal system having a common origin with the gender morphology in the nominal system. However, this does not hold for masculine gender, as discussed below in §5.1.5.3. 5.1.5.3 PKmn 3sg.m *haɗ(i) A 3SG.M pronoun *haɗ(i) can be reconstructed to PKMN through the reflexes in Table 117. Note the regular sound correspondence of medial/final Chali Uduk /ɗ/ to /d/ in Yabus Uduk and /r/ in the remaining languages (see §3.2.3.4 for reconstruction of PKMN *ɗ). Bender (1994:37) reconstructs *har but does not give an explanation for reconstructing final *r rather than *ɗ. Table 118 Reflexes of PKMN 3SG.M *haɗ(i) PKMN Gwama Komo Uduk Dana Opo Morpheme type áɗī† 3SG.M independent pronoun hār hàr hār wàr§ hádī‡ -nɪ ́ -r -(V)ɗ ār= 3SG.M bound S/A/P verb 3SG.M *haɗ(i) =ār -ɛ̀ -ār -ǎɗi =ār affixes píɗī† =dɛ́ bɪŕ =ɪr̄ =ɪŕ 3SG.M possessive pronoun pádī‡ † This form is from the Chali Uduk variety. ‡ This form is from the Yabus Uduk variety. § This form is from the Pame Opo variety and cannot occur independently. Differently from PKMN *haɓ 3SG.F, which is most likely related to the PKMN nominal gender morpheme *ɓa F, PKMN *haɗ(i) 3SG.M does not appear to have any related morphology in the PKMN nominal gender system. There are no vestiges of *ɗ or any other coronal consonants marking masculine gender outside of the pronominal system. 287 This suggests that a possible Pre-Proto-Koman gender system was binary: the feminine was coded by reflexes of *ɓ(a), which remain feminine modernly, and the masculine was coded by reflexes of the PKMN neuter morphology, *han 3SG.N, which is most likely related to the PKMN neuter nominal gender morpheme *à N. Evidence for this is the variation in loss of neuter marking throughout Koman: a vestige in the Gwama verb system and a peripheral neuter gender in Komo. Chali and Bonya Uduk retain presumably cognate à= CL2 morphemes. Only Dana and Opo retain a strong three-way gender system. PKMN could have innovated a masculine gender pronoun of the shape *hàɗ(i) though the source for this is unknown. 5.1.6 PKmn 3PL *hʊn(i) A 3PL independent pronoun *hʊn(i) can be reconstructed to PKMN via the cognates in Table 119. Reflexes of PKMN *hʊn(i) are attested in all branches as independent pronouns, bound verbal affixes and possessive pronouns. One exception is Proto-Opo (POP), which innovated a new series of third person independent pronouns, including 3PL. Nevertheless, Opo retains reflexes of *hʊn(i) in the possessive pronouns and bound verbal indexing forms.174 Bender (1994) reconstructs *un, though he does not discuss its shape (i.e. the absence of initial /h/). 174 The source for the POP innovated 3PL is discussed in §5.2.2. 288 Table 119 Reflexes of PKMN 3PL *hʊn PKMN Gwama Komo Uduk Dana Opo Morpheme type hʊ̄n hʊ̀n húnī hār – 3PL independent pronoun -n ʊ̄n= 3PL bound S/A/P verb 3SG.M *hʊn -ʊ̀n – =ʊ̄n -ʊ̄n =ʊ̄n affixes =bʊ́n bʊ́n bǔnī =mʊ̄n =mʊ́n 3PL possessive pronoun 5.2 Reconstruction of Koman nominal gender/number morphology Koman languages display varying degrees of nominal gender morphology, most of which can be reconstructed to PKMN.175 We saw in §5.1.5 that the PKMN pronominal system exhibited a three-way gender distinction in the third person. Some of the morphology employed in the pronominal system is cognate with some of the morphology in the nominal gender system, particularly in the Opo varieties, which have the most elaborated nominal gender marking systems in Koman. Koman languages that do not synchronically display productive nominal gender marking nevertheless retain vestiges of the historical gender morphology in kin terms. The following subsections reconstruct aspects of the PKMN nominal gender system. 5.2.1 PKMN singular nominal morphology There is enough evidence to reconstruct a three-way gender system for singular referents to PKMN through the cognates in Table 120. The clearest morpheme to reconstruct is a feminine *ɓa F, with reflexes retained in all languages. Note that Proto- Opo employed a reflex of this morpheme to innovate a 3SG.F pronoun (§5.1.5.2). 175 See §2.2.2 for synchronic descriptions of Koman nominal gender systems. 289 Table 120 PKMN reconstructed singular nominal morphology PKMN Gwama Komo Uduk Dana Opo M *ʊ ʊ̄= – – – ʊ̀= ɓā(bī)= F *ɓa pʼà† – ɓā† ɓā- bā† à† N *à à=‡ à= à- à† nঠhঠ† No longer productive, lexicalized ‡ Restricted distribution (cf. §2.2.2.1.3) ¦ This form cannot occur independently. A masculine gender marker *ʊ can be reconstructed to PKMN via the Gwama and Opo reflexes, but there are losses in the Komo-Uduk branch and in Dana. Interestingly, *ʊ does not appear to have any relation to masculine encoding in the pronominal system, which can be reconstructed to PKMN *haɗ(i). A neuter morpheme *à can be reconstructed to PKMN. All languages retain L tone in the reflexes of *à. Further, *à seems related to the Opo neuter morphemes /nà/ and /hà/, which also exhibit L tone. There are no /h : n/ sound correspondences to suggest a common origin for Opo /nà/ and /hà/ though these could be reflexes of PKMN *han 3N. Whether or not the PKMN *à neuter morpheme is related to PKMN *han has yet to be determined. There are two more reconstructable singular nominal gender morphemes, which do not bear any resemblance to the morphemes in Table 120. These morphemes, seen in Table 121, only have reflexes in Komo and Dana (see §2.2.2.2.3 for Komo and §2.2.2.4.3 for Dana). Thus, if these gender morphemes were to be reconstructed, it would have to 290 be to Proto-Central Koman given the lack of a reflex in Gwama. I tentatively reconstruct two morphemes to PCTRL: masculine *jE and feminine *jʊ. Table 121 PCTRL singular nominal morphology? PCTRL Komo Uduk Dana Opo M *jE jī= – jɛ̀= – F *jʊ – – jʊ̀= – 5.2.2 PKMN plural nominal gender morphology The synchronic Koman plural nominal proclitics are in Table 122. Only Gwama maintains a masculine/feminine gender distinction in the plural. Table 122 Koman plural nominal (gender) morphology Gwama Komo Uduk Dana Opo M mà= gʊ̀= ī= kɛ=̄ bɪ̀= F ɪ=̄ Two challenges result from the array of morphemes in Table 122: (i) whether to reconstruct gender in the plural system, and (ii) which morpheme(s) actually reconstruct. It appears that at least *ɪ can be reconstructed as a plural morpheme to PKMN given the reflexes in Gwama and Uduk. Whether the PKMN *i plural morpheme is related to the vestiges of an old plural suffix of the shape /-i/ seen in Komo and Dana in (193), is yet to be determined. 291 (193) Gloss Komo Dana ‘teeth’ ʃɛ̀ʔí ʃɛʔ̄ɛ ́ ‘goats’ mɛʔ̄í mɛʔ̄ɛ ́ We saw in §5.1.2 that the synchronic Gwama /mà/ M.PL morpheme may be a reflex of a Pre-Proto-Koman plural morpheme *ma employed to form a 1PL pronoun *mana. Thus, it may be possible that PKMN exhibited nominal gender marking in the plural, and this system is retained in Gwama. Turning to the remaining plural morphemes in Table 122, it is unclear whether Komo /gʊ̀=/ and Dana /kɛ=̄/ are related. I cannot account for the initial consonant correspondence and it does not follow the pattern seen in PKMN voiced onsets that became devoiced before *L tone in Proto-Dana-Opo (§3.1.2). As such, Komo /gʊ̀=/ and Dana /kɛ=̄/ could be independent innovations. Lastly, a PKMN plural morpheme *bɪ should be reconstructed based on the following cognates: Opo /bɪ̀=/ plural and Gwama /-bɪ/́ 3PL S/A argument indexing verb suffix. To summarize, the three possible PKMN plural morphemes are presented in Table 123. I have glossed gender as a possibility for *ma and *ɪ. The fact that three nominal plural morphemes can be reconstructed (two with a possible gender distinction) parallels the three-way gender distinction in the singular; perhaps PKMN *bɪ was a plural neuter morpheme. Table 123 PKMN plural nominal (gender) morphology PKMN Meaning *ma PL.(M?) *ɪ PL.(F?) *bɪ PL 292 5.3 Reconstruction of PKMN demonstrative elements No definitive demonstrative forms can be reconstructed for PKMN. Bender (1994:40) came to a similar conclusion but he reconstructs the demonstrative “archiforms” *n ‘this’ and *t ‘that’ for what I understand to mean proximal distance from an origo and medial/distal distance from an origo, respectively. My data somewhat confirms Bender’s analysis. The Koman proximal demonstrative elements are in Table 21. Note that only Gwama and Komo distinguish M and F gender and only Gwama further distinguishes number (i.e. the Gwama M and F forms are singular). We do see a pattern of /n/ in all of the proximal forms if we include the Gwama plural form. Table 124 Koman proximal demonstrative enclitics Gwama Komo Uduk Dana Opo M =ɛ̀ =nɪ ̌ -n-, -ns- =nɪ̀, =īnà =ínɔ̄ F =ɔ̀ =nɔ̀ɪ ̄ PL =nʊ̀n – – – – Demonstratives elements encoding a medial distance from an origo are in Table 125. In these forms, we see /t, d, ɗ, d̪/ consonants across the modern forms. Table 125 Koman medial demonstrative enclitics Gwama Komo Uduk Dana Opo M =tɛ̀ =dɪ ̌ -ɗ- =ínī F =tɔ̀ =dʊ̀ɪ ̄ -t- =d̪à -nt- =íntīn PL =tʊ̀n – – – – 293 In sum, while demonstrative morphemes cannot be easily reconstructed to PKMN, there appears to be remnants of consonants indicating distance from an origo: possibly *n for a proximal distance and *D (which includes alveolar and interdental obstruents) for a non-proximal distance. 5.4 Reconstruction of Koman Deictic Directional (DD) verb morphology Deictic directional (DD) suffixes form a core part of modern Koman verb morphological systems (see §2.2.3.2 for description). The Koman DD morphemes are strictly suffixing and occur immediately after verb root, before any additional morpohology.176 The fact that the Koman DDs occur deep within the verbal complex suggests they are old morphemes. Added to this is the fact that the forms are quite varied in in shape and exhibit nuances of semantic function, which also suggest a greater time depth. The Koman DD morphemes carry a heavy functional load across the languages, encoding direction of motion (including spatial orientation without motion), associated motion, and in some cases, aspect (Otero 2018a, accepted). It is safe to reconstruct some PKMN directional morphology which may have first coded a binary directed motion opposition with the speaker as the deictic center: motion towards a speaker (GOAL) and motion away from the speaker (SOURCE). I will refer to this opposition as ‘ventive’ and ‘itive’ when describing a possible historical evolution of these morphemes. Table 44 presents synchronic forms expressing ventive motion towards the speaker; I refer to these in Table 44 as ‘deictic directional 1’ in part because of meaning 176 The only exception is Gwama -gɪ́ DD2, which occurs after S/A suffixes on the verb. This morpheme may have been a later innovation. 294 extensions beyond clear ‘ventive motion’. Note that all of the languages display [+high, –ATR] vowels /ɪ, ʊ/ with H tone for the DD1 morphemes. Some languages have two forms, such as Chali Uduk, which employs /-í/ in finite verbs and /-ú/ in non-finite verbs (Killian 2015). Further, Modin Opo exhibits variation in deictic ‘come’ verbs that appear to contain both morphemes: dʒʊ̄ ‘come.SG’ and dʒāɪ ́‘come.PL’. Table 126 Koman Deictic Directional 1 (DD1) morphemes Gwama Komo Uduk Dana Opo Opo (Chali) (Bilugu, Pame) (Modin) -ʊ́ -ú -ʊ́ -ʊ́ DD1 -ɪ ́ -í -ɪ ́ -ɪ ́ From this evidence, we can either reconstruct two forms *ʊ́ and *ɪ,́ or a composite of both forms *ʊ́ɪ ́~ *íʊ́. At present I cannot find any diachronic sources for DD1. Further, many of the synchronic deictic motion verbs for ‘come’ appear to employ a historical root meaning either deictic ‘go’ or non–deictic ‘move’ which have lexicalized the synchronic DD1 morphemes. Thus, positing a ‘come’ verb (at least the synchronic ‘come’ verbs) as a possible source for PKMN ventive morphology is challenging at best. Given that all Koman languages also employ DD1 to express subsequent associated motion to the speaker (i.e. VERB then ‘come’) on non-motion lexical roots, this associated motion semantic sense may have also existed at a PKMN stage. Table 127 presents forms related to two additional deictic elements, that I call DD2 and DD∅. Synchronically, it is extremely important to note that the DD2 morphemes express motion towards the addressee. While this implies motion away from a source, the DD2 morphemes do not express this inherently. Only Komo exhibits additional non- 295 ventive directional suffixes, labeled DD∅. The DD∅ morphemes are inert and do not exhibit any semantic material: /-í/ occurs on finite verbs and /-á/ on non-finite verbs. Nevertheless, I suspect that the Komo /-í, á/ DD∅ morphemes are cognate with /-(j)á/ DD2 in and Dana and Opo and these morphemes could be reflexes of a PKMN (or PCTRL) itive morpheme. First, I turn to the synchronic DD2 morphemes. Table 127 Koman Deictic Directional DD∅ and DD2 morphemes Gwama Komo Uduk Dana Opo Opo (Chali) (Bilugu) (Pame) -úk, -kú, -kú DD2 -gɪ ́ -(j)á -(j)á -(j)á (kí)† -kí -í DD∅ – – – – – -á – indicates a bare or unmarked form of the verb †This form is lexicalized. While the Gwama and the Uduk DD2 forms contain velar onsets, I do not suspect that the Komo-Uduk forms are cognate with Gwama, though this remains to be thoroughly investigated.177 The Komo and Uduk DD2 forms are almost certainly cognate. In Chali Uduk, /-kí/ is employed on finite verbs and /-kú/ on non-finite verbs (Killian 2015). In Komo /-kú/ is employed on finite verbs and /-úk/ on non-finite verbs. Komo has a handful of verbs which appear to have lexicalized /-kí/, though this morpheme is no longer productive if it ever was at all. 177 Gwama /-gɪ/́ DD2 does not occur immediately on the verb root, but rather after S/A argument indexing suffixes, which suggests a possible Gwama innovation (see Hellenthal 2018 for discussion). 296 If the Gwama DD2 morpheme is not cognate with the Uduk and Komo forms, then one possible scenario is that an itive marker was a PCTRL innovation, sourced in the grammaticalization of a motion verb: *D̪a ‘go, move’ (see §3.2.10.1 for discussion). Reflexes of this possible PCTRL *D̪a itive are seen as DD2 in the Dana-Opo branch (with the semantic extension of motion towards the addressee).178 The PCTRL itive may have been lost in the Komo-Uduk branch though vestiges may be retained in the Komo DD∅ morphemes /-í, -á/, which are synchronically inert. After the loss, Proto-Komo-Uduk (PKOUD) possibly innovated a distinct itive form *-kV containing a velar onset (the vowel cannot be definitively reconstructed). PKOUD*-kV may have functioned as an itive and later became extended to express motion towards the addressee, or it may have been innovated initially as indicating motion towards the addressee. Whether this PKOUD innovation *-kV is cognate with the Gwama independent pronoun ɪk̄ 2SG remains a speculation, but these are the only forms with velar onsets and a meaning related to the second person. In conclusion, it seems very likely that PKMN exhibited a morphological opposition for encoding itive and ventive deictic direction on verbs: this is core Koman verb morphology. A PKMN ventive function can be clearly reconstructed possibly to a unitary morpheme (*ʊ́ɪ ́~ *íʊ́), or two distinct morphemes (*ɪ, *ʊ), though the former seems more likely to me. An itive form cannot be definitively reconstructed to PKMN given the lack of a trustworthy cognate in Gwama. It could be argued that following a loss of the PKMN itive, PCTRL innovated a new itive from a motion verb *D̪a, which is retained in the 178 See §3.2.10.1 for *D̪ correspondence set. 297 Dana-Opo branch and almost entirely lost in the Komo-Uduk branch (with a residual retention in Komo seen in the DD∅ forms). Gwama then later innovated -gɪ ́DD2. Another possibility is that a itive morpheme sourced in a motion verb *D̪a reconstructs to PKMN and was subsequently lost in Gwama. The question of how the synchronic DD2 forms, which are not cognate across the family, all code motion towards the addressee remains a mystery. This semantic extension may have occurred independently in Gwama, the KOUD branch and in the DAOP branch, or it could have resulted through contact across these groups. Nevertheless, motion towards the addressee (and even associated motion towards the addressee in the DAOP branch) is a unique feature of the synchronic Koman DD systems. 298 CHAPTER VI CONCLUSION 6 Conclusion This dissertation is a first pass at a reconstruction of the Koman language family. Much of the data in this reconstruction was collected first-hand from native speakers in Ethiopia and with refugee speakers in the U.S. In this reconstruction, I focused specifically on the living Koman languages whose genetic affiliation to Koman is undisputed (Gwama, Komo, Uduk, Opo). I also included data from Dana, a previously unrecognized Koman language, which forms a key part of the reconstruction as it retains many aspects of the Proto-Koman phonological system. Further, this study also included data from several Koman dialects, such as Yabus Uduk, which had very little, if any, prior description. In this study, I reconstructed parts of the phonology, lexicon and morphology of Proto-Koman. In the phonology, I reconstructed consonants, vowels and tone. I also discussed how historical tone in Pre-Koman impacted the development of syllable onset stop consonants. Another notable feature in the phonological reconstruction is the interaction between the Advanced Tongue Root [ATR] feature of high vowels and the realization of either alveolar or palatal fricatives in Gwama. I provide a wordlist of reconstructed etyma in Appendix D and indicate to which node each item reconstructs within Koman. Some morphological reconstructions include independent pronouns and deictic directional verb morphology. I have included all of the data employed in this study in the Appendices. 299 While Koman’s affiliation to the purported Nilo-Saharan super family is still under debate, my main aim was to provide a conservative reconstruction of Proto-Koman which will hopefully serve future Komans scholars as well as those interested in higher- level genetic classifications of East African languages. 300 APPENDIX A ABBREVIATIONS 7 APPENDIX A Abbreviations 1 first person 2 second person 3 third person A agent-like argument of a canonical transitive verb ACC accusative ADJZ adjectivizer ASS associative AUX auxiliary CL class COP copula DEF definite DEM demonstrative DEM.RT demonstrative root DD deictic directional DIST distal ERG ergative EX exclusive F feminine HUM human IN inclusive INT intentative (future) LOC locative M masculine MED medial N neuter/non-human NEG negative 301 P patient-like argument of a canonical transitive verb PL plural PLU pluractional PROX proximal RED reduplication REM remote Q polar question particle/marker S single argument of a canonical intransitive verb SG singular sp. species, type SV serial verb 302 APPENDIX B ETYMOLOGICAL WORDLIST 8 APPENDIX B Etymological wordlist This appendix contains all of the reconstructed lexica in this study. As different etyma can be reconstructed to different nodes, for each individual entry, I indicate the highest level (or node) to which it can be reconstructed. Each entry contains a reconstructed form, followed by the word class. Below each reconstructed word are the cognates in the daughter languages. The abbreviations for the language varieties are as follows: GwLo= Lowland Gwama, GwHi= Highland Gwama, UdYa= Yabus Uduk, Komo=Ethiopian Komo, UdCh= Chali Uduk, Dana=Dana, OpBi= Bilugu Opo, OpMo= Modin Opo, OpPa= Pame Opo, OpKi= Kigile Opo. I employ the grapheme <+> to indicate a historical or synchronic morpheme boundary. *pʰ(j)asʼ v. ‘laugh’: Reconstructs to: PKmn GwHi ɪs̄ʼ GwLo pāsʼ Komo pɛ̀sʼ UdYa pʰɛs̄ʼ UdCh pʰɛt̄ ̪̓ Dana pʰàsʼ OpBi pʰātʃʼ OpMo pʰātʃʼ OpPa pʰātʃʼ OpKi pʰātʃʼ *pʰā v. ‘bring_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi pʰā+ʊ́ 303 OpMo pʰā+jʊ́ OpPa pʰā+jʊ́ OpKi pʰā+jʊ́ *pʰàcʼ v. ‘soak_1’: Reconstructs to: PKmn GwHi pāsʼ GwLo pāsʼ Komo pàsʼ UdYa pʰāʃʼ UdCh pʰācʼ Dana - OpBi - OpMo - OpPa - OpKi - *pʰàd̪ v. ‘fly (v.)’: Reconstructs to: PKmn GwHi pāl GwLo pāj Komo pàj UdYa pʰāj UdCh pʰɛ ̄ Dana pʰàd̪ OpBi pʰāj OpMo pʰāj OpPa pʰāj OpKi pʰāj Most likely borrowed from or related to PNilotic *pär (Dimmendaal 1988:38) *pʰáɗ(a) v. ‘lay (v.)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo pár UdYa pʰád UdCh pʰáɗ Dana pʰʊ́ɗā OpBi pʰárá OpMo pʰárá OpPa pʰárá OpKi pʰárá 304 *pʰákʼ v. ‘plait or braid or weave_1’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi pʰákʼ OpMo pʰákʼ OpPa pʰákʼ OpKi pʰákʼ *pʰákʼá v. ‘chop_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa pʰáʔ UdCh - Dana pʰákʼá OpBi - OpMo - OpPa - OpKi - *pʰákʼá n. ‘shoe’: Reconstructs to: PKmn GwHi pák GwLo pák Komo páʔ UdYa pʰāʔ UdCh à+pʰāʔ Dana - OpBi pʰákʼá OpMo pʰákʼ OpPa pʰákʼ OpKi - *pʰal v. ‘come free and fall off’: Reconstructs to: PKmn GwHi pɪ ̌ GwLo pɪ ̌ Komo - UdYa - 305 UdCh - Dana - OpBi pʰál OpMo pʰál OpPa pʰál OpKi pʰál Loss of final lateral in Gwama: [al] → [ɪ] *pʰāl n. ‘young people_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo pāl UdYa - UdCh - Dana pʰāl OpBi pʰāl OpMo pʰāl OpPa pʰāl OpKi pʰāl *pʰárá n. ‘platform’: Reconstructs to: PKmn GwHi párá GwLo párá Komo párá UdYa - UdCh - Dana pʰárá OpBi pʰárá OpMo pʰárá OpPa pʰárá OpKi pʰárá Platform used to throw stones at birds when crops are ripe. *pʰɛɗ v. ‘peel, skin (with knife), peel, husk, peel’: Reconstructs to: PKoUd GwHi - GwLo - Komo pɛ̀t UdYa pʰítʰ UdCh - Dana - OpBi - 306 OpMo - OpPa - OpKi - *pʰɛɗ̄(ɛ) v. ‘untie, take out (quickly, e.g. out of fire)’: Reconstructs to: PCtrl GwHi - GwLo - Komo pɛ̀l UdYa pʰɛd̄ UdCh pʰɛɗ̄ Dana pʰɛ̀rɛ ́ OpBi - OpMo - OpPa - OpKi - *pʰɛk̄ʼɛʃ́ n. ‘Komo (ethnonym)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa pʰɛk̄ʼɛʃ́ UdCh pʰɛk̄ʼɛʃ́ Dana - OpBi - OpMo - OpPa - OpKi - *pʰí+tʼwā v. ‘kiss_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa pʰí+tʼwā UdCh pʰítʼwā Dana - OpBi - OpMo - OpPa - OpKi - 307 *pʰɪ ̄v. ‘stab_5’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi pʰɪ ̄ OpMo pʰɪ ̄ OpPa pʰɪ ̄ OpKi pʰɪ ̄ *(ɪ)pʰɪ ́v. ‘drink_SG’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɪṕ UdYa pʰí UdCh pʰí Dana ɪṕʰ OpBi pʰɪ ́ OpMo pʰɪ ́ OpPa ɪṕʰ OpKi ɪṕʰ *H tone in SG. *ɪ̀pʰ(ɪ) v. ‘drink_PL’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɪ̀p UdYa - UdCh - Dana ɪ̀pʰɪ ́ OpBi ɪp̄ʰɪ ̄ OpMo ɪp̄ʰɪ ̄ OpPa ɪp̄ʰ OpKi ɪp̄ʰ *L tone in PL. *pʰɔ́g n. ‘back_2’: Reconstructs to: PKoUd GwHi - GwLo - 308 Komo pɔ́g UdYa pʰɔ́ʔ UdCh à+pʰɔ́ʔ Dana - OpBi - OpMo - OpPa - OpKi - *pʰɔpʰ v. ‘soak_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana pʰɔ̀pʰ OpBi pʰɔ́pʰɔ́ OpMo pʰɔ́pʰ OpPa pʰɔ́pʰ OpKi pʰɔ́pʰ *pʰɔt ̪̓(ɔ) v. ‘light weight (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo pɔ̀tʼ UdYa pʰɔ̄d UdCh pʰɔ̄ɗ Dana pʰɔ́t ̪̓ OpBi pʰɔ́tʼɔ́ OpMo - OpPa pʰɔ́tʼ OpKi - *pʰú ~ pʰʊ́ v. ‘spoil, stink, smell bad_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo pʊ́ UdYa - UdCh - Dana - OpBi pʰűwú 309 OpMo pʰűjɪ ́ OpPa pʰűrű OpKi pʰűrű *pʰui(ki) v. ‘blow (with mouth)’: Reconstructs to: PKmn GwHi pɪk̄ɪ ̄ GwLo pɪ ̄ Komo pɪ̀ UdYa pʰûj UdCh pʰí Dana pʰùj OpBi pʰű OpMo pʰű OpPa pʰű OpKi pʰű *pʰuku n. ‘fig_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa pʰūʔ UdCh à+pʰūʔ Dana pʰúj OpBi - OpMo pʰűkū OpPa pűj OpKi pűj *pʰUl? v. ‘spicy (be)_1’: Reconstructs to: PKmn GwHi pʊ̀l GwLo pwì Komo - UdYa - UdCh - Dana pʰîl OpBi pʰıl̋ī OpMo pʰıl̋ī OpPa pʰıl̋ī OpKi pʰıl̋ī 310 *(ʃa)pʰuZa n. ‘burnt gound, soot’: Reconstructs to: PKmn GwHi ʃāpʊ́t GwLo ʃābʊ́t Komo kɪ+́pú UdYa bwà+pʰwí UdCh à+pʰíʔ Dana pùzàʔ OpBi pʰűjhá OpMo pʰűjhá OpPa pʰűjhá OpKi pʰűjhá *p⁽ʰ⁾ɪt́(̪á) v. ‘rise (oneself)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa pí UdCh pí Dana pɪt́á̪ OpBi pʰɪj́á OpMo pʰɪj́á OpPa - OpKi pʰɪʃ́á *p⁽ʰ⁾ɔ̀s v. ‘barren (be)’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana pɔ̀s OpBi pʰɔ̄s OpMo pʰɔ̄s OpPa pʰɔ̄s OpKi pʰɔ̄s *p⁽ʰ⁾ɔ̀t ̪v. ‘slippery, smooth (be)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - 311 UdCh - Dana pɔ̀t ̪h OpBi pʰɔ̄t OpMo pʰɔ̄t OpPa pʰɔ̄t OpKi pʰɔ̄t *p(ʰ)uɗ(i) v. ‘arrive_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa - UdCh pʰūɗ Dana pújī OpBi - OpMo - OpPa - OpKi - *p(ʰ)ʊ̄r v. ‘burn (hairs off of pig skin), roast next to fire’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa pūr UdCh pūr Dana - OpBi pʰʊ̄r OpMo pʰʊ̄r OpPa pʰʊ̄r OpKi - *pa n. ‘home, place_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa pā UdCh pā Dana - OpBi pà OpMo pà OpPa pà 312 OpKi - *páD v. ‘hunt_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa pátʰ UdCh pár Dana - OpBi - OpMo - OpPa - OpKi - *pàD v. ‘touch_1, crawl’: Reconstructs to: PKmn GwHi pāt GwLo pāt Komo pàt UdYa pātʰ UdCh pār Dana - OpBi - OpMo - OpPa - OpKi - Possible *d > t⁽ʰ⁾ word-finally given voiced reflexes in Chali Uduk. *pāgā v. ‘carry_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi pāgā OpMo pāgā OpPa pāgā OpKi pāgā 313 *páj n. ‘moon or month_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo páj UdYa páj UdCh à+pɛʔ́ Dana - OpBi - OpMo - OpPa - OpKi - *aj → /ɛ/ in Chali Uduk. *pǎjá n. ‘pottery, pot_2’: Reconstructs to: PGw GwHi pǎjá GwLo pǎjá Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pājā n. ‘side of body, rib_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana pājā OpBi - OpMo - OpPa zɪ+̄pāj OpKi - *pàl(Í) v. ‘court (v.), flirt with_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo pàlí 314 UdYa - UdCh - Dana pàlɪ ́ OpBi pàlɪ ́ OpMo pàlɪ ́ OpPa pàlɪ ́ OpKi pàlɪ ́ Unclear as to why it’s not voiced in Komo if initial was b and L tone. *pàndʒá n. ‘antelope_4’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa pàndʒá OpKi pàndʒá *pànsʼ n. ‘axe_4’: Reconstructs to: PGw GwHi pànsʼ GwLo pànsʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pàntʼ v. ‘plaster (v.), adhere_4’: Reconstructs to: PGw GwHi pàntʼ GwLo pàntʼ Komo - UdYa - UdCh - Dana - OpBi - 315 OpMo - OpPa - OpKi - *pǎŋgɔ̄ v. ‘lack (not have)_2’: Reconstructs to: PGw GwHi pǎŋ GwLo pǎŋgɔ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *parʃa ~ pard̪a n*. ‘horse_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo pǎrʃá UdYa - UdCh - Dana pàrd̪á OpBi pàrsá OpMo pàrsá OpPa pàrsá OpKi pàrsá Unknown /ʃ ~ d̪/ correspondence. Borrowing? *pǎʃá v. ‘transplant (e.g. plant)_4’: Reconstructs to: PGw GwHi pǎʃá GwLo pǎʃá Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 316 *pātʃá v. ‘choose_5, shut’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi pātʃá OpMo pātʃá OpPa pātʃá OpKi pātʃá *pàtʃʼɪ ́v. ‘pound (v.)_6’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi pàtʃʼɪ ́ OpMo pàtʃʼɪ ́ OpPa - OpKi - *pɛt̄í n. ‘bad (be)_2’: Reconstructs to: PGw GwHi pɛt̄í GwLo pɛt̄í Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pí v. ‘leave_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa pí 317 UdCh pí Dana - OpBi - OpMo - OpPa - OpKi - *pid(V) v. ‘shake (sth.)_1’: Reconstructs to: PKmn GwHi - GwLo pǐdí Komo pídá UdYa pítʰ UdCh pír Dana - OpBi - OpMo - OpPa - OpKi - *piɟ(i) v. ‘rub hands together (e.g. to make fire using stick)_3’: Reconstructs to: PCtrl GwHi - GwLo - Komo pìz UdYa - UdCh - Dana pʰìcʰ OpBi pīdʒí OpMo - OpPa pīdʒ OpKi - *pīs v. ‘satiated (be)_2’: Reconstructs to: PGw GwHi pīs GwLo pīs Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 318 OpKi - *písāk⁽ʼ⁾ɔ́ n. ‘star_3, firefly’: Reconstructs to: PCtrl GwHi - GwLo - Komo písākɔ́ UdYa - UdCh - Dana písākʼɔ́ OpBi pıs̋ákʼɔ̄ OpMo pıs̋ákʼɔ̄ OpPa pıs̋ákʼɔ̄ OpKi pıs̋ákʼɔ̄ *piʃ v. ‘disregard’: Reconstructs to: PKmn GwHi pīs GwLo pīs Komo píʃ UdYa - UdCh píʃ Dana píʃ OpBi pıs̋ OpMo pıs̋ OpPa pıs̋ OpKi pıs̋ *pít ̪h ~ pɪt́ ̪h n. ‘vagina_1’: Reconstructs to: PKmn GwHi pɪt́ GwLo pɪt́ Komo pít UdYa - UdCh - Dana pít ̪h OpBi - OpMo - OpPa - OpKi - *pītí n. ‘ash_2’: Reconstructs to: POp GwHi - GwLo - 319 Komo - UdYa - UdCh - Dana - OpBi pītí OpMo pītí OpPa - OpKi - *pitʼɔn ~ pit⁽ʰ⁾ɔn n*. ‘flour_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana pìtʰɔ́n OpBi pìtʼɔ̀n OpMo pìtʼɔ̀n OpPa pìtʼɔ̀n OpKi pìtʼɔ̀n *pɪ ̄v. ‘give birth_2’: Reconstructs to: PGw GwHi pɪ ̄ GwLo pɪ ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pɪd̄ɪ̀N n. ‘stone or rock_2’: Reconstructs to: PGw GwHi pɪd̄ɪ̀l GwLo pīdìn Komo - UdYa - UdCh - Dana - OpBi - 320 OpMo - OpPa - OpKi - *pɪd̄ʒɪ ́v. ‘soft (be)_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi pɪd̄ʒɪ ́ OpMo pɪd̄ʒɪ ́ OpPa pɪd̄ʒɪ ́ OpKi pɪd̄ʒɪ ́ *pɪńzɪá́kʼɪʃ̄ɪ̀n n. ‘arrow_2’: Reconstructs to: PGw GwHi pɪńzɪá́kʼɪʃ̄ɪ̀n GwLo pɪńzɪá́kʼɪʃ̄ɪ̀n Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pɪʃ̄ n. ‘broom_3’: Reconstructs to: PGw GwHi pɪʃ̄ GwLo pɪʃ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 321 *pɪťɪ̀ n. ‘hunger_2’: Reconstructs to: PGw GwHi pɪťɪ̀ GwLo pɪďɪ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pɔ́gɔ́ n. ‘river_3’: Reconstructs to: PGw GwHi pɔ́gɔ́ GwLo pɔ́gɔ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pɔ́kʼ v. ‘satiated (be)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo pɔ́g UdYa pɔ́kʼ UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pɔ́ʃ v. ‘dive_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa pɔ́ʃ 322 UdCh pɔ́ʃ Dana - OpBi - OpMo - OpPa - OpKi - *pɔ̌ʃɔ̀ v. ‘arrive_4’: Reconstructs to: PGw GwHi pɔ̌ʃɔ̀ GwLo pɔ̌ʃɔ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pùr n. ‘Dana (ethnonym)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa pùr UdCh pùr Dana - OpBi - OpMo - OpPa - OpKi - *put⁽ʰ⁾ v. ‘satiated (be)_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana pútʰ OpBi pǔt OpMo pǔt OpPa pǔt 323 OpKi pǔt *pʊ́gʊ́n v. ‘dive_1’: Reconstructs to: PGw GwHi pʊ́gʊ́n GwLo pʊ́gʊ́n Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pʊʃ ~ puʃ v. ‘wrap’: Reconstructs to: PKmn GwHi pʊ̄ʃ GwLo pʊ̄ʃ Komo - UdYa pūʃ UdCh pūʃ Dana púʃá OpBi pūsá OpMo pūsá OpPa pūsá OpKi pūsá *pwàʃ n. ‘food_2’: Reconstructs to: PGw GwHi pwàʃ GwLo pwāʃ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bàb v. ‘bury_1’: Reconstructs to: PCtrl GwHi - GwLo - 324 Komo bàb UdYa - UdCh - Dana - OpBi - OpMo - OpPa pàbà OpKi pàbà *bàbá n. ‘father_1’: Reconstructs to: PKmn GwHi - GwLo bàbá Komo bǎ UdYa à+bàbá UdCh à+bàbá Dana - OpBi àbá OpMo àbá OpPa àbá OpKi àbá *baɗa ~ paɗa n. ‘waist, hip_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo bàr UdYa pād UdCh pāɗ Dana pàɗáʔ OpBi - OpMo - OpPa pìpārá OpKi pìpāráʔ *bàgál n. ‘horse_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa bàgál UdCh bàgál Dana - OpBi - 325 OpMo - OpPa - OpKi - *bàj n. ‘elephant_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi bàj OpMo bàj OpPa bàj OpKi bàj *bàj ~ ɓàj v. ‘wide (be)’: Reconstructs to: PKmn GwHi pàj GwLo pàj Komo ɓàjá UdYa ɓàn UdCh bɛ̀ Dana - OpBi pàj OpMo pàj OpPa pàj OpKi pàj *baja v. ‘fast (from drinking or eating)’: Reconstructs to: PGw GwHi bājá GwLo bājā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bàk n. ‘hair_2’: Reconstructs to: PGw 326 GwHi bàk GwLo bàk Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bàmbá n. ‘drum_B’: Reconstructs to: PKmn GwHi pàmbà GwLo pàmbà Komo - UdYa bàmbá UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bǎmbàr n. ‘chair_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa bǎmbàr UdCh bǎmbàr Dana - OpBi - OpMo - OpPa - OpKi - *bángwà n. ‘cannabis_1’: Reconstructs to: PKmn GwHi báŋgà GwLo báŋgà Komo bánkɔ̀ UdYa bǎŋgɔ̀ UdCh bǎŋgɔ̀ 327 Dana - OpBi - OpMo - OpPa - OpKi - *bár n. ‘bird_heron’: Reconstructs to: PKmn GwHi - GwLo ʊ̄+bár Komo à+bár UdYa - UdCh - Dana à+bár OpBi à+bár OpMo à+bár OpPa à+bár OpKi à+bár *bàr v. ‘sweep_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi bàr OpMo bàr OpPa bàr OpKi bàr *bàs n. ‘blood_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo bàʃ UdYa bàs UdCh à+bàs Dana - OpBi - OpMo - OpPa - OpKi - 328 *basʼ v. ‘follow_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa bǎsʼ UdCh bǎt ̪̓ Dana pàt ̪h ɪ ́ OpBi - OpMo - OpPa - OpKi - *bàsʼ n. ‘milk_1’: Reconstructs to: PGw GwHi bàsʼ GwLo bàsʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bāsʼ v. ‘precede_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa bāsʼ UdCh bāt ̪h Dana - OpBi - OpMo - OpPa - OpKi - *bɛs n. ‘dream_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo bɛš 329 UdYa - UdCh - Dana - OpBi bɛs̄ OpMo bɛs̄ OpPa bɛs̄ OpKi bɛs̄ *bɛsɛr v. ‘slip (v.)_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo bɛ̀sɛ̀r UdYa - UdCh - Dana - OpBi bɛšɛŕ OpMo bɛšɛŕ OpPa bɛšɛŕ OpKi bɛšɛŕ *bɛʃ́ɛ̀ v. ‘bypass_2’: Reconstructs to: PGw GwHi bɛʃ́ɛ̀ GwLo bɛʃ́ɛ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *à+bí n. ‘cloth, clothes_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi à+bí OpMo à+bí 330 OpPa - OpKi - *bì+pāj+dʒàw n. ‘Komo (ethnonym)_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi bì+pāj+dʒàw OpMo bì+pāj+dʒàw OpPa bì+pāj+zàw OpKi bì+pāj+ʃàwà *bìb ~ bìb n*. ‘cow_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo bɪ̀b UdYa bìʔ UdCh bìpʰ Dana pìʔ OpBi pìb OpMo pì OpPa pì OpKi pì *bīrītʼ v. ‘stretch’: Reconstructs to: PGw GwHi bīrītʼ GwLo bīrītʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 331 *bīsʼàn n. ‘firefly_1’: Reconstructs to: PGw GwHi bīsʼàn GwLo bīsʼàn Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bīsʼàn n. ‘star_2’: Reconstructs to: PGw GwHi bīsʼàn GwLo bīsʼàn Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bíʃ v. ‘shiver_3’: Reconstructs to: PGw GwHi bíʃ GwLo bíʃ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bīt n. ‘large bird’: Reconstructs to: PGw GwHi bīt GwLo bīt Komo - UdYa - 332 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bit⁽ʰ⁾ v. ‘toss, throw away, fall over’: Reconstructs to: PKmn GwHi pɪ̀t GwLo pɪ̀t Komo bìt UdYa bǐtʰ UdCh bǐtʰ Dana - OpBi - OpMo - OpPa - OpKi - *Bɪncʼ n. ‘fishhook’: Reconstructs to: PKmn GwHi bɪn̄sʼ GwLo bɪn̄sʼ Komo bɪńsʼ UdYa bǐʃʼ UdCh à+bǐcʼ Dana bɪćʼ OpBi ɓītʃʼ OpMo ɓītʃʼ OpPa ɓītʃʼ OpKi ɓītʃʼ Cannot account for initial /ɓ/ in Opo. *bɪśʼɪńɪ ̄n. ‘four_2’: Reconstructs to: PGw GwHi bɪśʼɪńɪ ̄ GwLo bɪśʼɪńɪ ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - 333 OpPa - OpKi - *bɔ̀b ~ bɔ̀ɓ v. ‘hide, skulk’: Reconstructs to: PKmn GwHi pɔ̀pʼ GwLo pɔ̀pʼ Komo bɔ̀b UdYa - UdCh - Dana - OpBi pɔ̀j OpMo pɔ̀j OpPa pɔ̀j OpKi - *bɔ̀g(ɔ̀) v. ‘play_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo bɔ̀g UdYa - UdCh - Dana pɔ̀gɔ̀ʔ OpBi pɔ̀gɔ̀ OpMo pɔ̀gɔ̀ OpPa pɔ̀gɔ̀ OpKi pɔ̀gɔ̀ *bɔ̀r n. ‘chest_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa bɔ̀r UdCh à+bɔ̀r Dana - OpBi - OpMo - OpPa - OpKi - *bū ~ pū n*. ‘sesame_1’: Reconstructs to: PKoUd GwHi - 334 GwLo - Komo bū UdYa pū UdCh à+pū Dana - OpBi - OpMo - OpPa - OpKi - *bùl n. ‘drum_A’: Reconstructs to: PCtrl GwHi - GwLo - Komo bùl UdYa - UdCh - Dana bùl OpBi bùl OpMo bùl OpPa bùl OpKi bùl *bùl v. ‘spoil, stink, smell bad_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa bùl UdCh bùl Dana - OpBi - OpMo - OpPa - OpKi - *bUma v. ‘pregnant (be)’: Reconstructs to: PKmn GwHi pʊ̀m GwLo pʊ̀m Komo ɓú UdYa pwá UdCh bwà Dana pùmá 335 OpBi pǔmá OpMo pǔmá OpPa pǔmá OpKi pǔmá Cannot account for /ɓ/ in Komo and /p/ in Yabus Uduk. *bùmà n. ‘belly or stomach_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa bwàʔ UdCh - Dana mwà OpBi pùmà OpMo pùmà OpPa pùmà OpKi pùmà *buɲɛ n. ‘forehead_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa bwìɲ UdCh bùjɛ̀ Dana - OpBi - OpMo - OpPa - OpKi - *burbuɗ n. ‘dust, sand_1’: Reconstructs to: PKmn GwHi bùrbūt GwLo bùrbūt Komo - UdYa bǔtʰ UdCh bǔɗ Dana būrkʼùs OpBi būrkʼùs OpMo būrkʼùs OpPa būrkʼùs OpKi būrkʼùs 336 Also bùrbūtʼ in Gwama. Addition of *kʼus appears to be a PDaOp innovation though source is unknown.. *busʼ v. ‘choke, strangle_1’: Reconstructs to: PKmn GwHi bǔsʼ GwLo bǔsʼ Komo - UdYa bùsʼ UdCh bùt ̪̓ Dana - OpBi pʰǔtʃʼ OpMo - OpPa - OpKi - *bǔʃú v. ‘fart_2’: Reconstructs to: PGw GwHi bǔʃú GwLo bǔʃú Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bùʃùl n. ‘belly or stomach_1’: Reconstructs to: PKmn GwHi bùʃùl GwLo bùʃì Komo bùʃ UdYa - UdCh bùʃ Dana - OpBi pùsà OpMo pùsà OpPa pùsà OpKi pùsà *bʊ̀ n. ‘hole_2’: Reconstructs to: PGw GwHi bʊ̀ 337 GwLo bʊ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bʊkʰ v. ‘extract tooth_1, barking (of dog)’: Reconstructs to: PKmn GwHi pʊ̌ GwLo pʊ̌ Komo - UdYa - UdCh bǔkʰ Dana - OpBi - OpMo - OpPa - OpKi - *bʊ̌rà n. ‘cat_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo bʊ̌rà UdYa bǔrá UdCh à+bǔrá Dana - OpBi - OpMo - OpPa - OpKi - *bʊraɲ n. ‘cloth, clothes_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo bʊ̄lɛ̀n UdYa bǔrɛ̀ɲ UdCh bùrìɲɛ̀ Dana à+bʊ́râŋ 338 OpBi - OpMo - OpPa bʊ̀lɛ̀ŋ OpKi bʊ̀lɛ̀n *bʊ̀tʊ̀ v. ‘clear land (for planting)_2’: Reconstructs to: PGw GwHi bʊ̀tʊ̀ GwLo bʊ̀tʊ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bVr(màn) n. ‘root_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo bɪ̀l UdYa bǐr UdCh bǐrmàn Dana bìl+mà+cá OpBi - OpMo - OpPa - OpKi - *bwàhāɗgī(da) n. ‘palate_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa bwàgídà UdCh bwàhāɗgī Dana - OpBi - OpMo - OpPa - OpKi - 339 *bwàkʼ v. ‘hide (oneself)’: Reconstructs to: PCtrl GwHi - GwLo - Komo bɔ̀ʔ UdYa bàʔ UdCh bàkʼ Dana bàkʼ OpBi - OpMo - OpPa - OpKi bɔ̀kʼ *bwaŋ(a) v. ‘path, road_1, towards’: Reconstructs to: PKmn GwHi pwǎŋà GwLo pwǎŋà Komo - UdYa - UdCh bwàj Dana - OpBi - OpMo - OpPa - OpKi - *bwǎʃà n. ‘snake_3’: Reconstructs to: PGw GwHi bwǎʃà GwLo bwǎʃà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *bwí n. ‘arm_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa bwí 340 UdCh à+bǐ Dana - OpBi - OpMo - OpPa - OpKi - *bɔŋkʼɔ́ n. ‘frog_1’: Reconstructs to: PKmn GwHi pʼà+bɔ̌ŋgɔ́ GwLo pʼà+bɔ̌ŋgɔ́ Komo bā+ɓɔ̌ŋkʼɔ́ UdYa - UdCh - Dana à+ɓɔ̄ŋkʼɔ́ʔ OpBi à+ɓɔ̄ŋkʼɔ́ OpMo à+bɔ̄ŋkʼɔ́ OpPa à+bɔ̄ŋkʼɔ́ OpKi à+bɔ̄ŋkʼɔ́ *ɓ → b /V_V in Proto-Gwama *ɓ(w)áʃ v. ‘polygamous (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɓáʃ UdYa ɓwáʃ UdCh ɓáʃ Dana ɓáʃ OpBi - OpMo - OpPa - OpKi - *ɓā+tın̋ɛ ̄v. ‘lie down, sleep_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi ɓā+tın̋ɛ ̄ OpMo ɓā+tín 341 OpPa ɓā+tín OpKi ɓā+tɪ̂n *ɓácʼ n. ‘thigh_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɓásʼ UdYa - UdCh - Dana ɓácʼ OpBi ɓátʃʼ OpMo ɓátʃʼ OpPa ɓátʃʼ OpKi ɓátʃʼ *ɓakʼum n. ‘daughter’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɓā UdYa ɓwāʔ UdCh ɓāʔúm Dana - OpBi ɓākʼ OpMo ɓākʼ OpPa ɓākʼ OpKi ɓākʼ Most leikely related to, or source of 3sg.F morpheme *ɓa. *ɓalilaj n. ‘palate_1’: Reconstructs to: PKmn GwHi pʼàlílì GwLo pʼàlílí Komo ɓàlílá UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 342 *ɓānápʰà n. ‘girl_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ɓānápʰà OpBi ɓānàpà OpMo ɓānàpà OpPa ɓānàpà OpKi ɓānàpà *ɓáŋgɔ̄ n. ‘hyrax_1’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi ɓáŋgɔ̄ OpMo ɓáŋgɔ̄ OpPa ɓáŋgɔ̄ OpKi ɓáŋgɔ̄ *ɓāpʰā n. ‘woman’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ɓāpʰā OpBi ɓāpʰà OpMo ɓāpʰā OpPa ɓāpʰā OpKi ɓāpʰā *ɓár adv. ‘all_5, finish’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ɓár 343 UdCh ɓár Dana - OpBi - OpMo - OpPa - OpKi - *ɓàsʼ v. ‘hot (be)_1, ill (be)’: Reconstructs to: PKoUd GwHi - GwLo - Komo ɓàsʼ UdYa ɓàsʼ UdCh ɓāt ̪̓ Dana - OpBi - OpMo - OpPa - OpKi - *ɓaʃa n. ‘tree_sp._1’: Reconstructs to: PKmn GwHi pʼàʃ GwLo - Komo ɓàʃ UdYa ɓàʃà UdCh - Dana ɓàʃà OpBi ɓāsā OpMo ɓāsā OpPa ɓāsā OpKi - *ɓàtʃʼ n. ‘buffalo_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi ɓàtʃʼ OpMo ɓàtʃʼ OpPa ɓàtʃʼ 344 OpKi ɓàtʃʼ *ɓāʔɔ̄ n. ‘girl_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo bāʔɔ̄ UdYa - UdCh - Dana - OpBi ɓāɔ́ OpMo ɓāɔ́ OpPa ɓāɔ́ OpKi ɓāɔ́ *ɓɛr̄ v. ‘arrive_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi ɓɛr̄ OpMo ɓɛr̄ OpPa ɓɛr̄ OpKi ɓɛr̄ *ɓId̪a n. ‘neck’: Reconstructs to: PKmn GwHi pʼɪl̄ GwLo pʼɪ ̄ Komo ɓàʔ UdYa ɓāʔ UdCh ɓāʔ Dana ɓìd̪à OpBi ɓījā OpMo ɓījā OpPa ɓījā OpKi ɓījā *ɓɪb́ɪ ̄n. ‘wound_2’: Reconstructs to: PDaOp GwHi - GwLo - 345 Komo - UdYa - UdCh - Dana ɓɪb́ɪ ̄ OpBi ɓɪb́ɪ ̄ OpMo ɓɪb́ɪ ̄ OpPa ɓɪb́ɪ ̄ OpKi ɓɪb́ɪ ̄ Cognate with Komo /ɓî/ eye? Komo traditional scarring on cheeks can be circular. *ā+ɓīgɪń adv. ‘all_1’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi ā+ɓīgɪń OpMo ā+ɓīɡɪń OpPa ā+ɓīgɪń OpKi ā+ɓʊ̄gà *ɓɪs̄à n. ‘crocodile_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɓɪs̄à UdYa - UdCh - Dana - OpBi ɓīsā OpMo ɓɪs̄ā OpPa ɓɪs̄ā OpKi - *ɓɪśʼ ~ ɓísʼ v. ‘strong (be)_1’: Reconstructs to: PKmn GwHi pʼɪ ́ GwLo pʼɪ ́ Komo ɓísʼ UdYa ɓísʼ UdCh ɓít ̪̓ Dana - 346 OpBi - OpMo - OpPa - OpKi - *ɓOm(it⁽ʰ⁾) n. ‘woman’: Reconstructs to: PKoUd GwHi - GwLo - Komo ɓāmít UdYa ɓûm UdCh à+ɓɔ́m Dana - OpBi - OpMo - OpPa - OpKi - *ɓɔ́r v. ‘good (be)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo ɓɔ́l UdYa ɓɔ́r UdCh ɓɔ́r Dana - OpBi - OpMo - OpPa - OpKi - *ɓūntʃʼú n. ‘sap_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi ɓūntʃʼú OpMo ɓūntʃʼú OpPa ɓūntʃʼú OpKi ɓūntʃʼú 347 *ɓʊ́t ̪̓ v. ‘soft (be)_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ɓʊ́t ̪̓ OpBi ɓʊ̋tʼ OpMo ɓʊ̋tʼ OpPa ɓʊ̋tʼ OpKi ɓʊ̋tʼ *pʼá v. ‘dance_2’: Reconstructs to: PGw GwHi pʼá GwLo pʼá Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pʼàcʰ v. ‘sour (be)_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa pʼàcʰ UdCh pʼàcʰ Dana - OpBi - OpMo - OpPa - OpKi - *pʼátʼá v. ‘white (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo pʼátʼá UdYa - 348 UdCh - Dana pʼátʼá OpBi pʼátʼá OpMo pʼátʼá OpPa pʼátʼá OpKi pʼátʼá *pʼātʼà v. ‘help_2’: Reconstructs to: PGw GwHi pʼātʼà GwLo pʼātʼà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *pʼɛń n. ‘buttocks_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo pʼɛń UdYa pʼɛń UdCh pʼɛń Dana - OpBi - OpMo - OpPa - OpKi - *pʼɛŕ v. ‘red (be)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo pʼɛĺ UdYa pʼɛŕ UdCh pʼɛŕ Dana - OpBi - OpMo - OpPa - 349 OpKi - *(tʼi)pʼIkʼIɲ(a) n. ‘ash_1’: Reconstructs to: PKmn GwHi pʼɪk̄ʼɪń GwLo pʼɪk̄ʼɪń Komo pʼīn UdYa - UdCh tʼīpʼīɲ Dana pʼɪn̄ā OpBi - OpMo - OpPa pʼɪn̄ā OpKi pʼɪn̄ā *pʼi(n)Cʼ? v. ‘peel, husk_1’: Reconstructs to: PKmn GwHi pʼɪńsʼ GwLo pʼɪńsʼ Komo - UdYa - UdCh pʼíɗ Dana pʼɪt́ʼ OpBi - OpMo - OpPa - OpKi - Unsure of glottalized coda consonant. *pʼɔt ̪̓(a) v. ‘pick_1’: Reconstructs to: PKmn GwHi pʊ́t GwLo pʊ́t Komo pʼɔ̀tʼ UdYa - UdCh - Dana pʼɔ̀t ̪h à OpBi pʼɔ̄tʼā OpMo pʼɔ̄tʼā OpPa pʼɔ̄tʼā OpKi pʼɔ̄tʼā Loss of glottalization in final *t ̪̓ Dana and Gwama? 350 *pʼʊ̄mʊ̀tʼ n. ‘flour_2’: Reconstructs to: PGw GwHi pʼʊ̄mʊ̀tʼ GwLo pʼʊ̄mʊ̀tʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *t ̪h (w)ád̪ n. ‘hunger_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo twáj UdYa tʰɔ́j UdCh tʰɛʔ́ Dana t ̪h ád̪ OpBi tʰáj OpMo tʰáj OpPa tʰáj OpKi tʰáj *t ̪h áɓ v. ‘kick’: Reconstructs to: PKmn GwHi tápʼ GwLo tápʼ Komo táb UdYa tʰáb UdCh tʰáɓ Dana t ̪h áp OpBi tʰáp OpMo tʰáp OpPa tʰáp OpKi tʰáp *t ̪h ásʼ v. ‘pound (v.)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa tʰásʼ 351 UdCh t ̪h át ̪̓ Dana - OpBi - OpMo - OpPa - OpKi - *t ̪h at ̪̓ v. ‘mediate_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa - UdCh t ̪h āt ̪̓ Dana t ̪h ājd̪á OpBi - OpMo - OpPa - OpKi tʰájá In Chali Uduk the meaning is ‘come towards as many people’. *t ̪h ipʰ v. ‘raise (a child)_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa sīpʰ UdCh tī̪pʰ Dana t ̪h ì OpBi - OpMo - OpPa - OpKi - *t ̪h u v. ‘spit (v.)_1’: Reconstructs to: PKmn GwHi tū GwLo tū Komo - UdYa - UdCh - Dana t ̪h úwà OpBi tʰűjhá OpMo tʰűjhá 352 OpPa tʰűjhá OpKi tʰűjhá *t ̪h ùbá v. ‘show_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana t ̪h ùbá OpBi tʰūbá OpMo tʰūbá OpPa tʰūbá OpKi tʰūbá *t ̪h ùd̪ v. ‘dry out_1, roast next to fire, aim at’: Reconstructs to: PCtrl GwHi - GwLo - Komo tùd UdYa tʰūd UdCh tʰūr Dana t ̪h ùd̪ OpBi - OpMo - OpPa - OpKi - Meaning: ‘aim at’ and ‘roast next to fire’ in Dana and Uduk. *tʰaʃ ~ t ̪h aʃ v. ‘make go away’: Reconstructs to: PKmn GwHi tāʃ GwLo tāʃ Komo tàʃ UdYa tʰáʃ UdCh tʰáʃ Dana - OpBi - OpMo - OpPa - OpKi - 353 *tá̪j n. ‘sun_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tɛ ́ UdYa tíndí UdCh à+tɛńtɛ ́ Dana tá̪j OpBi táj OpMo táj OpPa táj OpKi táj Also means ‘season’ and ‘year’ in Dana. Reduplication in Uduk? *tá̪kʼāláj n. ‘half_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tákálá UdYa - UdCh - Dana tá̪kʼāláj OpBi tálá OpMo tálá OpPa tálá OpKi tákàlāj *ta̪m v. ‘pray, beg_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo dàm UdYa - UdCh tá̪m Dana tá̪má OpBi tāmá OpMo tāmá OpPa tāmá OpKi tāmá *tá̪pàj n. ‘hyena_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tápàjí 354 UdYa - UdCh - Dana à+tá̪pàjī OpBi à+tápàjɪ ́ OpMo à+tápàjɪ ́ OpPa à+tápàjɪ ́ OpKi à+tápàjɪ ́ *tá̪pʼí n. ‘flea_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tápí UdYa - UdCh - Dana tá̪pʼí OpBi - OpMo - OpPa - OpKi - *tà̪rá n. ‘drum_C’: Reconstructs to: PKmn GwHi - GwLo tàrá Komo tàrá UdYa - UdCh - Dana tà̪rá OpBi tàrá OpMo tàrá OpPa tàrá OpKi tàrá Borrowing from Arabic darabukka? *tÉ̪ v. ‘know or be able_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa - UdCh tɛ̪ ́ Dana tɪ̪ ́ OpBi - 355 OpMo - OpPa - OpKi - *tÉ̪ɗ(á) v. ‘swim_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana tɛ̪ɗ́á OpBi tɪ ́ OpMo tɪ ́ OpPa tɪ ́ OpKi tɪ ́ *tɛ̪ɗ́ v. ‘shallow (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa tɛd̄ UdCh tɛɗ̄ Dana tɛ̪ɗ́ OpBi tɛŕɛ ́ OpMo tɛŕ OpPa tɛŕ OpKi tɛŕ *tɛ̪ĺ v. ‘pelt_1, drive in, pound, hammer (something)’: Reconstructs to: PCtrl GwHi - GwLo - Komo tɛĺ UdYa - UdCh tɛ̪ĺ Dana tɛ̪ĺ OpBi tɛĺ OpMo tɛĺ OpPa tɛĺ OpKi tɛĺ 356 *tɛ̪̀mɛ̀ v. ‘try (test)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana tɛ̪̀mɛ̀ OpBi tɛ̀mɛ̀ OpMo - OpPa - OpKi tɛmɛ *tɛ̪n(ɛ) n. ‘hail, ice_2, melt’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa - UdCh tɛ̪n̄ Dana tɛ̪ń OpBi tɛńɛ ́ OpMo - OpPa tɛń OpKi tɛń In Chali Uduk the meaning is ‘melt’. *tɛ̪r̄ v. ‘urinate_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa - UdCh tɛ̪r̄ Dana - OpBi - OpMo - OpPa - OpKi - *tì̪mìs n. ‘beer filter_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - 357 UdYa - UdCh - Dana tì̪mìs OpBi tīmīs OpMo tīmīs OpPa tīmīs OpKi tīmīʃ *tɪ̪̀p⁽ʰ⁾a n. ‘sesame_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana tɪ̪̀pʰà OpBi - OpMo - OpPa - OpKi tɪp̄á *tɪ̪t́ ̪v. ‘roughen (stone for grinding)’: Reconstructs to: PKmn GwHi sɪt́ GwLo sɪt́ Komo tɪt́ UdYa títʰ UdCh tír Dana tɪ̪t́ ̪h OpBi tɪt́ɪ ́ OpMo tɪt́ʰ OpPa tɪt́ʰ OpKi tɪt́ʰ *(ɔ)t⁽̪ʰ⁾ɔɗ v. ‘grind (second grind)’: Reconstructs to: PKmn GwHi - GwLo twɛj́ Komo tɔ́ UdYa - UdCh tɔ̪́ɗ Dana ɔ̀t ̪h ɔ́ OpBi ɔ́tɔ́ OpMo ɔ́tʰ 358 OpPa ɔ́tʰ OpKi ɔ́tʰ Meaning in Chali Uduk is ‘scrape grindings into container’. *tU̪b(a) v. ‘pierce_1’: Reconstructs to: PKmn GwHi sʊ́ GwLo sʊ́ Komo - UdYa tūpʰ UdCh tūpʰ Dana tú̪bá OpBi - OpMo - OpPa tʰūbá OpKi tʰūbá *(pɪ)tʊ̪j n. ‘farm (n.)_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana tʊ̪̂j OpBi pɪ̀tɪ̀ OpMo pɪ̀tʊ̀j OpPa pɪ̀tʊ̀j OpKi pàtɔ̀j *tú̪k⁽ʰ⁾(u) n. ‘acacia’: Reconstructs to: PKmn GwHi tʊ̌kʊ̀ GwLo - Komo túk UdYa túkʰ UdCh à+tú̪kʰ Dana tú̪kʰ OpBi tűkʰ OpMo tűkʰ OpPa tűkʰ OpKi tűkʰ 359 *tU̪s n. ‘cotton, thread, spider web_1’: Reconstructs to: PKmn GwHi dʊ̄ʃ GwLo dʊ̄ʃ Komo tūʃ UdYa tūs UdCh tū̪s Dana - OpBi - OpMo - OpPa - OpKi - Cannot account for initial /d/ in Gwama. *tú̪s v. ‘extinguish_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo túʃ UdYa ūs UdCh ūs Dana tú̪s OpBi tűs OpMo tűs OpPa tűs OpKi tűs *tā̪gāj ~ tā̪gāj n. ‘giraffe_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tāgɪ ̄ UdYa - UdCh - Dana - OpBi tāgāj OpMo tāgāj OpPa tāgāj OpKi - *tí̪t(á) ~ tít(á) v. ‘precede_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tít 360 UdYa - UdCh - Dana - OpBi tıt̋á OpMo tıt̋á OpPa tıt̋á OpKi tıt̋á *ti̪til ~ titil v. ‘cut (meat into one long piece)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tìtíl UdYa - UdCh - Dana - OpBi - OpMo - OpPa tītíl OpKi tītíl *tù̪n ~ tùn v. ‘raise (a child)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tùn UdYa - UdCh - Dana - OpBi tūn OpMo tūn OpPa tūn OpKi tún *d̪(w)àràcʼ n. ‘urine_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo dàràsʼ UdYa dwàràʃ UdCh d̪àràcʼ Dana - OpBi - OpMo - 361 OpPa - OpKi - *d̪àm n. ‘bee, honey’: Reconstructs to: Pkmn GwHi tàm GwLo tàm Komo dàm UdYa dàm UdCh à+d̪àm Dana tâ̪m OpBi tàm OpMo tàm OpPa tàm OpKi tàm *d̪àm(V) v. ‘old (be)_1’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa zàmì UdCh d̪àmɔ̀ Dana - OpBi - OpMo - OpPa - OpKi - *d̪àn v. ‘big (be), male, elder_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa zàn UdCh d̪àn Dana - OpBi - OpMo - OpPa - OpKi - *d̪àn+cɔm n. ‘grandfather_2’: Reconstructs to: PUd GwHi - 362 GwLo - Komo - UdYa zàn+cɔ̂m UdCh à+d̪àn+cɔ̄m Dana - OpBi - OpMo - OpPa - OpKi - *d̪ana n. ‘Dana (ethnonym)_3’: Reconstructs to: PKmn GwHi dànā GwLo dànā Komo dāná UdYa - UdCh - Dana d̪àná OpBi dànā OpMo dànā OpPa dànā OpKi dànā *d̪Ekwād̪à n. ‘kidney_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana d̪ɛ̀kʰwād̪à OpBi dɪ̀kwādà OpMo dɪ̀kwādà OpPa dɪ̀kwādà OpKi dɪ̀kwādà *d̪ɛl v. ‘sweet (be)_1’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa zɛl̄ UdCh d̪ɛľɛl̄ Dana - 363 OpBi - OpMo - OpPa - OpKi - *d̪ɛ̀rɪń v. ‘tear (shred)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo sɛ̀rɪń UdYa zɛř UdCh d̪ɛř Dana - OpBi - OpMo - OpPa - OpKi - *d̪ɛrkʼɛsʼ v. ‘slip (v.)_1’: Reconstructs to: PKmn GwHi dɛr̄gɛśʼ GwLo dɛr̄gɛśʼ Komo - UdYa dɛ̀rɛ̀s UdCh d̪ɛ̀rɛ̀s Dana d̪ɛŕkʼɛś OpBi - OpMo - OpPa - OpKi - *d̪ìbà n. ‘rain (precipitation)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana d̪ìbà OpBi - OpMo - OpPa - OpKi dìbà 364 *d̪ìm v. ‘strain (solids from liquid)_1’: Reconstructs to: PKmn GwHi - GwLo zìŋā Komo zìm UdYa zǐm UdCh d̪ìm Dana tì̪m OpBi tìm OpMo tìm OpPa tìm OpKi tìm *ad̪imɛ n. ‘child_3’: Reconstructs to: PCtrl GwHi - GwLo - Komo āt UdYa - UdCh - Dana àd̪ín OpBi àdım̋ɛ ́ OpMo àdím OpPa àdím OpKi àdím *d̪ìnd̪àl n. ‘beer filter_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo zɪ̀nzàl UdYa zìnzàl UdCh à+d̪ìnd̪àl Dana - OpBi - OpMo - OpPa - OpKi - Spirantization in Yabus Uduk and Komo. *d̪is ~ d̪ɪs v. ‘new_1’: Reconstructs to: PKmn GwHi dīʃ GwLo dīʃ Komo zɪ̀ʃ 365 UdYa tīs UdCh tī̪s Dana - OpBi - OpMo - OpPa - OpKi - *d̪ìsʼ v. ‘sweep_1’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa dìsʼ UdCh d̪ìt ̪̓ Dana - OpBi - OpMo - OpPa - OpKi - *d̪ɔ̀bɔ n. ‘lion_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo dɔ̀b UdYa zɔ̀pʰ UdCh à+d̪ɔ̀pʰ Dana - OpBi tɔ̀bɔ̀ OpMo tɔ̀bɔ̀ OpPa tɔ̀b OpKi tɔ̀b *d̪ɔ̀sʼ v. ‘suck_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo dɔ̀sʼ UdYa dɔ̀sʼ UdCh d̪ɔ̀t ̪̓ Dana tɔ̪̀sʼ OpBi tɔ̀tʃʼɔ̀ OpMo tɔ̀tʃʼ 366 OpPa tɔ̀tʃʼ OpKi tɔ̀tʃʼ *d̪ɔ̀t ̪h (á) v. ‘tie up (tether)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana d̪ɔ̀t ̪h á OpBi dɔ̄tʰ OpMo dɔ̄tʰ OpPa dɔ̄tʰ OpKi dɔ̄tʰ *d̪U(ru)s(E) v. ‘fart_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tùʃ UdYa - UdCh wū+d̪ùrùs Dana tʊ̪̂s OpBi tīsí OpMo tīsí OpPa tʊ̄sɛ ́ OpKi tʊ̄ʃɪ ́ *d̪ǔgà n. ‘bean_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa zúgàʔ UdCh à+d̪ǔgà Dana - OpBi - OpMo - OpPa - OpKi - *t ̪̓ ākʰ v. ‘spit (v.)_2’: Reconstructs to: PUd GwHi - 367 GwLo - Komo - UdYa tʼākʰ UdCh t ̪̓ ākʰ Dana - OpBi - OpMo - OpPa - OpKi - *t ̪̓ ākʼán v. ‘abstain from_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana t ̪̓ ākʼán OpBi tʼākʼá OpMo tʼākʼá OpPa tʼākʼá OpKi - *t ̪̓ ākʼʊ́mʊ́ n. ‘calf of leg’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana t ̪̓ ākʼʊ́mʊ́ OpBi tʼākʼʊ̋mʊ̋ OpMo tʼākʼʊ̋mʊ̋ OpPa tʼākʼʊ̋mʊ̋ OpKi tʼākʼʊ̋mʊ̋ *t ̪̓àm v. ‘rinse face_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼà+ɓî UdYa - UdCh - Dana t ̪̓àm 368 OpBi tʼām OpMo tʼām OpPa tʼām OpKi tʼām *t ̪̓at ̪̓ v. ‘empty (be)_1, dull’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼɪ ̌ UdYa - UdCh tʼíʔ Dana t ̪̓àt ̪̓ OpBi tʼáj OpMo tʼáj OpPa tʼáj OpKi tʼáj *t ̪̓ ɛɗ v. ‘lick_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼɛr̄ UdYa tʼɛd̄ UdCh tʼɛɗ̄ Dana t ̪̓ ɛɗ́ OpBi tʼɛŕɛ ́ OpMo tʼɛŕ OpPa tʼɛŕ OpKi tʼɛŕ *t ̪̓ ɛn adv. ‘alone, abstain from, not want to do’: Reconstructs to: PKmn GwHi - GwLo sʼɪn̄ Komo sʼɛń UdYa tʼɛń UdCh tʼɛ ́ Dana gà+t ̪̓ ɛń OpBi ā+tʼɛń OpMo ā+tʼɛń OpPa ā+tʼɛń OpKi ā+tʼɛń 369 *t ̪̓in v. ‘stretch’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼìn UdYa tʼīnīn+īs UdCh tʼīnīn+īs Dana t ̪̓ìn OpBi tʼín OpMo tʼín OpPa tʼín OpKi tʼín *t ̪̓ ɪ(t ̪̓V) v. ‘thin (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼɪ̀̀tʼá UdYa tʼī UdCh tʼī Dana - OpBi tʼɪt̄ʼɪ ́ OpMo tʼɪt̄ʼɪ ́ OpPa tʼɪt̄ʼɪ ́ OpKi tʼɪt̄ʼɪ ́ *t ̪̓O v. ‘grab with fingers, pinch_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa - UdCh tʼú Dana t ̪̓ ɔ́wá OpBi tʼʊ́ OpMo tʼʊ́ OpPa tʼʊ́ OpKi tʼʊ́ *t ̪̓Osʼ v. ‘choke, strangle_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼúsʼ UdYa - 370 UdCh tʼɔ́cʰ Dana t ̪̓ ɔ́sʼ OpBi - OpMo - OpPa - OpKi - *t ̪̓ ɔ̀ɗ v. ‘slaughter, cross boundary, chop’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼɔ̀r UdYa - UdCh - Dana t ̪̓ ɔ̀ɗ OpBi - OpMo - OpPa tʼɔ̄r OpKi tʼɔ̄r *t ̪̓wa n. ‘mouth’: Reconstructs to: PKmn GwHi tʼwā GwLo tʼwā Komo tʼā UdYa tʼwā UdCh tʼwā Dana t ̪̓ āʔá OpBi tʼā OpMo tʼā OpPa tʼā OpKi tʼā *t ̪̓wI v. ‘enter_2, sprout’: Reconstructs to: PKmn GwHi tʼwɪ ́ GwLo tʼwɪ ́ Komo - UdYa - UdCh - Dana t ̪̓wī OpBi - OpMo - OpPa - 371 OpKi - Meaning in ‘sprout’ in Gwama. *tʰágá v. ‘mediate_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi tʰágá OpMo tʰágá OpPa tʰágá OpKi - *tʰánā ~ tʰánɪ ̄n. ‘turtle_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo bā+tánɪ ̄ UdYa - UdCh - Dana tʰánā OpBi tʰánā OpMo tʰánā OpPa tʰánā OpKi tʰánā *tʰáp v. ‘plaster (v.), adhere_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana tʰápʰ OpBi tʰáp OpMo tʰáp OpPa tʰáp OpKi tʰáp *tʰát ~ t ̪h át v. ‘transplant (e.g. plant)_1’: Reconstructs to: PCtrl GwHi - 372 GwLo - Komo tát UdYa - UdCh - Dana - OpBi tʰát OpMo tʰát OpPa tʰát OpKi tʰát *tʰi(aj) n. ‘pottery, pot_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sɪ ̄ UdYa - UdCh - Dana tījā OpBi tʰɪ ̌ OpMo tʰɪ ̌ OpPa tʰɪ ̌ OpKi tʰɪ ̌ *tʰíŋ+kùm n. ‘grandmother_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa tʰíŋ+kùm UdCh à+tʰíŋ+kūm Dana - OpBi - OpMo - OpPa - OpKi - *tʰır̋ v. ‘pour_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - 373 OpBi tʰır̋ OpMo tʰır̋ OpPa tʰır̋ OpKi tʰır̋ *tʰɔ́ɗ v. ‘plait or braid or weave_5’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa tʰɔ́d UdCh tʰɔ́ɗ Dana - OpBi - OpMo - OpPa - OpKi - *tʰɔ́r v. ‘show_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa tʰɔ́r UdCh tʰɔ́r Dana - OpBi - OpMo - OpPa - OpKi - *tʰú(i) v. ‘give birth_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tú UdYa - UdCh tʰú Dana tʰwí OpBi tʰű OpMo tʰű OpPa tʰű OpKi tʰű Chali Uduk meaning is ‘last child born’. 374 *tʰul n. ‘gourd_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo túl UdYa - UdCh tʰúl Dana - OpBi tʰɔ̀j OpMo tʰɔ̀j OpPa tʰɔ̀j OpKi - *tʰúlá n. ‘brain_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa tʰúlá UdCh à+tʰúláʔ Dana - OpBi - OpMo - OpPa - OpKi - *tʰaʃ ~ t ̪h aʃ v. ‘make go away’: Reconstructs to: PKmn GwHi tāʃ GwLo tāʃ Komo tàʃ UdYa tʰáʃ UdCh tʰáʃ Dana - OpBi - OpMo - OpPa - OpKi - *t⁽ʰ⁾(w)àg n. ‘forehead_1’: Reconstructs to: PKmn GwHi twã GwLo twã ̄ Komo tàg 375 UdYa - UdCh - Dana tàg OpBi bī+tʰāg OpMo - OpPa pī+tʰāg OpKi - *tà v. ‘be, do’: Reconstructs to: PKmn GwHi - GwLo ta Komo tà UdYa tā UdCh tā Dana - OpBi tā OpMo - OpPa tā OpKi - Gwama ‘be far’ Opo ‘do/work’ *tā v. ‘abstain from_2’: Reconstructs to: PGw GwHi tā GwLo tā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tā+dʒàj v. ‘fight_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi tā+dʒàj 376 OpMo tā+dʒàj OpPa tā+zàj OpKi tā+ʃàj *tābūk v. ‘mediate_4’: Reconstructs to: PGw GwHi tābūk GwLo tābūk Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *táɗā n. ‘mother_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo dâ UdYa à+tádā UdCh táɗā Dana - OpBi - OpMo - OpPa - OpKi - *takal n. ‘saliva_3’: Reconstructs to: PGw GwHi tākāl GwLo tāgɪ ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 377 *tǎŋá n. ‘bamboo_1’: Reconstructs to: PGw GwHi tǎŋá GwLo tǎŋá Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tàpʰà n. ‘chief_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa tàpʰà UdCh tàpʰà Dana - OpBi - OpMo - OpPa - OpKi - *tǎpʼ v. ‘follow_4’: Reconstructs to: PGw GwHi tǎpʼ GwLo tǎpʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *taʃ n. ‘mosquito_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa tāʃ 378 UdCh à+tāʃ Dana tāʃ OpBi tās OpMo tās OpPa tās OpKi tǎʃ *táʃ v. ‘strain (solids from liquid)_2’: Reconstructs to: PGw GwHi táʃ GwLo táʃ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tát n. ‘belly or stomach_3’: Reconstructs to: PGw GwHi tát GwLo tát Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tàwàn v. ‘hot (be)_3’: Reconstructs to: PGw GwHi tàwàn GwLo tàwàn Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 379 OpKi - *tàzà n. ‘beer filter_1’: Reconstructs to: PGw GwHi tàzà GwLo tàzà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tEŋ(g)(E) v. ‘shake (sth.)_2’: Reconstructs to: PKmn GwHi tɪḡɪ ̄ GwLo tɪḡɪ ̄ Komo - UdYa - UdCh - Dana tɛŋ́ OpBi - OpMo tɪŋ́há OpPa tɪŋ́há OpKi tɪŋ́há *tɛr v. ‘carry_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo tɛŕ UdYa tɛr̄ UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tíbí v. ‘listen_3’: Reconstructs to: PGw GwHi tíbí GwLo tíbí 380 Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tiŋ(a) ~ tin(a) n. ‘cheek_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana tīŋ OpBi tǐná OpMo tǐná OpPa tǐná OpKi tǐná Related to or borrowed from ‘breast’ in W. Nilotic (e.g. /ti̪n/ in Mayak)? *tiritiri n. ‘bird_dove (African mourning)’: Reconstructs to: PKmn GwHi - GwLo títìtī Komo - UdYa - UdCh - Dana - OpBi à+tìrıt̋ır̋ì OpMo à+tìrıt̋ır̋ì OpPa à+tìrıt̋ır̋ì OpKi à+tìrıt̋ır̋ì *tìʃàr n. ‘lion_2’: Reconstructs to: PGw GwHi tìʃàl GwLo tìʃàr Komo - UdYa - UdCh - Dana - 381 OpBi - OpMo - OpPa - OpKi - *tɪ ́v. ‘give’: Reconstructs to: PGw GwHi tɪ ́ GwLo tɪ ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tɪm̌ v. ‘protect_3’: Reconstructs to: PGw GwHi tɪm̌ GwLo tɪm̌ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tɪʃ́ n. ‘milk_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana tɪʃ́ OpBi tɪś OpMo tɪś OpPa tɪś OpKi - 382 *tjǎʃá n. ‘thigh_2’: Reconstructs to: PGw GwHi tjǎʃá GwLo tjǎʃá Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tɔ̀dɔ́ v. ‘slurp_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi tɔ̀dɔ́ OpMo tɔ̀dɔ́ OpPa tɔ̀dɔ́ OpKi - *tɔn v. ‘big (be), male, elder_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tɔ̀n UdYa - UdCh - Dana tɔ́n OpBi tɔ́n OpMo tɔ́n OpPa tɔ́n OpKi tɔ́n *tɔ̀ŋàs ~ twàŋgàs n. ‘chest_3’: Reconstructs to: PGw GwHi tɔ̀ŋàs GwLo twàŋgàs Komo - UdYa - 383 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tùl v. ‘dwell (live, reside)_1’: Reconstructs to: PGw GwHi tùl GwLo tùl Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tūl v. ‘short (be)_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi tūl OpMo tūl OpPa tūl OpKi tūl *tUr v. ‘long or tall (be)_1’: Reconstructs to: PKmn GwHi tǔ GwLo tǔ Komo tʊ́l UdYa túr UdCh túr Dana - OpBi - OpMo - OpPa - 384 OpKi - *tʊ̀nsʊ̀s v. ‘blame (somebody)_2’: Reconstructs to: PGw GwHi tʊ̀nzʊ̀ GwLo tʊ̀nsʊ̀s Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʊ̄rʊ̀m v. ‘scootch (move over)_5’: Reconstructs to: PGw GwHi tʊ̄rʊ̀m GwLo tʊ̄lʊ̀m Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *twī v. ‘call_4’: Reconstructs to: PGw GwHi twī GwLo twī Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dà n. ‘fat (from animals)_3’: Reconstructs to: PGw GwHi dà GwLo dà 385 Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dà n. ‘oil (organic substance)_2’: Reconstructs to: PGw GwHi dà GwLo dà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dáɟV n. ‘Dazu (S. Sudan)’: Reconstructs to: PKmn GwHi dázɔ̀ GwLo dázɔ̀ Komo dázʊ̀ UdYa dǎɟɔ̀ UdCh dǎɟɔ̀ Dana dáɟɔ̀ OpBi dádʒʊ̀ OpMo dádʒʊ̀ OpPa dádʒʊ̀ OpKi dádʒʊ̀ *dak v. ‘finish_1’: Reconstructs to: PKmn GwHi - GwLo dak Komo dàg UdYa dàkʰ UdCh dàkʰ Dana dâk OpBi dàk 386 OpMo - OpPa - OpKi - Initial /d/ did not devoice in DaOp due to tone not being L? Dana exhibits F tone which could mean dâk in Proto-Dana-Opo. *dāŋā v. ‘not know (how)_2’: Reconstructs to: PGw GwHi dāŋā GwLo dāŋā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *daʃV ~ ta(n)ʃV n. ‘snake_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo dàʃʊ́ UdYa tānʒí UdCh tāʃá Dana - OpBi - OpMo - OpPa - OpKi - It is unclear whether Komo t > d or PUd d > t. *dàwàʔ n. ‘baboon_3’: Reconstructs to: PKoUd GwHi - GwLo - Komo dàw UdYa - UdCh à+dàwà Dana - OpBi - OpMo - OpPa - 387 OpKi - *dɛǩʰ v. ‘tie (bundle)_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa dɛǩʰ UdCh dɛǩʰ Dana - OpBi - OpMo - OpPa - OpKi - *dɛm̄ɛśʼ v. ‘belch_2’: Reconstructs to: PGw GwHi dɛm̄ɛśʼ GwLo dɛm̄ɛśʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dɛŋ v. ‘choke, strangle_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana dɛŋ́ OpBi - OpMo dɛŋ̄ OpPa dɛŋ̄ OpKi dɛŋ̄ *dɛ̀ŋ v. ‘count_1’: Reconstructs to: PCtrl GwHi - GwLo - 388 Komo dɛ̀n UdYa dɛ̀ŋ+ɛ ̄ UdCh dɛ̀ŋ+ɛ ̄ Dana - OpBi - OpMo - OpPa dɛ̀n OpKi dɛ̀n Cannot account for lack of devoicing /d/ → /t/ in Opo. *dìd v. ‘dwell (live, reside)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa dìd UdCh dì Dana - OpBi - OpMo - OpPa - OpKi - *dìd v. ‘heavy (be)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo dìd UdYa dìtʰ UdCh dìtʰ Dana - OpBi - OpMo - OpPa - OpKi - *àdıj̋ɛ ̄n. ‘mother_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - 389 OpBi àdıj̋ɛ ̄ OpMo àdıj̋ɛ ̄ OpPa àdıj̋ɛ ̄ OpKi àdıj̋ɛ ̄ *dìncʼà n. ‘skin’: Reconstructs to: PCtrl GwHi - GwLo - Komo dìnsʼā UdYa - UdCh - Dana dìncʼà OpBi dìntʃʼà OpMo dìntʃʼà OpPa dìntʃʼà OpKi dìntʃʼà Cannot account for lack of expected devoicing d/ → t in Dana-Opo before L tone. *dǐpʰáɲ n. ‘flour_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa dípʰɛɲ́ UdCh à+dǐpʰáɲ Dana - OpBi - OpMo - OpPa - OpKi - *dìʃ(a) n. ‘mushroom_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo dìʃ UdYa dìʃìʔ UdCh à+dìʃ Dana tìʃà OpBi tìsà OpMo tìsà OpPa tìsà OpKi tìʃà 390 *dɪŋɪ n. ‘baboon_1, dog_2’: Reconstructs to: PKmn GwHi tɪ̀nɪ̀ GwLo tɪ̀nɪ̀ Komo - UdYa - UdCh - Dana tɛ̀ŋ OpBi tɪ̀nɪ̀ OpMo tɪ̀nɪ̀ OpPa tɪ̀n OpKi tɪ̀n Borrowing from Burun or W. Nilotic? cf. d̪ɛ̀ɛŋ ‘cow’ in Kurmuk (Andersen 2007:81) *dɪŋ́kā n. ‘fish_sp (electric)_1’: Reconstructs to: PKmn GwHi - GwLo ʊ̄+wàsʼ+dɪŋ́kā Komo dɪŋ́kā UdYa - UdCh - Dana - OpBi dıŋ̋ā OpMo dıŋ̋ā OpPa dıŋ̋ā OpKi dıŋ̋ā *djalIsʼ v. ‘stomp, step on_1’: Reconstructs to: PKmn GwHi dɛ̀lísʼ GwLo dàlísʼ Komo dɪ̀l UdYa dìl UdCh dìl Dana - OpBi tɪ̀lɪ ́ OpMo tɪ̀lɪ ́ OpPa tɪ̀lɪ ́ OpKi tɪ̀lɪ ́ *àdɔ̀j n. ‘moon or month_1’: Reconstructs to: PDaOp GwHi - GwLo - 391 Komo - UdYa - UdCh - Dana à+dɔ̀j OpBi à+dɔ̀j OpMo à+dɔ̀j OpPa à+dɔ̀j OpKi à+dɔ̀j *dɔl(ɔ) v. ‘shout_1’: Reconstructs to: PKmn GwHi dɔ́l GwLo dɔ́l Komo dɔ́l UdYa - UdCh - Dana ɔ̀lɔ́ OpBi - OpMo - OpPa - OpKi - *dɔlɔ n. ‘fish_sp (very small, scaled fish with a small rounded mouth)’: Reconstructs to: PKmn GwHi - GwLo dɔ̌lɔ́ Komo dɔ̀lɔ́ UdYa - UdCh - Dana dɔ́lɪʔ̄ OpBi dɔ̀lɔ́ OpMo dɔ̀lɔ́ OpPa dɔ̀lɔ́ OpKi dɔ́lɪ ̄ *dɔ̌lɔ́ n. ‘gourd_4’: Reconstructs to: PGw GwHi dɔ̌lɔ́ GwLo dɔ̌lɔ́ Komo - UdYa - UdCh - Dana - 392 OpBi - OpMo - OpPa - OpKi - *dɔ̀ŋ v. ‘bite (by animal)_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana dɔ̀ŋ OpBi dɔ̀ŋ OpMo dɔ̀ŋ OpPa dɔ̀ŋ OpKi dɔ̀ŋ *dɔ̀ŋgɔ̀l n. ‘chicken_2’: Reconstructs to: PGw GwHi dɔ̀ŋgɔ̀l GwLo dɔ̀ŋɪ ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dɔ̀ŋgɔ̀n n. ‘four_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo dɔ̀gɔ̀n UdYa dɔ̀ŋɔ̀n UdCh dɔ̀ŋɔ̀n Dana - OpBi - OpMo - OpPa - OpKi - 393 *dɔr v. ‘hit_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa dɔ̌r UdCh dɔ̌r Dana tɔ̀dɔ́r OpBi tɔ̀r OpMo tɔ̀r OpPa tɔ̀r OpKi tɔ̀r *dɔ̀ʃ v. ‘stand_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo dɔ̀ʃ UdYa dɔ̀ʃ UdCh dɔ̀ʃ Dana - OpBi - OpMo - OpPa - OpKi - *dɔ̌ʃ v. ‘court (v.), flirt with_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa dɔ̌ʃ UdCh dɔ̌ʃ Dana - OpBi - OpMo - OpPa - OpKi - *dɔ̀zɔ̀ v. ‘teach_2’: Reconstructs to: PGw GwHi dɔ̀zɔ̀ GwLo dɔ̀zɔ̀ Komo - UdYa - 394 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dùcʼá v. ‘urine_1, urinate_1’: Reconstructs to: PKmn GwHi dùsʼ GwLo tùsʼ Komo dʊ̀sʼ UdYa - UdCh - Dana tùcʼáʔ OpBi - OpMo - OpPa - OpKi - *dùm v. ‘pound (v.)_5’: Reconstructs to: PCtrl GwHi - GwLo - Komo dùm UdYa - UdCh - Dana dùm OpBi dùm OpMo dùm OpPa dùm OpKi dùm Cannot account for expected lack of devoicing d → /t/ in Dana-Opo with L tone. *dùmàj n. ‘tree (sp.)(sausage tree_Kigelia africana)’: Reconstructs to: PKmn GwHi - GwLo ū+dùmɪ̀ Komo dùmɛ̀ UdYa - UdCh - Dana dùmáj OpBi dùmàj OpMo dùmàj 395 OpPa dùmàj OpKi dùmàj *dǔrVcʼ n. ‘young people_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa dǔrīcʼ UdCh dǔrūcʼ Dana - OpBi - OpMo - OpPa - OpKi - *dʊ̄jʊ̄ n. ‘fist_3’: Reconstructs to: PGw GwHi dʊ̄jʊ̄ GwLo dʊ̄jʊ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dʊ́ɟɛ̀ n. ‘pipe (for smoking)_1’: Reconstructs to: PKmn GwHi dʊ́zɛ̀ GwLo dʊ́zɛ̀ Komo dʊ́zɛ̀ UdYa - UdCh - Dana dʊ́ɟɛ̀ OpBi - OpMo - OpPa - OpKi - *dV v. ‘shiver_4’: Reconstructs to: PUd GwHi - 396 GwLo - Komo - UdYa dɔ̄ UdCh dɛ ̌ Dana - OpBi - OpMo - OpPa - OpKi - *dwā n. ‘girl_3’: Reconstructs to: PGw GwHi dwà GwLo dwā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dwakʰ n. ‘bird_weaver’: Reconstructs to: PKmn GwHi ʊ̄+dwâk GwLo ʊ̄+dɔ́k Komo à+dɔ̂k UdYa - UdCh - Dana à+dwákʰ OpBi à+dwa̋kʰ OpMo à+dwa̋kʰ OpPa à+dwa̋kʰ OpKi à+dwa̋kʰ *dwákɪ̀ n. ‘hyrax_2’: Reconstructs to: PGw GwHi dwákɪ̀ GwLo dwákɪ̀ Komo - UdYa - UdCh - Dana - 397 OpBi - OpMo - OpPa - OpKi - *dwì v. ‘buy_1’: Reconstructs to: PGw GwHi dwì GwLo dwì Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *dwì v. ‘sell_1’: Reconstructs to: PGw GwHi dwì GwLo dwì Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *Dàsʼ v. ‘grind wet (first grind)’: Reconstructs to: PKmn GwHi dàsʼ GwLo dàsʼ Komo nàsʼ UdYa nàsʼ UdCh nàt ̪̓ Dana - OpBi - OpMo - OpPa - OpKi - Unsure of /n : d/ correspondence but it seems cognate. 398 *dOt v. ‘ask (inquire)_1’: Reconstructs to: PKmn GwHi tʊ̌t GwLo tʊ̌t Komo dɔ̀t UdYa dɔ̌tʰ UdCh dɔ̌tʰ Dana - OpBi tɔ̄tɔ́ OpMo tɔ̄tɔ́ OpPa - OpKi - *dwáɗ v. ‘hunt (in group)’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa wátʰ UdCh wáɗ Dana - OpBi dwàr OpMo dwàr OpPa dwàr OpKi - *Dangal v. ‘roll_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa ɗáŋgàl UdCh ɗáŋgál Dana nàŋgàl OpBi làŋgàl OpMo làŋgàl OpPa làŋgàl OpKi làŋgàl *ɗ(w)ankʼI n. ‘scorpion’: Reconstructs to: PKmn GwHi tʼwānkʼɪt́ʼwānkʼ GwLo tʼwānkʼɪt́ʼwānkʼ Komo bāɗāgíʔ 399 UdYa ɗwākʰ UdCh à+ɗākʰ Dana ɗāgí OpBi ɗāgı ̋ OpMo ɗāgı ̋ OpPa ɗāgı ̋ OpKi ɗāgı ̋ *ɗab(a) v. ‘adhere’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɗàb UdYa ɗāpʰ UdCh ɗāpʰ Dana ɗábá OpBi ɗábá OpMo ɗábá OpPa ɗábá OpKi ɗábá *ɗagi n. ‘bean_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo ɗɛ̀gí UdYa ɗàgì UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ɗAm n. ‘ladle_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo ɗɔ́m UdYa ɗám UdCh ɗám Dana - OpBi - OpMo - 400 OpPa - OpKi - *ɗar(a) v. ‘send someone_1’: Reconstructs to: PKmn GwHi tʼálà GwLo tʼájà Komo ɗàr UdYa ɗɛt̄ʰ UdCh - Dana ɗɛ̀d OpBi - OpMo ɗɛr̄ OpPa - OpKi ɗɛr̄ *ɗAs v. ‘dry out_3’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɗàʃ UdYa ɗās UdCh ɗās Dana ɗɛ̀s OpBi - OpMo - OpPa - OpKi - *ɗɛ n. ‘one (1)_1, alone_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɗɛ ́ UdYa ɗɛ ́ UdCh ɗɛ ́ Dana ɗɛ̀dɛ ́ OpBi ɗjān OpMo ɗjān OpPa ɗɛd̄ɛ ̄ OpKi ɗɛd̄ɛ ̄ Opo most likely bimorphemic ɗɛ+ān (3N suffix). 401 *ɗɛm v. ‘to stew (food)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɗɛ̀m UdYa - UdCh - Dana ɗɛ̀m OpBi ɗɛm̄ OpMo ɗɛm̄ OpPa ɗɛm̄ OpKi ɗɛm̄ *ɗiʃa v. ‘near_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɗìʃ UdYa ɗīʃ UdCh ɗīʃ Dana ɗìʃà OpBi īsā OpMo īsí OpPa ɗīsā OpKi ɗīʃā *ɗitʰ ? n. ‘bird’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɗɪ̂w UdYa ɗūwì UdCh ɗī Dana ɗītʰ OpBi ɗɪʊ̀̄ OpMo ɗɪʊ̀̄ OpPa ɗɪʊ̀̄ OpKi ɗɪʊ̀̄ *ɗɪĺá v. ‘rise (oneself)_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - 402 UdCh - Dana - OpBi - OpMo - OpPa ɗɪĺá OpKi ɗɪĺá *(tɔ)ɗɔ v. ‘carry many things_1’: Reconstructs to: PKmn GwHi tɔ̀dɔ̀ GwLo - Komo ɗɔ̀ UdYa ɗɔ̄ UdCh ɗɔ̄ Dana - OpBi ɗɔ̄ OpMo ɗɔ̀ OpPa ɗɔ̀ OpKi ɗɔ̀ *ɗɔ̄ŋgɔ̄rɔ̄ n. ‘salt_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ɗɔ̄ŋgɔ̄rɔ̄ UdCh à+ɗɔ̄ŋgɔ̄rɔ̄ Dana - OpBi - OpMo - OpPa - OpKi - *ɗúbá v. ‘tasty (be)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ɗúbá OpBi ɗűbá OpMo ɗűbá OpPa ɗűbá 403 OpKi ɗűbá *ɗwá n. ‘frog_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ɗwá UdCh à+ɗwá Dana - OpBi - OpMo - OpPa - OpKi - *tʼákʼál n. ‘tongue_2’: Reconstructs to: PGw GwHi tʼákʼál GwLo tʼákɪ ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʼápán v. ‘lick_2’: Reconstructs to: PGw GwHi tʼápán GwLo tʼápán Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʼaʃ n. ‘salt_(made from ash of a particular plant/tree)_2’: Reconstructs to: PKmn GwHi tʼāʃ GwLo tʼāʃ 404 Komo tʼàʃ UdYa tʼāʃ UdCh tʼāʃ Dana tʼàʃ OpBi tʼās OpMo - OpPa - OpKi - meaning is ‘stew/soup’ in Uduk. *tʼáʃá n. ‘mushroom_2’: Reconstructs to: PGw GwHi tʼáʃá GwLo tʼáʃá Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʼɪḱʼá v. ‘heavy (be)_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana tʼɪḱʼá OpBi tʼɪk̋ʼá OpMo tʼɪk̋ʼá OpPa tʼɪk̋ʼá OpKi tʼɪk̋ʼá *tʼɪŕá n. ‘rope_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - 405 OpBi tʼɪŕá OpMo tʼɪŕá OpPa tʼɪŕá OpKi tʼɪŕá *tʼɪʃ́ ~ tʼíʃ v. ‘grind (fine)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼɪʃ́ UdYa tʼíʃ UdCh tʼíʃ Dana tʼìʃíʔ OpBi tʼıs̋ OpMo tʼıs̋ OpPa tʼıs̋ OpKi - *tʼɔ̄pʼ v. ‘drink’: Reconstructs to: PGw GwHi tʼɔ̄pʼ GwLo tʼɔ̄pʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʼɔ̄ʃ v. ‘pain, be hurt_3’: Reconstructs to: PGw GwHi tʼɔ̄ʃ GwLo tʼɔ̄ʃ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 406 *tʼɔ́tʼɔ́ v. ‘black (be)_2’: Reconstructs to: PGw GwHi tʼɔ́tʼɔ́ GwLo tʼɔ́tʼɔ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʼɔ́zí v. ‘precede_3’: Reconstructs to: PGw GwHi tʼɔ́zí GwLo tʼɔ́zí Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʼum(á) n. ‘fist_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼùmá UdYa tʼúm+mɛ̀ɗ UdCh tʼúm+mɛ̀ɗ Dana tʼùmá OpBi tʼūmá OpMo tʼūmá OpPa tʼūmá OpKi tʼūmá *tʼʊ̀mʊ̀ n. ‘home, place_3’: Reconstructs to: PGw GwHi tʼʊ̀mʊ̀ GwLo tʼʊ́mʊ̄ Komo - UdYa - 407 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʼVʃ ~ t ̪̓Vʃ v. ‘forbid_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo tʼáʃ UdYa tʼáʃ UdCh tʼáʃ Dana - OpBi tʼɪśɪ ̄ OpMo tʼɪśɪ ̄ OpPa - OpKi - *tʼwālā v. ‘strong (be)_4’: Reconstructs to: PGw GwHi tʼwālā GwLo tʼwājā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *tʼwaŋ(k)a ~ t ̪̓waŋ(k)a v. ‘cut (split in half lengthwise)_1’: Reconstructs to: PKmn GwHi tʼāŋà GwLo tʼāŋà Komo - UdYa tʼwákʰ UdCh tʼwákʰ Dana - OpBi - OpMo - OpPa - 408 OpKi - *sʼ(j)am v. ‘sweet (be)_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼām UdYa - UdCh - Dana sʼɛḿ OpBi tʃʼɛḿ OpMo tʃʼɛḿ OpPa tʃʼɛḿ OpKi tʃʼɛḿ *sʼa v. ‘light (ignite)’: Reconstructs to: PKmn GwHi sʼā GwLo sʼā Komo sʼà UdYa sʼā UdCh t ̪̓ ā Dana sʼʊ́wà OpBi tʃʼǎ OpMo tʃʼǎ OpPa tʃʼǎ OpKi tʃʼǎ *sʼá v. ‘annoint (with oil)_1’: Reconstructs to: PGw GwHi sʼá GwLo sʼá Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sʼám v. ‘cold(be)_2, wet, sweet (be)’: Reconstructs to: PCtrl GwHi - GwLo - 409 Komo sʼám UdYa sʼám UdCh t ̪̓ám Dana sʼɛḿ OpBi tʃʼɛḿ+sɛ ̄ OpMo tʃʼɛḿ+sɛ ̄ OpPa tʃʼɛḿ+sɛ ̄ OpKi tʃʼɛḿ+sɛ ̄ Meaning is ‘wind (n.)’ in Uduk *sʼámá n. ‘blood_1’: Reconstructs to: PKmn GwHi sʼám GwLo sʼám Komo - UdYa - UdCh - Dana sʼámáʔ OpBi tʃʼámá OpMo tʃʼámá OpPa tʃʼámá OpKi tʃʼámá *sʼānsʼ n. ‘root_3’: Reconstructs to: PGw GwHi sʼānsʼ GwLo sʼānsʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sʼEU n. ‘bamboo_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼɪ̂w UdYa sʼì UdCh à+t ̪̓ ɛ ̄ Dana sʼɪ̂w 410 OpBi - OpMo - OpPa - OpKi - *sʼɛ ́v. ‘curse_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo sʼɛ ́ UdYa - UdCh tʼɛ ́ Dana - OpBi - OpMo - OpPa - OpKi - *sʼɛ̀d̪ ~ sʼɪ̀d̪ v. ‘shave’: Reconstructs to: PKmn GwHi sʼɪl̄ GwLo sʼɪ ̄ Komo sʼɛ̀ UdYa sʼī UdCh cʼɛ ̄ Dana sʼɛ̀d̪ OpBi tʃʼɛ ̄ OpMo tʃʼɛ ̄ OpPa tʃʼɛ ̄ OpKi tʃʼɛ ̄ *sʼɛḱʰ v. ‘rub hands together (e.g. to make fire using stick)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa sʼɛḱʰ UdCh t ̪̓ ɛḱʰ Dana - OpBi - OpMo - OpPa - OpKi - 411 *sʼɛ̀sʼ(ɛ̀) n. ‘termite_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼɛ̀sʼ UdYa - UdCh - Dana sʼɛ̀sʼ OpBi tʃʼɛt̄ʃʼɛ ̄ OpMo tʃʼɛt̄ʃʼɛ ̄ OpPa tʃʼɛt̄ʃʼ OpKi tʃʼɛt̄ʃʼ *sʼɛw̄àn ~ sʼjāwàn n. ‘moon or month_3’: Reconstructs to: PGw GwHi sʼɛw̄àn GwLo sʼjāwàn Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sʼí v. ‘die_2’: Reconstructs to: PGw GwHi sʼí GwLo sʼí Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sʼíd̪ v. ‘black (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼí UdYa sʼíʔ 412 UdCh t ̪̓íʔ Dana sʼíd̪ OpBi tʃʼı ̋ OpMo tʃʼı ̋ OpPa tʃʼı ̋ OpKi tʃʼı ̋ Also means ‘diarreah’ in KoUd *sʼÍk n. ‘rat_1’: Reconstructs to: PKmn GwHi sʼɪ ́ GwLo sʼɪ ́ Komo sʼík UdYa sʼíʔ UdCh à+t ̪̓íkʰ Dana - OpBi tʃʼıg̋ı ̋ OpMo tʃʼıg̋ı ̋ OpPa tʃʼıg̋ı ̋ OpKi tʃʼıg̋ı ̋ *sʼíl(ít⁽ʰ⁾) n. ‘thatch’: Reconstructs to: PKoUd GwHi - GwLo - Komo sʼílít UdYa sʼíl UdCh t ̪̓íl Dana - OpBi - OpMo - OpPa - OpKi - *sʼOpʰ v. ‘light (the way with torch/flashlight)_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼɔ̀p UdYa - UdCh - Dana sʼʊ̄pʰ OpBi tʃʼʊ̄ OpMo tʃʼʊ̄ 413 OpPa tʃʼʊ̄ OpKi tʃʼʊ̄ *sʼɔ̄ v. ‘pray, beg_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa sʼɔ̄ UdCh t ̪̓ ɔ̄ Dana - OpBi - OpMo - OpPa - OpKi - *sʼú v. ‘plait or braid or weave_3’: Reconstructs to: PGw GwHi sʼú GwLo sʼú Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sʼú v. ‘rub hands together (e.g. to make fire using stick)_1’: Reconstructs to: PGw GwHi sʼú GwLo sʼú Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 414 *sʼUB(V)(n) v. ‘dip food in sauce with fingers_1’: Reconstructs to: PKmn GwHi - GwLo sʼʊ́pʊ́n Komo sʼūb UdYa sʼúpʰ UdCh t ̪̓úpʰ Dana sʼúbá OpBi - OpMo - OpPa - OpKi - *sʼUDI v. ‘defecate_1, diarrea’: Reconstructs to: PKmn GwHi úsʼ GwLo úsʼ Komo sʼɪ ́ UdYa - UdCh t ̪̓í Dana sʼʊ̀dʊ́ OpBi tʃʼʊ̄jɪ ́ OpMo tʃʼʊ̄jɪ ́ OpPa tʃʼʊ̄jɪ ́ OpKi tʃʼʊ̄jɪ ́ *sʼuɗ v. ‘kiss_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼùr UdYa - UdCh - Dana sʼúɗ OpBi tʃʼűr OpMo tʃʼűr OpPa tʃʼűr OpKi tʃʼűr *sʼunsʼ v. ‘bite_2’: Reconstructs to: PGw GwHi sūnsʼ GwLo sʼṹnsʼ Komo - UdYa - 415 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *SʼUn(tʼ)a n. ‘nosebleed_2’: Reconstructs to: PKmn GwHi - GwLo tʼʊ̀ntʼ Komo - UdYa - UdCh - Dana sʼùnáʔ OpBi - OpMo - OpPa - OpKi - *sʼúpʼ v. ‘suck_2’: Reconstructs to: PGw GwHi sʼúpʼ GwLo sʼúpʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sʼʊ́p v. ‘cold (be)_1’: Reconstructs to: PKmn GwHi sʼʊ́p GwLo sʼʊ́p Komo - UdYa sʼúpʰ UdCh t ̪̓úpʰ Dana - OpBi - OpMo - OpPa - 416 OpKi - *sʼʊ́pʊ́n v. ‘kiss_2’: Reconstructs to: PGw GwHi sʼʊ́pʊ́n GwLo sʼʊ́pʊ́n Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sʼʊ̄pʼ n. ‘breast, milk_2’: Reconstructs to: PGw GwHi sʼʊ̄pʼ GwLo sʼʊ̄pʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sʼʊ́sʼ v. ‘scratch_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo sʼʊ́sʼ UdYa sʼúsʼ UdCh t ̪̓út ̪h Dana - OpBi - OpMo - OpPa - OpKi - *sʼʊ́sʼʊ́mɔ̀ n. ‘calf of leg’: Reconstructs to: PGw GwHi sʼʊ́sʼʊ́mɔ̀ GwLo sʼʊ́sʼʊ́mɔ̀ 417 Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *c(w)ā v. ‘big (be)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo swà UdYa cā UdCh cā Dana - OpBi - OpMo - OpPa - OpKi - *caja adv. ‘very, many_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sà UdYa - UdCh cājá Dana - OpBi sáj OpMo sáj OpPa sáj OpKi sáj Cannot account for /s/ in Opo, would expect /tʃ/. *càkʰO n. ‘grandfather_1’: Reconstructs to: PKmn GwHi - GwLo sàkʊ́ Komo sàkʊ́ UdYa - UdCh - Dana - 418 OpBi tʃàkʰɔ́ OpMo tʃàkʰɔ́ OpPa tʃàkʰɔ́ OpKi tʃàkʰɔ́ *cān v. ‘poor (be)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana cān OpBi tʃān OpMo tʃān OpPa tʃān OpKi tʃān *cɛŋ v. ‘curse_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana cɛŋ́ OpBi tʃɛŋ̄ OpMo tʃɛŋ̄ OpPa tʃɛŋ̄ OpKi tʃɛŋ̄ *cɛʃ̄ v. ‘protect_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa cɛʃ̄ UdCh cɛʃ̄ Dana - OpBi - OpMo - OpPa - OpKi - 419 *cikʼa ~ cɪkʼa v. ‘listen_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sɪ̀g UdYa ʃīkʼ UdCh cīkʰ Dana sʼìkʼà OpBi - OpMo - OpPa - OpKi - In Komo, final *kʼ → [g] /V_V then loss of final vowel. Dana later glottalizes initial /s/ → /sʼ/ to harmonize with glottal /kʼ/? *Cisʼ v. ‘warm up (sth.)’: Reconstructs to: PKmn GwHi ʃɪʃ̄ GwLo ʃɪʃ̄ Komo jíz UdYa jísʼ UdCh jít ̪h Dana hízá OpBi ıs̋á OpMo ıs̋á OpPa ıs̋á OpKi - Gwama may not be cognate. *cɔkʰ v. ‘sit_3’: Reconstructs to: PKoUd GwHi - GwLo - Komo sɔ̀k UdYa - UdCh cɔ́kʰ Dana - OpBi - OpMo - OpPa - OpKi - *cɔ̀m n. ‘father_2’: Reconstructs to: PCtrl GwHi - 420 GwLo - Komo sɔ̀m UdYa à+cɔ̄m UdCh cɔ̄m Dana sɔ̀m OpBi tʃɔ̀mɔ̀ OpMo tʃɔ̀mɔ̀ OpPa tʃɔ̀m OpKi tʃɔ̀m Initial /s/ in Dana due to contact with Komo? *cúl n. ‘firefly_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa - UdCh à+cúl Dana - OpBi - OpMo - OpPa - OpKi - *cut v. ‘whistle’: Reconstructs to: PKmn GwHi - GwLo ʃút Komo ʃɪt́ UdYa - UdCh cúwā Dana cʊ́ɪ̀ OpBi tʃúwì OpMo tʃúwì OpPa tʃúwì OpKi tʃúwì *cʊ̄k⁽ʰ⁾(ɪ) n. ‘hoof’: Reconstructs to: PKoUd GwHi - GwLo - Komo sʊ̄kɪ ́ UdYa - UdCh à+cūkʰ 421 Dana - OpBi - OpMo - OpPa - OpKi - *CVkʻUm v. ‘rinse mouth_1’: Reconstructs to: PKmn GwHi ʃʊ̀kʊ́m GwLo ʃʊ̀kʊ́m Komo zùkʼúm UdYa - UdCh - Dana ɟákʼʊ́m OpBi tʃākʼʊ́má OpMo tʃākʼʊ́má OpPa sākʼʊ́má OpKi tʃākʼʊ́má Cannot account for voiced initial consonant in Komo and Dana. *cwálá n. ‘tree’: Reconstructs to: PKmn GwHi swálá GwLo swájá Komo sá UdYa ʃwá UdCh cwá Dana cá OpBi tʃá OpMo tʃá OpPa tʃá OpKi tʃá *tʃápūm n. ‘waist, hip_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi tʃápūm OpMo tʃápūm OpPa - 422 OpKi - *tʃìm v. ‘bury_4’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi tʃìm OpMo tʃìm OpPa - OpKi - *tʃɔ̄t v. ‘pierce_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi tʃɔ̄t OpMo tʃɔ̄t OpPa - OpKi - *dʒìbàj n. ‘cloud, fog_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi dʒìbàj OpMo dʒìbàj OpPa dʒìbàj OpKi dʒìbàj *ɟà v. ‘dig_1’: Reconstructs to: PKmn GwHi ʃà GwLo ʃà 423 Komo - UdYa - UdCh - Dana cà OpBi tʃà OpMo tʃà OpPa tʃà OpKi tʃà *ɟàn n. ‘dream_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa zàn UdCh ɟàn Dana - OpBi - OpMo - OpPa - OpKi - *ɟana n. ‘sorghum, millet’: Reconstructs to: PKmn GwHi sjànà GwLo sjànà Komo zɛ̀nā UdYa - UdCh - Dana - OpBi dʒɛ̀ná OpMo - OpPa dʒɛ̀ná OpKi - *ɟaŋg(ɔ)aj n. ‘Nuer (ethnonym)_1’: Reconstructs to: PKmn GwHi zǎgʊ́ GwLo zǎgʊ́ Komo zǎgʊ́ UdYa ʒwǎŋgì UdCh - Dana ɟàŋwɛ̀j OpBi dʒāŋɔ́ 424 OpMo - OpPa zāŋwɛ ́ OpKi ʃàŋwɛ̀ *ɟarɛ n. ‘baboon_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ɟārɛ ̄ OpBi tʃàrɛ̀ OpMo tʃàrɛ̀ OpPa tʃàrɛ̀ OpKi tʃàrì *ɟàrú n. ‘bird_stork (maribou)’: Reconstructs to: PKmn GwHi zɛ̀rú GwLo zɛ̀rú Komo zɛ̀rú UdYa ʒàrú UdCh - Dana ɟɛ̀rú OpBi dʒɛ̀rű OpMo dʒɛ̀rű OpPa dʒɛ̀rű OpKi dʒɛ̀rű *ɟɛ̀ n. ‘elephant_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa - UdCh ɟɛ̀ Dana - OpBi - OpMo - OpPa - OpKi - 425 *ɟìbí v. ‘slurp_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo zìbí UdYa - UdCh - Dana ɟìbí OpBi - OpMo - OpPa - OpKi tʃibi *ɟíkʰì v. ‘forbid_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ɟíkʰì OpBi - OpMo - OpPa zɪk̄ʰɪ ́ OpKi ʃɪk̄ʰɪ ́ *ɟVtʰid n. ‘sweat (substance)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo zàtít UdYa ɟìtʰíd UdCh ɟìtʰí Dana - OpBi - OpMo - OpPa - OpKi - *cʼ(w)ɛs̄ʼ v. ‘tear (shred)_3’: Reconstructs to: PKmn GwHi sʼwɛ̃ ̌ GwLo sʼwɛ̃ ̌ Komo - UdYa - 426 UdCh cʼɛt̄ ̪̓ Dana - OpBi - OpMo - OpPa - OpKi - *cʼacʼ n. ‘chest_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼàsʼ UdYa - UdCh - Dana cʼàcʼ OpBi tʃʼātʃʼ OpMo tʃʼātʃʼ OpPa tʃʼātʃʼ OpKi sʼāsʼ *cʼákʼúmú n. ‘ladle_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana cʼákʼúmú OpBi tʃʼákʼűmű OpMo tʃʼákʼűmű OpPa tʃʼákʼűmű OpKi tʃʼákʼűmű *cʼɛ n. ‘ear’: Reconstructs to: PKmn GwHi sʼɛ ̄ GwLo sʼɛ ̄ Komo sʼɛ ̄ UdYa ʃʼɛ ́ UdCh cʼɛ ́ Dana kʼɛ ̄ OpBi tʃʼɛ̀ OpMo tʃʼɛ̀ OpPa tʃʼɛ̀ 427 OpKi tʃʼɛ̀ Cannot account for the initial /kʼ/ in Dana. *cʼɛk̄ʼ n. ‘termite_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa cʼɛk̄ʼ UdCh à+cʼɛk̄ʼ Dana - OpBi - OpMo - OpPa - OpKi - *cʼɛʃ́ n. ‘earth, soil, ground, floor_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa cʼɛʃ́ UdCh à+cʼɛʃ́ Dana - OpBi - OpMo - OpPa - OpKi - *à+cʼí n. ‘child_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa àɟí UdCh à+cʼí Dana - OpBi - OpMo - OpPa - OpKi - 428 *cʼīsân v. ‘sneeze_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ɟìsân UdCh cʼīsân Dana - OpBi - OpMo - OpPa - OpKi - *cʼɪkʼ v. ‘sour (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼɪk̄ʼ UdYa - UdCh - Dana cʼɪ̀kʼ OpBi tʃʼɪk̄ʼ+dʒɛ̀ OpMo tʃʼɪk̄ʼ+dʒɛ̀ OpPa tʃʼɪk̄ʼ+zɛ̀ OpKi tʃʼɪk̄ʼ+sɛ̀ *cʼɪḱʼɪŕ n. ‘fingernail, toenail, claw, hoof_3’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼɪḱʼɪĺ UdYa - UdCh - Dana kʼɪḱʼɪŕ OpBi tʃʼɪḱʼɪŕ OpMo tʃʼɪḱʼɪŕ OpPa tʃʼɪḱʼɪŕ OpKi tʃʼɪḱʼɪŕ Dana initial /kʼ/ instaed of expected /cʼ/ cannot be accounted for (cf. ‘ear’ in Dana). *cʼɔ(tʼɔ)l v. ‘drip (fall in globules)_1’: Reconstructs to: PKmn GwHi sʼɔ́tʼɔ́ GwLo sʼɔ́tʼɔ́ Komo sʼʊ̀lɪl̄ 429 UdYa - UdCh cʼɔ̄cʼɔ̄lɔ̄cʼ Dana cʼɔ̀ʔ OpBi tʃʼɔ̄ OpMo tʃʼɔ̄ OpPa tʃʼɔ̄ OpKi tʃʼɔ̄ *cʼúg v. ‘untie, undress_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼúk UdYa - UdCh cʼúkʰ Dana - OpBi tʃʼűg OpMo tʃʼűg OpPa tʃʼűg OpKi tʃʼűg *cʼúm~cʼúm v. ‘suck_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ʃʼúʒùm UdCh cʼúɲcʼúm Dana - OpBi - OpMo - OpPa - OpKi - *cʼVmaj n. ‘dregs’: Reconstructs to: PCtrl GwHi - GwLo - Komo sʼímá UdYa ʃʼɔ́má UdCh à+cʼúmá Dana cʼɪḿáj OpBi tʃʼım̋áj OpMo tʃʼım̋áj 430 OpPa tʃʼım̋áj OpKi tʃʼım̋áj *kʰā n. ‘chicken_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kʰā OpBi kʰā OpMo kʰā OpPa kʰā OpKi kʰā *kʰab n. ‘wind (n.)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa - UdCh hāpāpʰ Dana kʰàb OpBi kʰāb OpMo kʰāb OpPa kʰāb OpKi kʰāb *kʰaɓ v. ‘repair_1’: Reconstructs to: PKmn GwHi áp GwLo - Komo áb UdYa - UdCh - Dana ʊ́bā OpBi kʰápʼ OpMo kʰápʼ OpPa kʰápʼ OpKi kʰápʼ *kʰacʼ v. ‘shut_2’: Reconstructs to: PKmn GwHi kǎʃ 431 GwLo kǎʃ Komo - UdYa kʰáʃʼ UdCh kʰácʼ Dana - OpBi - OpMo - OpPa - OpKi - *kʰáɗ(a) v. ‘open’: Reconstructs to: PKmn GwHi kálá GwLo kájá Komo kár UdYa kʰǎd UdCh kʰǎɗ Dana kʰátá OpBi kʰátá OpMo kʰátá OpPa kʰátá OpKi kʰátá *kʰáɗúm n. ‘roof_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kárúm UdYa - UdCh - Dana kʰáɗúm OpBi kʰa̋rűm OpMo kʰa̋rűm OpPa kʰa̋rűm OpKi kʰa̋rűm *kʰāg v. ‘leave_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kʰāgɪ+́d̪ɛ̀ 432 OpBi - OpMo kʰāgɪ+́dʒɪ̀ OpPa kʰāgɪ+́zɪ̀ OpKi kʰāgɪ+́ʃɪ̀ *kʰāgá v. ‘hot (be)_4’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʰāgá OpMo kʰāgá OpPa kʰāgá OpKi - *kʰaj v. ‘herd (v.)’: Reconstructs to: PKmn GwHi kɛ̃ʔ̄ GwLo kɛ̃ʔ̄ Komo kàʔí UdYa - UdCh - Dana kʰàjɪʔ́ OpBi kʰājɪ ́ OpMo kʰājɪ ́ OpPa kʰājɪ ́ OpKi kʰājɪ ́ Lowland Gwama meaning is ‘herd (of animals)’ . *kʰakʼa v. ‘bitter, sour (be)’: Reconstructs to: PKmn GwHi kāgā GwLo kāgā Komo kàʔ UdYa kʰāʔ UdCh kʰāʔ Dana kʰàkʼà OpBi kʰākʼā OpMo kʰākʼā OpPa kʰākʼā OpKi kʰākʼā 433 *kʰal v. ‘bring_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kà-ʊ̄ UdYa kʰál+í UdCh kʰál+ú Dana kùjí OpBi - OpMo - OpPa - OpKi - Possibly borrowed from W.Nilotic (cf. Kurmuk /kal/ ‘steal’) *kʰal v. ‘carry_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kàj UdYa kʰál UdCh kʰál Dana kʰál OpBi kʰál OpMo kʰál OpPa kʰál OpKi kʰál *kʰál v. ‘carry on head_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʰál UdCh kʰál Dana - OpBi - OpMo - OpPa - OpKi - *kʰālí n. ‘sheep_1’: Reconstructs to: PCtrl GwHi - GwLo - 434 Komo kālí UdYa - UdCh - Dana kʰālí OpBi kʰālı ̋ OpMo kʰālı ̋ OpPa kʰālı ̋ OpKi kʰālı ̋ *kʰàɲ v. ‘light (the way with torch/flashlight)_1’: Reconstructs to: PKmn GwHi kɪn̄ GwLo kɪn̄ Komo - UdYa kʰāɲ UdCh kʰāɲ Dana - OpBi - OpMo - OpPa - OpKi - *kʰáŋgá n. ‘cannabis_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kʰáŋgá OpBi kʰáŋgá OpMo kʰáŋgá OpPa kʰáŋgá OpKi kʰáŋgá *kʰāpʰā v. ‘red (be)_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kʰāpʰā OpBi kʰāpā 435 OpMo kʰāpā OpPa kʰāpā OpKi kʰāpā *kʰar v. ‘pour_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kál UdYa - UdCh - Dana kʰár OpBi kʰārɪ ́ OpMo kʰārɪ ́ OpPa kʰārɪ ́ OpKi kʰārɪ ́ *kʰɛ ́v. ‘arrive_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo kɛ ́ UdYa cʰɛ ́ UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʰɛb̄ɛ ̄n. ‘vagina_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʰɛb̄ɛ ̄ OpMo kʰɛb̄ɛ ̄ OpPa kʰɛb̄ɛ ̄ OpKi kʰɛb̄ɛ ̄ *kʰɛm v. ‘pierce, cut (grass with sickle)’: Reconstructs to: PCtrl 436 GwHi - GwLo - Komo kɛḿ UdYa - UdCh - Dana - OpBi kʰɛ̀m OpMo kʰɛ̀m OpPa kʰɛ̀m OpKi kʰɛ̀m *kʰīd̪í n. ‘rat_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kʰīd̪í OpBi kʰıʔ̋ OpMo kʰıʔ̋ OpPa kʰıʔ̋ OpKi kʰıʔ̋ *kʰɪ ́v. ‘give’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɪ+̌k UdYa cʰíʔ UdCh cʰí Dana kʰɪ+́wā OpBi kʰɪʔ́ OpMo kʰɪʔ́ OpPa kʰɪʔ́ OpKi kʰɪʔ́ *kʰɪ̀l n. ‘root_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - 437 Dana - OpBi kʰɪ̀l OpMo kʰɪ̀l OpPa kʰɪ̀l OpKi kʰɪ̀l *kʰɪśʼ v. ‘burn_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɪśʼ UdYa cʰísʼ UdCh cʰít ̪̓ Dana kʰísʼā OpBi kʰɪt̋ʃʼā OpMo kʰɪt́ʃʼā OpPa kʰɪt́ʃʼā OpKi kʰɪt́ʃʼā *kʰO(r)nOn v. ‘snore’: Reconstructs to: PKmn GwHi kɔ̀nɔ̀n GwLo kɔ̀nɔ̀n Komo kɔ̀nɔ̀n UdYa kʰūnūn UdCh àkʰɔ̄rnɛʔ̄ Dana kʰʊ̀nʊ̀n OpBi kʰʊ̀nʊ̀n OpMo kʰʊ̀nʊ̀n OpPa kʰʊ̀nʊ̀n OpKi kʰʊ̀nʊ̀n *kʰOba n. ‘chair_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kʊ́bá UdYa - UdCh - Dana - OpBi kʰɔ́bá OpMo kʰɔ́bá OpPa kʰɔ́bá OpKi kʰɔ́bá 438 *kʰOG n. ‘giraffe_2’: Reconstructs to: PKmn GwHi kɔ́k GwLo kɔ́k Komo - UdYa kʰúʔ UdCh kʰúʔ Dana kʰɛǵ OpBi - OpMo - OpPa - OpKi kɛ̂k Related to Proto-Nilotic *ko-RI ‘giraffe’ (Dimmendaal 1988:40)? *kʰɔ́ n. ‘bad (be)_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʰɔ́ OpMo kʰɔ́ OpPa kʰɔ́ OpKi kʰɔ́ *kʰɔba n. ‘maize_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa kʰɔ́bà UdCh à+kʰɔ́bà Dana kʰɔ́bā OpBi kʰɔ̀bà OpMo kʰɔ̀bà OpPa kʰɔ̀bà OpKi kʰɔ̀bà *kʰɔɓ v. ‘bale out (water)’: Reconstructs to: PCtrl GwHi - GwLo - 439 Komo kɔ́p UdYa kʰɔ̄b UdCh kʰɔ̄ɓ Dana kʰɔ́pʰ OpBi kʰɔ́pʰá OpMo kʰɔ́pʰá OpPa kʰɔ́pʰá OpKi kʰɔ́pʰá *kʰɔm v. ‘pound (v.)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kʰɔ̀m OpBi - OpMo - OpPa kʰɔ́m OpKi kʰɔ́m *kʰɔ̀m v. ‘follow_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʰɔ̀m OpMo kʰɔ̀m OpPa kʰɔ̀m OpKi kʰɔ̀m *kʰɔr v. ‘stumble_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɔ̀r+ɔ̀ʃ UdYa - UdCh - Dana kʰɔ̀kʰɔ̀r OpBi kʰɔ̄kʰɔ̄r 440 OpMo kʰɔ̄kʰɔ̄r OpPa kʰɔ̄kʰɔ̄r OpKi kʰɔ̄kʰɔ̄r *kʰɔ́T v. ‘to prepare earth for farming_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɔ́t UdYa kʰɔ́d UdCh kʰɔ́r Dana kʰɔ́tʰ OpBi kʰɔ̄tɔ́ OpMo kʰɔ̄tɔ́ OpPa kʰɔ́t OpKi kʰɔ́t *kʰúɲ v. ‘dig_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʰúɲ UdCh kʰǔɲ Dana - OpBi - OpMo - OpPa - OpKi - *kʰūr n. ‘skin, hide (of animal), bark of tree_5’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa - UdCh kʰūr Dana - OpBi - OpMo - OpPa - OpKi - 441 *kʰuʃ v. ‘smoke out (e.g an animal out of a hole)_1’: Reconstructs to: PKmn GwHi kʊ̌ʃ GwLo kʊ̌ʃ Komo ūs UdYa - UdCh - Dana húʃ OpBi űs OpMo űs OpPa űs OpKi űʃ Also means ‘evaporate’ in Dana. *kʰʊ́sʼ v. ‘dry (be)’: Reconstructs to: PKmn GwHi kúsʼ GwLo kúsʼ Komo kʊ́sʼ UdYa kʰúsʼ UdCh kʰút ̪̓ Dana kʰʊ́sʼ OpBi kʰʊ́tʃʼʊ́ OpMo kʰʊ́tʃʼʊ́ OpPa kʰʊ́tʃʼʊ́ OpKi - *kʰwaGʼ v. ‘fear (be afraid)’: Reconstructs to: PKmn GwHi kwāgà GwLo kwāgà Komo kɔ̄g UdYa kʰɔ̄ʔ UdCh kʰɔ̄kʼ Dana kʰɔ́k ~ kʰɔ̀gɔ́ OpBi kʰɔ̄gɔ́ OpMo kʰɔ̄gɔ́ OpPa kʰɔ̄gɔ́ OpKi kʰɔ̄gɔ́ *kʰwal v. ‘steal_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo kwál 442 UdYa kʰwāl UdCh kʰwāl Dana kʰwàlà OpBi - OpMo - OpPa - OpKi kʰwālā *kʰwālàɲ n. ‘fat (from animals)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʰwālàɲ UdCh kʰwālàɲ Dana - OpBi - OpMo - OpPa - OpKi - *kʰwāʃkʰām n. ‘turtle_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʰwāʃkʰām UdCh à+kʰwāʃkʰām Dana - OpBi - OpMo - OpPa - OpKi - *k⁽ʰ⁾a v. ‘refuse_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kâ OpBi kʰà OpMo kʰà 443 OpPa kʰà OpKi kʰà *k⁽ʰ⁾a(ŋ)k⁽ʼ⁾a v. ‘thorn, sharp (be)’: Reconstructs to: PKmn GwHi kāŋà GwLo kāŋà Komo kákā UdYa kāʔ UdCh kǎkā Dana kʰákʰāʔ OpBi kʰákʼā OpMo kʰákʼā OpPa kʰákʼā OpKi kʰákʼā *k⁽ʰ⁾akʼas ~ k⁽ʰ⁾asakʼ n. ‘porcupine_1’: Reconstructs to: PKmn GwHi kákʼàʃ GwLo kákʼàʃ Komo káʃākʼ UdYa - UdCh - Dana kásākʼ OpBi kʰásākʼ OpMo kʰásākʼ OpPa kʰásākʼ OpKi kʰáʃākʼ *k⁽ʰ⁾ɪd̪ n. ‘horn (anatomy)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɪ ̄ UdYa cú UdCh cɛ ́ Dana kɪd̪̄ OpBi kʰɪw̄ OpMo kʰɪw̄ OpPa kʰɪw̄ OpKi kʰɪw̄ 444 *k⁽ʰ⁾ɪńáj n. ‘Opo (ethnonym)_1’: Reconstructs to: PKmn GwHi kɪńá GwLo kɪńá Komo kɪńá UdYa cʰínáj UdCh - Dana kɪńáj OpBi - OpMo - OpPa - OpKi - *k⁽ʰ⁾ír(a) v. ‘tear (shred)_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kír OpBi kʰírá OpMo kʰírá OpPa kʰírá OpKi kʰírá *k⁽ʰ⁾ɪ̀s adv. ‘new_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kɪ̀s OpBi kʰɪ̀s OpMo kʰɪ̀s OpPa kʰɪ̀s OpKi kʰɪ̀s *k⁽ʰ⁾ɪs n. ‘tree_sp. (mahogany, Trichilia emetica)’: Reconstructs to: PKmn GwHi kɪśɪ̀ GwLo ʊ̄+kɪs̄ Komo kɪs̄ UdYa cɛś 445 UdCh cʰís Dana kɪś OpBi kʰɪś OpMo kʰɪś OpPa kʰɪś OpKi kʰɪś *k⁽ʰ⁾ísʼ ~ kʼísʼ v*. ‘cut_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa cʼísʼ UdCh cʼít ̪h Dana kʰísʼ OpBi kʰıt̋ʃʼ OpMo kʰıt̋ʃʼ OpPa kʰıt̋ʃʼ OpKi kʰıt̋ʃʼ Meaning is ‘sting’ in Dana-Opo. Possibly not cognate with Uduk. *k⁽ʰ⁾ɪ̀t⁽̪ʰ⁾(à) adv. ‘right (direction)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kɪ̀t ̪h à OpBi kʰɪt̄ OpMo kʰɪt̄ OpPa kʰɪt̄ OpKi kɪt̄ā *k⁽ʰ⁾O v. ‘say_1’: Reconstructs to: PKmn GwHi kɔ̄ GwLo kɔ̄ Komo ʊ́ UdYa ɔ́ UdCh ɔ́ Dana - OpBi - OpMo - 446 OpPa - OpKi - *k⁽ʰ⁾Orɛcʼ n. ‘hoe (n.)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa kūrɛ ́ UdCh kʰɔ̄rɛ̂cʼ Dana kɔ̄rɛ ́ OpBi kɔ̄rɛ ́ OpMo kɔ̄rɛ ́ OpPa kɔ̄rɛ ́ OpKi kɔ̄rɛ ́ *k⁽ʰ⁾wal ~ kʼwal v*. ‘want_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kʼwàl UdYa - UdCh - Dana kʰwāj OpBi kwāj OpMo kʰwáj OpPa kʰwáj OpKi kʰwáj *kájá n. ‘sun_2’: Reconstructs to: PGw GwHi kájá GwLo kájá Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 447 *kājā n. ‘day (24 hours)_2’: Reconstructs to: PGw GwHi kājā GwLo kājā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kál v. ‘bypass_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kāl UdYa kál UdCh kál Dana kál OpBi kál OpMo kál OpPa kál OpKi kál *kàn v. ‘bury_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo kàn UdYa kān UdCh kān Dana kànà OpBi - OpMo - OpPa - OpKi - *kānā n. ‘dog_1’: Reconstructs to: PGw GwHi kānā GwLo kānā Komo - UdYa - 448 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kāná+wɔ̀nɛ̀ v. ‘cross legs_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kāná+wɔ̀nɛ̀ OpMo kāná+wɔ̀n OpPa kāná+wɔ̀n OpKi kāná+wɔ̀n *kānsʼ v. ‘pelt_2’: Reconstructs to: PGw GwHi kānsʼ GwLo kānsʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kāp v. ‘bring_3’: Reconstructs to: PGw GwHi kāp GwLo kāp Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 449 OpKi - *kàrà v. ‘listen_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kàrà OpMo - OpPa kàrà OpKi kàrà *kàrɔ̀ n. ‘salt_(from ash of a particular plant/tree)_1’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi - OpMo kàrɔ̀ OpPa kàrɔ̀ OpKi kàrɔ̀ *(dʒì+)kārɔ̄ n. ‘soup_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi - OpMo dʒì+kārɔ̄ OpPa zì+kārɔ̄ OpKi ʃì+kārɔ̄ *kàs v. ‘forbid_2’: Reconstructs to: PGw GwHi kàs GwLo kɪ̀s 450 Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kàsʼmǎná n. ‘hyena_1’: Reconstructs to: PGw GwHi kàsʼmǎná GwLo kàsʼmǎná Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kát⁽̪ʰ⁾ v. ‘protect_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kát ̪h OpBi - OpMo - OpPa kát OpKi kát *kāw v. ‘strong (be)_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kāw 451 OpMo kāw OpPa kāw OpKi kāw *kEɲ n. ‘bird_(cattle egret)’: Reconstructs to: PKmn GwHi kɪl̄ GwLo kɪl̄ Komo à+kɪĺ UdYa cɛćɛńā UdCh à+cɛɲ́ Dana à+kɪĺ OpBi à+kɪĺ OpMo à+kɪĺ OpPa à+kɪĺ OpKi à+kɪĺ *kɛ ̌v. ‘sweep_4’: Reconstructs to: PGw GwHi kɛ ̌ GwLo kɛ ̌ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɛńtɛʃ́ v. ‘stumble_2’: Reconstructs to: PGw GwHi kɛt́ɛʃ́ GwLo kɛńtɛʃ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɛńzɛ ́v. ‘offend_2’: Reconstructs to: PGw 452 GwHi kɛńzɛ ́ GwLo kɛńzɛ ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɛś(ɛ)́ v. ‘roast or fry_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɛʃ́ UdYa cɛś UdCh cɛś Dana kɛś OpBi kɛśɛ ́ OpMo kɛśɛ ́ OpPa kɛś OpKi kɛśɛ ́ *kɛʃ̄ v. ‘curse_3’: Reconstructs to: PGw GwHi kɛʃ̄ GwLo kɛʃ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʰàkʰá n. ‘grandmother_1’: Reconstructs to: PKmn GwHi - GwLo kàkáʔ Komo kàká UdYa - UdCh - 453 Dana kʰàkʰá OpBi kàká OpMo kàká OpPa kàká OpKi kàká Unclear as to why it’s not voiced in Komo if initial was b and L tone. *kija n. ‘bed_3’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa cī UdCh à+cī Dana kījā OpBi - OpMo - OpPa - OpKi - *kíl v. ‘sharpen_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kíl UdYa cíl UdCh cíl Dana - OpBi kıl̋ OpMo kıl̋ OpPa kıl̋ OpKi kıl̋ *kiʃ ~ kɪʃ n*. ‘antelope_gazelle’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɪʃ̄ UdYa cíʃ UdCh à+cíʃ Dana àkīʃ OpBi kʰīs OpMo kʰīs OpPa kʰīs 454 OpKi kīʃ Aspiration likely a POp innovation. *kíw v. ‘shout_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa cú UdCh cú Dana - OpBi kıw̋ OpMo kıw̋ OpPa kıw̋ OpKi kıw̋ *kɪl n. ‘star_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo kɪľ UdYa cúl UdCh à+cúl Dana - OpBi - OpMo - OpPa - OpKi - *kɪm(V) n. ‘day (24 hours)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɪḿɪ ́ UdYa cím UdCh à+cím Dana kɪ̀mà OpBi - OpMo - OpPa - OpKi - *kɪňtʼ n. ‘dew_3’: Reconstructs to: PGw GwHi kɪňtʼ 455 GwLo kɪňtʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɪs̄ɪ ̄v. ‘near_2’: Reconstructs to: PGw GwHi kɪs̄ɪ ̄ GwLo kɪs̄ɪ ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɪs̄ʼ v. ‘set (heavenly bodies)_2’: Reconstructs to: PGw GwHi kɪs̄ʼ GwLo kɪs̄ʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɪʃ̄ v. ‘spoil (become spoiled)_3’: Reconstructs to: PGw GwHi kɪʃ̄ GwLo kɪʃ̄ Komo - UdYa - UdCh - Dana - 456 OpBi - OpMo - OpPa - OpKi - *kjana adv. ‘tomorrow_1’: Reconstructs to: PKmn GwHi - GwLo gɪ̀+kjānā Komo - UdYa - UdCh - Dana - OpBi dʒì+kɛń OpMo dʒì+kɛń OpPa zì+kɛń OpKi ʃì+kɛń *kjankʼa v. ‘cluck (of hen)’: Reconstructs to: PKmn GwHi kǎkā GwLo kjāŋkʼā Komo kágá UdYa - UdCh - Dana kágà OpBi kɛn̄ OpMo kɛn̄ OpPa kɛn̄ OpKi kɛn̄ *kO(j) v. ‘cry’: Reconstructs to: PKmn GwHi kū GwLo kū Komo kɔ̀ UdYa kɔ̄ UdCh kɔ̄ Dana kɔ̀j OpBi kwɛ ̄ OpMo kwɛ ̄ OpPa kwɛ ̄ OpKi kwɛ ̄ 457 *kOpʰ v. ‘carry on head_1’: Reconstructs to: PKmn GwHi kǔ GwLo kǔ Komo - UdYa - UdCh - Dana kɔ̄pʰ OpBi kɔ̄p OpMo kɔ̄p OpPa kɔ̄p OpKi kɔ̄p *kOr n. ‘chief_1’: Reconstructs to: PKmn GwHi ʊ̄+kʊ̄l GwLo ʊ̄+kwɪ̀ Komo jī+kwɪ̀ UdYa - UdCh - Dana jɛ̀+kɔ̄rɔ̄ OpBi - OpMo - OpPa - OpKi - *kOr > kOl > kOj > kwɪ *dʒɪňɪ+́kɔ̄rɛ ́n. ‘sesame_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi dʒɪňɪ+́kɔ̄rɛ ́ OpMo dʒɪňɪ+́kɔ̄rɛ ́ OpPa zɪn̄ɪ+́kɔ̄rɛ ́ OpKi - *kOta̪n n. ‘head pad (for head carrying)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - 458 UdYa - UdCh - Dana kʊ̀t ̪h áj OpBi kɔ̄tìn OpMo kɔ̄tìn OpPa kɔ̄tìn OpKi - Dana aspirates intervocalically. *kɔ́ v. ‘roast or fry_2’: Reconstructs to: PGw GwHi kɔ́ GwLo kɔ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɔ́d̪ n. ‘breast, milk_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kɔ́ UdYa kɔ́ UdCh à+kɔ́ Dana kɔ́d̪ OpBi kɔ́j OpMo kɔ́j OpPa kɔ́j OpKi kɔ́j *kɔ́ndíl v. ‘count_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kɔ́ndíl 459 OpMo kɔ́ndíl OpPa - OpKi - *kɔ́ŋɔ̀ n. ‘chair_2’: Reconstructs to: PGw GwHi kɔ́ŋɔ̀ GwLo kɔ́ŋɔ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɔ́pʰ v. ‘pound (v.)_4’: Reconstructs to: PKmn GwHi kɔ́p GwLo kɔ́p Komo - UdYa kúpʰ UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɔ̄pʰ n. ‘antelope_waterbuck’: Reconstructs to: PKoUd GwHi - GwLo - Komo kɔ̄p UdYa kɔ̄pʰ UdCh kɔ̄pʰ Dana - OpBi - OpMo - OpPa - OpKi - 460 *kɔ́s v. ‘sew_1’: Reconstructs to: PKmn GwHi ɔ́s GwLo ɔ́s Komo - UdYa - UdCh - Dana kɔ́s OpBi kɔ́s OpMo kɔ́s OpPa kɔ́s OpKi kɔ́s *kɔ̀ʃì n. ‘skin, hide (of animal), bark of tree_4’: Reconstructs to: PGw GwHi kɔ̀ʃì GwLo kɔ̀ʃì Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kɔ̀tʰ v. ‘have’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kɔ̀tʰ OpBi kɔ̀tʰ OpMo - OpPa kɔ̀tʰ OpKi kɔ̀tʰ *kūcʼ v. ‘defecate_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kūʃʼ 461 UdCh kūcʼ Dana - OpBi - OpMo - OpPa - OpKi - *kuɗ n. ‘smoke (exhaust)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kūr UdYa kúd UdCh à+kúɗ Dana kūɗāʔ OpBi kűrā OpMo kűrā OpPa kűrā OpKi kűrā *kúkʰ n. ‘vagina_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kúʔ UdCh à+kúkʰ Dana - OpBi - OpMo - OpPa - OpKi - *kum(bi) v. ‘cover (v.)’: Reconstructs to: PKmn GwHi kùmbì GwLo kùmbì Komo kúm UdYa kūm UdCh kūm Dana kúmā OpBi kűmá OpMo kűmá OpPa kűmá 462 OpKi kűmá *kúmú n. ‘egg_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kúmú OpBi kűmű OpMo kűmű OpPa kűmű OpKi kűmű *kúnā v. ‘help_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kúnā OpBi kűná OpMo kűná OpPa kűná OpKi kűná *kūp n. ‘bamboo_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kūp OpMo kūp OpPa kūp OpKi kūp *kúʃ v. ‘white (be)_3’: Reconstructs to: PUd GwHi - GwLo - 463 Komo - UdYa kúʃ UdCh kúʃ Dana - OpBi - OpMo - OpPa - OpKi - *kʊ̀ v. ‘brood (v.)_2’: Reconstructs to: PGw GwHi kʊ̀ GwLo kʊ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʊ̀dʊ́s n. ‘pipe (for smoking)_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʊ̀dʊ́s OpMo kʊ̀dʊ́s OpPa kʊ̀dʊ́s OpKi kʊ̀dʊ́ʃ *kʊ̀ɪ ̄v. ‘rise (oneself)_2’: Reconstructs to: PGw GwHi kʊ̀ɪ ̄ GwLo kʊ̀ɪ ̄ Komo - UdYa - UdCh - Dana - OpBi - 464 OpMo - OpPa - OpKi - *kʊ̋ká n. ‘joint_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʊ̋ká OpMo kʊ̋ká OpPa kʊ̋ká OpKi kʊ̋ká *kʊ́má v. ‘carry on back_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʊ́má OpMo kʊ́má OpPa kʊ́má OpKi kʊ́má *kʊman n. ‘mother, female’: Reconstructs to: PKmn GwHi - GwLo kūm Komo kʊ̄mán UdYa kūmán UdCh kūm ~ kūmán Dana kʊ̀m ~ kwàn OpBi kʊ̄mán OpMo kʊ̄mán OpPa kʊ̄mán OpKi kʊ̄mán *Kʊ̄ʃʊ̀n n. ‘shadow_4’: Reconstructs to: PGw 465 GwHi gʊ̄ʃʊ̀n GwLo kʊ̄ʃʊ̀n Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *KUtʰ v. ‘short (be)_1’: Reconstructs to: PKmn GwHi gʊ̄t GwLo gʊ̄t Komo kùt UdYa kūtʰ UdCh kūtʰ Dana - OpBi - OpMo - OpPa - OpKi - Cannot explain initial /g/ in Gwama *kwabOʃ v. ‘steal_1’: Reconstructs to: PKmn GwHi kɔ̄bɔ́ʃ GwLo kɔ̄bɔ́ʃ Komo - UdYa - UdCh - Dana - OpBi kǎbús OpMo kǎbús OpPa kǎbús OpKi - *kwak v. ‘cut (split in half lengthwise)_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo kwàk UdYa - 466 UdCh - Dana kwɪ ́ OpBi kwāk OpMo kwāk OpPa kwāk OpKi kwāk *kwáɲ v. ‘pick up (small things), peck at_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kwán UdYa kwáɲ UdCh kwáɲ Dana kwán OpBi kwán OpMo kwán OpPa kwán OpKi - *kwàp n. ‘horn (anatomy)_2’: Reconstructs to: PGw GwHi kwàp GwLo kwàp Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *(gɔ)kwar(a) v. ‘swagger, arrogant (be)’: Reconstructs to: PKmn GwHi - GwLo kwárá Komo gɔ̀kɔ̀l UdYa kār+īs UdCh kār+īs Dana kʰɔ̄r OpBi kʰɔ̄r OpMo kʰɔ̄r OpPa kʰɔ̄r 467 OpKi kʰɔ̄r *ʊ=kam kamʊ n. ‘brother’: Reconstructs to: PKmn GwHi kwám GwLo kwám Komo kàm UdYa kām UdCh à+kām Dana āmʊ́ OpBi hàm OpMo hàm OpPa hàm OpKi hàm Possibly ʊ+kam → kwam → kam or *kamʊ → kwam → kam *(n(j)a)gaD v. ‘replace’: Reconstructs to: PKmn GwHi nāgát GwLo njāgát Komo gàdá UdYa gàs UdCh gàr Dana gàtá OpBi kàrá OpMo kàrá OpPa kàrá OpKi kàrá *gàɗ(am) v. ‘belch_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo gàl UdYa gàrà UdCh gǎɗ Dana gàràm OpBi kʼɛr̄ɛm̄ OpMo gàràm OpPa gàràm OpKi gàràm Cannot account for lack of expected word-inital devoicing befreo *L in Dana-Opo. Cannot account for initial /kʼ/ in Bilugu Opo. 468 *gāgáʔ n. ‘beeswax_1’: Reconstructs to: PKmn GwHi gāgá GwLo gāgá Komo - UdYa - UdCh - Dana gāgá OpBi gāgá OpMo gāgá OpPa gāgá OpKi gāgá *gàj n. ‘termite mound_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo gàʔ UdYa kʼɔ́pʰāgàj UdCh kʼúpʰāɟɛ̀ Dana kʼʊ́kʼàj OpBi kàj OpMo kàj OpPa kàj OpKi kàj *gàɟɪ ́v. ‘sow seeds (by throwing)_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana gàɟɪ ́ OpBi gàdʒɪ ́ OpMo gàdʒɪ ́ OpPa gàdʒɪ ́ OpKi gàdʒɪ ́ *gakʼ v. ‘fast (from drinking or eating)’: Reconstructs to: PKoUd GwHi - GwLo - Komo gáʔ UdYa gǎ 469 UdCh gǎkʼ Dana - OpBi - OpMo - OpPa - OpKi - *gàŋ(a) v. ‘smell (v.)’: Reconstructs to: PKmn GwHi kɛ̃ ̄ GwLo kɛ̃ ̄ Komo gàg UdYa - UdCh - Dana kàŋà OpBi kàŋà OpMo kàŋà OpPa kàŋà OpKi kàŋà *gàŋgārà n. ‘side of body, rib_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi gàŋgārà OpMo gàŋgārà OpPa - OpKi - *gǎŋú n. ‘horse_2’: Reconstructs to: PGw GwHi gǎŋú GwLo gǎŋú Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 470 OpKi - *gàs v. ‘celebrate’: Reconstructs to: PKoUd GwHi - GwLo - Komo gàs UdYa - UdCh gàs Dana - OpBi - OpMo - OpPa - OpKi - *g(w)àsʼ n. ‘person_1, man’: Reconstructs to: PCtrl GwHi - GwLo - Komo jī+gwàz UdYa wàzí UdCh wàtí̪ Dana jɛ+̄kàz OpBi ʊ̀+kàdʒ OpMo ʊ̀+kàdʒ OpPa ʊ̀+kàdʒ OpKi ʊ̀+kàdʒ sʼ > z intervocalically then loss of final V sʼ > cʼ in Opo then voicing to [dʒ] intervocalically with loss of final V. *g(w)àsʼ n. ‘person_2, man’: Reconstructs to: PCtrl GwHi - GwLo - Komo gwàz UdYa gwàsʼ UdCh gwàt ̪h Dana kwàz OpBi - OpMo - OpPa - OpKi - 471 *gàsʼ n. ‘husband_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo gàz UdYa à+kàsʼ UdCh kàt ̪h Dana kàzūpʼ OpBi kàdʒʊ́m OpMo kàdʒʊ́m OpPa kàdʒʊ́m OpKi kàdʒʊ́m *sʼ > z /V_V in Dana. *gaʃa n. ‘belt, sash’: Reconstructs to: PKmn GwHi gáʃà GwLo gáʃà Komo - UdYa gǎʃ UdCh gāʃá Dana gâʃ OpBi gātʃ OpMo gātʃ OpPa gātʃ OpKi gâʃ *gɛɗɪʃ n. ‘broom_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa ɟɛ̀dìʃ UdCh ɟɛ̀ɗɛ̀ʃ Dana kɛ̀ɗɛ̀ʃ OpBi kʰɛ̀rɪ̀s OpMo kʰɛ̀rɪ̀s OpPa kʰɛ̀rɪ̀s OpKi kʰɛ̀rɪ̀ʃ *gɛnd̪(V)(l) n. ‘beehive basket’: Reconstructs to: PKmn GwHi gɛńdɛĺ GwLo gɛ̀ndɪ ́ Komo kɛ̀ndɛ ̄ 472 UdYa - UdCh - Dana gɛ̀nd̪á OpBi gɛ̀ndá OpMo gɛ̀ndá OpPa gɪ̀ndá OpKi gɛ̀ndá *gì(n)sʼ v. ‘tie (bundle)_1’: Reconstructs to: PKmn GwHi kɪ̀nsʼ GwLo kɪ̀nsʼ Komo gìsʼ UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *gìcʼ v. ‘enter_1’: Reconstructs to: PKmn GwHi kɪšʼ GwLo kɪšʼ Komo gìz UdYa - UdCh cīcʼ Dana - OpBi kìtú OpMo kìtú OpPa kìtú OpKi kìtú Alveolar /t/ in Opo cannot be accounted for. *gìrì v. ‘poor (be)_4’: Reconstructs to: PGw GwHi gìrì GwLo gìrì Komo - UdYa - UdCh - Dana - OpBi - 473 OpMo - OpPa - OpKi - *gɪ̀m v. ‘pack in, stuff into container_1’: Reconstructs to: PKmn GwHi gɪ̀m GwLo gɪ̀m Komo - UdYa - UdCh ɟìm Dana - OpBi - OpMo - OpPa - OpKi - *gɪ̀m v. ‘stuff into_2’: Reconstructs to: PGw GwHi gɪ̀m GwLo gɪ̀m Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *gjapaj n. ‘dew_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa ɟápɛʔ̄ UdCh ɟápɛʔ̄ Dana kʰìpʰɔ̂j OpBi gīpàj OpMo gīpàj OpPa gīpàj OpKi gīpàj 474 *gOmpʰVja n. ‘bird_stork (abdim)’: Reconstructs to: PCtrl GwHi - GwLo - Komo à+gɔ̀mpíjá UdYa - UdCh - Dana - OpBi à+kʊ̀mpʰūjá OpMo - OpPa à+kʊ̀mpʰūjá OpKi - *gɔ́kʼɔ̀m v. ‘rough (be)_3’: Reconstructs to: PGw GwHi gɔ́kʼɔ̀m GwLo gɔ́kʼɔ̀m Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *gɔ̀l v. ‘stumble_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa gɔ̀l UdCh gɔ̀l Dana - OpBi - OpMo - OpPa - OpKi - *gɔ̀n v. ‘scratch_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - 475 UdCh - Dana gɔ̀n OpBi gɔ̀n OpMo gɔ̀n OpPa gɔ̀n OpKi gɔ̀n *gɔnkʼ(ɔʃ) n. ‘skin, hide (of animal), bark of tree_1’: Reconstructs to: PKmn GwHi gɔ̀kʼɔ́ʃ GwLo gɔ̀kʼɔ́ʃ Komo gɔ̀nkʼí UdYa gɔ̀kʰ UdCh - Dana kɔ̀gɔ̀ OpBi kɔ̀gɔ̀ OpMo kɔ̀gɔ̀ OpPa kɔ̀gɔ̀ OpKi gwàŋgɪ ́ *gɔ̀r v. ‘climb_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo gɔ̀l UdYa - UdCh - Dana - OpBi kɔ̀r OpMo kɔ̀r OpPa kɔ̀r OpKi kɔ̀r *gùɓ(V) n. ‘house_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo gùbí UdYa gùb UdCh gùɓ Dana kùʔú OpBi kù OpMo kù OpPa kù 476 OpKi kù *gUd̪Um n. ‘pig_1’: Reconstructs to: PKmn GwHi kʊ̀rʊ́m GwLo kʊ̀rʊ́m Komo gùdúm UdYa - UdCh - Dana gùd̪ùm OpBi kùdùmà OpMo kùdùmà OpPa kùdùmà OpKi kùdùmà Would expect initial /k/ in Dana before *L (cf. Gwama). *gùr n. ‘fish_sp (big and fat sized fish)’: Reconstructs to: PCtrl GwHi - GwLo - Komo gùr UdYa - UdCh - Dana gùr OpBi gùr OpMo gùr OpPa gùr OpKi gùr *gÙs v. ‘run (SG)_1, flow, bleed’: Reconstructs to: PKmn GwHi gʊ̀s GwLo gʊ̀s Komo gùʃ UdYa gùs UdCh gùs Dana - OpBi - OpMo - OpPa - OpKi - 477 *gUsʼ v. ‘swallow_1’: Reconstructs to: PKmn GwHi gùsʼ GwLo kùsʼ Komo gʊ̀sʼ UdYa - UdCh - Dana kʊ̄sʼá OpBi kʊ̀tʃʼá OpMo kʊ̀tʃʼá OpPa kʊ̀tʃʼá OpKi kʊ̀tʃʼá *gʊlɪla n. ‘bird (yellow-billed kite or black kite)’: Reconstructs to: PKmn GwHi - GwLo gʊ́lːā Komo bāgʊ́lɪ̀lā UdYa - UdCh - Dana ɓāgʊ̂lːā OpBi - OpMo ɓāgʊ́lʊ́l OpPa ɓāgʊ́lʊ́l OpKi ɓāgʊ́là *gʊ̌pʊ̀ v. ‘pound (v.)_7’: Reconstructs to: PGw GwHi gʊ̌pʊ̀ GwLo gʊ̌pʊ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *gVs n. ‘hole_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo gìʃ UdYa ɟìs 478 UdCh ɟìs Dana kùs OpBi kùs OpMo kùs OpPa kùs OpKi kùs Vowel must have been /i/ in PKoUd to account for PUd *g > ɟ. *gwàj n. ‘elephant_1’: Reconstructs to: PKmn GwHi kwɪ̀ GwLo kwɪ̀ Komo gwà UdYa gwàj UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *gwàj n. ‘name_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa gwàj UdCh gwàj Dana - OpBi - OpMo - OpPa - OpKi - *gwàlɪ ́n. ‘bean_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana gwàlí OpBi gwàlɪ ́ OpMo gwàlɪ ́ 479 OpPa gwàlɪ ́ OpKi gwàlɪ ́ *gwama n. ‘Gwama (ethnonym)’: Reconstructs to: PKmn GwHi gwàmá GwLo kwāmà Komo gwàmá UdYa - UdCh - Dana gɔ̀má OpBi gɔ̀má OpMo gɔ̀má OpPa gɔ̀má OpKi gɔ̀má *gwǎpʰī n. ‘fingernail, toenail, claw, hoof_1’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa gwǎpʰī UdCh à+gwǎpʰī Dana - OpBi - OpMo - OpPa - OpKi - *gwǎr n. ‘side of body, rib_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa gwǎr UdCh gwǎr Dana - OpBi - OpMo - OpPa - OpKi - 480 *gwàrásʼ n. ‘broom_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo gwàrásʼ UdYa - UdCh - Dana gwàrásʼ OpBi gwàrátʃʼ OpMo gwàrátʃʼ OpPa gwàrátʃʼ OpKi gwàrátʃʼ *gwàtʰá n. ‘gourd_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana gwàtʰá OpBi - OpMo - OpPa - OpKi gwàtʰá *gwatʰV n. ‘head pad (for head carrying)_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo gwàtʊ́ UdYa gwǎtʰɛ ̄ UdCh gwǎtʰɛ ̄ Dana - OpBi - OpMo - OpPa - OpKi - *gàm v. ‘find, meet’: Reconstructs to: PKmn GwHi kàm GwLo kàm Komo gàm UdYa gàm 481 UdCh gàm Dana kàm OpBi kàm OpMo kàm OpPa kàm OpKi kàm *gjama? n. ‘wound_1’: Reconstructs to: PKmn GwHi kāmā GwLo kāmā Komo zàmā UdYa ɟàmá UdCh ɟàmá Dana - OpBi - OpMo - OpPa - OpKi - *gɔ̀ɗɔkʼ v. ‘deep (be)_1’: Reconstructs to: PKmn GwHi kwì GwLo kwǐ Komo gʊ̀r UdYa - UdCh - Dana kɔ̀ɗɔ̀kʼ OpBi kɔ̄rɔ́ OpMo kɔ̄rɔ́ OpPa kɔ̄rɔ́ OpKi kɔ̄rɔ́ *k(j)asʼVN n. ‘earth, soil, ground, floor_1’: Reconstructs to: PKmn GwHi kʼjáŋásʼ GwLo kʼɛs̄ʼɛń Komo kʼāsʼɪ̀ UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 482 OpKi - *kʼ(w)ás n. ‘back_1’: Reconstructs to: PKmn GwHi kʼwás GwLo kʼwás Komo kʼǎw UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼá(n)cʼír v. ‘chop_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kʼácʼír OpBi kʼántʃʼír OpMo kʼántʃʼír OpPa kʼántʃʼír OpKi kʼántʃʼír *kʼáɗ v. ‘soft (be)_1’: Reconstructs to: PKmn GwHi kʼátʼ GwLo kʼátʼ Komo kʼátʼ UdYa kʼád UdCh kʼáɗ Dana - OpBi - OpMo - OpPa - OpKi - *kʼáɗ v. ‘sprout (verb)’: Reconstructs to: PCtrl GwHi - GwLo - 483 Komo kʼáʔ UdYa kʼáj UdCh kʼáɗ Dana kʼájī OpBi kʼáʔ OpMo kʼáʔ OpPa kʼáʔ OpKi kʼáʔ *kʼáj v. ‘good (be)_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʼáj OpMo kʼáj OpPa kʼáj OpKi kʼáj *kʼājà v. ‘grab with fingers, pinch_2’: Reconstructs to: PGw GwHi kʼājà GwLo kʼājà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼalala n. ‘tonsils_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kʼálálá UdYa - UdCh - Dana kʼàlálá OpBi kǎltʼā 484 OpMo kǎltʼā OpPa kǎltʼā OpKi kǎltʼā *kʼama v. ‘eat (hard food)’: Reconstructs to: PKmn GwHi kʼā GwLo kʼā Komo kʼá UdYa kʼáʔ UdCh kʼá Dana kʼámá OpBi kʼámá OpMo kʼámá OpPa kʼámá OpKi kʼámá *kʼándí n. ‘head pad (for head carrying)_3’: Reconstructs to: PGw GwHi kʼándí GwLo kʼándí Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼaʃ v. ‘lack (not have)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kʼáʃà UdYa kʼáʃ UdCh kʼáʃ Dana kʼàʃ OpBi kʼās OpMo kʼās OpPa kʼās OpKi kʼāʃ 485 *kʼāʃ v. ‘red (be)_3’: Reconstructs to: PGw GwHi kʼāʃ GwLo kʼāʃ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼáw ~ kʼwá n*. ‘dog_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo kʼáw UdYa kʼwáʔ UdCh à+kʼá Dana - OpBi ʔwáj OpMo - OpPa - OpKi - *kʼɛɗ v. ‘break (destroy or get destroyed)’: Reconstructs to: PKoUd GwHi - GwLo - Komo kʼɛ̀r UdYa cʼɛd́ UdCh cʼɛɗ́ Dana - OpBi - OpMo - OpPa - OpKi - *kʼɛm v. ‘sing_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa cʼɛm̄ 486 UdCh - Dana kʼɛ̀m OpBi - OpMo - OpPa - OpKi - *kʼɛ̀r(ɛ)́ v. ‘clear land (for planting)_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana kʼɛ̀rɛ ́ OpBi kʼɛr̄ OpMo kʼɛr̄ OpPa kʼɛr̄ OpKi kʼɛr̄ *kʼI(m)Isʼ n. ‘charcoal or coal’: Reconstructs to: PKmn GwHi sʼɪśʼɪń GwLo sʼɪśʼɪń Komo kìsʼísʼìʔ UdYa cʼɛl̄ɛs̄ʼ UdCh cʼīlāt ̪̓ Dana kʼɪ̀sʼ OpBi kʼīmītʃʼ OpMo kʼīmītʃʼ OpPa kʼɪt̄ʃʼ OpKi kʼɪt̄ʃʼ *kʼɪ(mɪ)sʼ v. ‘chop_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo kʼɪśʼ UdYa cʼīmīsʼ UdCh cʼīmīt ̪̓ Dana - OpBi - OpMo - OpPa - 487 OpKi - *kʼOʃ v. ‘kill, fight’: Reconstructs to: PKmn GwHi kʼʊ́ʃ GwLo kʼʊ́ʃ Komo kʼɔ́ʃ UdYa kʼɔ́ʃ UdCh kʼɔ́ʃ Dana kʼɔ́ʃ OpBi kʼɔ́sɔ́ OpMo kʼɔ́sɔ́ OpPa kʼɔ́s OpKi kʼɔ́ʃ *kʼɔ́ v. ‘sit_1, dwell (live, reside)_PL’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa kʼɔ́ UdCh kʼɔ́ Dana kʼɔ́ʔ OpBi kʼɔ́ OpMo kʼɔ́ OpPa kʼɔ́ OpKi kʼɔ́ *kʼɔ̀d+ʃɛ̀ʔ n. ‘gums_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʼɔ̀d+ʃɛ̀ʔ UdCh kʼɔ̀ɗ+ʃɛ̀ʔ Dana - OpBi - OpMo - OpPa - OpKi - *kʼɔj v. ‘peel_1’: Reconstructs to: PCtrl GwHi - GwLo - 488 Komo kʼɪ ́ UdYa cʼɛ ́ UdCh cʼɛ ́ Dana kʼɔ̀j OpBi kʼɛ ́ OpMo - OpPa - OpKi - *kʼɔ́kʼɔ́l n. ‘cheek_2’: Reconstructs to: PGw GwHi kʼɔ́kʼɔ́l GwLo kʼɔ́kí Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼɔlɔ n. ‘hand_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo kʼɔ̀lɔ̀ UdYa - UdCh - Dana - OpBi kʼɔ̄j OpMo kʼɔ̄j OpPa kʼwī OpKi - *kʼɔ́tʃʼɔ́ n. ‘earth, soil, ground, floor_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi kʼɔ́tʃʼɔ́ 489 OpMo kʼɔ́tʃʼɔ́ OpPa kʼɔ́tʃʼɔ́ OpKi kʼɔ́tʃʼɔ́ *kʼúɓ v. ‘finish_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo kúpʼ UdYa - UdCh kʼúɓ Dana - OpBi - OpMo - OpPa - OpKi - *kʼūcʰūr ~ kʼūcʼūr v. ‘rinse mouth_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʼúʒūr UdCh kʼūcʰūr Dana - OpBi - OpMo - OpPa - OpKi - *kʼúl v. ‘deep (be)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʼúl UdCh kʼúl Dana - OpBi - OpMo - OpPa - OpKi - 490 *kʼúmú n. ‘navel, umbilical cord’: Reconstructs to: PCtrl GwHi - GwLo - Komo kʼúmú UdYa - UdCh - Dana kʼúmú OpBi kʼűmű OpMo kʼűmű OpPa kʼűmű OpKi kʼűmű *kʼúnsʼ v. ‘bite (by animal)_2’: Reconstructs to: PGw GwHi kʼúnsʼ GwLo kʼúnsʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼūɲ v. ‘tasty (be)_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʼūɲ UdCh kʼūɲ Dana - OpBi - OpMo - OpPa - OpKi - *kʼúpʰ v. ‘to stew (food)_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʼúpʰ 491 UdCh kʼúpʰ Dana - OpBi - OpMo - OpPa - OpKi - *kʼuT̪ v. ‘clear land (for planting)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo kʼúʃ UdYa kʼús UdCh kʼút ̪h Dana - OpBi - OpMo - OpPa - OpKi - *kʼʊkʼʊl v. ‘crow (verb)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo kʼʊ̀kʼʊ̀l UdYa - UdCh - Dana kʼʊ́kʼʊ́l OpBi kʼʊ́kʼʊ́lʊ́ OpMo kʼʊ́kʼʊ́lʊ́ OpPa kʼʊ́kʼʊ́lʊ́ OpKi kʼʊ́kʼʊ́lʊ́ *kʼʊ́mkɛ̃ n. ‘termite_3’: Reconstructs to: PGw GwHi kʼʊ́mkɛ̃ GwLo kʼʊ́mkɛ̃ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 492 OpKi - *kʼʊ̄mpʼ n. ‘fingernail, toenail, claw, hoof_2’: Reconstructs to: PGw GwHi kʼʊ̄mpʼ GwLo kʼʊ̄mpʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼʊ́p n. ‘head’: Reconstructs to: PKmn GwHi kʼʊ́p GwLo ʔʊ́p Komo kʼʊ́p UdYa kʼúpʰ UdCh kʼúpʰ Dana kʼʊ́pʰ OpBi kʼʊ́p OpMo kʼʊ́p OpPa kʼʊ́p OpKi kʼʊ́p *kʼʊ́s n. ‘throat’: Reconstructs to: PKmn GwHi k’úʃ GwLo k’úʃ Komo kʼʊ́ʃ UdYa kʼús UdCh kʼús Dana kʼʊ́s OpBi kʼʊ́sʊ́ OpMo kʼʊ́s OpPa kʼʊ́s OpKi kʼʊ́s *kʼʊ́s(V) n. ‘river_1’: Reconstructs to: PDaOp GwHi - GwLo - 493 Komo - UdYa - UdCh - Dana kʼʊ́sī OpBi kʼʊ́sʊ́+dʒì OpMo kʼʊ́sʊ́+dʒì OpPa kʼʊ́sɪ+́zì OpKi kʼʊ́sɪ+́sì *kʼʊ̃̄sʼ v. ‘scratch_3’: Reconstructs to: PGw GwHi kʼʊ̃̄sʼ GwLo kʼʊ̃̄sʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼʊ́t v. ‘cut_2’: Reconstructs to: PGw GwHi kʼʊ́t GwLo kʼʊ́t Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *kʼʊt ̪kʼut ̪v*. ‘cough’: Reconstructs to: PKmn GwHi kʼʊ́kʼʊ́t GwLo kʼʊ́kʼʊ́t Komo kʼùt UdYa kʼútʰ UdCh kʼútʰ Dana kʼút ̪h OpBi kʼútʼù 494 OpMo kʼútʼù OpPa kʼútʼù OpKi kʼútʼù *kʼVkʼVr v. ‘rough (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa ʃʼɛǵɛr̄ UdCh - Dana kʼàkʼàr OpBi kʼɛk̄ʼɛr̄ OpMo kʼɛk̄ʼɛr̄ OpPa kʼɛk̄ʼɛr̄ OpKi kʼákʼár If a, vowel must have shifted to /ɛ/ in Proto- Uduk to account for kʼ > cʼ > ʃʼ shift in Yabus Uduk. *kʼwà n. ‘gourd_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa kʼwà UdCh kʼwà Dana - OpBi - OpMo - OpPa - OpKi - *kʼwàd̪(ɪ)́ v. ‘dig (for water)_1’: Reconstructs to: PKmn GwHi kʼwī GwLo kʼwī Komo kʼɔ̀ UdYa kʼwā UdCh - Dana kʼwàd̪ɪ ́ OpBi kʼɔ̄j OpMo wārɪ ́ OpPa wādɪ ́ OpKi kʼwārɪ ́ 495 Loss of initial /kʼ/ in Modin and Pame Opo. *kʼwànt ̪̓ n. ‘tick’: Reconstructs to: PKmn GwHi kʼwāntʼɪ ́̓kwāntʼ GwLo kʼwāntʼɪ ́̓kwāntʼ Komo kʼwàtʼ UdYa kʼwātʰ UdCh à+kʼwāɗ Dana kʼwàt ̪̓ OpBi kʼwātʼ OpMo kʼwātʼ OpPa kʼwātʼ OpKi kʼwātʼ *kʼwǎʃà n. ‘bean_2’: Reconstructs to: PGw GwHi kʼwāʃà GwLo kʼwǎʃà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *s(w)à v. ‘dance_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo ʃwà UdYa sā UdCh sā Dana - OpBi - OpMo - OpPa - OpKi - *s(w)am v. ‘warm oneself’: Reconstructs to: PKmn GwHi sɔ́m GwLo sɔ́m 496 Komo ʃɔ́m UdYa sām UdCh sām Dana sɔ̀m OpBi sɔ̄m OpMo sɔ̄m OpPa sɔ̄m OpKi ʃɔ̄m *sā v. ‘court (v.), flirt with_2’: Reconstructs to: PGw GwHi sā GwLo sā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sáɗ(á) n. ‘calf of leg’: Reconstructs to: PKoUd GwHi - GwLo - Komo ʃárá UdYa sád UdCh à+sáɗ Dana - OpBi - OpMo - OpPa - OpKi - *sāmpʼ n. ‘side of body, rib_5’: Reconstructs to: PGw GwHi sāmpʼ GwLo sāmpʼ Komo - UdYa - UdCh - Dana - OpBi - 497 OpMo - OpPa - OpKi - *sàmún n. ‘maize_3’: Reconstructs to: PGw GwHi sàmún GwLo sàmún Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sáŋkʼ v. ‘swim_3’: Reconstructs to: PGw GwHi sáŋkʼ GwLo sáŋkʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sàzà v. ‘dry out_2’: Reconstructs to: PGw GwHi sàzà GwLo sàzà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 498 *sE v. ‘sow seeds (by planting)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃɛ̀ʔ UdYa sī UdCh sī Dana sɛ̀ OpBi sɛ ̄ OpMo sɛ ̄ OpPa sɛ ̄ OpKi - *sɛl v. ‘climb_1’: Reconstructs to: PKmn GwHi sál GwLo sɛl̄ Komo - UdYa sɛ ̄ UdCh sɛ ̄ Dana - OpBi - OpMo - OpPa - OpKi - *sɛńɛʔ̄ n. ‘one (1)_2’: Reconstructs to: PGw GwHi sɛńɛʔ̄ GwLo sɛńɛʔ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sí v. ‘lay (v.)_2’: Reconstructs to: PGw GwHi sí GwLo sí Komo - UdYa - 499 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sīɓ n. ‘sand_1’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa sīb UdCh à+sīɓ Dana - OpBi - OpMo - OpPa - OpKi - *sìl v. ‘heavy (be)_2’: Reconstructs to: PGw GwHi sìl GwLo sìʔ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *símpʼ n. ‘egg_2’: Reconstructs to: PGw GwHi símpʼ GwLo símpʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 500 OpKi - *sīnkʼ n. ‘smoke (exhaust)_2’: Reconstructs to: PGw GwHi sīnkʼ GwLo sīnkʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sír v. ‘dive_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana sír OpBi sır̋ OpMo sır̋ OpPa sır̋ OpKi sır̋ *sIrE n. ‘tooth (canine)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃɪ̀l UdYa - UdCh - Dana sírɛ ́ OpBi sır̋á OpMo sır̋á OpPa sır̋á OpKi sır̋á *sīzì n. ‘crocodile_2’: Reconstructs to: PGw GwHi sīzì GwLo sīzì 501 Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sɪɗ́(a) v. ‘hang up_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃɪŕ UdYa ʃír UdCh ʃír Dana sɪɗ́ā OpBi sɪŕ OpMo sɪŕ OpPa sɪŕ OpKi sɪŕ *sɪĺ(ɪ) v. ‘cut (meat into one long piece)_2’: Reconstructs to: PKmn GwHi ʃɪĺɪ̀ GwLo ʃɪ ́ Komo - UdYa - UdCh - Dana - OpBi sɪ ́ OpMo - OpPa - OpKi - *sɪ̀t ̪̓ v. ‘far (be)_1’: Reconstructs to: PKmn GwHi - GwLo ʃɪt̄ʼ Komo ʃɪ̀tʼ UdYa sīd UdCh sīɗ Dana sɪ̀t ̪̓ OpBi sɪt̄ʼ 502 OpMo sɪt̄ʼ OpPa sɪt̄ʼ OpKi sɪt̄ʼ *sɪt́ʼɪn̄ n. ‘back_4’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi sɪt́ʼɪn̄ OpMo sɪt́ʼɪn̄ OpPa sɪt́ʼɪn̄ OpKi sɪt́ʼɪn̄ *sjã n. ‘penis_2’: Reconstructs to: PGw GwHi sjã GwLo sjã Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *sjãkʼúʃ n. ‘tonsils_2’: Reconstructs to: PGw GwHi sjãkʼúʃ GwLo sjãkʼúʃ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 503 *sɔ v. ‘run_PL_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa sɔ̄ UdCh - Dana - OpBi sɔ́ OpMo sɔ́ OpPa sɔ́ OpKi sɔ́ *sɔ́D v. ‘offend_3’: Reconstructs to: PKoUd GwHi - GwLo - Komo ʃɔ́t UdYa sɔ́d UdCh sɔ́r Dana - OpBi - OpMo - OpPa - OpKi - *sɔ́p v. ‘stab_1’: Reconstructs to: PKmn GwHi sɔ́ GwLo sɔ́ Komo - UdYa - UdCh - Dana - OpBi sɔ́p OpMo - OpPa sɔ́p OpKi - *sɔ̄sɔ́r v. ‘poor (be)_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - 504 UdCh - Dana - OpBi sɔ̄sɔ́r OpMo sɔ̄sɔ́r OpPa sɔ̄sɔ́r OpKi sɔ̄sɔ́r *sud̪(i) n. ‘beer’: Reconstructs to: PKmn GwHi ʃʊ́l GwLo ʃwɪ ́ Komo ʃùʔí UdYa sū UdCh à+sū Dana sùd̪ OpBi sī OpMo swī OpPa swī OpKi swī Gwama presumably u > ʊ and then s > ʃ *sūkʼ v. ‘stab_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa sūkʼ UdCh sūkʼ Dana - OpBi - OpMo - OpPa - OpKi - *sūsá ~ sʊ̄sá v*. ‘push_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana sūsá OpBi sʊ̄sá OpMo sʊ̄sá 505 OpPa sʊ̄sá OpKi sʊ̄sá *sūtʼā v. ‘dip food in sauce with fingers_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi sūtʼā OpMo sūtʼā OpPa sūtʼā OpKi sūtʼā *sʊkʼa adv. ‘two’: Reconstructs to: PKmn GwHi swījā GwLo swījā Komo sʊ̄ UdYa sú UdCh sú Dana sʊ́kʼà OpBi sʊ̄kʼá OpMo sʊ̄kʼá OpPa sʊ̋kʼá OpKi sʊ̋kʼá *sʊm n. ‘python_1’: Reconstructs to: PKmn GwHi ʃʊ̄ʃʊ́m GwLo ʃʊ̄ʃʊ́m Komo ʃʊ̌m UdYa - UdCh súm Dana sʊ̄mʊ́ OpBi sʊ̄mʊ́ OpMo sʊ̄mʊ́ OpPa sʊ̄mʊ́ OpKi sʊ̄mʊ́ 506 *sʊs v. ‘lead (guide)_1’: Reconstructs to: PKmn GwHi ʃʊ́ʃ GwLo ʃʊ́ʃ Komo ʃʊ́ʃ UdYa sús UdCh sús Dana sʊ́ʔ OpBi - OpMo - OpPa - OpKi - *swāl n. ‘house_2’: Reconstructs to: PGw GwHi swāl GwLo swɪ ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *swɪ ́v. ‘pour_3’: Reconstructs to: PGw GwHi swɪ ́ GwLo swɪ ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ʃa v. ‘eat (soft food)_SG’: Reconstructs to: PKmn GwHi ʃā GwLo ʃā Komo ʃá UdYa ʃwá 507 UdCh ʃwá Dana úʃā OpBi sá OpMo sá OpPa sá OpKi ʃá PCtrl innovates *uʃa for plural, Uduk metathesizes plural for both singular and plural, Dana changes tone for sg/pl opposition. *ʃ(w)at ̪̓En n. ‘children_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃwàtʼɪń UdYa - UdCh - Dana ʊ̀+ʃàt ̪̓ ɛń OpBi bì+tʼí OpMo - OpPa - OpKi - *ʃàdɪ ́v. ‘hot (be)_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ʃàdɪ ́ OpBi - OpMo - OpPa - OpKi ʃàrɪ ́ *ʃákáná v. ‘empty (be)_3’: Reconstructs to: PGw GwHi ʃákáná GwLo ʃákáná Komo - UdYa - UdCh - Dana - OpBi - 508 OpMo - OpPa - OpKi - *ʃàkʼà v. ‘carry_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ʃàkʼà OpBi sākʼā OpMo sākʼā OpPa sākʼā OpKi sākʼā *ʃali n. ‘cloud, fog_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃɛ̀lí UdYa ʃīlá UdCh ʃīláʔ Dana àpàjʃàlíʔ OpBi - OpMo - OpPa - OpKi - *ʃalɪ v. ‘offend_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ʃàlɪ ́ OpBi sálɪ ̄ OpMo sálɪ ̄ OpPa sálɪ ̄ OpKi ʃál 509 *ʃám v. ‘love_4’: Reconstructs to: PGw GwHi ʃám GwLo ʃám Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ʃám v. ‘want_2’: Reconstructs to: PGw GwHi ʃám GwLo ʃám Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ʃàpɔ́ n. ‘salt_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana à+ʃàbɔ́ OpBi à+sàpɔ́ OpMo à+sàpɔ́ OpPa à+sàpɔ́ OpKi à+ʃàbɔ́ʔ *ʃāpʼ v. ‘rain (v.)_2’: Reconstructs to: PGw GwHi ʃāpʼ GwLo ʃāpʼ Komo - UdYa - 510 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ʃāʃ(ā) n. ‘birth pangs (pain)’: Reconstructs to: PKmn GwHi ʃɛ ́ GwLo ʃɛ ́ Komo ʃāʃ UdYa - UdCh - Dana ʃāʃā OpBi sāsā OpMo sāsā OpPa sāsā OpKi ʃāʃā *ʃáw v. ‘pain, be hurt_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃáw UdYa - UdCh - Dana ʃáw OpBi sáw OpMo sáw OpPa sáw OpKi ʃáw *ʃāwā v. ‘slippery, smooth (be)_1’: Reconstructs to: PGw GwHi ʃāwā GwLo ʃāwā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 511 OpKi - *ʃE n. ‘tooth’: Reconstructs to: PKmn GwHi ʃī GwLo ʃī Komo ʃɛ̀ UdYa ʃɛ ̄ UdCh ʃɛ ̄ Dana ʃɛ ̄ OpBi sɛ ̄ OpMo sɛ ̄ OpPa sɛ ̄ OpKi ʃɛ ̄ *ʃɛ̀ɗ v. ‘sweep_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃɛ̀r UdYa - UdCh - Dana ʃɛ̀ɗ OpBi - OpMo - OpPa - OpKi - *ʃɛŋ́gɛĺ v. ‘light weight (be)_2’: Reconstructs to: PGw GwHi ʃɛŋ́gɛĺ GwLo ʃɛŋ́ɪʔ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ʃɛt̄(ʰ) n. ‘antelope_1’: Reconstructs to: PKmn GwHi ʊ̄ʃɛt̄ GwLo ʃɛt̄ 512 Komo ʃɛt̄ UdYa ʃɛt̄ʰ UdCh à+ʃɛt̄ʰ Dana - OpBi - OpMo - OpPa - OpKi - *ʃíɗ v. ‘tie up (tether)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ʃíd UdCh ʃīɗ Dana - OpBi - OpMo - OpPa - OpKi - *ʃij(a) v. ‘shiver_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ɛ̀s+ʃìjà OpBi ɪs̄+sıj̋á OpMo ɪs̄+sıj̋á OpPa ɪs̄+sıj̋á OpKi ɪs̄+sıj̋á *ʃìjà v. ‘dance_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ʃìjà OpBi sījā 513 OpMo sījā OpPa sījā OpKi ʃījā Most likely not cognate with ‘dance_1’ as /ʃ/ in Dana would indicate *ʃ and /ʃ/ in Uduk is lacking. *ʃīkānā n. ‘tooth (canine)_2’: Reconstructs to: PGw GwHi ʃī+kānā GwLo ʃīkānā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ʃɪḱʼ v. ‘long or tall (be)_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi sɪḱʼ OpMo sɪḱʼ OpPa sɪḱʼ OpKi ʃɪḱʼ *ʃin v. ‘pelt_3, throw’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃìn UdYa ʃīn UdCh ʃīn Dana ʃìnà OpBi sīnā OpMo sīnā OpPa - OpKi - 514 *ʃìpʰá n. ‘knife’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃìpá UdYa - UdCh - Dana ʃìpʰá OpBi - OpMo - OpPa - OpKi - *ʃɪ ́v. ‘rain (v.)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃɪ ́ UdYa - UdCh - Dana ʃɪ ́ OpBi sɪ ́ OpMo sɪ ́ OpPa sɪ ́ OpKi sɪ ́ *ʃɪǵ n. ‘bad (be)_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo ʃɪǵ UdYa ʃíʔ UdCh ʃíʔ Dana - OpBi - OpMo - OpPa - OpKi - *ʃɪntʼ ~ ʃintʼ v. ‘blow nose’: Reconstructs to: PKmn GwHi ʃɪn̄tʼ GwLo ʃɪn̄tʼ Komo ʃīn+ʃʊ̀nʃ 515 UdYa ʃīn UdCh ʃīn Dana ʃìnà+ʃʊ̀ʃ OpBi sın̋á OpMo sın̋á OpPa sın̋á OpKi sın̋á *ʃɪʃ́ v. ‘extinguish_2’: Reconstructs to: PGw GwHi ʃɪʃ́ GwLo ʃɪʃ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ʃO n. ‘grass’: Reconstructs to: PKmn GwHi ʃʊ́ʃʊ̄ GwLo ʃʊ́ʃʊ̄ʃʊ̄ Komo ʃɔ̀ʔí UdYa ʃɔ̄ UdCh à+ʃɔ̄ Dana ʃɔ̄ʔɔ́ OpBi sɔ̀ OpMo sɔ̀ OpPa sɔ̀ OpKi sɔ̀ Possibly reduplicated in Gwama. *ʃOkʼ(VN) n. ‘louse_1’: Reconstructs to: PKmn GwHi - GwLo ʃʊ̄gʊ̀n Komo ʃùwɛ̀n UdYa ʃɔ̄kʼɔ̄m UdCh à+ʃɔ̄kʼɔ̄m Dana ʃùkʼnáj OpBi sūkʼɛń 516 OpMo sūkʼɛń OpPa sūkʼɛń OpKi ʃūkʼɛń *ʃOl(V) v. ‘stuff into_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ʃɔ̀lɔ́ OpBi sʊ̄lɪ ́ OpMo sʊ̄lɪ ́ OpPa sʊ̄lɪ ́ OpKi ʃʊ̄lɪ ́ *ʃɔ́ v. ‘sew_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo ʃɔ́ UdYa ʃɔ́ UdCh ʃɔ́ Dana - OpBi - OpMo - OpPa - OpKi - *ʃɔ̀d v. ‘scootch (move over)_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo ʃɔ̀d UdYa ʃɔ̄r UdCh ʃɔ̄r Dana - OpBi - OpMo - OpPa - OpKi - 517 *ʃɔkʼ n. ‘rain (precipitation)_2’: Reconstructs to: PKmn GwHi - GwLo ʃʊ́ Komo ɔ́ UdYa ʃɔ̄kʼ UdCh à+ʃɔ̄kʼ Dana - OpBi hɔ́ OpMo hɔ̌ OpPa hɔ̌ OpKi - *ʃ > s > h in Opo? *ʃɔnkʼ n. ‘foot or leg_1’: Reconstructs to: PKmn GwHi sɔ̄ŋkʼ GwLo sɔ̄ntʼ Komo ʃɔ̀g UdYa ʃɔ̄kʼ UdCh ʃɔ̄ʔ Dana ʃɔ̀g OpBi - OpMo - OpPa - OpKi - Cannot account for final /tʼ in Lowland Gwama. *ʃɔŋ(V) v. ‘pack in, stuff into container_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ʃɔ̀ŋɔ́ OpBi sɔ̄ŋɪ ́ OpMo sɔ̄ŋɪ ́ OpPa sɔ̄ŋɪ ́ OpKi ʃɔ̄ŋɔ́ *ʃudaj n. ‘hair_3’: Reconstructs to: PDaOp GwHi - GwLo - 518 Komo - UdYa - UdCh - Dana ʃùdáj OpBi sūjɛ ́ OpMo sūjɛ ́ OpPa sūjɛ ́ OpKi sūjɛ ́ *ʃUImakʼ n. ‘bone’: Reconstructs to: PKmn GwHi sí GwLo sí Komo ʃúmákʼ UdYa sīmāʔ UdCh à+sīmāʔ Dana ʃʊ́j OpBi sʊ́j OpMo sʊ́j OpPa sʊ́j OpKi sʊ́j *ʃukʼ(i) v. ‘wake (trs.)_1’: Reconstructs to: PKmn GwHi sūgì GwLo sūgì Komo ʃùg UdYa ʃūkʼ UdCh - Dana ʃùg OpBi sūg OpMo sūg OpPa sūg OpKi ʃūg *ʃukʼin v. ‘breathe’: Reconstructs to: PKmn GwHi ʃɪňʃɪ ́ GwLo ʃɪn̄ʃɪ̀ Komo ʃùʔɛn̄ UdYa ʃīʔīn UdCh ʃīʔīn Dana ʃìkʼ OpBi sīkʼ 519 OpMo sīkʼ OpPa sīkʼ OpKi sīkʼ *ʃum(a) n. ‘meat, animal’: Reconstructs to: PKmn GwHi sūm GwLo sūm Komo ʃùm UdYa ʃūm UdCh ʃūm Dana ʃùmà OpBi sūmā OpMo sūmā OpPa sūmā OpKi ʃūmā *ʃúm+īs v. ‘stuff into_1’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ʃúm+īs UdCh ʃúm+īs Dana - OpBi - OpMo - OpPa - OpKi - *ʃúmā v. ‘brood (v.)_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ʃúmā OpBi sűmá OpMo sűmá OpPa sűmá OpKi sűmá 520 *ʃutʼ ? n. ‘rope_1’: Reconstructs to: PKmn GwHi ʃʊ̄dʊ̀l GwLo ʃwītʼìn Komo ʃʊ̀ʔɪ ́ UdYa ʃí UdCh à+ʃí Dana ʃʊ́l OpBi - OpMo - OpPa - OpKi - *ʃʊ v. ‘buy_3, sell’: Reconstructs to: PCtrl GwHi - GwLo - Komo ʃʊ̀ UdYa - UdCh - Dana ʃʊ̀ OpBi sʊ̄ OpMo sʊ̄ OpPa sʊ̄ OpKi sʊ̄ *ʃʊ(n)kʼ n. ‘tendon, vein’: Reconstructs to: PKmn GwHi ʃʊ̄nkʼ GwLo ʃʊ̄nkʼ Komo ʃʊ̄ UdYa ʃúʔ UdCh à+ʃúʔ Dana ʃùkʼ OpBi sʊ̄kʼ OpMo sʊ̄kʼ OpPa sʊ̄kʼ OpKi ʃūkʼ *ʃʊ́ɓ v. ‘copulate (animal)_3’: Reconstructs to: PKmn GwHi ʃʊ́pʼ GwLo ʃʊ́pʼ Komo - UdYa ʃúb 521 UdCh ʃúɓ Dana - OpBi - OpMo - OpPa - OpKi - *ʃʊnʃ n. ‘nose’: Reconstructs to: PKmn GwHi ʃʊ̃ʃ̄ GwLo ʃʊ̃ʃ̄ Komo ʃʊ̀nʃ UdYa ʃūʃ UdCh ʃūʃ Dana ʃʊ̀ʃ OpBi sʊ̀sʊ̀ OpMo sʊ̀s OpPa sʊ̀s OpKi ʃʊ̀ʃ *ʃʊ̄nʊ̄ n. ‘brain_3’: Reconstructs to: PGw GwHi ʃʊ̄nʊ̄ GwLo ʃʊ̄nʊ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ʃʊ́ʃʊ́ v. ‘polygamous (be)_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi sʊ́sʊ́ OpMo sʊ́sʊ́ OpPa sʊ́s 522 OpKi ʃʊ́ʃ *ha v. ‘come, come_SG’: Reconstructs to: PKmn GwHi hɔ̌ GwLo hɔ̌ Komo hà + ʊ́ UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *haɗ v. ‘pull, pull off, drag’: Reconstructs to: PCtrl GwHi - GwLo - Komo hàd UdYa - UdCh hǎɗ Dana hàdí OpBi hādɪ ́ OpMo hādɪ ́ OpPa hādɪ ́ OpKi hādɪ ́ *háɗìkʼ v. ‘hiccough_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa - UdCh àhǎɗkīʔ Dana háɗìkʼ OpBi ha̋rīkʼ OpMo hárìkʼ OpPa hárìkʼ OpKi - *hag(a) v. ‘have sex_1’: Reconstructs to: PKmn GwHi hã́ʔ GwLo hã́ʔ 523 Komo hág UdYa - UdCh háʔ Dana - OpBi hágá OpMo hágá OpPa hágá OpKi hágá *ham v. ‘yawn_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo hám UdYa hám UdCh à+hāmɛ ̄ Dana hâm OpBi hām OpMo hām OpPa hām OpKi - *hàntʼà v. ‘big (be)_2’: Reconstructs to: PGw GwHi hàntʼà GwLo hàntʼà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *hasʼ v. ‘trample, ruminate’: Reconstructs to: PKmn GwHi - GwLo hāsʼɪ̀ Komo hásʼ UdYa hásʼ UdCh hát ̪̓ Dana hásʼ OpBi hátʃʼ 524 OpMo hátʃʼ OpPa hátʃʼ OpKi hátʃʼ *hāsʼ+kʼɔ̀d v. ‘poor (be)_5’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa hāsʼ+kʼɔ̀d UdCh hāt ̪̓+kʼɔ̀ɗ Dana - OpBi - OpMo - OpPa - OpKi - *hat ̪̓is v. ‘sneeze_1’: Reconstructs to: PKmn GwHi hǎtʼìʃ GwLo - Komo - UdYa - UdCh - Dana hát ̪̓ìs OpBi ha̋tʼīs OpMo ha̋tʼīs OpPa ha̋tʼīs OpKi ha̋tʼīs *hāwā v. ‘yawn_2’: Reconstructs to: PGw GwHi hāwā GwLo hāwā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 525 *hɛǵɛ̀ v. ‘frighten_1’: Reconstructs to: PGw GwHi hɛǵɛ̀ GwLo hɛǵɛ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *hīgìm v. ‘hiccough_3’: Reconstructs to: PGw GwHi hīgìm GwLo hīgìm Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *hǐmì v. ‘fight_3’: Reconstructs to: PGw GwHi hǐmì GwLo hǐmì Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *hísʼ ~ jísʼ v*. ‘dress up_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo jíz UdYa jísʼ 526 UdCh hít ̪̓ Dana hízá OpBi hıd̋ʒá OpMo hıd̋ʒá OpPa hıd̋ʒá OpKi hıd̋ʒá *hɪ̀ntʼ v. ‘lost (be)_2’: Reconstructs to: PGw GwHi hɪ̀ntʼ GwLo hɪ̀ntʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *hɔ̄ v. ‘go’: Reconstructs to: PGw GwHi hɔ̄ GwLo hɔ̄ Komo hà UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *hɔcʼ v. ‘bite_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo wɔ́sʼ UdYa wɔ̌ʃʼ UdCh wɔ̌cʼ Dana hɔ́cʼ OpBi hɔ̄tʃʼɔ̄ OpMo hɔ̄tʃʼɔ̄ OpPa hɔ̄tʃʼ 527 OpKi hɔ̄tʃʼ *hɔ̄j v. ‘hunt_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo hɔ̄j UdYa - UdCh - Dana - OpBi hɔ̄j OpMo hɔ̄j OpPa hɔ̄j OpKi hɔ̄j *hɔ̀sà ~ wɔ̀sà v. ‘frighten_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana wɔ̀sà OpBi hɔ̄sā OpMo hɔ̄sā OpPa hɔ̄sā OpKi hɔ̄sā *hɔ̄ʃɔ́ʃ n. ‘sap_3’: Reconstructs to: PGw GwHi hɔ̄ʃɔ́ʃ GwLo hɔ̄ʃɔ́ʃ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *hwaj v. ‘love_1’: Reconstructs to: PCtrl GwHi - GwLo - 528 Komo hwāj UdYa - UdCh - Dana ɔ́j OpBi ɔ́já OpMo ɔ́já OpPa ɔ́já OpKi ɔ́já *ràkʰ n. ‘cloud, fog_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ràkʰ UdCh ràkʰ Dana - OpBi - OpMo - OpPa - OpKi - *rOk⁽ʰ⁾Oɲ n. ‘corner’: Reconstructs to: PKmn GwHi rɔ́kɔ̀n GwLo rɔ́kɔ̀n Komo rɔ́kɔ̀n UdYa rúgùɲ UdCh à+rúkūɲ Dana rɔ́kʰɔ̀n OpBi rʊ́kʊ̄n OpMo rʊ́kʊ̄n OpPa rʊ́kʊ̄n OpKi rʊ́kʊ̄n *RUKa n. ‘basket’: Reconstructs to: PKmn GwHi lúkà GwLo lúkà Komo lùg UdYa rǔgà UdCh - Dana rɔ̀kʰà OpBi rùwà 529 OpMo rɔ̀kà OpPa rɔ̀kà OpKi rɔ̀kà *rukʼ(i) v. ‘tie (knot)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa rūkʼ UdCh rūh Dana rwí OpBi rwì OpMo rwì OpPa rwì OpKi rwì *rúm v. ‘frighten_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa rúm UdCh rúm Dana - OpBi - OpMo - OpPa - OpKi - *lala v. ‘crawl_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana lálá OpBi lālá OpMo lālá OpPa lālá OpKi lālá 530 *làlɪ ́v. ‘annoint (with oil)_5’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana làlɪ ́ OpBi làlɪ ́ OpMo làlɪ ́ OpPa làlɪ ́ OpKi làlɪ ́ *lám+ɛ ̄v. ‘rinse face_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa lám+ɛ ̄ UdCh lám+ʔɛ ̄ Dana - OpBi - OpMo - OpPa - OpKi - *langarɛt n. ‘bed_1’: Reconstructs to: PKmn GwHi ángàr GwLo - Komo ángàr UdYa áŋgàr UdCh - Dana - OpBi làŋgàrɛt́ OpMo àŋgàrípʰ OpPa àŋgàrípʰ OpKi àŋgàrípʰ *latɔ̪n n. ‘cotton, thread, spider web_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - 531 UdCh - Dana lát ̪h ɔ̄n OpBi lǎtɔ́n OpMo lǎtɔ́n OpPa lǎtɔ́n OpKi lǎtɔ́n Dana aspirates voiceless interdental intervocalically. *làwà v. ‘cut (meat into one long piece)_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi làwà OpMo làwà OpPa - OpKi - *lEt ̪̓a n. ‘tongue_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo lɛ̀tʼ UdYa lɛ̀d UdCh à+lɛ̀ɗ Dana lɪ̀t ̪̓á OpBi lītʼá OpMo lītʼá OpPa lītʼá OpKi lītʼá *lɛ̀pʰɛ ́n. ‘claves (instrument)_1’: Reconstructs to: PKmn GwHi lɛ̀pɛ ́ GwLo àlàpɛ ́ Komo lɛ̀pɛ ́ UdYa lɛ̀pʰɛ ́ UdCh - Dana lɛ̀pʰɛ ́ OpBi àlɛ̀pʰɛ ́ OpMo àlɛ̀pʰɛ ́ 532 OpPa àlɛ̀pʰɛ ́ OpKi àlɛ̀pʰɛ ́ *lili v. ‘sink (descend)’: Reconstructs to: PKmn GwHi līlí GwLo līlí Komo lìlí UdYa - UdCh - Dana līl OpBi līlí OpMo līlí OpPa līlı ̋ OpKi līlí *lìtʰ n. ‘large bird’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi lìtʰ OpMo lìtʰ OpPa lìtʰ OpKi lìtʰ *lɪ ̄v. ‘extract tooth_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi lɪ ̄ OpMo lɪ ̄ OpPa lɪ ̄ OpKi lɪ ̄ *lɔ̀b v. ‘play_3’: Reconstructs to: PUd GwHi - 533 GwLo - Komo - UdYa lɔ̀b UdCh lɔ̀ɓ Dana - OpBi - OpMo - OpPa - OpKi - *lɔ̀gɔ́ v. ‘tell_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana lɔ̀gɔ́ OpBi - OpMo - OpPa - OpKi lɔ̀gɔ́ *lɔ̀j n. ‘antelope_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi lɔ̀j OpMo lɔ̀j OpPa - OpKi - *lɔ̀lɔ̀kʼ n. ‘brain_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo lɔ̀lɔ̀kʼ UdYa - UdCh - Dana lɔ̀lɔ̀ʔ 534 OpBi lɔ̀lɔ̀kʼ OpMo lɔ̀lɔ̀kʼ OpPa lɔ̀lɔ̀kʼ OpKi lɔ̀lɔ̀kʼ *lɔ̀ŋɔ̀ n. ‘testicles_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana lɔ̀ŋɔ̀ OpBi lɔ̀ŋɔ̀ OpMo lɔ̀ŋɔ̀ OpPa lɔ̀ŋɔ̀ OpKi lɔ̀ŋɔ̀ *lɔ́s v. ‘swallow_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa lɔ́s UdCh lɔ́s Dana - OpBi - OpMo - OpPa - OpKi - *lud̪(a) v. ‘copulate (animal)_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo lùd UdYa - UdCh - Dana lùd̪á OpBi lūdá OpMo lūdá OpPa lūdá OpKi lūdá 535 *lus ~ rus v*. ‘hide (sth.)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo lùs UdYa - UdCh - Dana rùs OpBi lùs OpMo lùs OpPa rùs OpKi - *lut⁽̪ʰ⁾u n. ‘trunk (of elephant)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo lūd UdYa - UdCh - Dana lùt ̪h ú OpBi - OpMo - OpPa - OpKi - Komo voices intervocalically: *t ̪> d /V_V. Subsequent loss of final vowel in Komo and retention of word-final /d/. *lùt ̪̓ ~ lʊ̀t ̪̓ n. ‘testicles_1’: Reconstructs to: PKmn GwHi dūtʼ GwLo dūtʼ Komo lʊ̀tʼ UdYa lùd UdCh à+lùɗ Dana - OpBi - OpMo - OpPa - OpKi - Reduplication *lùt ̪̓ > lùt ̪̓ūt ̪̓ > lùd̪ūt ̪̓ > lùdūtʼ > dutʼ in Gwama? 536 *lʊm n. ‘bird_vulture (white-backed)_1’: Reconstructs to: PKmn GwHi ʊ̄+lʊ̌m GwLo - Komo lʊ̌m UdYa lùʔ UdCh - Dana lʊ̄m OpBi lʊ̌m OpMo lʊ̌m OpPa lʊ̌m OpKi lʊ̌m *mà(ʔV) n. ‘food_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo mɛ̀ʔí UdYa mà UdCh mà Dana màʔá OpBi mǎ OpMo mǎ OpPa mǎ OpKi mǎ *mā+kɪḱjātà n. ‘women_2’: Reconstructs to: PGw GwHi mā+kɪḱjātà GwLo mā+kɪḱjātà Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *mǎjā v. ‘criticize_2’: Reconstructs to: PGw GwHi mǎjā GwLo mǎjā Komo - UdYa - 537 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *mákʰ n. ‘fox_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa mákʰ UdCh à+mákʰ Dana - OpBi - OpMo - OpPa - OpKi - *mal(at) ~ mɔl v*. ‘not know (how)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo màlàt UdYa mɔ̀l UdCh mɔ̀l Dana - OpBi tā+mál OpMo tā+mál OpPa tā+mál OpKi tā+mál *mal(i) v. ‘fish (v.)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo māl UdYa mālí UdCh - Dana màl OpBi màl OpMo màl OpPa màl 538 OpKi màl *màm(a) v. ‘carry on back_1’: Reconstructs to: PKmn GwHi màm GwLo màm Komo màmá UdYa màm UdCh màm Dana màmá OpBi màmā OpMo màmā OpPa màmā OpKi màmā *mǎn n. ‘children_3’: Reconstructs to: PGw GwHi mǎn GwLo mǎn Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *mǎn n. ‘young people_2’: Reconstructs to: PGw GwHi mǎn GwLo mǎn Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *maɲa n. ‘saliva_1’: Reconstructs to: PKoUd GwHi - GwLo - 539 Komo mālà UdYa màɲà UdCh màɲà Dana - OpBi - OpMo - OpPa - OpKi - *màʃ n. ‘wife_1, marry, wedding’: Reconstructs to: PKmn GwHi - GwLo pʼā+màʃ Komo màʃ UdYa màʃ UdCh màʃ Dana màʃ OpBi màs OpMo màs OpPa màs OpKi màs Range of meanings: ‘wife’ in Gwama, ‘wedding’ in Komo, wife and ‘marry’ in Dana, wife in Opo *mɛ n. ‘goat_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo mɛ ́ UdYa mɛ̀ UdCh mì Dana mɛ ̄ OpBi mɛ̀ OpMo mɛ̀ OpPa mɛ̀ OpKi mɛ̀ *mɛʃ́ɛ ́v. ‘sharpen_2’: Reconstructs to: PGw GwHi mɛʃ́ɛ ́ GwLo mɛʃ́ɛ ́ Komo - UdYa - UdCh - 540 Dana - OpBi - OpMo - OpPa - OpKi - *mɛt(̪i) v. ‘chase’: Reconstructs to: PCtrl GwHi - GwLo - Komo mɛ̀tí UdYa - UdCh - Dana mɛt́ ̪h OpBi mɛt́ī OpMo mɛt́ī OpPa mɛt́ī OpKi mɛt́ī *mɛ̀t ̪̓ n. ‘hand_1’: Reconstructs to: PKmn GwHi bɪ̀tʼ GwLo mɪ̀tʼ Komo - UdYa mɛ̀d UdCh mɛ̀ɗ Dana mɛ̀t ̪̓ OpBi mɪ̀tʼɪ ́ OpMo mɪ̀tʼɪ ́ OpPa - OpKi mɛ̀tʼ *mililu ~ milili n*. ‘bird_cormant (whistling duck)’: Reconstructs to: PCtrl GwHi - GwLo - Komo à+mīlílū UdYa - UdCh - Dana mīlílū OpBi mīlílì OpMo mīlílì OpPa mīlílì OpKi mīlílì 541 *mítʼī v. ‘grind (fine)_2’: Reconstructs to: PGw GwHi mítʼī GwLo mítʼī Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *mɪmɪ n. ‘mosquito_2, firefly’: Reconstructs to: PKmn GwHi mɪm̌ɪ ́ GwLo mɪm̌ɪ ́ Komo mɪm̌ɪ ́ UdYa mǐmī UdCh - Dana - OpBi - OpMo - OpPa - OpKi - Meaning is ‘firefly’ in Yabus Uduk. *mɪn̄ v. ‘tie (bundle)_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi mɪn̄ OpMo mɪn̄ OpPa mɪn̄ OpKi mɪn̄ 542 *mɪś n. ‘sky_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo mɪʃ́ UdYa mís UdCh mís Dana - OpBi - OpMo - OpPa - OpKi - *Kɪ̀-mɪś v. ‘rise (oneself)_4’: Reconstructs to: PKoUd GwHi - GwLo - Komo gɪ̀mɪʃ́ UdYa cì+mís UdCh cì+mís Dana - OpBi - OpMo - OpPa - OpKi - Unsure whether initial consonant was voiced or voiceless. *mɪʃ v. ‘know or be able_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo mɪ̀ʃ UdYa míʃ UdCh míʃ Dana - OpBi - OpMo - OpPa - OpKi - *mɔ̀nd̪ɛ̀ɗ n. ‘farm (n.)_2’: Reconstructs to: PUd GwHi - GwLo - Komo - 543 UdYa mɔ̀nzɛ̀d UdCh mɔ̀nd̪ɛ̀ɗ Dana - OpBi - OpMo - OpPa - OpKi - *mɔra n. ‘fat (from animals)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo mɔ̌rá UdYa - UdCh - Dana mɔ́rá OpBi mɔ̄rá OpMo mɔ̄rá OpPa mɔ̄rá OpKi mɔ̄rá *mɔ̀s v. ‘clan_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo mɔ̀s UdYa - UdCh - Dana mɔ̀s OpBi mɔ̀s OpMo mɔ̀s OpPa mɔ̀s OpKi mɔ̀s *mùr n. ‘hair_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo mùl UdYa mùr UdCh à+mùr Dana - OpBi - OpMo - 544 OpPa - OpKi - *mùsʼ v. ‘rinse face_3’: Reconstructs to: PGw GwHi mùsʼ GwLo mùsʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *mùsʼ n. ‘soup_2’: Reconstructs to: PGw GwHi mùsʼ GwLo mùsʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *mùtʰa̋ v. ‘finish_3’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi mùtʰa̋ OpMo mùtʰa̋ OpPa mùtʰá OpKi mùtʰá 545 *mʊ̀sʼ v. ‘wring_3’: Reconstructs to: PGw GwHi mʊ̀sʼ GwLo mʊ̀sʼ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *mVsʼ(a) v. ‘shut_1, close eyes’: Reconstructs to: PKmn GwHi mɪ̀sʼ GwLo mɪ̀sʼ Komo mʊ̀sʼ UdYa músʼ+ɛ ̄ UdCh mút ̪̓+ɛ ̄ Dana mìsʼà OpBi mʊ̀tʃʼà OpMo mʊ̀tʃʼà OpPa mʊ̀tʃʼà OpKi mʊ̀tʃʼà Cannot account for the variation in vowels. *nǎ v. ‘pick up (small things), peck at_2’: Reconstructs to: PGw GwHi nǎ GwLo nǎ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *nàbɔ̀ng(w)à n. ‘bird_pelican (brown)’: Reconstructs to: PKmn GwHi - GwLo ʊ̄+nàbɔ̀ŋà Komo nàbɔ̀ŋà 546 UdYa - UdCh - Dana nàbɔ̀ŋgɔ̀ OpBi nàbɔ̀ŋgɔ̀ OpMo nàbɔ̀ŋgɔ̀ OpPa nàbɔ̀ŋgɔ̀ OpKi nàbɔ̀ŋgɔ̀ *nàm v. ‘ask (inquire)_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana nàm OpBi - OpMo - OpPa nàm OpKi nàm *nāmā v. ‘trade or barter_2’: Reconstructs to: PGw GwHi nāmā GwLo nāmā Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *nap(a) v. ‘disabled (be), angry (be)’: Reconstructs to: PKmn GwHi nǎpá GwLo nǎpá Komo nǎpá UdYa - UdCh nápʰɛ ̄ Dana nápʰ OpBi nāp OpMo nāp 547 OpPa nāp OpKi nāp *nɛḱʰ n. ‘bird_hammerkopf_1’: Reconstructs to: PKmn GwHi - GwLo ʊ̄+nɛḱ Komo à+nɛḱ UdYa nɛḱʰ UdCh à+nɛḱʰ Dana à+nɪḱʰ OpBi - OpMo - OpPa à+nɪḱʰ OpKi à+nɪḱʰ *nɪ̀ n. ‘antelope (dikdik, small deer)_2’: Reconstructs to: PGw GwHi ʊ̄+nɪ̀ GwLo nɪ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *nɪn̄ɪ ̄v. ‘hide (sth.)_2’: Reconstructs to: PGw GwHi nɪ̃ ̄ GwLo nɪn̄ɪ ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 548 *njansʼɛsʼ n. ‘honey badger’: Reconstructs to: PKmn GwHi jānsʼɛśʼ GwLo ɲɛn̄zɛśʼ Komo nɛ̀z UdYa nɛ̀sʼ UdCh à+nɛ̀t ̪̓ Dana nɛ̀s OpBi nɛ̀dʒɛ̀ OpMo - OpPa nɛ̀dʒɛ̀ OpKi - *nokʼa v. ‘spoil (become spoiled)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana nɔ́kʼà OpBi nɔ̌kʼá OpMo nɔ̌kʼá OpPa nɔ̌kʼá OpKi nɔ̌kʼá *nɔg v. ‘blame (somebody)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo nɔ̀g UdYa - UdCh - Dana nɔ̂g OpBi nɔ̀gá OpMo nɔ̀gá OpPa nɔ̀gá OpKi nɔ̀gá *nɔ́kɔ́ v. ‘good (be)_3’: Reconstructs to: PGw GwHi nɔ́kɔ́ GwLo nɔ́kɔ́ Komo - UdYa - 549 UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *nɔ́s n. ‘pottery, pot_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa nɔ́s UdCh à+nɔ́s Dana - OpBi - OpMo - OpPa - OpKi - *nʊpʼ ~ nʊmpʼ v*. ‘bury (sideways)’: Reconstructs to: PKmn GwHi ʊ̄mpʼ GwLo ʊ̄mpʼ Komo - UdYa - UdCh - Dana nʊ̀pʼá OpBi - OpMo - OpPa - OpKi - *ànùrɛ̀ n. ‘bird_hammerkopf_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi ànùrɛ̀ OpMo ànùrɛ̀ OpPa - 550 OpKi - *ɲa(n) n. ‘goat_2’: Reconstructs to: PKmn GwHi njã ̄ GwLo njã ̄ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ɲàrá n. ‘girl_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ɲàrá UdCh ɲàrá Dana - OpBi - OpMo - OpPa - OpKi - *ɲáw n. ‘cat_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ɲáw UdCh à+ɲáw Dana - OpBi - OpMo - OpPa - OpKi - *ɲɔ̀kʰ v. ‘try (test)_3’: Reconstructs to: PUd GwHi - GwLo - 551 Komo - UdYa ɲɔ̀kʰ UdCh ɲɔ̀kʰ Dana - OpBi - OpMo - OpPa - OpKi - *ɲɔ̀r v. ‘pain, be hurt_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ɲɔ̀r UdCh ɲɔ̀r Dana - OpBi - OpMo - OpPa - OpKi - *ɲú v. ‘wring_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ɲú UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *ɲùrúɲ n. ‘hyena_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ŋūríŋsʼíʔ UdCh à+ɲùrúɲ Dana - OpBi - 552 OpMo - OpPa - OpKi - *ŋápʰ v. ‘want_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa ŋápʰ UdCh ŋápʰ Dana - OpBi - OpMo - OpPa - OpKi - *ŋEɟ v. ‘teach_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ŋɪ̀ɟ OpBi ŋɛ̀tʃ OpMo - OpPa - OpKi ŋitʃ *àŋɛ̀ràŋ n. ‘arrow_1’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi àŋɛ̀ràŋ OpMo - OpPa àŋɛ̀ràŋ OpKi - 553 *ŋwan n. ‘four_3’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana ŋwān OpBi hwàn OpMo hwàn OpPa hwàn OpKi hwàn Borrowed from or related to *ŋwan Proto-Nilotic (Dimmendaal 1988:39). *(w)àʃ n. ‘wife_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo wàʃ UdYa āʃ UdCh āʃ Dana - OpBi - OpMo - OpPa - OpKi - *wa v. ‘break (v.)_1’: Reconstructs to: PKmn GwHi wǎ GwLo wǎ Komo wà UdYa wá UdCh wá Dana - OpBi - OpMo - OpPa - OpKi - *wà v. ‘dwell (live, reside)_5’: Reconstructs to: POp GwHi - GwLo - Komo - 554 UdYa - UdCh - Dana - OpBi wà OpMo wà OpPa wà OpKi wà *wa(n)t ̪̓(Iʃ) n. ‘fire or firewood’: Reconstructs to: PKmn GwHi āntʼ GwLo āntʼ Komo wàtʼíʃ UdYa ɔ̄d UdCh ɔ̄ɗ Dana ɔ̀t ̪̓ OpBi wʊ̄tʼɪ ́ OpMo wʊ̄tʼɪ ́ OpPa wʊ̄tʼɪ ́ OpKi ʊ̄tʼɪ ́ *wàɓ n. ‘boar (wild)_1’: Reconstructs to: PKmn GwHi wàpʼ GwLo wàpʼ Komo wàpʼ UdYa wàpʰ UdCh à+wàɓ Dana - OpBi - OpMo - OpPa - OpKi - *waC(a) adv. ‘fast (velocity)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɪ+́wàʃà UdYa ŋwácʰà UdCh wácʰà Dana wāt ̪h OpBi wāt OpMo wāt 555 OpPa wāt OpKi wāt *wàcʼà n. ‘fish (n.) general term’: Reconstructs to: PKmn GwHi wàsʼ GwLo wàsʼ Komo wàsʼ UdYa wàʃʼ UdCh wàcʼ Dana wàcʼà OpBi wàtʃʼà OpMo wàtʃʼà OpPa wàtʃʼà OpKi wàtʃʼà *wad̪ n. ‘home, place_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana wád̪ OpBi - OpMo - OpPa - OpKi wǎtʰ *wàg v. ‘criticize_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo wàg UdYa wàkʰ UdCh wàkʰ Dana wàg OpBi wàg OpMo wàg OpPa wàg OpKi wàg *wàkʰ n. ‘clan_4’: Reconstructs to: PUd GwHi - 556 GwLo - Komo - UdYa wàkʰ UdCh wàkʰ Dana - OpBi - OpMo - OpPa - OpKi - *waŋa n. ‘chicken_1’: Reconstructs to: PKmn GwHi wāŋā GwLo wāŋā Komo wàgá UdYa ŋwá UdCh à+ŋwá Dana - OpBi - OpMo - OpPa - OpKi - *wǎŋʊ́ n. ‘fox_2’: Reconstructs to: PGw GwHi wǎŋʊ́ GwLo wǎŋʊ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *wār n. ‘child_1’: Reconstructs to: PGw GwHi wār GwLo wāl Komo - UdYa - UdCh - Dana - 557 OpBi - OpMo - OpPa - OpKi - *wárkɪ̀n v. ‘shallow (be)_3’: Reconstructs to: PGw GwHi wárkɪ̀n GwLo wálkɪ̀n Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *was(ikʼ) v. ‘boil (of liquid)_1’: Reconstructs to: PKmn GwHi wǎʃ GwLo wàʃí Komo wàʃíkʼ UdYa - UdCh - Dana wás OpBi wās OpMo wās OpPa wās OpKi wās *wasakʼ n. ‘hail, ice_1’: Reconstructs to: PKmn GwHi wàsà GwLo wàsà Komo wàʃākʼ UdYa wàsáʔ UdCh à+wàsáʔ Dana - OpBi - OpMo - OpPa - OpKi - 558 *wata̪la n. ‘fox_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo wātálà UdYa - UdCh - Dana à+wāt ̪h álā OpBi à+wātálà OpMo à+wātálà OpPa à+wātálà OpKi à+wātálà *wɛ ́v. ‘dress up_2’: Reconstructs to: PGw GwHi wɛ ́ GwLo wɛ ́ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *wɔ v. ‘tell_3’: Reconstructs to: PKoUd GwHi - GwLo - Komo wɔ̀ UdYa ɔ́ UdCh ɔ́ Dana - OpBi - OpMo - OpPa - OpKi - *wɔ̀d̪ v. ‘help_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo wɔ̀l UdYa wɔ̀s 559 UdCh wɔ̀t ̪h Dana - OpBi - OpMo - OpPa - OpKi - *wɔ̀j n. ‘axe_5’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi wɔ̀j OpMo wɔ̀j OpPa wɔ̀j OpKi wɔ̀j *wɔ̀nɛ̀ n. ‘foot or leg_2’: Reconstructs to: POp GwHi - GwLo - Komo - UdYa - UdCh - Dana - OpBi wɔ̀nɛ̀ OpMo wɔ̀nɛ̀ OpPa wɔ̀n OpKi wɔ̀n *wɔ̀r n. ‘river_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa wɔ̀r UdCh wɔ̀r Dana - OpBi - OpMo - OpPa - 560 OpKi - *wɔ̀ʃ n. ‘stone or rock_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo ɔ̀ʃ UdYa wɔ̀ʃ UdCh wɔ̀ʃ Dana - OpBi - OpMo - OpPa - OpKi - *wɔ̀t ̪h v. ‘mediate_5, help’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa wòs UdCh wɔ̀t ̪h Dana - OpBi - OpMo - OpPa - OpKi - *wúpʰ v. ‘brood (v.)_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa wúpʰ UdCh wùpʰ Dana - OpBi - OpMo - OpPa - OpKi - *wUs n. ‘sky_1’: Reconstructs to: PKmn GwHi wús GwLo wús 561 Komo - UdYa - UdCh - Dana - OpBi wʊ̀s OpMo wʊ̀s OpPa wʊ̀s OpKi wʊ̀s *wúsʼ n. ‘enset_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo wúsʼ UdYa wúsʼ UdCh à+wút ̪̓ Dana - OpBi - OpMo - OpPa - OpKi - *wutʰ n. ‘bird_ostrich_1’: Reconstructs to: PKmn GwHi - GwLo wǔt Komo wǔt UdYa útʰ UdCh à+útʰ Dana - OpBi hűtʰ OpMo hűtʰ OpPa hűtʰ OpKi hűtʰ Possibly borrowed from W.Nilotic (cf. Kurmuk /ʔúuɗù/ (Andersen 2007:75)) *wVɗ v. ‘become, become angry’: Reconstructs to: PKmn GwHi wɛť GwLo wɛť Komo wàl UdYa wár UdCh wáɗ Dana wâl 562 OpBi - OpMo - OpPa - OpKi - *(pa)jasʼ v. ‘vomit_1’: Reconstructs to: PKmn GwHi pājàsʼ GwLo pājàsʼ Komo jàʔ UdYa ɟǎʔ UdCh jǎʔ Dana - OpBi - OpMo - OpPa - OpKi - *jàgàl n. ‘sweat (substance)_2’: Reconstructs to: PKmn GwHi jàgàl GwLo jàgɪ̀ Komo - UdYa - UdCh - Dana - OpBi dʒīkʼāj OpMo dʒīkʼāj OpPa zīkʼāj OpKi ʃīkʼāj Possibly not cognate between Opo and Gwama. Opo most likely compound water + good meaning ‘good water’. *jàhút ̪n. ‘fish_(small, small scales)’: Reconstructs to: PKmn GwHi jàhú GwLo jàhú Komo jàhú UdYa - UdCh - Dana àhút ̪h OpBi àhű OpMo àhű OpPa àhűwı ̋ 563 OpKi àhűj *jàkʼ n. ‘son’: Reconstructs to: PCtrl GwHi - GwLo - Komo jà UdYa jàʔ UdCh jàʔ Dana jàkʼ OpBi dʒàkʼ OpMo dʒàkʼ OpPa zàkʼ OpKi sàkʼ *jàn v. ‘pray, beg_2’: Reconstructs to: PGw GwHi jàn GwLo jàn Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *jàsʼ v. ‘wet_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo jɛ̀sʼ UdYa jàsʼ UdCh jàt ̪̓ Dana - OpBi - OpMo - OpPa - OpKi - 564 *jawa ? n. ‘stone or rock_3, grindstone (bottom)’: Reconstructs to: PCtrl GwHi - GwLo - Komo wàʔí UdYa jɔ̀ʔ UdCh à+jɔ̀ Dana jàwà OpBi dʒàw OpMo dʒàw OpPa zàw OpKi ʃàwà Narrowing of meaning to ‘grindstone’ in PKoUd. *jEkʰ v. ‘sow seeds (by throwing)_1’: Reconstructs to: PKmn GwHi jɪ̀ GwLo jɪ̀ Komo jɛ̀k UdYa jɛ̀kʰ UdCh jɛ̀kʰ Dana - OpBi - OpMo - OpPa - OpKi - *jEn(ɪ) n. ‘oil (organic substance)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo jɛ̀n UdYa jìn UdCh à+jìn Dana jɪń OpBi dʒɪn̄ɪ ́ OpMo dʒɪn̄ɪ ́ OpPa zɪn̄ OpKi ʃɪn̄ *jEsI v. ‘wet_1, slippery’: Reconstructs to: PKmn GwHi ɪʃ́ɪ̀ GwLo ɪʃ́ɪ̀ Komo jɛ̀ʃ 565 UdYa jɛ̀s UdCh jɛ̀s Dana sìʔ OpBi - OpMo - OpPa - OpKi - *jɛkʰ v. ‘extract tooth_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo jɛ̀ʔ UdYa ɛḱʰ UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *jɛ̀sʼ n. ‘tree_sp.’: Reconstructs to: PKoUd GwHi - GwLo - Komo jɛ̀sʼ UdYa jɛ̀sʼ UdCh à+jɛ̀t ̪̓ Dana - OpBi - OpMo - OpPa - OpKi - *jìcʼ n. ‘skin, hide (of animal), bark of tree_3’: Reconstructs to: PKoUd GwHi - GwLo - Komo jìsʼ UdYa jìnʒà UdCh jìcʼ Dana - OpBi - OpMo - 566 OpPa - OpKi - *cʼ > ʒ /V_V in Yabus Uduk. *jiɗ n. ‘skin, hide (of animal), bark of tree_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa jīd UdCh à+jǐɗ Dana jɪ̀ɗ+mà+ʃúmà OpBi - OpMo - OpPa - OpKi - *jiɗE n. ‘water’: Reconstructs to: PKmn GwHi ìjáʔ GwLo ìjáʔ Komo jǐ UdYa wùdíʔ UdCh jìɗɛʔ́ Dana jìʔí OpBi dʒì OpMo dʒì OpPa zì OpKi ʃìʔ Cannot explain initial /wu/ in Yabus Uduk. *jigi v. ‘protect_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo jìgíʔ UdYa - UdCh - Dana - OpBi dʒīgí OpMo dʒīgí OpPa - OpKi - 567 *jikin v. ‘shake (sth.)_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa jígìn UdCh jǐkʰīn Dana - OpBi - OpMo - OpPa - OpKi - *jíl v. ‘play_2’: Reconstructs to: PGw GwHi jíl GwLo jí Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *jílɔ́ŋ n. ‘shadow_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana jílɔ́ŋ OpBi - OpMo - OpPa - OpKi jıl̋ɔ́ŋ *jìman n. ‘sap_1’: Reconstructs to: PKoUd GwHi - GwLo - Komo ìmàn+sá UdYa - 568 UdCh jìmán Dana - OpBi - OpMo - OpPa - OpKi - *jìʃ n. ‘maggot_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa jìʃ UdCh à+jìʃ Dana - OpBi - OpMo - OpPa - OpKi - *jɪ v. ‘play (instrument)_2’: Reconstructs to: PKmn GwHi jɪ̀ GwLo jɪ̀ Komo - UdYa - UdCh - Dana - OpBi dʒɪ ̄ OpMo dʒɪ ̄ OpPa zɪ ̄ OpKi ʃɪ ̄ *jɪ ̌v. ‘leave_1’: Reconstructs to: PGw GwHi jɪ ̌ GwLo jɪ ̌ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - 569 OpKi - *jɪd̪̄ n. ‘axe_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo jɪʔ̄ UdYa - UdCh - Dana jɪd̪̄ OpBi - OpMo - OpPa - OpKi - *jɪp̄ʼ v. ‘stab_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo jɪp̄ʼ UdYa jípʰ UdCh jípʰ Dana - OpBi - OpMo - OpPa - OpKi - Deglottalization word-finally *pʼ > pʰ in Proto-Uduk? *jɔ́l v. ‘buy_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa jɔ́l UdCh jɔ́l Dana - OpBi - OpMo - OpPa - OpKi - 570 *jɔ́l v. ‘sell_2’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa jɔ́l UdCh jɔ́l Dana - OpBi - OpMo - OpPa - OpKi - *jɔ̀r v. ‘grind (dry)_4’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa jɔ̀r UdCh jɔ̀r Dana - OpBi - OpMo - OpPa - OpKi - *jú v. ‘pour_5’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa jú UdCh jú Dana - OpBi - OpMo - OpPa - OpKi - *juga v. ‘call_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa júkʰ 571 UdCh júkʰ Dana - OpBi dʒùgà OpMo dʒùgà OpPa zùgà OpKi ʃùgà Also means ‘sing’ in Chali Uduk *jVl v. ‘set (heavenly bodies)_3’: Reconstructs to: PUd GwHi - GwLo - Komo - UdYa jíl UdCh júl Dana - OpBi - OpMo - OpPa - OpKi - *T̪A(t ̪h ) n. ‘arrow_3’: Reconstructs to: PCtrl GwHi - GwLo - Komo sá UdYa sá UdCh à+tá̪ Dana sɛt̄ ̪h OpBi - OpMo tʃɛ ̄ OpPa - OpKi tʃɛ ̄ Cognacy of KoUd with DaOp is questionable. *T̪ál v. ‘resemble (reflect)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo sál UdYa sál UdCh tá̪l Dana hálà OpBi tʃál 572 OpMo tʃál OpPa tʃál OpKi tʃál *T̪in n. ‘tail (of animal)_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo sìn UdYa sīn UdCh tī̪n Dana - OpBi - OpMo - OpPa - OpKi - *(a)T̪ina ? n. ‘nosebleed_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo zīnà UdYa - UdCh - Dana - OpBi àtʃwìnà OpMo àtʃwìnà OpPa àtʃwìnà OpKi àtʃwìnà Cannot account for voicing in intial /z/ in Komo. *T̪is v. ‘lost (be)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa tís UdCh tí̪s Dana sís OpBi sísī OpMo sísī OpPa sísī OpKi sísī 573 *T̪isʼ(a) v. ‘urinate_3’: Reconstructs to: PCtrl GwHi - GwLo - Komo - UdYa - UdCh tī̪t ̪̓ Dana - OpBi tʃītʃʼá OpMo tʃītʃʼá OpPa tʃītʃʼá OpKi tʃītʃʼá *T̪Ol(a) v. ‘descend (land)’: Reconstructs to: PKmn GwHi - GwLo ʃʊ̀ Komo - UdYa sūl UdCh tū̪l Dana - OpBi tʃɔ́já OpMo tʃɔ́já OpPa tʃɔ́já OpKi tʃɔ́já *T̪út⁽̪ʰ⁾(a) v. ‘set (heavenly bodies)_4’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana sút ̪h ā OpBi - OpMo - OpPa - OpKi tʃútʰá *T̪ʊ(n)G n. ‘guinea fowl_1’: Reconstructs to: PKmn GwHi ʃʊ̌nkʼ GwLo ʃʊ̌nkʼ Komo zʊ̀g UdYa - 574 UdCh tū̪kʰ Dana sʊ̄kʰ OpBi tʃʊ̄gʊ́ OpMo tʃʊ̄gʊ́ OpPa tʃʊ̄g OpKi tʃʊ̄g *D̪a v. ‘go_SG’: Reconstructs to: PCtrl GwHi - GwLo - Komo jà UdYa jà UdCh jà Dana d̪ā OpBi dʒà OpMo dʒà OpPa zà OpKi ʃà *D̪akʼa v. ‘vomit_2’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana d̪ákʼà OpBi dʒākʼá OpMo dʒākʼá OpPa zākʼá OpKi ʃākʼá *D̪E n. ‘eye_1’: Reconstructs to: PKmn GwHi zì GwLo zì Komo - UdYa ī UdCh ɛ ̄ Dana d̪ɛ̀ OpBi dʒɛ̀ OpMo dʒɛ̀ OpPa zɛ̀ 575 OpKi ʃɛ̀ *D̪E(se) n. ‘seed_1’: Reconstructs to: PKmn GwHi zì GwLo zì Komo - UdYa - UdCh - Dana d̪ɛ̀sɛ̂ OpBi dʒɛ̀sɛ ̄ OpMo dʒɛ̀sɛ ̄ OpPa zɛ̀sɛ ̄ OpKi sɛ̀sɛ ̄ Cf. ‘eye’ *D̪i v. ‘doze off_1’: Reconstructs to: PKmn GwHi sɪśɪ ̄ GwLo sɪśɪ ̄ Komo zìzí UdYa - UdCh - Dana - OpBi dìsì OpMo dìsì OpPa zìz OpKi sìs *D̪ìbàl n. ‘bird_Quelea (Red-billed)’: Reconstructs to: PCtrl GwHi - GwLo - Komo dìbàl UdYa - UdCh - Dana - OpBi à+dʒìbà OpMo à+dʒìbà OpPa à+dʒìbà OpKi à+dʒìbà Cannot account for lack of expected initial /z/ in Komo. 576 *D̪ir ? v. ‘green_1’: Reconstructs to: PKmn GwHi zı̃ ̀ GwLo zĩ ̂ Komo zì UdYa zì UdCh d̪ì Dana sīsī OpBi tʃìr OpMo tʃìr OpPa tʃìr OpKi tʃìr *D̪ita̪ n. ‘person_2’: Reconstructs to: PKmn GwHi sīt GwLo sīt Komo - UdYa - UdCh - Dana d̪ítà̪ OpBi - OpMo - OpPa ʊ̀+zìtà OpKi ʊ̀+ʃìtà *D̪ǐtʰ n. ‘shadow_5’: Reconstructs to: PCtrl GwHi - GwLo - Komo zǐt UdYa - UdCh - Dana - OpBi dʒǐtʰ OpMo dʒǐtʰ OpPa zǐtʰ OpKi ʃǐtʰ *D̪ɔm v. ‘string (v.) (e.g. beads)_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɔ̀m UdYa wɔ̌m 577 UdCh hɔ́m Dana d̪ɔ̀m OpBi dʒɔ̀m OpMo dʒɔ̀m OpPa dʒɔ̀m OpKi ʃɔ̀m *d̪ɔt ̪h ɛn n. ‘snake_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana d̪ɔ̄t ̪h ɛ̀n OpBi dʒɔ̌ OpMo dʒɔ̌ OpPa zɔ̌ OpKi ʃɔ̌ *D̪uga n. ‘name_1’: Reconstructs to: PCtrl GwHi - GwLo - Komo zāgà UdYa - UdCh - Dana - OpBi dʒùgà OpMo - OpPa - OpKi ʃùgà *D̪ùkʼà v. ‘grind (dry)_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana dùkʼà OpBi dʒùkʼà OpMo dʒùkʼà OpPa zùkʼà 578 OpKi ʃùkʼà Would expect initial /d̪/ in Dana. *D̪Vʃ n. ‘maggot_1’: Reconstructs to: PDaOp GwHi - GwLo - Komo - UdYa - UdCh - Dana dìʃ OpBi dùsù OpMo dʒùs OpPa zùs OpKi ʃùʃ Would expect initial /d̪/ in Dana. *zàl v. ‘sit_2’: Reconstructs to: PGw GwHi zàl GwLo zɛ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *zɪńzɪ̀ v. ‘thin (be)_2’: Reconstructs to: PGw GwHi zɪńzɪ̀ GwLo zɪńzɪ̀ Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 579 *zjaɗa(j) n. ‘chili pepper_1’: Reconstructs to: PKmn GwHi zɛt́ GwLo zɛt́ Komo zjàntʼá UdYa - UdCh - Dana à+zāt ̪h ɛʔ̄ OpBi dʒɛ̀rà OpMo - OpPa - OpKi - *zūgū v. ‘stand_2’: Reconstructs to: PGw GwHi zūgū GwLo zūgū Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - *zʊj n. ‘antelope (dikdik, small deer)_2’: Reconstructs to: PCtrl GwHi - GwLo - Komo zwɪ̀ UdYa wí UdCh à+ʔúj Dana zʊ̀j OpBi à+ʔʊ́jʊ́ OpMo à+ʔʊ́ɪ ́ OpPa à+ʔʊ́ɪ ́ OpKi - *ʔja v. ‘go_PL’: Reconstructs to: PCtrl GwHi - GwLo - Komo ɪ̀ UdYa ī 580 UdCh ī Dana - OpBi ʔjá OpMo ʔjá OpPa ʔjá OpKi ʔjá *ʔjamVn n. ‘seed_2’: Reconstructs to: PKoUd GwHi - GwLo - Komo ʔjāmʊ́n UdYa ɛm̄ɛń UdCh ɛm̄ɛń Dana - OpBi - OpMo - OpPa - OpKi - Also means ‘kidney’ in Uduk. *ʔù v. ‘grind (dry)_3’: Reconstructs to: PGw GwHi ʔù GwLo ʔù Komo - UdYa - UdCh - Dana - OpBi - OpMo - OpPa - OpKi - 581 APPENDIX C TONE CORRESPONDENCE SETS 9 APPENDIX C Tone correspondence sets C.1 PKMN Tone Correspondence Set A Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi cannabis_1 *bángwà báŋgà báŋgà bánkɔ̀ bǎŋgɔ̀ bǎŋgɔ̀ - - - - - bird_heron *bár - ʊ̄+bár à+bár - - à+bár à+bár à+bár à+bár à+bár strong (be)_1 *ɓɪśʼ ~ ɓísʼ pʼɪ ́ pʼɪ ́ ɓísʼ ɓísʼ ɓít ̪̓ - - - - - whistle *cut - ʃút ʃɪt́ - cúwā cʊ́ɪ̀ tʃúwì tʃúwì tʃúwì tʃúwì tree *cwálá swálá swájá sá ʃwá cwá cá tʃá tʃá tʃá tʃá Dazu (S. Sudan) *dáɟV dázɔ̀ dázɔ̀ dázʊ̀ dǎɟɔ̀ dǎɟɔ̀ dáɟɔ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ shout_1 *dɔl(ɔ) dɔ́l dɔ́l dɔ́l - - ɔ̀lɔ́ - - - - pipe (for *dʊ́ɟɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ - - dʊ́ɟɛ̀ - - - - smoking)_1 have sex_1 *hag(a) hã́ʔ hã́ʔ hág - háʔ - hágá hágá hágá hágá trample, ruminate *hasʼ - hāsʼɪ̀ hásʼ hásʼ hát ̪̓ hásʼ hátʃʼ hátʃʼ hátʃʼ hátʃʼ Opo (ethnonym)_1 *k⁽ʰ⁾ɪńáj kɪńá kɪńá kɪńá cʰínáj - kɪńáj - - - - sew_1 *kɔ́s ɔ́s ɔ́s - - - kɔ́s kɔ́s kɔ́s kɔ́s kɔ́s pound (v.)_4 *kɔ́pʰ kɔ́p kɔ́p - kúpʰ - - - - - - repair_1 *kʰaɓ áp - áb - - ʊ́bā kʰápʼ kʰápʼ kʰápʼ kʰápʼ shut_2 *kʰacʼ kǎʃ kǎʃ - kʰáʃʼ kʰácʼ - - - - - open *kʰáɗ(a) kálá kájá kár kʰǎd kʰǎɗ kʰátá kʰátá kʰátá kʰátá kʰátá giraffe_2 *kʰOG kɔ́k kɔ́k - kʰúʔ kʰúʔ kʰɛǵ - - - kɛ̂k dry (be) *kʰʊ́sʼ kúsʼ kúsʼ kʊ́sʼ kʰúsʼ kʰút ̪̓ kʰʊ́sʼ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ - *k⁽ʰ⁾akʼas ~ porcupine_1 kákʼàʃ kákʼàʃ káʃākʼ - - kásākʼ kʰásākʼ kʰásākʼ kʰásākʼ kʰáʃākʼ k⁽ʰ⁾asakʼ soft (be)_1 *kʼát ̪̓ kʼátʼ kʼátʼ kʼátʼ kʼád kʼáɗ - - - - - 582 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi kill, fight *kʼOʃ kʼʊ́ʃ kʼʊ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́sɔ́ kʼɔ́sɔ́ kʼɔ́s kʼɔ́ʃ head *kʼʊ́p kʼʊ́p ʔʊ́p kʼʊ́p kʼúpʰ kʼúpʰ kʼʊ́pʰ kʼʊ́p kʼʊ́p kʼʊ́p kʼʊ́p throat *kʼʊ́s k’úʃ k’úʃ kʼʊ́ʃ kʼús kʼús kʼʊ́s kʼʊ́sʊ́ kʼʊ́s kʼʊ́s kʼʊ́s bird_hammerkopf_ *nɛḱʰ - ʊ̄+nɛḱ à+nɛḱ nɛḱʰ à+nɛḱʰ à+nɪḱʰ - - à+nɪḱʰ à+nɪḱʰ 1 shake (sth.)_1 *pid(V) - pǐdí pídá pítʰ pír - - - - - vagina_1 *pít ̪h ~ pɪt́ ̪h pɪt́ pɪt́ pít - - pít ̪h - - - - platform *pʰárá párá párá párá - - pʰárá pʰárá pʰárá pʰárá pʰárá peel, husk_1 *pʼi(n)Cʼ? pʼɪńsʼ pʼɪńsʼ - - pʼíɗ pʼɪt́ʼ - - - - corner *rOk⁽ʰ⁾Oɲ rɔ́kɔ̀n rɔ́kɔ̀n rɔ́kɔ̀n rúgùɲ à+rúkūɲ rɔ́kʰɔ̀n rʊ́kʊ̄n rʊ́kʊ̄n rʊ́kʊ̄n rʊ́kʊ̄n cut (meat into one *sɪĺ(ɪ) ʃɪĺɪ̀ ʃɪ ́ - - - - sɪ ́ - - - long piece)_2 stab_1 *sɔ́p sɔ́ sɔ́ - - - - sɔ́p - sɔ́p - lead (guide)_1 *sʊs ʃʊ́ʃ ʃʊ́ʃ ʃʊ́ʃ sús sús sʊ́ʔ - - - - blood_1 *sʼámá sʼám sʼám - - - sʼámá tʃʼámá tʃʼámá tʃʼámá tʃʼámá cold (be)_1 *sʼʊ́p sʼʊ́p sʼʊ́p - sʼúpʰ t ̪̓úpʰ - - - - - copulate (animal)_3 *ʃʊ́ɓ ʃʊ́pʼ ʃʊ́pʼ - ʃúb ʃúɓ - - - - - grind (second grind) *(ɔ)t⁽̪ʰ⁾ɔɗ - twɛj́ tɔ́ - tɔ̪́ɗ ɔ̀t ̪h ɔ́ ɔ́tɔ́ ɔ́tʰ ɔ́tʰ ɔ́tʰ roughen (stone for *tɪ̪t́ ̪ sɪt́ sɪt́ tɪt́ títʰ tír tɪ̪t́ ̪h tɪt́ɪ ́ tɪt́ʰ tɪt́ʰ tɪt́ʰ grinding) kick *t ̪h áɓ tápʼ tápʼ táb tʰáb tʰáɓ t ̪h áp tʰáp tʰáp tʰáp tʰáp C.2 PKMN Tone Correspondence Set B Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi neck *ɓid̪a ~ ɓɪd̪a pʼɪl̄ pʼɪ ̄ ɓàʔ ɓāʔ ɓāʔ ɓìd̪à ɓījā ɓījā ɓījā ɓījā new_1 *d̪is ~ d̪ɪs dīʃ dīʃ zɪ̀ʃ tīs tī̪s - - - - - send *ɗar(a) tʼálà tʼájà ɗàr ɗɛt̄ʰ - ɗɛ̀d - ɗɛr̄ - ɗɛr̄ someone_1 deep (be)_1 *gɔ̀ɗɔkʼ kwì kwǐ gʊ̀r - - kɔ̀ɗɔ̀kʼ kɔ̄rɔ́ kɔ̄rɔ́ kɔ̄rɔ́ kɔ̄rɔ́ 583 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi short (be)_1 *KÙtʼ gʊ̄t gʊ̄t kùt kūtʰ kūtʰ - - - - - cry *kO(j) kū kū kɔ̀ kɔ̄ kɔ̄ kɔ̀j kwɛ ̄ kwɛ ̄ kwɛ ̄ kwɛ ̄ herd (v.) *kʰaj kɛ̃ʔ̄ kɛ̃ʔ̄ kàʔí - - kʰàjɪʔ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ bitter, sour (be) *kʰakʼa kāgā kāgā kàʔ kʰāʔ kʰāʔ kʰàkʼà kʰākʼā kʰākʼā kʰākʼā kʰākʼā dig (for *kʼwàd̪(ɪ)́ kʼwī kʼwī kʼɔ̀ kʼwā - kʼwàd̪ɪ ́ kʼɔ̄j wārɪ ́ wādɪ ́ kʼwārɪ ́ water)_1 tick *kʼwànt ̪̓ kʼwāntʼɪ ́̓kwāntʼ kʼwāntʼɪ ́̓kwāntʼ kʼwàtʼ kʼwātʰ à+kʼwāɗ kʼwàt ̪̓ kʼwātʼ kʼwātʼ kʼwātʼ kʼwātʼ sink (descend) *lili līlí līlí lìlí - - līl līlí līlí līlı ̋ līlí *nʊpʼ ~ bury (sideways) ʊ̄mpʼ ʊ̄mpʼ - - - nʊ̀pʼá - - - - *nʊmpʼ touch_1, crawl *pàD pāt pāt pàt pātʰ pār - - - - - wrap *pʊʃ ~ puʃ pʊ̄ʃ pʊ̄ʃ - pūʃ pūʃ púʃá pūsá pūsá pūsá pūsá laugh *pʰ(j)asʼ ɪs̄ʼ pāsʼ pɛ̀sʼ pʰɛs̄ʼ pʰɛt̄ ̪̓ pʰàsʼ pʰātʃʼ pʰātʃʼ pʰātʃʼ pʰātʃʼ soak_1 *pʰàcʼ pāsʼ pāsʼ pàsʼ pʰāʃʼ pʰācʼ - - - - - fly (v.) *pʰàd̪ pāl pāj pàj pʰāj pʰɛ ̄ pʰàd̪ pʰāj pʰāj pʰāj pʰāj far (be)_1 *sɪ̀t ̪̓ - ʃɪt̄ʼ ʃɪ̀tʼ sīd sīɗ sɪ̀t ̪̓ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ shave *sʼɛ̀d̪ ~ sʼɪ̀d̪ sʼɪl̄ sʼɪ ̄ sʼɛ̀ sʼī cʼɛ ̄ sʼɛ̀d̪ tʃʼɛ ̄ tʃʼɛ ̄ tʃʼɛ ̄ tʃʼɛ ̄ louse_1 *ʃOkʼ(VN) - ʃʊ̄gʊ̀n ʃùwɛ̀n ʃɔ̄kʼɔ̄m à+ʃɔ̄kʼɔ̄m ʃùkʼnáj sūkʼɛ́n sūkʼɛ́n sūkʼɛ́n ʃūkʼɛ́n foot or leg_1 *ʃɔnkʼ sɔ̄ŋkʼ sɔ̄ntʼ ʃɔ̀g ʃɔ̄kʼ ʃɔ̄ʔ ʃɔ̀g - - - - wake (trs.)_1 *ʃukʼ(V) sūgì sūgì ʃùg ʃūkʼ - ʃùg sūg sūg sūg ʃūg meat, animal *ʃum(a) sūm sūm ʃùm ʃūm ʃūm ʃùmà sūmā sūmā sūmā ʃūmā forehead_1 *t⁽ʰ⁾(w)àg twã twã ̄ tàg - - tàg bī+tʰāg - pī+tʰāg - be, do *tà - ta tà tā tā - tā - tā - salt_(made from ash of a *tʼaʃ tʼāʃ tʼāʃ tʼàʃ tʼāʃ tʼāʃ tʼàʃ tʼās - - - particular plant/tree)_2 fire or firewood *wa(n)t ̪̓(Iʃ) āntʼ āntʼ wàtʼíʃ ɔ̄d ɔ̄ɗ ɔ̀t ̪̓ wʊ̄tʼɪ ́ wʊ̄tʼɪ ́ wʊ̄tʼɪ ́ ʊ̄tʼɪ ́ 584 C.3 PKMN Tone Correspondence Set C Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi father_1 *bàbá - bàbá bǎ à+bàbá à+bàbá - àbá àbá àbá àbá wide (be) *bàj ~ ɓàj pàj pàj ɓàjá ɓàn bɛ̀ - pàj pàj pàj pàj drum_B *bàmbá pàmbà pàmbà - bàmbá - - - - - - toss, throw away, *bit⁽ʰ⁾ pɪ̀t pɪ̀t bìt bǐtʰ bǐtʰ - - - - - fall over hide, skulk *bɔ̀b ~ bɔ̀ɓ pɔ̀pʼ pɔ̀pʼ bɔ̀b - - - pɔ̀j pɔ̀j pɔ̀j - pregnant (be) *bUma pʊ̀m pʊ̀m ɓú pwá bwà pùmá pǔmá pǔmá pǔmá pǔmá belly or *bùʃùl bùʃùl bùʃì bùʃ - bùʃ - pùsà pùsà pùsà pùsà stomach_1 path, road_1, *bwaŋ(a) pwǎŋà pwǎŋà - - bwàj - - - - - towards palate_1 *ɓalilaj pʼàlílì pʼàlílí ɓàlílá - - - - - - - grandfather_1 *càkʰO - sàkʊ́ sàkʊ́ - - - tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ dig_1 *ɟà ʃà ʃà - - - cà tʃà tʃà tʃà tʃà baboon_1, dog_2 *dɪŋɪ tɪ̀nɪ̀ tɪ̀nɪ̀ - - - tɛ̀ŋ tɪ̀nɪ̀ tɪ̀nɪ̀ tɪ̀n tɪ̀n stomp, step on_1 *djalIsʼ dɛ̀lísʼ dàlísʼ dɪ̀l dìl dìl - tɪ̀lɪ ́ tɪ̀lɪ ́ tɪ̀lɪ ́ tɪ̀lɪ ́ urine_1, *dùcʼá dùsʼ tùsʼ dʊ̀sʼ - - tùcʼáʔ - - - - urinate_1 tree (sp.)(sausage tree_Kigelia *dùmàj - ū+dùmɪ̀ dùmɛ̀ - - dùmáj dùmàj dùmàj dùmàj dùmàj africana) grind wet (first *Dàsʼ dàsʼ dàsʼ nàsʼ nàsʼ nàt ̪̓ - - - - - grind) Dana *d̪ana dànā dànā dāná - - d̪àná dànā dànā dànā dànā (ethnonym)_3 strain (solids *d̪ìm - zìŋā zìm zǐm d̪ìm tì̪m tìm tìm tìm tìm from liquid)_1 bee, honey *d̪àm tàm tàm dàm dàm à+d̪àm tâ̪m tàm tàm tàm tàm seed_1 *D̪E(se) zì zì - - - d̪ɛ̀sɛ̂ dʒɛ̀sɛ ̄ dʒɛ̀sɛ ̄ zɛ̀sɛ ̄ sɛ̀sɛ ̄ 585 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi replace *(n(j)a)gaD nāgát njāgát gàdá gàs gàr gàtá kàrá kàrá kàrá kàrá smell (v.) *gàŋ(a) kɛ̃ ̄ kɛ̃ ̄ gàg - - kàŋà kàŋà kàŋà kàŋà kàŋà tie (bundle)_1 *gì(n)sʼ kɪ̀nsʼ kɪ̀nsʼ gìsʼ - - - - - - - pack in, stuff into *gɪ̀m gɪ̀m gɪ̀m - - ɟìm - - - - - container_1 skin, hide (of animal), bark of *gɔnkʼ(ɔʃ) gɔ̀kʼɔ́ʃ gɔ̀kʼɔ́ʃ gɔ̀nkʼí gɔ̀kʰ - kɔ̀gɔ̀ kɔ̀gɔ̀ kɔ̀gɔ̀ kɔ̀gɔ̀ gwàŋgɪ ́ tree_1 pig_1 *gUd̪Um kʊ̀rʊ́m kʊ̀rʊ́m gùdúm - - gùd̪ùm kùdùmà kùdùmà kùdùmà kùdùmà run (SG)_1, flow, *gùs ~ gʊ̀s gʊ̀s gʊ̀s gùʃ gùs gùs - - - - - bleed swallow_1 *gUsʼ gùsʼ kùsʼ gʊ̀sʼ - - kʊ̄sʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá elephant_1 *gwàj kwɪ̀ kwɪ̀ gwà gwàj - - - - - - Gwama *gwama gwàmá kwāmà gwàmá - - gɔ̀má gɔ̀má gɔ̀má gɔ̀má gɔ̀má (ethnonym) find, meet *gàm kàm kàm gàm gàm gàm kàm kàm kàm kàm kàm sow seeds (by *jEkʰ jɪ̀ jɪ̀ jɛ̀k jɛ̀kʰ jɛ̀kʰ - - - - - throwing)_1 wet_1, slippery *jEsI ɪʃ́ɪ̀ ɪʃ́ɪ̀ jɛ̀ʃ jɛ̀s jɛ̀s sìʔ - - - - water *ji(ɗE) ìjáʔ ìjáʔ jǐ wùdíʔ jìɗɛʔ́ jìʔí dʒì dʒì zì ʃìʔ sorghum, millet *ɟana sjànà sjànà zɛ̀nā - - - dʒɛ̀ná - dʒɛ̀ná - bird_stork *ɟàrú zɛ̀rú zɛ̀rú zɛ̀rú ʒàrú - ɟɛ̀rú dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű (maribou) grandmother_1 *kʰàkʰá - kàká kàká - - kʰàkʰá kàká kàká kàká kàká claves *lɛ̀pʰɛ ́ lɛ̀pɛ ́ àlàpɛ ́ lɛ̀pɛ ́ lɛ̀pʰɛ ́ - lɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ (instrument)_1 carry on back_1 *màm(a) màm màm màmá màm màm màmá màmā màmā màmā màmā wife_1, marry, *màʃ - pʼā+màʃ màʃ màʃ màʃ màʃ màs màs màs màs wedding hand_1 *mɛ̀t ̪̓ bɪ̀tʼ mɪ̀tʼ - mɛ̀d mɛ̀ɗ mɛ̀t ̪̓ mɪ̀tʼɪ ́ mɪ̀tʼɪ ́ - mɛ̀tʼ 586 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi bird_pelican *nàbɔ̀ng(w)à - ʊ̄+nàbɔ̀ŋà nàbɔ̀ŋà - - nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ (brown) nosebleed_2 *SʼUn(tʼ)a - tʼʊ̀ntʼ - - - sʼùnáʔ - - - - drum_C *tà̪rá - tàrá tàrá - - tà̪rá tàrá tàrá tàrá tàrá boar (wild)_1 *wàɓ wàpʼ wàpʼ wàpʼ wàpʰ à+wàɓ - - - - - fish (n.) general *wàcʼà wàsʼ wàsʼ wàsʼ wàʃʼ wàcʼ wàcʼà wàtʃʼà wàtʃʼà wàtʃʼà wàtʃʼà term hail, ice_1 *wasakʼ wàsà wàsà wàʃākʼ wàsáʔ à+wàsáʔ - - - - - C.4 POP XH Correspondence Set Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi warm up (sth.) *Cisʻ ʃɪʃ̄ ʃɪʃ̄ jíz jísʼ jít ̪h hízá ıs̋á ıs̋á ıs̋á - fish_sp *dɪŋ́kā - ʊ̄+wàsʼ+dɪŋ́kā dɪŋ́kā - - - dıŋ̋ā dıŋ̋ā dıŋ̋ā dıŋ̋ā (electric)_1 bird_weaver *dwakʰ ʊ̄+dwâk ʊ̄+dɔ́k à+dɔ̂k - - à+dwákʰ à+dwa̋kʰ à+dwa̋kʰ à+dwa̋kʰ à+dwa̋kʰ sneeze_1 *hat ̪̓is hǎtʼìʃ - - - - hát ̪̓ìs ha̋tʼīs ha̋tʼīs ha̋tʼīs ha̋tʼīs fish_(small, *jàhút ̪ jàhú jàhú jàhú - - àhút ̪h àhű àhű àhűwı ̋ àhűj small scales) smoke out (e.g an animal out of *kʰuʃ kʊ̌ʃ kʊ̌ʃ ūs - - húʃ űs űs űs űʃ a hole)_1 cover (v.) *kum(bi) kùmbì kùmbì kúm kūm kūm kúmā kűmá kűmá kűmá kűmá blow (with *pʰui(ki) pɪk̄ɪ ̄ pɪ ̄ pɪ̀ pʰûj pʰí pʰùj pʰű pʰű pʰű pʰű mouth) spicy (be)_1 *pʰUl? pʊ̀l pwì - - - pʰîl pʰıl̋ī pʰıl̋ī pʰıl̋ī pʰıl̋ī burnt gound, *pʰuZa ʃāpʊ́t ʃābʊ́t kɪ+́pú bwà+pʰwí à+pʰíʔ pùzàʔ pʰűjhá pʰűjhá pʰűjhá pʰűjhá soot 587 disregard *piʃ pīs pīs píʃ - píʃ píʃ pıs̋ pıs̋ pıs̋ pıs̋ rat_1 *sʼÍk sʼɪ ́ sʼɪ ́ sʼík sʼíʔ à+t ̪̓íkʰ - tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ blow nose *ʃɪntʼ ~ ʃɪn̄tʼ ʃɪn̄tʼ ʃīn+ʃʊ̀nʃ ʃīn ʃīn ʃìnà+ʃʊ̀ʃ sın̋á sın̋á sın̋á sın̋á *ʃintʼ spit (v.)_1 *t ̪h ú tū tū - - - t ̪h úwà tʰűjhá tʰűjhá tʰűjhá tʰűjhá bird_dove (African *tiritiri - títìtī - - - - à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì mourning) acacia *tú̪k⁽ʰ⁾(u) tʊ̌kʊ̀ - túk túkʰ à+tú̪kʰ tú̪kʰ tűkʰ tűkʰ tűkʰ tűkʰ bird_ostrich_1 *wutʰ - wǔt wǔt útʰ à+útʰ - hűtʰ hűtʰ hűtʰ hűtʰ 588 APPENDIX D PROTO-KOMAN CONSONANT CORRESPONDENCE SETS 10 APPENDIX D PKmn consonant correspondence sets The abbreviations for the language varieties are as follows: GwLo= Lowland Gwama, GwHi= Highland Gwama, UdYa= Yabus Uduk, Komo=Ethiopian Komo, UdCh= Chali Uduk, Dana=Dana, OpBi= Bilugu Opo, OpMo= Modin Opo, OpPa= Pame Opo, OpKi= Kigile Opo. I employ the grapheme <+> to indicate a historical or synchronic morpheme boundary. D.1 PKMN Correspondence Set 1: *pʰ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi laugh *pʰ(j)asʼ ɪs̄ʼ pāsʼ pɛ̀sʼ pʰɛs̄ʼ pʰɛt̄ ̪̓ pʰàsʼ pʰātʃʼ pʰātʃʼ pʰātʃʼ pʰātʃʼ soak_1 *pʰàcʼ pāsʼ pāsʼ pàsʼ pʰāʃʼ pʰācʼ - - - - - fly (v.) *pʰàd̪ pāl pāj pàj pʰāj pʰɛ ̄ pʰàd̪ pʰāj pʰāj pʰāj pʰāj shoe *pʰákʼá pák pák páʔ pʰāʔ à+pʰāʔ - pʰákʼá pʰákʼ pʰákʼ - come free and *pʰal pɪ ̌ pɪ ̌ - - - - pʰál pʰál pʰál pʰál fall off platform *pʰárá párá párá párá - - pʰárá pʰárá pʰárá pʰárá pʰárá blow (with *pʰui(ki) pɪk̄ɪ ̄ pɪ ̄ pɪ̀ pʰûj pʰí pʰùj pʰű pʰű pʰű pʰű mouth) spicy (be)_1 *pʰUl? pʊ̀l pwì - - - pʰîl pʰıl̋ī pʰıl̋ī pʰıl̋ī pʰıl̋ī burnt ground, *pʰuZa ʃāpʊ́t ʃābʊ́t kɪ+́pú bwà+pʰwí à+pʰíʔ pùzàʔ pʰűjhá pʰűjhá pʰűjhá pʰűjhá soot pound (v.)_4 *kɔ́pʰ kɔ́p kɔ́p - kúpʰ - - - - - - claves *lɛ̀pʰɛ ́ lɛ̀pɛ ́ àlàpɛ ́ lɛ̀pɛ ́ lɛ̀pʰɛ ́ - lɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ (instrument)_1 cold (be)_1 *sʼʊ́p sʼʊ́p sʼʊ́p - sʼúpʰ t ̪̓úpʰ - - - - - bathe *úpʰ ʊ́p ʊ́p úp úpʰ úpʰ úpʰ űpʰ űpʰ űpʰ űpʰ 589 D.2 PKMN Correspondence Set 2: *p Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi touch_1, crawl *pàD pāt pāt pàt pātʰ pār - - - - - shake (sth.)_1 *pid(V) - pǐdí pídá pítʰ pír - - - - - disregard *piʃ pīs pīs píʃ - píʃ píʃ pıs̋ pıs̋ pıs̋ pıs̋ vagina_1 *pít ̪h ~ pɪt́ ̪h pɪt́ pɪt́ pít - - pít ̪h - - - - wrap *pʊʃ ~ puʃ pʊ̄ʃ pʊ̄ʃ - pūʃ pūʃ púʃá pūsá pūsá pūsá pūsá carry on head_1 *kOpʰ kǔ kǔ - - - kɔ̄pʰ kɔ̄p kɔ̄p kɔ̄p kɔ̄p head *kʼʊ́p kʼʊ́p ʔʊ́p kʼʊ́p kʼúpʰ kʼúpʰ kʼʊ́pʰ kʼʊ́p kʼʊ́p kʼʊ́p kʼʊ́p disabled, angry (be) *nap(a) nǎpá nǎpá nǎpá - nápʰɛ ̄ nápʰ nāp nāp nāp nāp stab_1 *sɔ́p sɔ́ sɔ́ - - - - sɔ́p - sɔ́p - D.3 PKMN Correspondence Set 3: *b Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi father_1 *bàbá - bàbá bǎ à+bàbá à+bàbá - àbá àbá àbá àbá cannabis_1 *bángwà báŋgà báŋgà bánkɔ̀ bǎŋgɔ̀ bǎŋgɔ̀ - - - - - bird_heron *bár - ʊ̄+bár à+bár - - à+bár à+bár à+bár à+bár à+bár fishhook *Bɪncʼ bɪn̄sʼ bɪn̄sʼ bɪńsʼ bǐʃʼ à+bǐcʼ bɪćʼ ɓītʃʼ ɓītʃʼ ɓītʃʼ ɓītʃʼ dust, sand_1 *burbuɗ bùrbūt bùrbūt - bǔtʰ bǔɗ būrkʼùs būrkʼùs būrkʼùs būrkʼùs būrkʼùs steal_1 *kwabOʃ kɔ̄bɔ́ʃ kɔ̄bɔ́ʃ - - - - kǎbús kǎbús kǎbús - bird_pelican *nàbɔ̀ng(w)à - nàbɔ̀ŋà nàbɔ̀ŋà - - nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ (brown) dip food in sauce with *sʼUB(V)(n) - sʼʊ́pʊ́n sʼūb sʼúpʰ t ̪̓úpʰ sʼúbá - - - - fingers_1 *tʊ̪b(a) ~ pierce_1 sʊ́ sʊ́ - tūpʰ tūpʰ tú̪bá - - tʰūbá tʰūbá tu̪ba(a) 590 D.3a PKMN Correspondence Set 3a: *b Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi wide (be) *bàj ~ ɓàj pàj pàj ɓàjá ɓàn bɛ̀ - pàj pàj pàj pàj drum_B *bàmbá pàmbà pàmbà - bàmbá - - - - - - toss, throw away, fall *bit⁽ʰ⁾ pɪ̀t pɪ̀t bìt bǐtʰ bǐtʰ - - - - - over hide, skulk *bɔ̀b ~ bɔ̀ɓ pɔ̀pʼ pɔ̀pʼ bɔ̀b - - - pɔ̀j pɔ̀j pɔ̀j - pregnant (be) *bUma pʊ̀m pʊ̀m ɓú pwá bwà pùmá pǔmá pǔmá pǔmá pǔmá belly or stomach_1 *bùʃùl bùʃùl bùʃì bùʃ - bùʃ - pùsà pùsà pùsà pùsà extract tooth_1, *bʊkʰ pʊ̌ pʊ̌ - - bǔkʰ - - - - - barking (of dog) path, road_1, towards *bwaŋ(a) pwǎŋà pwǎŋà - - bwàj - - - - - choke, strangle_1 *busʼ bǔsʼ bǔsʼ - bùsʼ bùt ̪̓ - pʰǔtʃʼ - - - D.4 PKMN Correspondence Set 4: *ɓ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi frog_1 *bɔŋkʼɔ́ pʼà+bɔ̌ŋgɔ́ pʼà+bɔ̌ŋgɔ́ bā+ɓɔ̌ŋkʼɔ́ - - à+ɓɔ̄ŋkʼɔ́ à+ɓɔ̄ŋkʼɔ́ à+bɔ̄ŋkʼɔ́ à+bɔ̄ŋkʼɔ́ à+bɔ̄ŋkʼɔ́ palate_1 *ɓalilaj pʼàlílì pʼàlílí ɓàlílá - - - - - - - tree_sp._1 *ɓaʃa pʼàʃ - ɓàʃ ɓàʃà - ɓàʃà ɓāsā ɓāsā ɓāsā - *ɓid̪a ~ neck pʼɪl̄ pʼɪ ̄ ɓàʔ ɓāʔ ɓāʔ ɓìd̪à ɓījā ɓījā ɓījā ɓījā ɓɪd̪a strong *ɓɪśʼ ~ pʼɪ ́ pʼɪ ́ ɓísʼ ɓísʼ ɓít ̪̓ - - - - - (be)_1 ɓísʼ copulate *ʃʊ́ɓ ʃʊ́pʼ ʃʊ́pʼ - ʃúb ʃúɓ - - - - - (animal)_3 kick *t ̪h áɓ tápʼ tápʼ táb tʰáb tʰáɓ t ̪h áp tʰáp tʰáp tʰáp tʰáp boar *wàɓ wàpʼ wàpʼ wàpʼ wàpʰ à+wàɓ - - - - - (wild)_1 591 D.5 PKMN Correspondence Set 5: *pʼ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi ash_1 *(tʼi)pʼIkʼIɲ(a) pʼɪk̄ʼɪń pʼɪk̄ʼɪń pʼīn - tʼīpʼīɲ pʼɪn̄ā - - pʼɪn̄ā pʼɪn̄ā peel, husk_1 *pʼi(n)Cʼ? pʼɪńsʼ pʼɪńsʼ - - pʼíɗ pʼɪt́ʼ - - - - pick_1 *pʼɔt ̪̓(a) pʊ́t pʊ́t pʼɔ̀tʼ - - pʼɔ̀t ̪h à pʼɔ̄tʼā pʼɔ̄tʼā pʼɔ̄tʼā pʼɔ̄tʼā *nʊpʼ ~ bury (sideways) ʊ̄mpʼ ʊ̄mpʼ - - - nʊ̀pʼá - - - - nʊmpʼ sip (liquid) *wɔpʼ ~ hɔpʼ hɔ̌bɔ̀sʼhɔ̀pʼ hɔ̌bɔ̀sʼ wɔ́pʼ kʰɔ̄bɔ̄s - hɔ̀pʼ hɔ̄pʼɔ̄ hɔ̄pʼ hɔ̄pʼ hɔ̄pʼ D.6 PKMN Correspondence Set 6: *t ̪h Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi kick *t ̪h áɓ tápʼ tápʼ táb tʰáb tʰáɓ t ̪h áp tʰáp tʰáp tʰáp tʰáp make go away *tʰaʃ ~ t ̪h aʃ tāʃ tāʃ tàʃ tʰáʃ tʰáʃ - - - - - spit (v.)_1 *t ̪h ú tū tū - - - t ̪h úwà tʰűjhá tʰűjhá tʰűjhá tʰűjhá vagina_1 *pít ̪h ~ pɪt́ ̪h pɪt́ pɪt́ pít - - pít ̪h - - - - D.7 PKMN Correspondence Set 7: *t ̪ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi cotton, thread, *tʊ̪s ~ tu̪s dʊ̄ʃ dʊ̄ʃ tūʃ tūs tū̪s - - - - - spider web_1 drum_C *tà̪rá - tàrá tàrá - - tà̪rá tàrá tàrá tàrá tàrá acacia *tú̪k⁽ʰ⁾(u) tʊ̌kʊ̀ - túk túkʰ à+tú̪kʰ tú̪kʰ tűkʰ tűkʰ tűkʰ tűkʰ roughen (stone *tɪ̪t́ ̪ sɪt́ sɪt́ tɪt́ títʰ tír tɪ̪t́ ̪h tɪt́ɪ ́ tɪt́ʰ tɪt́ʰ tɪt́ʰ for grinding) pierce_1 *tʊ̪b(a) ~ tu̪ba(a) sʊ́ sʊ́ - tūpʰ tūpʰ tú̪bá - - tʰūbá tʰūbá grind (second *(ɔ)t⁽̪ʰ⁾ɔɗ - twɛj́ tɔ́ - tɔ̪́ɗ ɔ̀t ̪h ɔ́ ɔ́tɔ́ ɔ́tʰ ɔ́tʰ ɔ́tʰ grind) 592 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi person_2 *D̪ita̪ sīt sīt - - - d̪ítà̪ - - ʊ̀+zìtà ʊ̀+ʃìtà D.8 PKMN Correspondence Set 8: *d̪ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Dana *d̪ana dànā dànā dāná - - d̪àná dànā dànā dànā dànā (ethnonym)_3 slip (v.)_1 *d̪ɛrkʼɛsʼ dɛr̄gɛśʼ dɛr̄gɛśʼ - dɛ̀rɛ̀s d̪ɛ̀rɛ̀s d̪ɛŕkʼɛś - - - - *ɓid̪a ~ neck pʼɪl̄ pʼɪ ̄ ɓàʔ ɓāʔ ɓāʔ ɓìd̪à ɓījā ɓījā ɓījā ɓījā ɓɪd̪a beehive basket *gɛnd̪(V)(l) gɛńdɛĺ gɛ̀ndɪ ́ kɛ̀ndɛ ̄ - - gɛ̀nd̪á gɛ̀ndá gɛ̀ndá gɪ̀ndá gɛ̀ndá pig_1 *gUd̪Um kʊ̀rʊ́m kʊ̀rʊ́m gùdúm - - gùd̪ùm kùdùmà kùdùmà kùdùmà kùdùmà dig (for *kʼwàd̪(ɪ)́ kʼwī kʼwī kʼɔ̀ kʼwā - kʼwàd̪ɪ ́ kʼɔ̄j wārɪ ́ wādɪ ́ kʼwārɪ ́ water)_1 fly (v.) *pʰàd̪ pāl pāj pàj pʰāj pʰɛ ̄ pʰàd̪ pʰāj pʰāj pʰāj pʰāj beer *sud̪(i) ʃʊ́l ʃwɪ ́ ʃùʔí sū à+sū sùd̪ sī swī swī swī shave *sʼɛ̀d̪ ~ sʼɪ̀d̪ sʼɪl̄ sʼɪ ̄ sʼɛ̀ sʼī cʼɛ ̄ sʼɛ̀d̪ tʃʼɛ ̄ tʃʼɛ ̄ tʃʼɛ ̄ tʃʼɛ ̄ D.8 PKMN Correspondence Set 8a: *d̪ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi strain (solids from *d̪ìm - zìŋā zìm zǐm d̪ìm tì̪m tìm tìm tìm tìm liquid)_1 new_1 *d̪is ~ d̪ɪs dīʃ dīʃ zɪ̀ʃ tīs tī̪s - - - - - bee, honey *d̪àm tàm tàm dàm dàm à+d̪àm tâ̪m tàm tàm tàm tàm 593 D.9 PKMN Correspondence Set 9: *t ̪̓ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi sneeze_1 *hat ̪̓is hǎtʼìʃ - - - - hát ̪̓ìs ha̋tʼīs ha̋tʼīs ha̋tʼīs ha̋tʼīs alone, abstain from, not want to do *t ̪̓ ɛn - sʼɪn̄ sʼɛń tʼɛń tʼɛ ́ gà+t ̪̓ ɛń ā+tʼɛń ā+tʼɛń ā+tʼɛń ā+tʼɛń mouth *t ̪̓wa tʼwā tʼwā tʼā tʼwā tʼwā t ̪̓ āʔá tʼā tʼā tʼā tʼā enter_2, sprout *t ̪̓wI tʼwɪ ́ tʼwɪ ́ - - - t ̪̓wī - - - - D.9 PKMN Correspondence Set 9a: *t ̪̓ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi soft (be)_1 *kʼát ̪̓ kʼátʼ kʼátʼ kʼátʼ kʼád kʼáɗ - - - - - cough *kʼUt ̪̓ kʼʊ́kʼʊ́t kʼʊ́kʼʊ́t kʼùt kʼútʰ kʼútʰ kʼút ̪h kʼútʼù kʼútʼù kʼútʼù kʼútʼù kʼwāntʼɪ ́̓ kʼwāntʼɪ ́̓ tick *kʼwànt ̪̓ kʼwàtʼ kʼwātʰ à+kʼwāɗ kʼwàt ̪̓ kʼwātʼ kʼwātʼ kʼwātʼ kʼwātʼ kwāntʼ kwāntʼ testicles_1 *lùt ̪̓ ~ lʊ̀t ̪̓ dūtʼ dūtʼ lʊ̀tʼ lùd à+lùɗ - - - - - hand_1 *mɛ̀t ̪̓ bɪ̀tʼ mɪ̀tʼ - mɛ̀d mɛ̀ɗ mɛ̀t ̪̓ mɪ̀tʼɪ ́ mɪ̀tʼɪ ́ - mɛ̀tʼ pick_1 *pʼɔt ̪̓(a) pʊ́t pʊ́t pʼɔ̀tʼ - - pʼɔ̀t ̪h à pʼɔ̄tʼā pʼɔ̄tʼā pʼɔ̄tʼā pʼɔ̄tʼā far (be)_1 *sɪ̀t ̪̓ - ʃɪt̄ʼ ʃɪ̀tʼ sīd sīɗ sɪ̀t ̪̓ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ fire or *wa(n)t ̪̓(Iʃ) āntʼ āntʼ wàtʼíʃ ɔ̄d ɔ̄ɗ ɔ̀t ̪̓ wʊ̄tʼɪ ́ wʊ̄tʼɪ ́ wʊ̄tʼɪ ́ ʊ̄tʼɪ ́ firewood chili *zjat ̪̓a(j) zɛ́t zɛ́t zjàntʼá - - à+zāt ̪h ɛʔ̄ dʒɛ̀rà - - - pepper_1 594 D.10 PKMN Correspondence Set 10: *tʰ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi toss, throw away, fall *bit⁽ʰ⁾ pɪ̀t pɪ̀t bìt bǐtʰ bǐtʰ - - - - - over short (be)_1 *KÙtʰ gʊ̄t gʊ̄t kùt kūtʰ kūtʰ - - - - - antelope_1 *ʃɛt̄(ʰ) ʊ̄ʃɛt̄ ʃɛt̄ ʃɛt̄ ʃɛt̄ʰ à+ʃɛt̄ʰ - - - - - bird_ostrich_1 *wutʰ - wǔt wǔt útʰ à+útʰ - hűtʰ hűtʰ hűtʰ hűtʰ D.11 PKMN Correspondence Set 11: *t Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi forehead_1 *t⁽ʰ⁾(w)àg twã twã ̄ tàg - - tàg bī+tʰāg - pī+tʰāg - be, do *tà - ta tà tā tā - tā - tā - shake (sth.)_2 *tEŋ(g)(E) tɪḡɪ ̄ tɪḡɪ ̄ - - - tɛŋ́ - tɪŋ́há tɪŋ́há tɪŋ́há bird_dove (African *tiritiri - títìtī - - - - à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì mourning) long or tall *tur ~ tʊr tǔ tǔ tʊ́l túr túr - - - - - (be)_1 tell_1 *(w)ɔt(V) ɔ̌dɔ́ ɔ̌d - - - - ɔ́tá ɔ́tá ɔ́tá - D.12 PKMN Correspondence Set 12: *d Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Dazu (S. Sudan) *dáɟV dázɔ̀ dázɔ̀ dázʊ̀ dǎɟɔ̀ dǎɟɔ̀ dáɟɔ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ shout_1 *dɔl(ɔ) dɔ́l dɔ́l dɔ́l - - ɔ̀lɔ́ - - - - finish_1 *dak - dak dàg dàkʰ dàkʰ dâk dàk - - - 595 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi tree (sp.)(sausage tree_Kigelia *dùmàj - ū+dùmɪ̀ dùmɛ̀ - - dùmáj dùmàj dùmàj dùmàj dùmàj africana) grind wet ̪̓ (first grind) *Dàsʼ dàsʼ dàsʼ nàsʼ nàsʼ nàt - - - - - fish_sp *dɪŋ́kā - ʊ̄+wàsʼ (electric)_1 +dɪŋ́kā dɪŋ́kā - - - dıŋ̋ā dıŋ̋ā dıŋ̋ā dıŋ̋ā pipe (for smoking)_1 *dʊ́ɟɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ - - dʊ́ɟɛ̀ - - - - fish_sp (very small, scaled fish with a small *dɔlɔ - dɔ̌lɔ́ dɔ̀lɔ́ - - dɔ́lɪʔ̄ dɔ̀lɔ́ dɔ̀lɔ́ dɔ̀lɔ́ dɔ́lɪ ̄ rounded mouth) bird_weaver *dwakʰ ʊ̄+dwâk ʊ̄+dɔ́k à+dɔ̂k - - à+dwákʰ à+dwa̋kʰ à+dwa̋kʰ à+dwa̋kʰ à+dwa̋kʰ D.12 PKMN Correspondence Set 12a: *d Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi baboon_1, dog_2 *dɪŋɪ tɪ̀nɪ̀ tɪ̀nɪ̀ - - - tɛ̀ŋ tɪ̀nɪ̀ tɪ̀nɪ̀ tɪ̀n tɪ̀n stomp, step on_1 *djalIsʼ dɛ̀lísʼ dàlísʼ dɪ̀l dìl dìl - tɪ̀lɪ ́ tɪ̀lɪ ́ tɪ̀lɪ ́ tɪ̀lɪ ́ urine_1, urinate_1 *dùcʼá dùsʼ tùsʼ dʊ̀sʼ - - tùcʼáʔ - - - - ask (inquire)_1 *dOt tʊ̌t tʊ̌t dɔ̀t dɔ̌tʰ dɔ̌tʰ - tɔ̄tɔ́ tɔ̄tɔ́ - - D.13 PKMN Correspondence Set 13: *ɗ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi carry many *(tɔ)ɗɔ tɔ̀dɔ̀ - ɗɔ̀ ɗɔ̄ ɗɔ̄ - ɗɔ̄ ɗɔ̀ ɗɔ̀ ɗɔ̀ things_1 596 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi scorpion *ɗ(w)ankʼi ~ ɗ(w)ankʼɪ *tʼwānkʼɪ ́ *tʼwānkʼ bāɗāgí ɗwākʰ à+ɗākʰ ɗāgí ɗāgı ̋ ɗāgı ̋ ɗāgı ̋ ɗāgı ̋ send someone_1 *ɗar(a) tʼálà tʼájà ɗàr ɗɛt̄ʰ - ɗɛ̀d - ɗɛr̄ - ɗɛr̄ dust, sand_1 *burbuɗ bùrbūt bùrbūt - bǔtʰ bǔɗ būrkʼùs būrkʼùs būrkʼùs būrkʼùs būrkʼùs water *ji(ɗE) ìjáʔ ìjáʔ jǐ wùdíʔ jìɗɛʔ́ jìʔí dʒì dʒì zì ʃìʔ dust, sand_1 *burbuɗ bùrbūt bùrbūt - bǔtʰ bǔɗ būrkʼùs būrkʼùs būrkʼùs būrkʼùs būrkʼùs 3SG.M *haɗ(i) ʊ̄+hāl ʊ̄+hāj hàr hádī áɗī hār - - - - D.14 PKMN Correspondence Set 14: *tʼ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi salt_(made from ash of a particular *tʼaʃ tʼāʃ tʼāʃ tʼàʃ tʼāʃ tʼāʃ tʼàʃ tʼās - - - plant/tree)_2 cut (split in half *tʼwaŋ(k)a ~ lengthwise)_1 t ̪̓waŋ(k)a tʼāŋà tʼāŋà - tʼwákʰ tʼwákʰ - - - - - drip (fall in globules)_1 *cʼɔ(tʼɔ)l sʼɔ́tʼɔ́ sʼɔ́tʼɔ́ sʼʊ̀lɪl̄ - cʼɔ̄cʼɔ̄lɔ̄cʼ cʼɔ̀ʔ tʃʼɔ̄ tʃʼɔ̄ tʃʼɔ̄ tʃʼɔ̄ nosebleed_2 *SʼUn(tʼ)a - tʼʊ̀ntʼ - - - sʼùnáʔ - - - - rope_1 *ʃutʼ ? ʃʊ̄dʊ̀l ʃwītʼìn ʃʊ̀ʔɪ ́ ʃí à+ʃí ʃʊ́l - - - - D.15 PKMN Correspondence Set 15: *sʼ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi blood_1 *sʼámá sʼám sʼám - - - sʼámáʔ tʃʼámá tʃʼámá tʃʼámá tʃʼámá light (ignite) *sʼa sʼā sʼā sʼà sʼā t ̪̓ ā sʼʊ́wà tʃʼǎ tʃʼǎ tʃʼǎ tʃʼǎ cold (be)_1 *sʼʊ́p sʼʊ́p sʼʊ́p - sʼúpʰ t ̪̓úpʰ - - - - - shave *sʼɛ̀d̪ ~ sʼɪ̀d̪ sʼɪl̄ sʼɪ ̄ sʼɛ̀ sʼī cʼɛ ̄ sʼɛ̀d̪ tʃʼɛ ̄ tʃʼɛ ̄ tʃʼɛ ̄ tʃʼɛ ̄ 597 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi rat_1 *sʼÍk sʼɪ ́ sʼɪ ́ sʼík sʼíʔ à+t ̪̓íkʰ - tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ dip food in sauce with fingers_1 *sʼUB(V)(n) - sʼʊ́pʊ́n sʼūb sʼúpʰ t ̪̓úpʰ sʼúbá - - - - defecate_1, diarrhea *sʼUDI úsʼ úsʼ sʼɪ ́ - t ̪̓í sʼʊ̀dʊ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ nosebleed_2 *SʼUn(tʼ)a - tʼʊ̀ntʼ - - - sʼùnáʔ - - - - choke, strangle_1 *busʼ bǔsʼ bǔsʼ - bùsʼ bùt ̪̓ - pʰǔtʃʼ - - - strong (be)_1 *ɓɪśʼ ~ ɓísʼ pʼɪ ́ pʼɪ ́ ɓísʼ ɓísʼ ɓít ̪̓ - - - - - tear (shred)_3 *cʼ(w)ɛs̄ʼ sʼwɛ̃ ̌ sʼwɛ̃ ̌ - - cʼɛt̄ ̪̓ - - - - - grind wet (first *Dàsʼ dàsʼ dàsʼ nàsʼ nàsʼ nàt ̪̓ - - - - - grind) slip (v.)_1 *d̪ɛrkʼɛsʼ dɛr̄gɛśʼ dɛr̄gɛśʼ - dɛ̀rɛ̀s d̪ɛ̀rɛ̀s d̪ɛŕkʼɛś - - - - tie (bundle)_1 *gì(n)sʼ kɪ̀nsʼ kɪ̀nsʼ gìsʼ - - - - - - - swallow_1 *gUsʼ gùsʼ kùsʼ gʊ̀sʼ - - kʊ̄sʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá trample, ruminate *hasʼ - hāsʼɪ̀ hásʼ hásʼ hát ̪̓ hásʼ hátʃʼ hátʃʼ hátʃʼ hátʃʼ vomit_1 *(pa)jasʼ pājàsʼ pājàsʼ jàʔ ɟǎʔ jǎʔ - - - - - dry (be) *kʰʊ́sʼ kúsʼ kúsʼ kʊ́sʼ kʰúsʼ kʰút ̪̓ kʰʊ́sʼ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ - earth, soil, ground, floor_1 *k(j)asʼVN kʼjáŋásʼ kʼɛs̄ʼɛń kʼāsʼɪ̀ - - - - - - - charcoal or coal *kʼI(m)Isʼ sʼɪśʼɪń sʼɪśʼɪń kìsʼísʼìʔ cʼɛl̄ɛs̄ʼ cʼīlāt ̪̓ kʼɪ̀sʼ kʼīmītʃʼ kʼīmītʃʼ kʼɪt̄ʃʼ kʼɪt̄ʃʼ shut_1, close eyes *mVsʼ mɪ̀sʼ mɪ̀sʼ mʊ̀sʼ músʼ+ɛ ̄ mút ̪̓+ɛ ̄ mìsʼà mʊ̀tʃʼà mʊ̀tʃʼà mʊ̀tʃʼà mʊ̀tʃʼà honey badger *njansʼɛsʼ jānsʼɛśʼ ɲɛn̄zɛśʼ nɛ̀z nɛ̀sʼ à+nɛ̀t ̪̓ nɛ̀s nɛ̀dʒɛ̀ - nɛ̀dʒɛ̀ - laugh *pʰ(j)asʼ ɪs̄ʼ pāsʼ pɛ̀sʼ pʰɛs̄ʼ pʰɛt̄ ̪̓ pʰàsʼ pʰātʃʼ pʰātʃʼ pʰātʃʼ pʰātʃʼ peel, husk_1 *pʼi(n)Cʼ? pʼɪńsʼ pʼɪńsʼ - - pʼíɗ pʼɪt́ʼ - - - - warm up (sth.) *Cisʻ ʃɪʃ̄ ʃɪʃ̄ jíz jísʼ jít ̪h hízá ıs̋á ıs̋á ıs̋á - D.16 PKMN Correspondence Set 16: *c Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi grandfather_1 *càkʰO - sàkʊ́ sàkʊ́ - - - tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ 598 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi whistle *cut - ʃút ʃɪt́ - cúwā cʊ́ɪ̀ tʃúwì tʃúwì tʃúwì tʃúwì tree *cwálá swálá swájá sá ʃwá cwá cá tʃá tʃá tʃá tʃá rinse mouth_1 *CVkʻUm ʃʊ̀kʊ́m ʃʊ̀kʊ́m zùkʼúm - - ɟákʼʊ́m tʃākʼʊ́má tʃākʼʊ́má sākʼʊ́má tʃākʼʊ́má belt, sash *gaʃa gáʃà gáʃà - gǎʃ gāʃá gâʃ gātʃ gātʃ gātʃ gâʃ D.17 PKMN Correspondence Set 17: *ɟ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi dig_1 *ɟà ʃà ʃà - - - cà tʃà tʃà tʃà tʃà Nuer (ethnonym)_1 *ɟaŋg(ɔ)aj zǎgʊ́ zǎgʊ́ zǎgʊ́ ʒwǎŋgì - ɟàŋwɛ̀j dʒāŋɔ́ - zāŋwɛ ́ ʃàŋwɛ̀ sorghum, millet *ɟana sjànà sjànà zɛ̀nā - - - dʒɛ̀ná - dʒɛ̀ná - bird_stork (maribou) *ɟàrú zɛ̀rú zɛ̀rú zɛ̀rú ʒàrú - ɟɛ̀rú dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű Dazu (S. Sudan) *dáɟV dázɔ̀ dázɔ̀ dázʊ̀ dǎɟɔ̀ dǎɟɔ̀ dáɟɔ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ pipe (for smoking)_1 *dʊ́ɟɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ - - dʊ́ɟɛ̀ - - - - D.18 PKMN Correspondence Set 18: *cʼ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi tear (shred)_3 *cʼ(w)ɛs̄ʼ sʼwɛ̃ ̌ sʼwɛ̃ ̌ - - cʼɛt̄ ̪̓ - - - - - ear *cʼɛ sʼɛ ̄ sʼɛ ̄ sʼɛ ̄ ʃʼɛ ́ cʼɛ ́ kʼɛ ̄ tʃʼɛ̀ tʃʼɛ̀ tʃʼɛ̀ tʃʼɛ̀ drip (fall in globules)_1 *cʼɔ(tʼɔ)l sʼɔ́tʼɔ́ sʼɔ́tʼɔ́ sʼʊ̀lɪl̄ - cʼɔ̄cʼɔ̄lɔ̄cʼ cʼɔ̀ʔ tʃʼɔ̄ tʃʼɔ̄ tʃʼɔ̄ tʃʼɔ̄ fishhook *Bɪncʼ bɪn̄sʼ bɪn̄sʼ bɪńsʼ bǐʃʼ à+bǐcʼ bɪćʼ ɓītʃʼ ɓītʃʼ ɓītʃʼ ɓītʃʼ urine_1, urinate_1 *dùcʼá dùsʼ tùsʼ dʊ̀sʼ - - tùcʼáʔ - - - - enter_1 *gìcʼ kɪšʼ kɪšʼ gìz - cīcʼ - kìtú kìtú kìtú kìtú shut_2 *kʰacʼ kǎʃ kǎʃ - kʰáʃʼ kʰácʼ - - - - - soak_1 *pʰàcʼ pāsʼ pāsʼ pàsʼ pʰāʃʼ pʰācʼ - - - - - fish (n.) general term *wàcʼà wàsʼ wàsʼ wàsʼ wàʃʼ wàcʼ wàcʼà wàtʃʼà wàtʃʼà wàtʃʼà wàtʃʼà 599 D.19 PKMN Correspondence Set 19: *kʰ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi dry (be) *kʰʊ́sʼ kúsʼ kúsʼ kʊ́sʼ kʰúsʼ kʰút ̪̓ kʰʊ́sʼ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ - repair_1 *kʰaɓ áp - áb - - ʊ́bā kʰápʼ kʰápʼ kʰápʼ kʰápʼ shut_2 *kʰacʼ kǎʃ kǎʃ - kʰáʃʼ kʰácʼ - - - - - open *kʰáɗ(a) kálá kájá kár kʰǎd kʰǎɗ kʰátá kʰátá kʰátá kʰátá kʰátá light (the way with torch/flashlight)_1 *kʰàɲ kɪn̄ kɪn̄ - kʰāɲ kʰāɲ - - - - - snore *kʰO(r)nOn kɔ̀nɔ̀n kɔ̀nɔ̀n kɔ̀nɔ̀n kʰūnūn àkʰɔ̄rnɛʔ̄ kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n fear (be afraid) *kʰwaGʼ kwāgà kwāgà kɔ̄g kʰɔ̄ʔ kʰɔ̄kʼ kʰɔ́k ~ kʰɔ̀gɔ́ kʰɔ̄gɔ́ kʰɔ̄gɔ́ kʰɔ̄gɔ́ kʰɔ̄gɔ́ herd (v.) *kʰaj kɛ̃ʔ̄ kɛ̃ʔ̄ kàʔí - - kʰàjɪʔ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ bitter, sour (be) *kʰakʼa kāgā kāgā kàʔ kʰāʔ kʰāʔ kʰàkʼà kʰākʼā kʰākʼā kʰākʼā kʰākʼā tree_sp. (mahogany, Trichilia emetica) *k⁽ʰ⁾ɪs kɪśɪ̀ ʊ̄+kɪs̄ kɪs̄ cɛś cʰís kɪś kʰɪś kʰɪś kʰɪś kʰɪś extract tooth_1, barking (of dog) *bʊkʰ pʊ̌ pʊ̌ - - bǔkʰ - - - - - grandfather_1 *càkʰO - sàkʊ́ sàkʊ́ - - - tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ sow seeds (by throwing)_1 *jEkʰ jɪ̀ jɪ̀ jɛ̀k jɛ̀kʰ jɛ̀kʰ - - - - - bird_hammerkopf_1 *nɛḱʰ - ʊ̄+nɛḱ à+nɛḱ nɛḱʰ à+nɛḱʰ à+nɪḱʰ - - à+nɪḱʰ à+nɪḱʰ acacia *tú̪k⁽ʰ⁾(u) tʊ̌kʊ̀ - túk túkʰ à+tú̪kʰ tú̪kʰ tűkʰ tűkʰ tűkʰ tűkʰ bird_weaver *dwakʰ ʊ̄+dwâk ʊ̄+dɔ́k à+dɔ̂k - - à+dwákʰ à+dwa̋kʰ à+dwa̋kʰ à+dwa̋kʰ à+dwa̋kʰ D.20 PKMN Correspondence Set 20: *k Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi sew_1 *kɔ́s ɔ́s ɔ́s - - - kɔ́s kɔ́s kɔ́s kɔ́s kɔ́s pound (v.)_4 *kɔ́pʰ kɔ́p kɔ́p - kúpʰ - - - - - - tomorrow_1 *kjana - gɪ̀+kjānā - - - - dʒì+kɛń dʒì+kɛń zì+kɛń ʃì+kɛń cluck (of hen) *kjankʼa kǎkā kjāŋkʼā kágá - - kágà kɛn̄ kɛn̄ kɛn̄ kɛn̄ cry *kO(j) kū kū kɔ̀ kɔ̄ kɔ̄ kɔ̀j kwɛ ̄ kwɛ ̄ kwɛ ̄ kwɛ ̄ 600 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi carry on head_1 *kOpʰ kǔ kǔ - - - kɔ̄pʰ kɔ̄p kɔ̄p kɔ̄p kɔ̄p chief_1 *kOr ʊ̄+kʊ̄l ʊ̄+kwɪ̀ jī+kwɪ̀ - - jɛ̀+kɔ̄rɔ̄ - - - - cover (v.) *kum(bi) kùmbì kùmbì kúm kūm kūm kúmā kűmá kűmá kűmá kűmá steal_1 *kwabOʃ kɔ̄bɔ́ʃ kɔ̄bɔ́ʃ - - - - kǎbús kǎbús kǎbús - grandmother_ 1 *kʰàkʰá - kàká kàká - - kʰàkʰá kàká kàká kàká kàká brother *ʊ=kam ~ *kamʊ kwám kwám kàm kām à+kām āmʊ́ hàm hàm hàm hàm D.20 PKMN Correspondence Set 20a: *k Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Opo (ethnonym)_1 *k⁽ʰ⁾ɪńáj kɪńá kɪńá kɪńá cʰínáj - kɪńáj - - - - bird_(cattle egret) *kEɲ kɪl̄ kɪl̄ à+kɪĺ cɛćɛńā à+cɛɲ́ à+kɪĺ à+kɪĺ à+kɪĺ à+kɪĺ à+kɪĺ D.21 PKMN Correspondence Set 21: *g Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi beeswax_1 *gāgáʔ gāgá gāgá - - - gāgá gāgá gāgá gāgá gāgá belt, sash *gaʃa gáʃà gáʃà - gǎʃ gāʃá gâʃ gātʃ gātʃ gātʃ gâʃ bird (yellow- billed kite or *gʊlɪla - gʊ́lːā bāgʊ́lɪ̀lā - - ɓāgʊ̂lːā - ɓāgʊ́lʊ́l ɓāgʊ́lʊ́l ɓāgʊ́là black kite) run (SG)_1, flow, *gùs ~ bleed gʊ̀s gʊ̀s gʊ̀s gùʃ gùs gùs - - - - - Gwama (ethnonym) *gwama gwàmá kwāmà gwàmá - - gɔ̀má gɔ̀má gɔ̀má gɔ̀má gɔ̀má finish_1 *dak - dak dàg dàkʰ dàkʰ dâk dàk - - - cannabis_1 *bángwà báŋgà báŋgà bánkɔ̀ bǎŋgɔ̀ bǎŋgɔ̀ - - - - - have sex_1 *hag(a) hã́ʔ hã́ʔ hág - háʔ - hágá hágá hágá hágá giraffe_2 *kʰOG kɔ́k kɔ́k - kʰúʔ kʰúʔ kʰɛ́g - - - kɛ̂k 601 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi bed_1 *langarɛt ángàr - ángàr áŋgàr - - làŋgàrɛ́t àŋgàrípʰ àŋgàrípʰ àŋgàrípʰ forehead_1 *t⁽ʰ⁾(w)àg twã twã ̄ tàg - - tàg bī+tʰāg - pī+tʰāg - D.21 PKMN Correspondence Set 21a: *g Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi tie (bundle)_1 *gì(n)sʼ kɪ̀nsʼ kɪ̀nsʼ gìsʼ - - - - - - - swallow_1 *gUsʼ gùsʼ kùsʼ gʊ̀sʼ - - kʊ̄sʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá replace *(n(j)a)gaD nāgát njāgát gàdá gàs gàr gàtá kàrá kàrá kàrá kàrá enter_1 *gìcʼ kɪšʼ kɪšʼ gìz - cīcʼ - kìtú kìtú kìtú kìtú smell (v.) *gàŋ(a) kɛ̃ ̄ kɛ̃ ̄ gàg - - kàŋà kàŋà kàŋà kàŋà kàŋà skin, hide (of animal), *gɔnkʼ(ɔʃ) gɔ̀kʼɔ́ʃ gɔ̀kʼɔ́ʃ gɔ̀nkʼí gɔ̀kʰ - kɔ̀gɔ̀ kɔ̀gɔ̀ kɔ̀gɔ̀ kɔ̀gɔ̀ gwàŋgɪ ́ bark of tree_1 pig_1 *gUd̪Um kʊ̀rʊ́m kʊ̀rʊ́m gùdúm - - gùd̪ùm kùdùmà kùdùmà kùdùmà kùdùmà elephant_1 *gwàj kwɪ̀ kwɪ̀ gwà gwàj - - - - - - D.22 PKMN Correspondence Set 21: *kʼ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi back_1 *kʼ(w)ás kʼwás kʼwás kʼǎw - - - - - - - soft (be)_1 *kʼát ̪̓ kʼátʼ kʼátʼ kʼátʼ kʼád kʼáɗ - - - - - eat (hard food) *kʼama kʼā kʼā kʼá kʼáʔ kʼá kʼámá kʼámá kʼámá kʼámá kʼámá kill, fight *kʼOʃ kʼʊ́ʃ kʼʊ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́sɔ́ kʼɔ́sɔ́ kʼɔ́s kʼɔ́ʃ head *kʼʊ́p kʼʊ́p ʔʊ́p kʼʊ́p kʼúpʰ kʼúpʰ kʼʊ́pʰ kʼʊ́p kʼʊ́p kʼʊ́p kʼʊ́p throat *kʼʊ́s k’úʃ k’úʃ kʼʊ́ʃ kʼús kʼús kʼʊ́s kʼʊ́sʊ́ kʼʊ́s kʼʊ́s kʼʊ́s dig (for water)_1 *kʼwàd̪(ɪ)́ kʼwī kʼwī kʼɔ̀ kʼwā - kʼwàd̪ɪ ́ kʼɔ̄j wārɪ ́ wādɪ ́ kʼwārɪ ́ 602 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi tick *kʼwànt ̪̓ kʼwāntʼɪ ́̓ kʼwāntʼɪ ́̓ ̪̓kwāntʼ kwāntʼ kʼwàtʼ kʼwātʰ à+kʼwāɗ kʼwàt kʼwātʼ kʼwātʼ kʼwātʼ kʼwātʼ earth, soil, ground, floor_1 *k(j)asʼVN kʼjáŋásʼ kʼɛs̄ʼɛń kʼāsʼɪ̀ - - - - - - - cough *kʼUt ̪̓ kʼʊ́kʼʊ́t kʼʊ́kʼʊ́t kʼùt kʼútʰ kʼútʰ kʼút ̪h kʼútʼù kʼútʼù kʼútʼù kʼútʼù bitter, sour (be) *kʰakʼa kāgā kāgā kàʔ kʰāʔ kʰāʔ kʰàkʼà kʰākʼā kʰākʼā kʰākʼā kʰākʼā cluck (of hen) *kjankʼa kǎkā kjāŋkʼā kágá - - kágà kɛn̄ kɛn̄ kɛn̄ kɛn̄ porcupine_1 *k⁽ʰ⁾akʼas ~ k⁽ʰ⁾asakʼ kákʼàʃ kákʼàʃ káʃākʼ - - kásākʼ kʰásākʼ kʰásākʼ kʰásākʼ kʰáʃākʼ foot or leg_1 *ʃɔnkʼ sɔ̄ŋkʼ sɔ̄ntʼ ʃɔ̀g ʃɔ̄kʼ ʃɔ̄ʔ ʃɔ̀g - - - - wake (trs.)_1 *ʃukʼ(V) sūgì sūgì ʃùg ʃūkʼ - ʃùg sūg sūg sūg ʃūg frog_1 pʼà+bɔ̌ŋg pʼà+bɔ̌ŋg bā+ɓɔ̌ŋkʼ à+bɔ̄ŋkʼ*bɔŋkʼɔ́ - - à+ɓɔ̄ŋkʼɔ́ à+ɓɔ̄ŋkʼɔ́ à+bɔ̄ŋkʼɔ́ à+bɔ̄ŋkʼɔ́ ɔ́ ɔ́ ɔ́ ɔ́ rinse mouth_1 *CVkʻUm ʃʊ̀kʊ́m ʃʊ̀kʊ́m zùkʼúm - - ɟákʼʊ́m tʃākʼʊ́m tʃākʼʊ́m sākʼʊ́m tʃākʼʊ́má á á á *ɗ(w)ankʼi scorpion ~ *tʼwānkʼ *tʼwānkʼ bāɗāgíʔ ɗwākʰ à+ɗākʰ ɗāgí ɗāgı ̋ ɗāgı ̋ ɗāgı ̋ ɗāgı ̋ ɗ(w)ankʼɪ sweat (substance)_2 *jàgàl jàgàl jàgɪ̀ - - - - dʒīkʼāj dʒīkʼāj zīkʼāj ʃīkʼāj fear (be afraid) *kʰwaG kwāgà kwāgà kɔ̄g kʰɔ̄ʔ kʰɔ̄kʼ kʰɔ́k ~ kʰɔ̀gɔ́ kʰɔ̄gɔ́ kʰɔ̄gɔ́ kʰɔ̄gɔ́ kʰɔ̄gɔ́ shoe *pʰákʼá pák pák páʔ pʰāʔ à+pʰāʔ - pʰákʼá pʰákʼ pʰákʼ - two *sʊkʼa swījā swījā sʊ̄ sú sú sʊ́kʼà sʊ̄kʼá sʊ̄kʼá sʊ̋kʼá sʊ̋kʼá louse_1 *ʃOkʼ(VN) - ʃʊ̄gʊ̀n ʃùwɛ̀n ʃɔ̄kʼɔ̄ à+ʃɔ̄kʼɔ̄m m ʃùkʼnáj sūkʼɛń sūkʼɛń sūkʼɛń ʃūkʼɛń rain (precipitation)_ *ʃɔkʼ - ʃʊ́ ɔ́ ʃɔ̄kʼ à+ʃɔ̄kʼ - hɔ́ hɔ̌ hɔ̌ - 2 breathe *ʃukʼin ʃɪňʃɪ ́ ʃɪn̄ʃɪ̀ ʃùʔɛn̄ ʃīʔīn ʃīʔīn ʃìkʼ sīkʼ sīkʼ sīkʼ sīkʼ tendon, vein *ʃʊ(n)kʼ ʃʊ̄nkʼ ʃʊ̄nkʼ ʃʊ̄ ʃúʔ à+ʃúʔ ʃùkʼ sʊ̄kʼ sʊ̄kʼ sʊ̄kʼ ʃūkʼ 603 D.23 PKMN Correspondence Set 23: *s Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi lead (guide)_1 *sʊs ʃʊ́ʃ ʃʊ́ʃ ʃʊ́ʃ sús sús sʊ́ʔ - - - - cut (meat into one long *sɪĺ(ɪ) ʃɪĺɪ̀ ʃɪ ́ - - - - sɪ ́ - - - piece)_2 far (be)_1 *sɪ̀t ̪̓ - ʃɪt̄ʼ ʃɪ̀tʼ sīd sīɗ sɪ̀t ̪̓ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ beer *sud̪(i) ʃʊ́l ʃwɪ ́ ʃùʔí sū à+sū sùd̪ sī swī swī swī python_1 *sʊm ʃʊ̄ʃʊ́m ʃʊ̄ʃʊ́m ʃʊ̌m - súm sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ *d̪is ~ new_1 dīʃ dīʃ zɪ̀ʃ tīs tī̪s - - - - - d̪ɪs sneeze_1 *hat ̪̓is hǎtʼìʃ - - - - hát ̪̓ìs ha̋tʼīs ha̋tʼīs ha̋tʼīs ha̋tʼīs wet_1, slippery *jEsI ɪʃ́ɪ̀ ɪʃ́ɪ̀ jɛ̀ʃ jɛ̀s jɛ̀s sìʔ - - - - throat *kʼʊ́s k’úʃ k’úʃ kʼʊ́ʃ kʼús kʼús kʼʊ́s kʼʊ́sʊ́ kʼʊ́s kʼʊ́s kʼʊ́s lead (guide)_1 *sʊs ʃʊ́ʃ ʃʊ́ʃ ʃʊ́ʃ sús sús sʊ́ʔ - - - - cotton, thread, spider *tU̪s dʊ̄ʃ dʊ̄ʃ tūʃ tūs tū̪s - - - - - web_1 roast (something) *t ̪̓ ɔ́s tʼɔ̄ʃ tʼɔ̄ʃ tʼɔ́ʃ tʼɔ́s t ̪̓ ɔ́s - - - - - boil (of liquid)_1 *was(ikʼ) wǎʃ wàʃí wàʃíkʼ - - wás wās wās wās wās D.23 PKMN Correspondence Set 23a: *s Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi warm oneself *s(w)am sɔ́m sɔ́m ʃɔ́m sām sām sɔ̀m sɔ̄m sɔ̄m sɔ̄m ʃɔ̄m climb_1 *sɛl sál sɛl̄ - sɛ ̄ sɛ ̄ - - - - - stab_1 *sɔ́p sɔ́ sɔ́ - - - - sɔ́p - sɔ́p - body_1 *(j)Es - jɪs̄ ɪʃ̄ īs īs ɛ̀s ɛs̄ ɛs̄ ɛs̄ ɛs̄ run (SG)_1, flow, bleed *gùs ~ gʊ̀s gʊ̀s gʊ̀s gùʃ gùs gùs - - - - - ripen *ís ~ ís ɪś ɪś íʃ ís ís ísá ıs̋á ıs̋á ıs̋á ıt̋ʃá 604 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi sew_1 *kɔ́s ɔ́s ɔ́s - - - kɔ́s kɔ́s kɔ́s kɔ́s kɔ́s hail, ice_1 *wasakʼ wàsà wàsà wàʃākʼ wàsáʔ à+wàsáʔ - - - - - back_1 *kʼ(w)ás kʼwás kʼwás kʼǎw - - - - - - - tree_sp. (mahogany, Trichilia emetica) *k⁽ʰ⁾ɪs kɪśɪ̀ ʊ̄+kɪs̄ kɪs̄ cɛś cʰís kɪś kʰɪś kʰɪś kʰɪś kʰɪś sky_1 *wUs wús wús - - - - wʊ̀s wʊ̀s wʊ̀s wʊ̀s D.24 PKMN Correspondence Set 24: *ʃ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi birth pangs (pain) *ʃāʃ(ā) ʃɛ ́ ʃɛ ́ ʃāʃ - - ʃāʃā sāsā sāsā sāsā ʃāʃā warm up (sth.) *Cisʻ ʃɪʃ̄ ʃɪʃ̄ jíz jísʼ jít ̪h hízá ıs̋á ıs̋á ıs̋á - copulate (animal)_3 *ʃʊ́ɓ ʃʊ́pʼ ʃʊ́pʼ - ʃúb ʃúɓ - - - - - eat (soft food)_SG *ʃa ʃā ʃā ʃá ʃwá ʃwá úʃā sá sá sá ʃá tooth *ʃE ʃī ʃī ʃɛ̀ ʃɛ ̄ ʃɛ ̄ ʃɛ ̄ sɛ ̄ sɛ ̄ sɛ ̄ ʃɛ ̄ antelope_1 *ʃɛt̄(ʰ) ʊ̄ʃɛt̄ ʃɛt̄ ʃɛt̄ ʃɛt̄ʰ à+ʃɛt̄ʰ - - - - - blow nose *ʃIntʼ ʃɪn̄tʼ ʃɪn̄tʼ ʃīn+ʃʊ̀nʃ ʃīn ʃīn ʃìnà+ʃʊ̀ʃ sın̋á sın̋á sın̋á sın̋á grass *ʃO ʃʊ́ʃʊ̄ ʃʊ́ʃʊ̄ʃʊ̄ ʃɔ̀ʔí ʃɔ̄ à+ʃɔ̄ ʃɔ̄ʔɔ́ sɔ̀ sɔ̀ sɔ̀ sɔ̀ louse_1 *ʃOkʼ(VN) - ʃʊ̄gʊ̀n ʃùwɛ̀n ʃɔ̄kʼɔ̄m à+ʃɔ̄kʼɔ̄m ʃùkʼnáj sūkʼɛń sūkʼɛń sūkʼɛń ʃūkʼɛń rain (precipitation)_2 *ʃɔkʼ - ʃʊ́ ɔ́ ʃɔ̄kʼ à+ʃɔ̄kʼ - hɔ́ hɔ̌ hɔ̌ - breathe *ʃukʼin ʃɪňʃɪ ́ ʃɪn̄ʃɪ̀ ʃùʔɛn̄ ʃīʔīn ʃīʔīn ʃìkʼ sīkʼ sīkʼ sīkʼ sīkʼ tendon, vein *ʃʊ(n)kʼ ʃʊ̄nkʼ ʃʊ̄nkʼ ʃʊ̄ ʃúʔ à+ʃúʔ ʃùkʼ sʊ̄kʼ sʊ̄kʼ sʊ̄kʼ ʃūkʼ rope_1 *ʃutʼ ? ʃʊ̄dʊ̀l ʃwītʼìn ʃʊ̀ʔɪ ́ ʃí à+ʃí ʃʊ́l - - - - nose *ʃʊnʃ ʃʊ̃ʃ̄ ʃʊ̃ʃ̄ ʃʊ̀nʃ ʃūʃ ʃūʃ ʃʊ̀ʃ sʊ̀sʊ̀ sʊ̀s sʊ̀s ʃʊ̀ʃ belly or stomach_1 *bùʃùl bùʃùl bùʃì bùʃ - bùʃ - pùsà pùsà pùsà pùsà tree_sp._1 *ɓaʃa pʼàʃ - ɓàʃ ɓàʃà - ɓàʃà ɓāsā ɓāsā ɓāsā - lie down, sleep_1 *iʃ ìʃ ìʃ ìʃ īʃ īʃ īʃá - - - - 605 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi smoke out (e.g an animal out of a *kʰuʃ kʊ̌ʃ kʊ̌ʃ ūs - - húʃ űs űs űs űʃ hole)_1 kill, fight *kʼOʃ kʼʊ́ʃ kʼʊ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́sɔ́ kʼɔ́sɔ́ kʼɔ́s kʼɔ́ʃ wife_1, marry, wedding *màʃ - pʼā+màʃ màʃ màʃ màʃ màʃ màs màs màs màs wrap *pʊʃ ~ puʃ pʊ̄ʃ pʊ̄ʃ - pūʃ pūʃ púʃá pūsá pūsá pūsá pūsá nose *ʃʊnʃ ʃʊ̃ʃ̄ ʃʊ̃ʃ̄ ʃʊ̀nʃ ʃūʃ ʃūʃ ʃʊ̀ʃ sʊ̀sʊ̀ sʊ̀s sʊ̀s ʃʊ̀ʃ make go away *tʰaʃ ~ t ̪h aʃ tāʃ tāʃ tàʃ tʰáʃ tʰáʃ - - - - - salt_(made from ash of a particular *tʼaʃ tʼāʃ tʼāʃ tʼàʃ tʼāʃ tʼāʃ tʼàʃ tʼās - - - plant/tree)_2 buttocks_1 *ʊʃ īʃ ʊ̄ʃ - - - ʊ̄ʃ ʊ̄s ʊ̄s ʊ̄s ʊ̄ʃ birth pangs (pain) *ʃāʃ(ā) ʃɛ ́ ʃɛ ́ ʃāʃ - - ʃāʃā sāsā sāsā sāsā ʃāʃā D.24 PKMN Correspondence Set 24a: *ʃ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi foot or leg_1 *ʃɔnkʼ sɔ̄ŋkʼ sɔ̄ntʼ ʃɔ̀g ʃɔ̄kʼ ʃɔ̄ʔ ʃɔ̀g - - - - wake (trs.)_1 *ʃukʼ(V) sūgì sūgì ʃùg ʃūkʼ - ʃùg sūg sūg sūg ʃūg bone *ʃUImakʼ sí sí ʃúmákʼ sīmāʔ à+sīmāʔ ʃʊ́j sʊ́j sʊ́j sʊ́j sʊ́j meat, animal *ʃum(a) sūm sūm ʃùm ʃūm ʃūm ʃùmà sūmā sūmā sūmā ʃūmā disregard *piʃ pīs pīs píʃ - píʃ píʃ pıs̋ pıs̋ pıs̋ pıs̋ D.25 PKMN Correspondence Set 25: *h Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi have sex_1 *hag(a) hã́ʔ hã́ʔ hág - háʔ - hágá hágá hágá hágá trample, ruminate *hasʼ - hāsʼɪ̀ hásʼ hásʼ hát ̪̓ hásʼ hátʃʼ hátʃʼ hátʃʼ hátʃʼ fish_(small, small scales) *jàhút ̪ jàhú jàhú jàhú - - àhút ̪h àhű àhű àhűwı ̋ àhűj 606 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi come, come_SG *ha hɔ̌ hɔ̌ hà + ʊ́ - - - - - - - sip (liquid) *wɔpʼ(ɔsʼ) ~ hɔpʼ(ɔsʼ) hɔ̌bɔ̀sʼ hɔ̌bɔ̀sʼ wɔ́pʼ kʰɔ̄bɔ̄s - hɔ̀pʼ hɔ̄pʼɔ̄ hɔ̄pʼ hɔ̄pʼ hɔ̄pʼ D.26 PKMN Correspondence Set 26: *r Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi corner *rOk⁽ʰ⁾Oɲ rɔ́kɔ̀n rɔ́kɔ̀n rɔ́kɔ̀n rúgùɲ à+rúkūɲ rɔ́kʰɔ̀n rʊ́kʊ̄n rʊ́kʊ̄n rʊ́kʊ̄n rʊ́kʊ̄n swagger, arrogant (be) *(gɔ)kwar(a) - kwárá gɔ̀kɔ̀l kār+īs kār+īs kʰɔ̄r kʰɔ̄r kʰɔ̄r kʰɔ̄r kʰɔ̄r bird_heron *bár - ʊ̄+bár à+bár - - à+bár à+bár à+bár à+bár à+bár green_1 *D̪ir zı̃ ̀ zĩ ̂ zì zì d̪ì sīsī tʃìr tʃìr tʃìr tʃìr bird_stork (maribou) *ɟàrú zɛ̀rú zɛ̀rú zɛ̀rú ʒàrú - ɟɛ̀rú dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű chief_1 *kOr ʊ̄+kʊ̄l ʊ̄+kwɪ̀ jī+kwɪ̀ - - jɛ̀+kɔ̄rɔ̄ - - - - platform *pʰárá párá párá párá - - pʰárá pʰárá pʰárá pʰárá pʰárá drum_C *tà̪rá - tàrá tàrá - - tà̪rá tàrá tàrá tàrá tàrá D.27 PKMN Correspondence Set 27: *l Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi bed_1 *langarɛt ángàr - ángàr áŋgàr - - làŋgàrɛt́ àŋgàrípʰ àŋgàrípʰ àŋgàrípʰ sink (descend) *lili līlí līlí lìlí - - līl līlí līlí līlı ̋ līlí testicles_1 *lùt ̪̓ ~ lʊ̀t ̪̓ dūtʼ dūtʼ lʊ̀tʼ lùd à+lùɗ - - - - - bird_vulture (white- *lʊm ʊ̄lʊ̌m - lʊ̌m lùʔ - lʊ̄m lʊ̌m lʊ̌m lʊ̌m lʊ̌m backed)_1 claves (instrument)_ *lɛ̀pʰɛ ́ lɛ̀pɛ ́ àlàpɛ ́ lɛ̀pɛ ́ lɛ̀pʰɛ ́ - lɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ àlɛ̀pʰɛ ́ 1 palate_1 *ɓalilaj pʼàlílì pʼàlílí ɓàlílá - - - - - - - 607 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi stomp, step on_1 *djalIsʼ dɛ̀lísʼ dàlísʼ dɪ̀l dìl dìl - tɪ̀lɪ ́ tɪ̀lɪ ́ tɪ̀lɪ ́ tɪ̀lɪ ́ shout_1 *dɔl(ɔ) dɔ́l dɔ́l dɔ́l - - ɔ̀lɔ́ - - - - fish_sp (very small, scaled fish with a *dɔlɔ - dɔ̌lɔ́ dɔ̀lɔ́ - - dɔ́lɪʔ̄ dɔ̀lɔ́ dɔ̀lɔ́ dɔ̀lɔ́ dɔ́lɪ ̄ small rounded mouth) bird (yellow- billed kite or *gʊlɪla - gʊ́lːā bāgʊ́lɪ̀lā - - ɓāgʊ̂lːā - ɓāgʊ́lʊ́l ɓāgʊ́lʊ́l ɓāgʊ́là black kite) ululate_1 *ɪlɪl ~ ilil ɪĺɪ̀l ɪĺɪ̀l ílīl - - ílíl ıl̋ıl̋ ıl̋ıl̋ ıl̋ıl̋ ıl̋ıl̋ come free and fall off *pʰal pɪ ̌ pɪ ̌ - - - - pʰál pʰál pʰál pʰál spicy (be)_1 *pʰUl pʊ̀l pwì - - - pʰîl pʰıl̋ī pʰıl̋ī pʰıl̋ī pʰıl̋ī climb_1 *sɛl sál sɛl̄ - sɛ ̄ sɛ ̄ - - - - - D.28 PKMN Correspondence Set 28: *m Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi shut_1, close eyes *mVsʼ mɪ̀sʼ mɪ̀sʼ mʊ̀sʼ músʼ+ɛ ̄ mút ̪̓+ɛ ̄ mìsʼà mʊ̀tʃʼà mʊ̀tʃʼà mʊ̀tʃʼà mʊ̀tʃʼà carry on back_1 *màm(a) màm màm màmá màm màm màmá màmā màmā màmā màmā mosquito_2, firefly *mɪmɪ mɪm̌ɪ ́ mɪm̌ɪ ́ mɪm̌ɪ ́ mǐmī - - - - - - wife_1, marry, wedding *màʃ - pʼā+màʃ màʃ màʃ màʃ màʃ màs màs màs màs hand_1 *mɛ̀t ̪̓ bɪ̀tʼ mɪ̀tʼ - mɛ̀d mɛ̀ɗ mɛ̀t ̪̓ mɪ̀tʼɪ ́ mɪ̀tʼɪ ́ - mɛ̀tʼ drum_B *bàmbá pàmbà pàmbà - bàmbá - - - - - - pregnant (be) *bUma pʊ̀m pʊ̀m ɓú pwá bwà pùmá pǔmá pǔmá pǔmá pǔmá tree (sp.)(sausage tree_Kigelia africana) *dùmàj - ū+dùmɪ̀ dùmɛ̀ - - dùmáj dùmàj dùmàj dùmàj dùmàj strain (solids from liquid)_1 *d̪ìm - zìŋā zìm zǐm d̪ìm tì̪m tìm tìm tìm tìm bee, honey *d̪àm tàm tàm dàm dàm à+d̪àm tâ̪m tàm tàm tàm tàm 608 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Gwama (ethnonym) *gwama gwàmá kwāmà gwàmá - - gɔ̀má gɔ̀má gɔ̀má gɔ̀má gɔ̀má find, meet *gàm kàm kàm gàm gàm gàm kàm kàm kàm kàm kàm pack in, stuff into container_1 *gɪ̀m gɪ̀m gɪ̀m - - ɟìm - - - - - wound_1 *gjama? kāmā kāmā zàmā ɟàmá ɟàmá - - - - - cover (v.) *kum(bi) kùmbì kùmbì kúm kūm kūm kúmā kűmá kűmá kűmá kűmá brother *ʊ=kam *kamʊ kwám kwám kàm kām à+kām āmʊ́ hàm hàm hàm hàm eat (hard food) *kʼama kʼā kʼā kʼá kʼáʔ kʼá kʼámá kʼámá kʼámá kʼámá kʼámá bird_vulture (white- backed)_1 *lʊm ʊ̄lʊ̌m - lʊ̌m lùʔ - lʊ̄m lʊ̌m lʊ̌m lʊ̌m lʊ̌m warm oneself *s(w)am sɔ́m sɔ́m ʃɔ́m sām sām sɔ̀m sɔ̄m sɔ̄m sɔ̄m ʃɔ̄m python_1 *sʊm ʃʊ̄ʃʊ́m ʃʊ̄ʃʊ́m ʃʊ̌m - súm sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ blood_1 *sʼámá sʼám sʼám - - - sʼámáʔ tʃʼámá tʃʼámá tʃʼámá tʃʼámá bone *ʃUImakʼ sí sí ʃúmákʼ sīmāʔ à+sīmāʔ ʃʊ́j sʊ́j sʊ́j sʊ́j sʊ́j meat, animal *ʃum(a) sūm sūm ʃùm ʃūm ʃūm ʃùmà sūmā sūmā sūmā ʃūmā D.29 PKMN Correspondence Set 29: *n Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi honey badger *njansʼɛsʼ jānsʼɛ́sʼ ɲɛn̄zɛ́sʼ nɛ̀z nɛ̀sʼ à+nɛ̀t ̪̓ nɛ̀s nɛ̀dʒɛ̀ - nɛ̀dʒɛ̀ - disabled (be), angry *nap(a) nǎpá nǎpá nǎpá - nápʰɛ ̄ nápʰ nāp nāp nāp nāp (be) bird_hammerkopf_1 *nɛ́kʰ - ʊ̄+nɛ́k à+nɛ́k nɛ́kʰ à+nɛ́kʰ à+nɪḱʰ - - à+nɪḱʰ à+nɪḱʰ *nʊpʼ ~ bury (sideways) ʊ̄mpʼ ʊ̄mpʼ - - - nʊ̀pʼá - - - - *nʊmpʼ bird_pelican *nàbɔ̀ng(w)à - ʊ̄+nàbɔ̀ŋà nàbɔ̀ŋà - - nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ (brown) Dana *d̪ana dànā dànā dāná - - d̪àná dànā dànā dànā dànā (ethnonym)_3 609 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Opo *k⁽ʰ⁾ɪńáj kɪńá kɪńá kɪńá cʰínáj - kɪńáj - - - - (ethnonym)_1 sorghum, millet *ɟana sjànà sjànà zɛ̀nā - - - dʒɛ̀ná - dʒɛ̀ná - snore *kʰO(r)nOn kɔ̀nɔ̀n kɔ̀nɔ̀n kɔ̀nɔ̀n kʰūnūn àkʰɔ̄rnɛ ̄ kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n tomorrow_1 *kjana - gɪ̀+kjānā - - - - dʒì+kɛń dʒì+kɛń zì+kɛń ʃì+kɛń *ʃɪntʼ ~ blow nose ʃɪn̄tʼ ʃɪn̄tʼ ʃīn+ʃʊ̀nʃ ʃīn ʃīn ʃìnà+ʃʊ̀ʃ sın̋á sın̋á sın̋á sın̋á *ʃintʼ alone, abstain from, not want to *t ̪̓ ɛn - sʼɪn̄ sʼɛń tʼɛń tʼɛ ́ gà+t ̪̓ ɛń ā+tʼɛń ā+tʼɛń ā+tʼɛń ā+tʼɛń do seed_2 *ʔjamVn - - ʔjāmʊ́n ɛm̄ɛń ɛm̄ɛń - - - - - D.30 PKMN Correspondence Set 30: *ɲ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi ash_1 *(tʼi)pʼIkʼIɲ(a) pʼɪk̄ʼɪń pʼɪk̄ʼɪń pʼīn - tʼīpʼīɲ pʼɪn̄ā - - pʼɪn̄ā pʼɪn̄ā bird_(cattle egret) *kEɲ kɪl̄ kɪl̄ à+kɪĺ cɛćɛńā à+cɛɲ́ à+kɪĺ à+kɪĺ à+kɪĺ à+kɪĺ à+kɪĺ light (the way with *kʰàɲ kɪn̄ kɪn̄ - kʰāɲ kʰāɲ - - - - - torch/flashlight)_1 corner *rOk⁽ʰ⁾Oɲ rɔ́kɔ̀n rɔ́kɔ̀n rɔ́kɔ̀n rúgùɲ à+rúkūɲ rɔ́kʰɔ̀n rʊ́kʊ̄n rʊ́kʊ̄n rʊ́kʊ̄n rʊ́kʊ̄n D.31 PKMN Correspondence Set 31: *ŋ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi path, road_1, towards *bwaŋ(a) pwǎŋà pwǎŋà - - bwàj - - - - - baboon_1, dog_2 *dɪŋɪ tɪ̀nɪ̀ tɪ̀nɪ̀ - - - tɛ̀ŋ tɪ̀nɪ̀ tɪ̀nɪ̀ tɪ̀n tɪ̀n 610 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi fish_sp ʊ̄+wàsʼ (electric)_1 *dɪŋ́kā - +dɪŋ́kā dɪŋ́kā - - - dıŋ̋ā dıŋ̋ā dıŋ̋ā dıŋ̋ā smell (v.) *gàŋ(a) kɛ̃ ̄ kɛ̃ ̄ gàg - - kàŋà kàŋà kàŋà kàŋà kàŋà Nuer (ethnonym)_1 *ɟaŋg(ɔ)aj zǎgʊ́ zǎgʊ́ zǎgʊ́ ʒwǎŋgì - ɟàŋwɛ̀j dʒāŋɔ́ - zāŋwɛ ́ ʃàŋwɛ̀ shake (sth.)_2 *tEŋ(g)(E) tɪḡɪ ̄ tɪḡɪ ̄ - - - tɛŋ́ - tɪŋ́há tɪŋ́há tɪŋ́há cut (split in half *tʼwaŋ(k)a lengthwise)_1 ~ t ̪̓waŋ(k)a tʼāŋà tʼāŋà - tʼwákʰ tʼwákʰ - - - - - chicken_1 *waŋa wāŋā wāŋā wàgá ŋwá à+ŋwá - - - - - D.32 PKMN Correspondence Set 32: *w Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi boil (of liquid)_1 *was(ikʼ) wǎʃ wàʃí wàʃíkʼ - - wás wās wās wās wās hail, ice_1 *wasakʼ wàsà wàsà wàʃākʼ wàsáʔ à+wàsáʔ - - - - - break (v.)_1 *wa wǎ wǎ wà wá wá - - - - - fire or firewood *wa(n)t ̪̓(Iʃ) āntʼ āntʼ wàtʼíʃ ɔ̄d ɔ̄ɗ ɔ̀t ̪̓ wʊ̄tʼɪ ́ wʊ̄tʼɪ ́ wʊ̄tʼɪ ́ ʊ̄tʼɪ ́ boar (wild)_1 *wàɓ wàpʼ wàpʼ wàpʼ wàpʰ à+wàɓ - - - - - fish (n.) general term *wàcʼà wàsʼ wàsʼ wàsʼ wàʃʼ wàcʼ wàcʼà wàtʃʼà wàtʃʼà wàtʃʼà wàtʃʼà chicken_1 *waŋa wāŋā wāŋā wàgá ŋwá à+ŋwá - - - - - sky_1 *wUs wús wús - - - - wʊ̀s wʊ̀s wʊ̀s wʊ̀s bird_ostrich_1 *wutʰ - wǔt wǔt útʰ à+útʰ - hűtʰ hűtʰ hűtʰ hűtʰ become, become angry *wVɗ wɛť wɛť wàl wár wáɗ wâl - - - - D.33 PKMN Correspondence Set 33: *j Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi wet_1, slippery *jEsI ɪʃ́ɪ̀ ɪʃ́ɪ̀ jɛ̀ʃ jɛ̀s jɛ̀s sìʔ - - - - 611 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi vomit_1 *(pa)jasʼ pājàsʼ pājàsʼ jàʔ ɟǎʔ jǎʔ - - - - - play (instrument)_2 *jɪ jɪ̀ jɪ̀ - - - - dʒɪ ̄ dʒɪ ̄ zɪ ̄ ʃɪ ̄ sow seeds (by throwing)_1 *jEkʰ jɪ̀ jɪ̀ jɛ̀k jɛ̀kʰ jɛ̀kʰ - - - - - water *ji(ɗE) ìjáʔ ìjáʔ jǐ wùdíʔ jìɗɛʔ́ jìʔí dʒì dʒì zì ʃìʔ sweat (substance)_2 *jàgàl jàgàl jàgɪ̀ - - - - dʒīkʼāj dʒīkʼāj zīkʼāj ʃīkʼāj elephant_1 *gwàj kwɪ̀ kwɪ̀ gwà gwàj - - - - - - herd (v.) *kʰaj kɛ̃ʔ̄ kɛ̃ʔ̄ kàʔí - - kʰàjɪʔ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ cry *kO(j) kū kū kɔ̀ kɔ̄ kɔ̄ kɔ̀j kwɛ ̄ kwɛ ̄ kwɛ ̄ kwɛ ̄ blow (with mouth) *pʰui(ki) pɪk̄ɪ ̄ pɪ ̄ pɪ̀ pʰûj pʰí pʰùj pʰű pʰű pʰű pʰű D.34 PKMN Correspondence Set 34: *T̪ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi descend (land) *T̪Ol(a) - ʃʊ̀ - sūl tū̪l - tʃɔ́já tʃɔ́já tʃɔ́já tʃɔ́já guinea fowl_1 *T̪ʊ(n)G ʃʊ̌nkʼ ʃʊ̌nkʼ zʊ̀g - tū̪kʰ sʊ̄kʰ tʃʊ̄gʊ́ tʃʊ̄gʊ́ tʃʊ̄g tʃʊ̄g D.35 PKMN Correspondence Set 34: *D̪ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi eye_1 *D̪E zì zì - ī ɛ ̄ d̪ɛ̀ dʒɛ̀ dʒɛ̀ zɛ̀ ʃɛ̀ doze off_1 *D̪i sɪśɪ ̄ sɪśɪ ̄ zìzí - - - dìsì dìsì zìz sìs green_1 *D̪ir zı̃ ̀ zĩ ̂ zì zì d̪ì sīsī tʃìr tʃìr tʃìr tʃìr person_2 *D̪ita̪ sīt sīt - - - d̪ítà̪ - - ʊ̀+zìtà ʊ̀+ʃìtà seed_1 *D̪E(se) zì zì - - - d̪ɛ̀sɛ̂ dʒɛ̀sɛ ̄ dʒɛ̀sɛ ̄ zɛ̀sɛ ̄ sɛ̀sɛ ̄ 612 APPENDIX E PROTO-KOMAN VOWEL CORRESPONDENCE SETS 11 APPENDIX E PKmn vowel correspondence sets E.1 Correspondence Set V1: PKMN *i1 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi green_1 *D̪ir zı̃ ̀ zĩ ̂ zì zì d̪ì sīsī tʃìr tʃìr tʃìr tʃìr person_2 *D̪ita̪ sīt sīt - - - d̪ítà̪ - - ʊ̀+zìtà ʊ̀+ʃìtà lie down, sleep_1 *iʃ ìʃ ìʃ ìʃ īʃ īʃ īʃá - - - - water *ji(ɗE) ìjáʔ ìjáʔ jǐ wùdíʔ jìɗɛʔ́ jìʔí dʒì dʒì zì ʃìʔ sink (descend) *lili līlí līlí lìlí - - līl līlí līlí līlı ̋ līlí shake (sth.)_1 *pid(V) - pǐdí pídá pítʰ pír - - - - - strain (solids from liquid)_1 *d̪ìm - zìŋā zìm zǐm d̪ìm tì̪m tìm tìm tìm tìm bird_dove (African mourning) *tiritiri - títìtī - - - - à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì disregard *piʃ pīs pīs píʃ - píʃ píʃ pıs̋ pıs̋ pıs̋ pıs̋ E.2 Correspondence Set V2: PKMN *ɪ1 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi ash_1 *(tʼi)pʼIkʼIɲ(a) pʼɪk̄ʼɪń pʼɪk̄ʼɪń pʼīn - tʼīpʼīɲ pʼɪn̄ā - - pʼɪn̄ā pʼɪn̄ā fishhook *Bɪncʼ bɪn̄sʼ bɪn̄sʼ bɪńsʼ bǐʃʼ à+bǐcʼ bɪćʼ ɓītʃʼ ɓītʃʼ ɓītʃʼ ɓītʃʼ Opo (ethnonym)_1 *k⁽ʰ⁾ɪńáj kɪńá kɪńá kɪńá cʰínáj - kɪńáj - - - - pack in, stuff into *gɪ̀m gɪ̀m gɪ̀m - - ɟìm - - - - - container_1 play (instrument)_2 *jɪ jɪ̀ jɪ̀ - - - - dʒɪ ̄ dʒɪ ̄ zɪ ̄ ʃɪ ̄ bird_(cattle egret) *kEɲ kɪl̄ kɪl̄ à+kɪĺ cɛćɛńā à+cɛɲ́ à+kɪĺ à+kɪĺ à+kɪĺ à+kɪĺ à+kɪĺ 613 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi charcoal or coal *kʼI(m)Isʼ sʼɪśʼɪń sʼɪśʼɪń kìsʼísʼìʔ cʼɛl̄ɛs̄ʼ cʼīlāt ̪̓ kʼɪ̀sʼ kʼīmītʃʼ kʼīmītʃʼ kʼɪt̄ʃʼ kʼɪt̄ʃʼ mosquito_2, firefly *mɪmɪ mɪm̌ɪ ́ mɪm̌ɪ ́ mɪm̌ɪ ́ mǐmī - - - - - - peel, husk_1 *pʼi(n)Cʼ? pʼɪńsʼ pʼɪńsʼ - - pʼíɗ pʼɪt́ʼ - - - - cut (meat into one long piece)_2 *sɪĺ(ɪ) ʃɪĺɪ̀ ʃɪ ́ - - - - sɪ ́ - - - far (be)_1 *sɪ̀t ̪̓ - ʃɪt̄ʼ ʃɪ̀tʼ sīd sīɗ sɪ̀t ̪̓ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ roughen (stone for grinding) *tɪ̪t́ ̪ sɪt́ sɪt́ tɪt́ títʰ tír tɪ̪t́ ̪h tɪt́ɪ ́ tɪt́ʰ tɪt́ʰ tɪt́ʰ tree_sp. (mahogany, Trichilia emetica) *k⁽ʰ⁾ɪs kɪśɪ̀ ʊ̄+kɪs̄ kɪs̄ cɛś cʰís kɪś kʰɪś kʰɪś kʰɪś kʰɪś E.2 Correspondence Set V2a: PKMN *ɪ2 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi strong (be)_1 *ɓɪśʼ ~ ɓísʼ pʼɪ ́ pʼɪ ́ ɓísʼ ɓísʼ ɓít ̪̓ - - - - - toss, throw away, fall over *bit⁽ʰ⁾ pɪ̀t pɪ̀t bìt bǐtʰ bǐtʰ - - - - - doze off_1 *D̪i sɪśɪ ̄ sɪśɪ ̄ zìzí - - - dìsì dìsì zìz sìs fish_sp (electric)_1 *dɪŋ́kā - ʊ̄+wàsʼ +dɪŋ́kā dɪŋ́kā - - - dıŋ̋ā dıŋ̋ā dıŋ̋ā dıŋ̋ā tie (bundle)_1 *gì(n)sʼ kɪ̀nsʼ kɪ̀nsʼ gìsʼ - - - - - - - enter_1 *gìcʼ kɪšʼ kɪšʼ gìz - cīcʼ - kìtú kìtú kìtú kìtú LOC, BE LOC *ɪ ́~ *í - ɪ ́ ɪ ́ - í - ı ̋ - - - ululate_1 *ɪlɪl ~ ilil ɪĺɪ̀l ɪĺɪ̀l ílīl - - ílíl ıl̋ıl̋ ıl̋ıl̋ ıl̋ıl̋ ıl̋ıl̋ ripen *ís ~ ís ɪś ɪś íʃ ís ís ísá ıs̋á ıs̋á ıs̋á ıt̋ʃá vagina_1 *pít ̪h ~ pɪt́ ̪h pɪt́ pɪt́ pít - - pít ̪h - - - - warm up (sth.) *Cisʻ ʃɪʃ̄ ʃɪʃ̄ jíz jísʼ jít ̪h hízá ıs̋á ıs̋á ıs̋á - blow nose *ʃɪntʼ ~ *ʃintʼ ʃɪn̄tʼ ʃɪn̄tʼ ʃīn+ʃʊ̀nʃ ʃīn ʃīn ʃìnà+ʃʊ̀ʃ sın̋á sın̋á sın̋á sın̋á rat_1 *sʼÍk sʼɪ ́ sʼɪ ́ sʼík sʼíʔ à+t ̪̓íkʰ - tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ 614 E.3 Correspondence Set V3: PKMN *u1 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi choke, strangle_1 *busʼ bǔsʼ bǔsʼ - bùsʼ bùt ̪̓ - pʰǔtʃʼ - - - dust, sand_1 *burbuɗ bùrbūt bùrbūt - bǔtʰ bǔɗ būrkʼùs būrkʼùs būrkʼùs būrkʼùs būrkʼùs belly or stomach_1 *bùʃùl bùʃùl bùʃì bùʃ - bùʃ - pùsà pùsà pùsà pùsà extract tooth_1, barking (of dog) *bʊkʰ pʊ̌ pʊ̌ - - bǔkʰ - - - - - urine_1, urinate_1 *dùcʼá dùsʼ tùsʼ dʊ̀sʼ - - tùcʼáʔ - - - - tree (sp.)(sausage tree_Kigelia africana) *dùmàj - ū+dùmɪ̀ dùmɛ̀ - - dùmáj dùmàj dùmàj dùmàj dùmàj fish_(small, small scales) *jàhút ̪ jàhú jàhú jàhú - - àhút ̪h àhű àhű àhűwı ̋ àhűj wake (trs.)_1 *ʃukʼ(V) sūgì sūgì ʃùg ʃūkʼ - ʃùg sūg sūg sūg ʃūg cover (v.) *kum(bi) kùmbì kùmbì kúm kūm kūm kúmā kűmá kűmá kűmá kűmá smoke out (e.g an animal out of a *kʰuʃ kʊ̌ʃ kʊ̌ʃ ūs - - húʃ űs űs űs űʃ hole)_1 follow_2 *uD ū ū út úr úr - - - - - spit (v.)_1 *t ̪h ú tū tū - - - t ̪h úwà tʰűjhá tʰűjhá tʰűjhá tʰűjhá meat, animal *ʃum(a) sūm sūm ʃùm ʃūm ʃūm ʃùmà sūmā sūmā sūmā ʃūmā bird_ostrich_1 *wutʰ - wǔt wǔt útʰ à+útʰ - hűtʰ hűtʰ hűtʰ hűtʰ E.3a Correspondence Set V3a: PKMN *u2 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi pregnant (be) *bUma pʊ̀m pʊ̀m ɓú pwá bwà pùmá pǔmá pǔmá pǔmá pǔmá pig_1 *gUd̪Um kʊ̀rʊ́m kʊ̀rʊ́m gùdúm - - gùd̪ùm kùdùmà kùdùmà kùdùmà kùdùmà short (be)_1 *KÙtʰ gʊ̄t gʊ̄t kùt kūtʰ kūtʰ - - - - - run (SG)_1, flow, bleed *gùs ~ gʊ̀s gʊ̀s gʊ̀s gùʃ gùs gùs - - - - - cough *kʼUt ̪̓ kʼʊ́kʼʊ́t kʼʊ́kʼʊ́t kʼùt kʼútʰ kʼútʰ kʼút ̪h kʼútʼù kʼútʼù kʼútʼù kʼútʼù 615 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi wrap *pʊʃ ~ puʃ pʊ̄ʃ pʊ̄ʃ - pūʃ pūʃ púʃá pūsá pūsá pūsá pūsá burnt gound, soot *pʰuZa ʃāpʊ́t ʃābʊ́t kɪ+́pú bwà+pʰwí à+pʰíʔ pùzàʔ pʰűjhá pʰűjhá pʰűjhá pʰűjhá dip food in sauce with *sʼUB(V)(n) - sʼʊ́pʊ́n sʼūb sʼúpʰ t ̪̓úpʰ sʼúbá - - - - fingers_1 nosebleed_2 *SʼUn(tʼ)a - tʼʊ̀ntʼ - - - sʼùnáʔ - - - - bathe *úpʰ ʊ́p ʊ́p úp úpʰ úpʰ úpʰ űpʰ űpʰ űpʰ űpʰ cotton, thread, spider web_1 *tʊ̪s ~ tu̪s dʊ̄ʃ dʊ̄ʃ tūʃ tūs tū̪s - - - - - acacia *tú̪k⁽ʰ⁾(u) tʊ̌kʊ̀ - túk túkʰ à+tú̪kʰ tú̪kʰ tűkʰ tűkʰ tűkʰ tűkʰ pierce_1 *tʊ̪b(a) ~ tu̪ba(a) sʊ́ sʊ́ - tūpʰ tūpʰ tú̪bá - - tʰūbá tʰūbá E.4 Correspondence Set V4: PKMN *ʊ1 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi pipe (for smoking)_1 *dʊ́ɟɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ - - dʊ́ɟɛ̀ - - - - bird (yellow-billed kite or black kite) *gʊlɪla - gʊ́lːā bāgʊ́lɪ̀lā - - ɓāgʊ̂lːā - ɓāgʊ́lʊ́l ɓāgʊ́lʊ́l ɓāgʊ́là head *kʼʊ́p kʼʊ́p ʔʊ́p kʼʊ́p kʼúpʰ kʼúpʰ kʼʊ́pʰ kʼʊ́p kʼʊ́p kʼʊ́p kʼʊ́p bird_vulture (white- backed)_1 *lʊm ʊ̄lʊ̌m - lʊ̌m lùʔ - lʊ̄m lʊ̌m lʊ̌m lʊ̌m lʊ̌m bury (sideways) *nʊpʼ ~ *nʊmpʼ ʊ̄mpʼ ʊ̄mpʼ - - - nʊ̀pʼá - - - - guinea fowl_1 *T̪ʊ(n)G ʃʊ̌nkʼ ʃʊ̌nkʼ zʊ̀g - tū̪kʰ sʊ̄kʰ tʃʊ̄gʊ́ tʃʊ̄gʊ́ tʃʊ̄g tʃʊ̄g nose *ʃʊnʃ ʃʊ̃ʃ̄ ʃʊ̃ʃ̄ ʃʊ̀nʃ ʃūʃ ʃūʃ ʃʊ̀ʃ sʊ̀sʊ̀ sʊ̀s sʊ̀s ʃʊ̀ʃ lead (guide)_1 *sʊs ʃʊ́ʃ ʃʊ́ʃ ʃʊ́ʃ sús sús sʊ́ʔ - - - - python_1 *sʊm ʃʊ̄ʃʊ́m ʃʊ̄ʃʊ́m ʃʊ̌m - súm sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ cold (be)_1 *sʼʊ́p sʼʊ́p sʼʊ́p - sʼúpʰ t ̪̓úpʰ - - - - - copulate (animal)_3 *ʃʊ́ɓ ʃʊ́pʼ ʃʊ́pʼ - ʃúb ʃúɓ - - - - - 616 E.4a Correspondence Set V4a: PKMN *ʊ2 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi throat *kʼʊ́s k’úʃ k’úʃ kʼʊ́ʃ kʼús kʼús kʼʊ́s kʼʊ́sʊ́ kʼʊ́s kʼʊ́s kʼʊ́s swallow_1 *gUsʼ gùsʼ kùsʼ gʊ̀sʼ - - kʊ̄sʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá dry (be) *kʰʊ́sʼ kúsʼ kúsʼ kʊ́sʼ kʰúsʼ kʰút ̪̓ kʰʊ́sʼ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ - testicles_1 *lùt ̪̓ ~ lʊ̀t ̪̓ dūtʼ dūtʼ lʊ̀tʼ lùd à+lùɗ - - - - - long or tall (be)_1 *tur ~ tʊr tǔ tǔ tʊ́l túr túr - - - - - defecate_1, diarrea *sʼUDI úsʼ úsʼ sʼɪ ́ - t ̪̓í sʼʊ̀dʊ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ sky_1 *wUs wús wús - - - - wʊ̀s wʊ̀s wʊ̀s wʊ̀s E.5 Correspondence Set V5: PKMN *ɛ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi ear *cʼɛ sʼɛ ̄ sʼɛ ̄ sʼɛ ̄ ʃʼɛ ́ cʼɛ ́ kʼɛ ̄ tʃʼɛ̀ tʃʼɛ̀ tʃʼɛ̀ tʃʼɛ̀ seed_1 *D̪E(se) zì zì - - - d̪ɛ̀sɛ̂ dʒɛ̀sɛ ̄ dʒɛ̀sɛ ̄ zɛ̀sɛ ̄ sɛ̀sɛ ̄ slip (v.)_1 *d̪ɛrkʼɛsʼ dɛr̄gɛśʼ dɛr̄gɛśʼ - dɛ̀rɛ̀s d̪ɛ̀rɛ̀s d̪ɛŕkʼɛś - - - - bird_stork (maribou) *ɟàrú zɛ̀rú zɛ̀rú zɛ̀rú ʒàrú - ɟɛ̀rú dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű beehive basket *gɛnd̪(V)(l) gɛńdɛĺ gɛ̀ndɪ ́ kɛ̀ndɛ ̄ - - gɛ̀nd̪á gɛ̀ndá gɛ̀ndá gɪ̀ndá gɛ̀ndá climb_1 *sɛl sál sɛl̄ - sɛ ̄ sɛ ̄ - - - - - antelope_1 *ʃɛt̄(ʰ) ʊ̄ʃɛt̄ ʃɛt̄ ʃɛt̄ ʃɛt̄ʰ à+ʃɛt̄ʰ - - - - - alone, abstain from, not want to do *t ̪̓ ɛn - sʼɪn̄ sʼɛń tʼɛń tʼɛ ́ gà+t ̪̓ ɛń ā+tʼɛń ā+tʼɛń ā+tʼɛń ā+tʼɛń E.6 Correspondence Set V6: PKMN *ɔ Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi carry many things_1 *(tɔ)ɗɔ tɔ̀dɔ̀ - ɗɔ̀ ɗɔ̄ ɗɔ̄ - ɗɔ̄ ɗɔ̀ ɗɔ̀ ɗɔ̀ tell_1 *(w)ɔt(V) ɔ̌dɔ́ ɔ̌d - - - - ɔ́tá ɔ́tá ɔ́tá - 617 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi hide, skulk *bɔ̀b ~ bɔ̀ɓ pɔ̀pʼ pɔ̀pʼ bɔ̀b - - - pɔ̀j pɔ̀j pɔ̀j - drip (fall in globules)_1 *cʼɔ(tʼɔ)l sʼɔ́tʼɔ́ sʼɔ́tʼɔ́ sʼʊ̀lɪl̄ - cʼɔ̄cʼɔ̄lɔ̄cʼ cʼɔ̀ʔ tʃʼɔ̄ tʃʼɔ̄ tʃʼɔ̄ tʃʼɔ̄ shout_1 *dɔl(ɔ) dɔ́l dɔ́l dɔ́l - - ɔ̀lɔ́ - - - - fish_sp (very small, scaled fish with a small rounded *dɔlɔ - dɔ̌lɔ́ dɔ̀lɔ́ - - dɔ́lɪʔ̄ dɔ̀lɔ́ dɔ̀lɔ́ dɔ̀lɔ́ dɔ́lɪ ̄ mouth) skin, hide (of animal), bark of *gɔnkʼ(ɔʃ) gɔ̀kʼɔ́ʃ gɔ̀kʼɔ́ʃ gɔ̀nkʼí gɔ̀kʰ - kɔ̀gɔ̀ kɔ̀gɔ̀ kɔ̀gɔ̀ kɔ̀gɔ̀ gwàŋgɪ ́ tree_1 foot or leg_1 *ʃɔnkʼ sɔ̄ŋkʼ sɔ̄ntʼ ʃɔ̀g ʃɔ̄kʼ ʃɔ̄ʔ ʃɔ̀g - - - - roast (something) *t ̪̓ ɔ́s tʼɔ̄ʃ tʼɔ̄ʃ tʼɔ́ʃ tʼɔ́s t ̪̓ ɔ́s - - - - - pound (v.)_4 *kɔ́pʰ kɔ́p kɔ́p - kúpʰ - - - - - - frog_1 *bɔŋkʼɔ́ pʼà+ pʼà+ bā+ - - à+ à+ à+ à+ à+ bɔ̌ŋgɔ́ bɔ̌ŋgɔ́ ɓɔ̌ŋkʼɔ́ ɓɔ̄ŋkʼɔ́ʔ ɓɔ̄ŋkʼɔ́ bɔ̄ŋkʼɔ́ bɔ̄ŋkʼɔ́ bɔ̄ŋkʼɔ́ deep (be)_1 *gɔ̀ɗɔkʼ kwì kwǐ gʊ̀r - - kɔ̀ɗɔ̀kʼ kɔ̄rɔ́ kɔ̄rɔ́ kɔ̄rɔ́ kɔ̄rɔ́ stab_1 *sɔ́p sɔ́ sɔ́ - - - - sɔ́p - sɔ́p - sew_1 *kɔ́s ɔ́s ɔ́s - - - kɔ́s kɔ́s kɔ́s kɔ́s kɔ́s grind (second grind) *(ɔ)t⁽̪ʰ⁾ɔɗ - twɛj́ tɔ́ - tɔ̪́ɗ ɔ̀t ̪h ɔ́ ɔ́tɔ́ ɔ́tʰ ɔ́tʰ ɔ́tʰ sip (liquid) *wɔpʼ ~ hɔpʼ hɔ̌bɔ̀sʼhɔ̀pʼ hɔ̌bɔ̀sʼ wɔ́pʼ kʰɔ̄bɔ̄s - hɔ̀pʼ hɔ̄pʼɔ̄ hɔ̄pʼ hɔ̄pʼ hɔ̄pʼ E.7 Correspondence Set V7: PKMN *a Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi father_1 *bàbá - bàbá bǎ à+bàbá à+bàbá - àbá àbá àbá àbá finish_1 *dak - dak dàg dàkʰ dàkʰ dâk dàk - - - wide (be) *bàj ~ ɓàj pàj pàj ɓàjá ɓàn bɛ̀ - pàj pàj pàj pàj bird_heron *bár - ʊ̄+bár à+bár - - à+bár à+bár à+bár à+bár à+bár drum_B *bàmbá pàmbà pàmbà - bàmbá - - - - - - 618 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi bitter, sour *kʰakʼa kāgā kāgā kàʔ kʰāʔ kʰāʔ kʰàkʼà kʰākʼā kʰākʼā kʰākʼā kʰākʼā (be) palate_1 *ɓalilaj pʼàlílì pʼàlílí ɓàlílá - - - - - - - grandfather_1 *càkʰO - sàkʊ́ sàkʊ́ - - - tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ tree_sp._1 *ɓaʃa pʼàʃ - ɓàʃ ɓàʃà - ɓàʃà ɓāsā ɓāsā ɓāsā - dig_1 *ɟà ʃà ʃà - - - cà tʃà tʃà tʃà tʃà grind wet *Dàsʼ dàsʼ dàsʼ nàsʼ nàsʼ nàt ̪̓ - - - - - (first grind) cannabis_1 *bángwà báŋgà báŋgà bánkɔ̀ bǎŋgɔ̀ bǎŋgɔ̀ - - - - - Dazu (S. *dáɟV dázɔ̀ dázɔ̀ dázʊ̀ dǎɟɔ̀ dǎɟɔ̀ dáɟɔ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ Sudan) sneeze_1 *hat ̪̓is hǎtʼìʃ - - - - hát ̪̓ìs ha̋tʼīs ha̋tʼīs ha̋tʼīs ha̋tʼīs have sex_1 *hag(a) hã́ʔ hã́ʔ hág - háʔ - hágá hágá hágá hágá belt, sash *gaʃa gáʃà gáʃà - gǎʃ gāʃá gâʃ gātʃ gātʃ gātʃ gâʃ Dana *d̪ana dànā dànā dāná - - d̪àná dànā dànā dànā dànā (ethnonym)_3 *k⁽ʰ⁾akʼas ~ porcupine_1 kákʼàʃ kákʼàʃ káʃākʼ - - kásākʼ kʰásākʼ kʰásākʼ kʰásākʼ kʰáʃākʼ k⁽ʰ⁾asakʼ trample, *hasʼ - hāsʼɪ̀ hásʼ hásʼ hát ̪̓ hásʼ hátʃʼ hátʃʼ hátʃʼ hátʃʼ ruminate replace *(n(j)a)gaD nāgát njāgát gàdá gàs gàr gàtá kàrá kàrá kàrá kàrá vomit_1 *(pa)jasʼ pājàsʼ pājàsʼ jàʔ ɟǎʔ jǎʔ - - - - - find, meet *gàm kàm kàm gàm gàm gàm kàm kàm kàm kàm kàm come, *ha hɔ̌ hɔ̌ hà + ʊ́ - - - - - - - come_SG Nuer *ɟaŋg(ɔ)aj zǎgʊ́ zǎgʊ́ zǎgʊ́ ʒwǎŋgì - ɟàŋwɛ̀j dʒāŋɔ́ - zāŋwɛ́ ʃàŋwɛ̀ (ethnonym)_1 boil (of *was(ikʼ) wǎʃ wàʃí wàʃíkʼ - - wás wās wās wās wās liquid)_1 repair_1 *kʰaɓ áp - áb - - ʊ́bā kʰápʼ kʰápʼ kʰápʼ kʰápʼ shut_2 *kʰacʼ kǎʃ kǎʃ - kʰáʃʼ kʰácʼ - - - - - open *kʰáɗ(a) kálá kájá kár kʰǎd kʰǎɗ kʰátá kʰátá kʰátá kʰátá kʰátá herd (v.) *kʰaj kɛ̃ʔ̄ kɛ̃ʔ̄ kàʔí - - kʰàjɪʔ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ soft (be)_1 *kʼát ̪̓ kʼátʼ kʼátʼ kʼátʼ kʼád kʼáɗ - - - - - 619 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi *ɗ(w)ankʼi scorpion *tʼwānkʼ *tʼwānkʼ bāɗāgíʔ ɗwākʰ à+ɗākʰ ɗāgí ɗāgı ̋ ɗāgı ̋ ɗāgı ̋ ɗāgı ̋ ~ ɗ(w)ankʼɪ platform *pʰárá párá párá párá - - pʰárá pʰárá pʰárá pʰárá pʰárá bed_1 *langarɛt ángàr - ángàr áŋgàr - - làŋgàrɛ́ àŋgàríp àŋgàríp àŋgàríp t ʰ ʰ ʰ eat (hard *kʼama kʼā kʼā kʼá kʼáʔ kʼá kʼámá kʼámá kʼámá kʼámá kʼámá food) forehead_1 *t⁽ʰ⁾(w)àg twã twã ̄ tàg - - tàg bī+tʰāg - pī+tʰāg - grandmother_ *kʰàkʰá - kàká kàká - - kʰàkʰá kàká kàká kàká kàká 1 carry on *màm(a) màm màm màmá màm màm màmá màmā màmā màmā màmā back_1 wife_1, marry, *màʃ - pʼā+màʃ màʃ màʃ màʃ màʃ màs màs màs màs wedding disabled (be), *nap(a) nǎpá nǎpá nǎpá - nápʰɛ ̄ nápʰ nāp nāp nāp nāp angry (be) soak_1 *pʰàcʼ pāsʼ pāsʼ pàsʼ pʰāʃʼ pʰācʼ - - - - - come free and *pʰal pɪ ̌ pɪ ̌ - - - pʰál pʰál pʰál pʰál fall off - fly (v.) *pʰàd̪ pāl pāj pàj pʰāj pʰɛ ̄ pʰàd̪ pʰāj pʰāj pʰāj pʰāj blood_1 *sʼámá sʼám sʼám - - - sʼámáʔ tʃʼámá tʃʼámá tʃʼámá tʃʼámá light (ignite) *sʼa sʼā sʼā sʼà sʼā t ̪̓ ā sʼʊ́wà tʃʼǎ tʃʼǎ tʃʼǎ tʃʼǎ beeswax_1 *gāgáʔ gāgá gāgá - - - gāgá gāgá gāgá gāgá gāgá bird_pelican *nàbɔ̀ng(w) nàbɔ̀ŋg nàbɔ̀ŋg nàbɔ̀ŋg nàbɔ̀ŋg - ʊ̄+nàbɔ̀ŋà nàbɔ̀ŋà - - nàbɔ̀ŋgɔ̀ (brown) à ɔ̀ ɔ̀ ɔ̀ ɔ̀ make go *tʰaʃ ~ t ̪h aʃ tāʃ tāʃ tàʃ tʰáʃ tʰáʃ - - - - - away thorn, sharp *k⁽ʰ⁾a(ŋ)k⁽ʼ⁾a kāŋà kāŋà kákā kāʔ kǎkā kʰákʰāʔ kʰákʼā kʰákʼā kʰákʼā kʰákʼā (be) salt_(made from ash of a *tʼaʃ tʼāʃ tʼāʃ tʼàʃ tʼāʃ tʼāʃ tʼàʃ tʼās - - - particular plant/tree)_2 be, do *tà - ta tà tā tā - tā - tā - drum_C *tà̪rá - tàrá tàrá - - tà̪rá tàrá tàrá tàrá tàrá 620 Meaning PKmn GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi à+wàsá hail, ice_1 *wasakʼ wàsà wàsà wàʃākʼ wàsáʔ - - - - - ʔ touch_1, *pàD pāt pāt pàt pātʰ pār - - - - - crawl kick *t ̪h áɓ tápʼ tápʼ táb tʰáb tʰáɓ t ̪h áp tʰáp tʰáp tʰáp tʰáp birth pangs *ʃāʃ(ā) ʃɛ́ ʃɛ́ ʃāʃ - - ʃāʃā sāsā sāsā sāsā ʃāʃā (pain) back_1 *kʼ(w)ás kʼwás kʼwás kʼǎw - - - - - - - break (v.)_1 *wa wǎ wǎ wà wá wá - - - - - boar (wild)_1 *wàɓ wàpʼ wàpʼ wàpʼ wàpʰ à+wàɓ - - - - - fish (n.) *wàcʼà wàsʼ wàsʼ wàsʼ wàʃʼ wàcʼ wàcʼà wàtʃʼà wàtʃʼà wàtʃʼà wàtʃʼà general term chicken_1 *waŋa wāŋā wāŋā wàgá ŋwá à+ŋwá - - - - - bee, honey *d̪àm tàm tàm dàm dàm à+d̪àm tâ̪m tàm tàm tàm tàm 621 APPENDIX F COMPLETE DATASET 12 APPENDIX F Full dataset This appendix contains the complete dataset used in this reconstruction. It is organized alphabetically by meaning. The first column (NODE) indicates the highest node to which a particular etymon can be reconstructed. The *proto column contains all reconstructed forms. If it is the case that a meaning has only one entry, then I do not reconstruct a word for this meaning and the proto column is left empty. The abbreviations for the language varieties are as follows: GwLo= Lowland Gwama, GwHi= Highland Gwama, UdYa= Yabus Uduk, Komo=Ethiopian Komo, UdCh= Chali Uduk, Dana=Dana, OpBi= Bilugu Opo, OpMo= Modin Opo, OpPa= Pame Opo, OpKi= Kigile Opo. I employ the grapheme <+> to indicate a historical or synchronic morpheme boundary. NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi abstain PDaOp *t ̪̓ ākʼán - - - - - t ̪̓ ākʼán tʼākʼá tʼākʼá tʼākʼá - from_1 abstain PGw *tā tā tā - - - - - - - - from_2 PKmn acacia *tú̪k⁽ʰ⁾(u) tʊ̌kʊ̀ - túk túkʰ à+tú̪kʰ tú̪kʰ tűkʰ tűkʰ tűkʰ tűkʰ PCtrl adhere *ɗab(a) - - ɗàb ɗāpʰ ɗāpʰ ɗábá ɗábá ɗábá ɗábá ɗábá 622 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi POp all_1 *ā+ɓīgɪń - - - - - - ā+ɓīgɪń ā+ɓīɡɪń ā+ɓīgɪń ā+ɓʊ̄gà Komo all_2 - - ɪ+́ʃîn - - - - - - - PGw all_3 - mʊkɪn - - - - - - - - Dana all_4 - - - - - gà+pʰān - - - - PUd all_5, finish *ɓár - - - ɓár ɓár - - - - - alone, abstain PKmn *t ̪̓ ɛn - sʼɪn̄ sʼɛ́n tʼɛ́n tʼɛ́ gà+t ̪̓ ɛ́n ā+tʼɛ́n ā+tʼɛ́n ā+tʼɛ́n ā+tʼɛ́n from, not want to do annoint PGw *sʼá sʼá sʼá - - - - - - - - (with oil)_1 annoint Komo - - mʊ̀ - - - - - - - (with oil)_2 annoint kʼɔ́ʃ+jì PUd - - - - - - - - - (with oil)_3 n annoint PUd - - - - cūr - - - - - (with oil)_4 annoint PDaOp *làlɪ ́ - - - - - làlɪ ́ làlɪ ́ làlɪ ́ làlɪ ́ làlɪ ́ (with oil)_5 623 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi antelope (dikdik, PGw *nɪ̀ ʊ̄+nɪ̀ nɪ̀ - - - - - - - - small deer)_2 antelope (dikdik, PCtrl *zʊj - - zwɪ̀ wí à+ʔúj zʊ̀j à+ʔʊ́jʊ́ à+ʔʊ́ɪ ́ à+ʔʊ́ɪ ́ - small deer)_2 PKmn antelope_1 *ʃɛt̄(ʰ) ʊ̄ʃɛt̄ ʃɛt̄ ʃɛt̄ ʃɛt̄ʰ à+ʃɛt̄ʰ - - - - - POp antelope_2 *lɔ̀j - - - - - - lɔ̀j lɔ̀j - - Dana antelope_3 - - - - - bùjɛ̀ - - - - POp antelope_4 *pàndʒá - - - - - - - - pàndʒá pàndʒá antelope_g PCtrl *kiʃ ~ *kɪʃ - - kɪʃ̄ cíʃ à+cíʃ àkīʃ kʰīs kʰīs kʰīs kīʃ azelle antelope_w PKoUd *kɔ̄pʰ - - kɔ̄p kɔ̄pʰ kɔ̄pʰ - - - - - aterbuck PGwKo arm_1 *kwapa kwǎpá kwǎpá kɔ̀p - - - - - - - PUd arm_2 *bwí - - - bwí à+bǐ - - - - - arm_3, Dana - - - - - kàlɔ̀ŋ - - - - shoulder 624 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PCtrl arrive_1 *p(ʰ)uɗ(i) - - - - pʰūɗ pújī - - - - PKoUd arrive_2 *kʰɛ́ - - kɛ́ cʰɛ́ - - - - - - POp arrive_3 *ɓɛr̄ - - - - - - ɓɛr̄ ɓɛr̄ ɓɛr̄ ɓɛr̄ PGw arrive_4 *pɔ̌ʃɔ̀ pɔ̌ʃɔ̀ pɔ̌ʃɔ̀ - - - - - - - - POp arrow_1 *àŋɛ̀ràŋ - - - - - - àŋɛ̀ràŋ - àŋɛ̀ràŋ - *pɪńzɪá́kʼɪʃ̄ɪ̀ pɪńzɪá́kʼɪʃ̄ɪ̀ pɪńzɪá́kʼɪʃ̄ɪ̀ PGw arrow_2 - - - - - - - - n n n PCtrl arrow_3 *T̪A(t ̪h ) - - sá sá à+tá̪ sɛt̄ ̪h - tʃɛ ̄ - tʃɛ ̄ *(tʼi)pʼIkʼIɲ( PKmn ash_1 pʼɪk̄ʼɪń pʼɪk̄ʼɪń pʼīn - tʼīpʼīɲ pʼɪn̄ā - - pʼɪn̄ā pʼɪn̄ā a) POp ash_2 *pītí - - - - - - pītí pītí - - PUd ash_3 - - - bútʰ - - - - - - ask PKmn *dOt tʊ̌t tʊ̌t dɔ̀t dɔ̌tʰ dɔ̌tʰ - tɔ̄tɔ́ tɔ̄tɔ́ - - (inquire)_1 ask PDaOp *nàm - - - - - nàm - - nàm nàm (inquire)_2 PCtrl avenge *ɔ́g - - ɔ́g - - ɔ́g ɔ́g ɔ́g ɔ́g ɔ́g PUd axe_1 - - - cʰīsʼ - - - - - - PCtrl axe_2 *jɪd̪̄ - - jɪʔ̄ - - jɪd̪̄ - - - - PUd axe_3 - - - - cʰút ̪h - - - - - 625 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PGw axe_4 *pànsʼ pànsʼ pànsʼ - - - - - - - - POp axe_5 *wɔ̀j - - - - - - wɔ̀j wɔ̀j wɔ̀j wɔ̀j baboon_1, PKmn *dɪŋɪ tɪ̀nɪ̀ tɪ̀nɪ̀ - - - tɛ̀ŋ tɪ̀nɪ̀ tɪ̀nɪ̀ tɪ̀n tɪ̀n dog_2 PDaOp baboon_2 *ɟarɛ - - - - - ɟārɛ ̄ tʃàrɛ̀ tʃàrɛ̀ tʃàrɛ̀ tʃàrì PKoUd baboon_3 *dàwàʔ - - dàw - à+dàwà - - - - - PUd baboon_4 - - - gàwàʃ - - - - - - PGw baboon_5 *ɛľɛ́ ɛľɛ́ ɛľɛ́ - - - - - - - - PKmn back_1 *kʼ(w)ás kʼwás kʼwás kʼǎw - - - - - - - PKoUd back_2 *pʰɔ́g - - pɔ́g pʰɔ́ʔ à+pʰɔ́ʔ - - - - - Dana back_3 - - - - - dàrsɔ̂g - - - - POp back_4 *sɪt́ʼɪn̄ - - - - - - sɪt́ʼɪn̄ sɪt́ʼɪn̄ sɪt́ʼɪn̄ sɪt́ʼɪn̄ PKoUd bad (be)_1 *ʃɪǵ - - ʃɪǵ ʃíʔ ʃíʔ - - - - - PGw bad (be)_2 *pɛt̄í pɛt̄í pɛt̄í - - - - - - - - POp bad (be)_3 *kʰɔ́ - - - - - - kʰɔ́ kʰɔ́ kʰɔ́ kʰɔ́ Dana bad (be)_4 - - - - - sít ̪h ā - - - - bale out PCtrl *kʰɔɓ - - kɔ́p kʰɔ̄b kʰɔ̄ɓ kʰɔ́pʰ kʰɔ́pʰá kʰɔ́pʰá kʰɔ́pʰá kʰɔ́pʰá (water) PGw bamboo_1 *tǎŋá tǎŋá tǎŋá - - - - - - - - PCtrl bamboo_2 *sʼEU - - sʼɪ̂w sʼì à+t ̪̓ ɛ ̄ sʼɪ̂w - - - - 626 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi POp bamboo_3 *kūp - - - - - - kūp kūp kūp kūp PDaOp barren (be) *p⁽ʰ⁾ɔ̀s - - - - - pɔ̀s pʰɔ̄s pʰɔ̄s pʰɔ̄s pʰɔ̄s PKmn basket *RUKa lúkà lúkà lùg rǔgà - rɔ̀kʰà rùwà rɔ̀kà rɔ̀kà rɔ̀kà PKmn bathe *úpʰ ʊ́p ʊ́p úp úpʰ úpʰ úpʰ űpʰ űpʰ űpʰ űpʰ PKmn be, do *tà - ta tà tā tā - tā - tā - PKoUd bean_1 *ɗagi - - ɗɛ̀gí ɗàgì - - - - - - PGw bean_2 *kʼwǎʃà kʼwāʃà kʼwǎʃà - - - - - - - - PDaOp bean_3 *gwàlɪ ́ - - - - - gwàlí gwàlɪ ́ gwàlɪ ́ gwàlɪ ́ gwàlɪ ́ PUd bean_4 *d̪ǔgà - - - zúgàʔ à+d̪ǔgà - - - - - become, PKmn become *wVɗ wɛť wɛť wàl wár wáɗ wâl - - - - angry àŋgàríp PKmn bed_1 *langarɛt ángàr - ángàr áŋgàr - - làŋgàrɛ́t àŋgàrípʰ àŋgàrípʰ ʰ PGw bed_2 - bámbàr - - - - - - - - PCtrl bed_3 *kija - - - cī à+cī kījā - - - - Pkmn bee, honey *d̪àm tàm tàm dàm dàm à+d̪àm tâ̪m tàm tàm tàm tàm beehive PKmn *gɛnd̪(V)(l) gɛ́ndɛ́l gɛ̀ndɪ ́ kɛ̀ndɛ ̄ - - gɛ̀nd̪á gɛ̀ndá gɛ̀ndá gɪ̀ndá gɛ̀ndá basket PKmn beer *sud̪(i) ʃʊ́l ʃwɪ ́ ʃùʔí sū à+sū sùd̪ sī swī swī swī 627 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi beer PGw *tàzà tàzà tàzà - - - - - - - - filter_1 beer PKoUd *d̪ìnd̪àl - - zɪ̀nzàl zìnzàl à+d̪ìnd̪àl - - - - - filter_2 beer PDaOp *tì̪mìs - - - - - tì̪mìs tīmīs tīmīs tīmīs tīmīʃ filter_3 PKmn beeswax_1 *gāgáʔ gāgá gāgá - - - gāgá gāgá gāgá gāgá gāgá Komo beeswax_2 - - ʃùtʼ - - - - - - - PUd beeswax_3 - - - jīgán - - - - - - PUd beeswax_4 - - - - ɗíŋgílá - - - - - PCtrl belch_1 *gàɗ(am) - - gàl gàrà gǎɗ gàràm kʼɛr̄ɛm̄ gàràm gàràm gàràm PGw belch_2 *dɛm̄ɛ́sʼ dɛm̄ɛ́sʼ dɛm̄ɛ́sʼ - - - - - - - - belly or PKmn *bùʃùl bùʃùl bùʃì bùʃ - bùʃ - pùsà pùsà pùsà pùsà stomach_1 belly or PCtrl *bùmà - - - bwàʔ - mwà pùmà pùmà pùmà pùmà stomach_2 belly or PGw *tát tát tát - - - - - - - - stomach_3 PKmn belt, sash *gaʃa gáʃà gáʃà - gǎʃ gāʃá gâʃ gātʃ gātʃ gātʃ gâʃ PKoUd big (be)_1 *c(w)ā - - swà cā cā - - - - - 628 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PGw big (be)_2 *hàntʼà hàntʼà hàntʼà - - - - - - - - big (be), PCtrl male, *tɔn - - tɔ̀n - - tɔ́n tɔ́n tɔ́n tɔ́n tɔ́n elder_1 big (be), PUd male, *d̪àn - - - zàn d̪àn - - - - - elder_2 PCtrl bird *ɗitʰ ? - - ɗɪ̂w ɗūwì ɗī ɗītʰ ɗɪʊ̀̄ ɗɪʊ̀̄ ɗɪʊ̀̄ ɗɪʊ̀̄ bird (yellow- PKmn billed kite *gʊlɪla - gʊ́lːā bāgʊ́lɪ̀lā - - ɓāgʊ̂lːā - ɓāgʊ́lʊ́l ɓāgʊ́lʊ́l ɓāgʊ́là or black kite) bird_(cattle PKmn *kEɲ kɪl̄ kɪl̄ à+kɪĺ cɛ́cɛ́nā à+cɛ́ɲ à+kɪĺ à+kɪĺ à+kɪĺ à+kɪĺ à+kɪĺ egret) bird_corma nt *mililu ~ PCtrl - - à+mīlílū - - mīlílū mīlílì mīlílì mīlílì mīlílì (whistling *milili duck) 629 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi bird_dove à+tìrıt̋ır̋ à+tìrıt̋ır̋ PKmn (African *tiritiri - títìtī - - - - à+tìrıt̋ır̋ì à+tìrıt̋ır̋ì ì ì mourning) bird_hamm PKmn *nɛ́kʰ - ʊ̄+nɛ́k à+nɛ́k nɛ́kʰ à+nɛ́kʰ à+nɪḱʰ - - à+nɪḱʰ à+nɪḱʰ erkopf_1 bird_hamm POp *ànùrɛ̀ - - - - - - ànùrɛ̀ ànùrɛ̀ - - erkopf_2 PKmn bird_heron *bár - ʊ̄+bár à+bár - - à+bár à+bár à+bár à+bár à+bár bird_ostric PKmn *wutʰ - wǔt wǔt útʰ à+útʰ - hűtʰ hűtʰ hűtʰ hűtʰ h_1 bird_ostric Dana - - - - - rɔ̀t ̪h áj - - - - h_2 bird_pelica *nàbɔ̀ng(w) nàbɔ̀ŋg PKmn - ʊ̄+nàbɔ̀ŋà nàbɔ̀ŋà - - nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ nàbɔ̀ŋgɔ̀ n (brown) à ɔ̀ bird_Quele PCtrl a (Red- *D̪ìbàl - - dìbàl - - - à+dʒìbà à+dʒìbà à+dʒìbà à+dʒìbà billed) bird_stork à+gɔ̀mpí à+kʊ̀mp à+kʊ̀m PCtrl *gOmpʰVja - - - - - - - (abdim) já ʰūjá pʰūjá 630 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi bird_stork PKmn *ɟàrú zɛ̀rú zɛ̀rú zɛ̀rú ʒàrú - ɟɛ̀rú dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű dʒɛ̀rű (maribou) bird_vultur PKmn e (white- *lʊm ʊ̄lʊ̌m - lʊ̌m lùʔ - lʊ̄m lʊ̌m lʊ̌m lʊ̌m lʊ̌m backed)_1 bird_vultur PGw e (white- - kʼʊ̌pʊ́ - - - - - - - - backed)_2 bird_vultur PUd e (white- - - - - kɔ́rɔ́ - - - - - backed)_2 bird_weave à+dwa̋k à+dwa̋k PKmn *dwakʰ ʊ̄+dwâk ʊ̄+dɔ́k à+dɔ̂k - - à+dwákʰ à+dwa̋kʰ à+dwa̋kʰ r ʰ ʰ birth pangs PKmn *ʃāʃ(ā) ʃɛ́ ʃɛ́ ʃāʃ - - ʃāʃā sāsā sāsā sāsā ʃāʃā (pain) bite (by Komo - - ɗʊ́r - - - - - - - animal)_1 bite (by PGw *kʼúnsʼ kʼúnsʼ kʼúnsʼ - - - - - - - - animal)_2 631 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi bite (by PUd - - - - ɟɛp̌ʰ - - - - - animal)_3 bite (by PDaOp *dɔ̀ŋ - - - - - dɔ̀ŋ dɔ̀ŋ dɔ̀ŋ dɔ̀ŋ dɔ̀ŋ animal)_4 PCtrl bite_1 *hɔcʼ - - wɔ́sʼ wɔ̌ʃʼ wɔ̌cʼ hɔ́cʼ hɔ̄tʃʼɔ̄ hɔ̄tʃʼɔ̄ hɔ̄tʃʼ hɔ̄tʃʼ PGw bite_2 *sʼunsʼ sūnsʼ sʼṹnsʼ - - - - - - - - bitter, sour PKmn *kʰakʼa kāgā kāgā kàʔ kʰāʔ kʰāʔ kʰàkʼà kʰākʼā kʰākʼā kʰākʼā kʰākʼā (be) PCtrl black (be)_1 *sʼíd̪ - - sʼí sʼíʔ t ̪̓íʔ sʼíd̪ tʃʼı ̋ tʃʼı ̋ tʃʼı ̋ tʃʼı ̋ PGw black (be)_2 *tʼɔ́tʼɔ́ tʼɔ́tʼɔ́ tʼɔ́tʼɔ́ - - - - - - - - blame PCtrl (somebody) *nɔg - - nɔ̀g - - nɔ̂g nɔ̀gá nɔ̀gá nɔ̀gá nɔ̀gá _1 blame PGw (somebody) *tʊ̀nsʊ̀s tʊ̀nzʊ̀ tʊ̀nsʊ̀s - - - - - - - - _2 PKmn blood_1 *sʼámá sʼám sʼám - - - sʼámáʔ tʃʼámá tʃʼámá tʃʼámá tʃʼámá PKoUd blood_2 *bàs - - bàʃ bàs à+bàs - - - - - blow (with PKmn *pʰui(ki) pɪk̄ɪ ̄ pɪ ̄ pɪ̀ pʰûj pʰí pʰùj pʰű pʰű pʰű pʰű mouth) 632 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PKmn blow nose *ʃɪntʼ ~ ʃintʼ ʃɪn̄tʼ ʃɪn̄tʼ ʃīn+ʃʊ̀nʃ ʃīn ʃīn ʃìnà+ʃʊ̀ʃ sın̋á sın̋á sın̋á sın̋á boar PKmn *wàɓ wàpʼ wàpʼ wàpʼ wàpʰ à+wàɓ - - - - - (wild)_1 boar POp - - - - - - tʰǎrkʼá - - - (wild)_2 PKmn body_1 *(j)Es - jɪs̄ ɪʃ̄ īs īs ɛ̀s ɛs̄ ɛs̄ ɛs̄ ɛs̄ PGw body_2 wúsīn - - - - - - - - - boil (of PKmn *was(ikʼ) wǎʃ wàʃí wàʃíkʼ - - wás wās wās wās wās liquid)_1 boil (of PUd - - - gàbùʃà - - - - - - liquid)_2 PKmn bone *ʃUImakʼ sí sí ʃúmákʼ sīmāʔ à+sīmāʔ ʃʊ́j sʊ́j sʊ́j sʊ́j sʊ́j PCtrl brain_1 *lɔ̀lɔ̀kʼ - - lɔ̀lɔ̀kʼ - - lɔ̀lɔ̀ʔ lɔ̀lɔ̀kʼ lɔ̀lɔ̀kʼ lɔ̀lɔ̀kʼ lɔ̀lɔ̀kʼ PUd brain_2 *tʰúlá - - - tʰúlá à+tʰúláʔ - - - - - PGw brain_3 *ʃʊ̄nʊ̄ ʃʊ̄nʊ̄ ʃʊ̄nʊ̄ - - - - - - - - break (destroy or PKoUd *kʼɛɗ - - kʼɛ̀r cʼɛ́d cʼɛ́ɗ - - - - - get destroyed) PKmn break (v.)_1 *wa wǎ wǎ wà wá wá - - - - - 633 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi POp break (v.)_2 - - - - - - pɛ̀má - - - Dana break (v.)_3 - - - - - táj - - - - breast, PCtrl *kɔ́d̪ - - kɔ́ kɔ́ à+kɔ́ kɔ́d̪ kɔ́j kɔ́j kɔ́j kɔ́j milk_1 breast, PGw *sʼʊ̄pʼ sʼʊ̄pʼ sʼʊ̄pʼ - - - - - - - - milk_2 PKmn breathe *ʃukʼin ʃɪňʃɪ ́ ʃɪn̄ʃɪ̀ ʃùʔɛn̄ ʃīʔīn ʃīʔīn ʃìkʼ sīkʼ sīkʼ sīkʼ sīkʼ PCtrl bring_1 *kʰal - - kà-ʊ̄ kʰál+í kʰál+ú kùjí - - - - POp bring_2 *pʰā - - - - - - pʰā+ʊ́ pʰā+jʊ́ pʰā+jʊ́ pʰā+jʊ́ PGw bring_3 *kāp kāp kāp - - - - - - - - Dana bring_4 - - - - - ɗɔ́jɛ ̄ - - - - Komo brood (v.)_1 - - ʃùg - - - - - - - PGw brood (v.)_2 *kʊ̀ kʊ̀ kʊ̀ - - - - - - - - PUd brood (v.)_3 *wúpʰ - - - wúpʰ wùpʰ - - - - - PDaOp brood (v.)_4 *ʃúmā - - - - - ʃúmā sűmá sűmá sűmá sűmá PCtrl broom_1 *gwàrásʼ - - gwàrásʼ - - gwàrásʼ gwàrátʃʼ gwàrátʃʼ gwàrátʃʼ gwàrátʃʼ PCtrl broom_2 *gɛɗɪʃ - - - ɟɛ̀dìʃ ɟɛ̀ɗɛ̀ʃ kɛ̀ɗɛ̀ʃ kʰɛ̀rɪ̀s kʰɛ̀rɪ̀s kʰɛ̀rɪ̀s kʰɛ̀rɪ̀ʃ PGw broom_3 *pɪʃ̄ pɪʃ̄ pɪʃ̄ - - - - - - - - *ʊ=kam PKmn brother kwám kwám kàm kām à+kām āmʊ́ hàm hàm hàm hàm *kamʊ 634 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi GwKo buffalo_1 *gwàsʼ kwàsʼ kwàsʼ gwàsʼ - - - - - - - POp buffalo_2 *ɓàtʃʼ - - - - - - ɓàtʃʼ ɓàtʃʼ ɓàtʃʼ ɓàtʃʼ ʃūmā+ ʃùmà+sʼí PCtrl buffalo_3 *ʃùmà+sʼíd̪ - - - - - - - - sʼíʔ d̪ burn (hairs off of pig PCtrl *p(ʰ)ʊ̄r - - - pūr pūr - pʰʊ̄r pʰʊ̄r pʰʊ̄r - skin), roast next to fire PCtrl burn_1 *kʰɪśʼ - - kɪśʼ cʰísʼ cʰít ̪̓ kʰísʼā kʰɪt̋ʃʼā kʰɪt́ʃʼā kʰɪt́ʃʼā kʰɪt́ʃʼā PGw burn_2 - tʼántʼ - - - - - - - - burnt bwà+p PKmn *(ʃa)pʰuZa ʃāpʊ́t ʃābʊ́t kɪ+́pú à+pʰíʔ pùzàʔ pʰűjhá pʰűjhá pʰűjhá pʰűjhá gound, soot ʰwí bury *nʊpʼ ~ PKmn ʊ̄mpʼ ʊ̄mpʼ - - - nʊ̀pʼá - - - - (sideways) *nʊmpʼ PCtrl bury_1 *bàb - - bàb - - - - - pàbà pàbà PCtrl bury_2 *kàn - - kàn kān kān kànà - - - - POp bury_4 *tʃìm - - - - - - tʃìm tʃìm - - PKmn buttocks_1 *ʊʃ īʃ ʊ̄ʃ - - - ʊ̄ʃ ʊ̄s ʊ̄s ʊ̄s ʊ̄ʃ PKoUd buttocks_2 *pʼɛ́n - - pʼɛ́n pʼɛ́n pʼɛ́n - - - - - PGw buy_1 *dwì dwì dwì - - - - - - - - 635 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PUd buy_2 *jɔ́l - - - jɔ́l jɔ́l - - - - - PCtrl buy_3, sell *ʃʊ - - ʃʊ̀ - - ʃʊ̀ sʊ̄ sʊ̄ sʊ̄ sʊ̄ PCtrl bypass_1 *kál - - kāl kál kál kál kál kál kál kál PGw bypass_2 *bɛ́ʃɛ̀ bɛ́ʃɛ̀ bɛ́ʃɛ̀ - - - - - - - - PKoUd calf of leg *sáɗ(á) - - ʃárá sád à+sáɗ - - - - - PGw calf of leg *sʼʊ́sʼʊ́mɔ̀ sʼʊ́sʼʊ́mɔ̀ sʼʊ́sʼʊ́mɔ̀ - - - - - - - - tʼākʼʊ̋m tʼākʼʊ̋m PDaOp calf of leg *t ̪̓ ākʼʊ́mʊ́ - - - - - t ̪̓ ākʼʊ́mʊ́ tʼākʼʊ̋mʊ̋ tʼākʼʊ̋mʊ̋ ʊ̋ ʊ̋ Dana call_1 - - - - - àr - - - - PCtrl call_2 *juga - - - júkʰ júkʰ - dʒùgà dʒùgà zùgà ʃùgà Komo call_3 - - ʊ́l - - - - - - - PGw call_4 *twī twī twī - - - - - - - - PKmn cannabis_1 *bángwà báŋgà báŋgà bánkɔ̀ bǎŋgɔ̀ bǎŋgɔ̀ - - - - - PDaOp cannabis_2 *kʰáŋgá - - - - - kʰáŋgá kʰáŋgá kʰáŋgá kʰáŋgá kʰáŋgá carry many PKmn *(tɔ)ɗɔ tɔ̀dɔ̀ - ɗɔ̀ ɗɔ̄ ɗɔ̄ - ɗɔ̄ ɗɔ̀ ɗɔ̀ ɗɔ̀ things_1 carry many PGw - pʊ̀l - - - - - - - - things_2 carry on PKmn *màm(a) màm màm màmá màm màm màmá màmā màmā màmā màmā back_1 636 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi carry on POp *kʊ́má - - - - - - kʊ́má kʊ́má kʊ́má kʊ́má back_2 carry on PKmn *kOpʰ kǔ kǔ - - - kɔ̄pʰ kɔ̄p kɔ̄p kɔ̄p kɔ̄p head_1 carry on PUd *kʰál - - - kʰál kʰál - - - - - head_2 carry on Komo - - kákà - - - - - - - head_3 PCtrl carry_1 *kʰal - - kàj kʰál kʰál kʰál kʰál kʰál kʰál kʰál PKoUd carry_2 *tɛr - - tɛ́r tɛr̄ - - - - - - POp carry_3 *pāgā - - - - - - pāgā pāgā pāgā pāgā PDaOp carry_4 *ʃàkʼà - - - - - ʃàkʼà sākʼā sākʼā sākʼā sākʼā PGw carry_5 - sūndí - - - - - - - - Dana carry_6 - - - - - kwà - - - - PUd carry_7 - - - - hāɓ - - - - - PKoUd cat_1 *bʊ̌rà - - bʊ̌rà bǔrá à+bǔrá - - - - - PUd cat_2 *ɲáw - - - ɲáw à+ɲáw - - - - - PKoUd celebrate *gàs - - gàs - gàs - - - - - PCtrl chair_1 *kʰOba - - kʊ́bá - - - kʰɔ́bá kʰɔ́bá kʰɔ́bá kʰɔ́bá PGw chair_2 *kɔ́ŋɔ̀ kɔ́ŋɔ̀ kɔ́ŋɔ̀ - - - - - - - - 637 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Dana chair_3 - - - - - kànáj - - - - bǎmbà PUd chair_4 *bǎmbàr - - - bǎmbàr - - - - - r charcoal or PKmn *kʼI(m)Isʼ sʼɪśʼɪń sʼɪśʼɪń kìsʼísʼìʔ cʼɛl̄ɛs̄ʼ cʼīlāt ̪̓ kʼɪ̀sʼ kʼīmītʃʼ kʼīmītʃʼ kʼɪt̄ʃʼ kʼɪt̄ʃʼ coal PCtrl chase *mɛt(̪i) - - mɛ̀tí - - mɛ́t ̪h mɛ́tī mɛ́tī mɛ́tī mɛ́tī Komo cheek_1 - - kʼɪ̀ʃ - - - - - - - PGw cheek_2 *kʼɔ́kʼɔ́l kʼɔ́kʼɔ́l kʼɔ́kí - - - - - - - - PUd cheek_3 - - - tʼīʃ - - - - - - *tiŋ(a) ~ PDaOp cheek_4 - - - - - tīŋ tǐná tǐná tǐná tǐná tin(a) PCtrl chest_1 *cʼacʼ - - sʼàsʼ - - cʼàcʼ tʃʼātʃʼ tʃʼātʃʼ tʃʼātʃʼ sʼāsʼ PUd chest_2 *bɔ̀r - - - bɔ̀r à+bɔ̀r - - - - - *tɔ̀ŋàs ~ PGw chest_3 tɔ̀ŋàs twàŋgàs - - - - - - - - twàŋgàs PKmn chicken_1 *waŋa wāŋā wāŋā wàgá ŋwá à+ŋwá - - - - - PGw chicken_2 *dɔ̀ŋgɔ̀l dɔ̀ŋgɔ̀l dɔ̀ŋɪ ́ - - - - - - - - PDaOp chicken_3 *kʰā - - - - - kʰā kʰā kʰā kʰā kʰā PKmn chief_1 *kOr ʊ̄+kʊ̄l ʊ̄+kwɪ̀ jī+kwɪ̀ - - jɛ̀+kɔ̄rɔ̄ - - - - PUd chief_2 *tàpʰà - - - tàpʰà tàpʰà - - - - - 638 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PGw child_1 *wār wār wāl - - - - - - - - PUd child_2 *à+cʼí - - - àɟí à+cʼí - - - - - PCtrl child_3 *ad̪imɛ - - āt - - àd̪ín àdım̋ɛ́ àdím àdím àdím PCtrl children_1 *ʃ(w)at ̪̓En - - ʃwàtʼɪń - - ʊ̀+ʃàt ̪̓ ɛ́n bì+tʼí - - - PUd children_2 *ū+cʰí - - - ū+ʃíʔ ū+cʰí - - - - - PGw children_3 *mǎn mǎn mǎn - - - - - - - - chili PKmn *zjaɗa(j) zɛ́t zɛ́t zjàntʼá - - à+zāt ̪h ɛʔ̄ dʒɛ̀rà - - - pepper_1 chili PUd *à+cʰɛ́d̪à - - - cʰɛ́zà à+cʰɛ́d̪à - - - - - pepper_2 choke, PKmn *busʼ bǔsʼ bǔsʼ - bùsʼ bùt ̪̓ - pʰǔtʃʼ - - - strangle_1 choke, PCtrl *t ̪̓Osʼ - - tʼúsʼ - tʼɔ́cʰ t ̪̓ ɔ́sʼ - - - - strangle_2 choke, PDaOp *dɛŋ - - - - - dɛ́ŋ - dɛŋ̄ dɛŋ̄ dɛŋ̄ strangle_3 GwKo choose_1 *lɔ̀s lɔ̀s lɔ̀s lɔ̀s - - - - - - - PUd choose_2 - - - gú - - - - - - Dana choose_3 - - - - - mɛ̂kʰ - - - - PUd choose_4 - - - - ɲìt ̪h +īs - - - - - 639 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi choose_5, POp *pātʃá - - - - - - pātʃá pātʃá pātʃá pātʃá shut PCtrl chop_1 *pʰákʼá - - - pʰáʔ - pʰákʼá - - - - PKoUd chop_2 *kʼɪ(mɪ)sʼ - - kʼɪśʼ cʼīmīsʼ cʼīmīt ̪̓ - - - - - PDaOp chop_3 *kʼá(n)cʼír - - - - - kʼácʼír kʼántʃʼír kʼántʃʼír kʼántʃʼír kʼántʃʼír PCtrl clan_1 *mɔ̀s - - mɔ̀s - - mɔ̀s mɔ̀s mɔ̀s mɔ̀s mɔ̀s PGw clan_2 tʼʊ̀tʼʊ̀mʊ̀ - - - - - - - - - PGw clan_3 - zɛr̄í - - - - - - - - PUd clan_4 *wàkʰ - - - wàkʰ wàkʰ - - - - - claves PKmn (instrument *lɛ̀pʰɛ́ lɛ̀pɛ́ àlàpɛ́ lɛ̀pɛ́ lɛ̀pʰɛ́ - lɛ̀pʰɛ́ àlɛ̀pʰɛ́ àlɛ̀pʰɛ́ àlɛ̀pʰɛ́ àlɛ̀pʰɛ́ )_1 claves PUd (instrument - - - - gɔ̌l - - - - - )_2 clear land PKoUd (for *kʼuT̪ - - kʼúʃ kʼús kʼút ̪h - - - - - planting)_1 640 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi clear land PGw (for *bʊ̀tʊ̀ bʊ̀tʊ̀ bʊ̀tʊ̀ - - - - - - - - planting)_2 clear land PDaOp (for *kʼɛ̀r(ɛ́) - - - - - kʼɛ̀rɛ́ kʼɛr̄ kʼɛr̄ kʼɛr̄ kʼɛr̄ planting)_3 PKmn climb_1 *sɛl sál sɛl̄ - sɛ ̄ sɛ ̄ - - - - - PCtrl climb_2 *gɔ̀r - - gɔ̀l - - - kɔ̀r kɔ̀r kɔ̀r kɔ̀r Dana climb_3 - - - - - kʰàc - - - - cloth, PCtrl *bʊraɲ - - bʊ̄lɛ̀n bǔrɛ̀ɲ bùrìɲɛ̀ à+bʊ́râŋ - - bʊ̀lɛ̀ŋ bʊ̀lɛ̀n clothes_1 cloth, PGw *ɔ̀lɔ̀ ɔ̀lɔ̀ ɔ̀jɔ̀ - - - - - - - - clothes_2 cloth, POp *à+bí - - - - - - à+bí à+bí - - clothes_3 cloud, PCtrl *ʃali - - ʃɛ̀lí ʃīlá ʃīláʔ àpàjʃàlíʔ - - - - fog_1 cloud, POp *dʒìbàj - - - - - - dʒìbàj dʒìbàj dʒìbàj dʒìbàj fog_2 641 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi cloud, PUd *ràkʰ - - - ràkʰ ràkʰ - - - - - fog_3 cloud, PGw *ʊkʊ ʊ̀kʊ̀ ʊ̌kʊ́ - - - - - - - - fog_4 cluck (of PKmn *kjankʼa kǎkā kjāŋkʼā kágá - - kágà kɛn̄ kɛn̄ kɛn̄ kɛn̄ hen) PKmn cold (be)_1 *sʼʊ́p sʼʊ́p sʼʊ́p - sʼúpʰ t ̪̓úpʰ - - - - - cold(be)_2, PCtrl wet, sweet *sʼám - - sʼám sʼám t ̪̓ám sʼɛ́m tʃʼɛ́m+sɛ ̄ tʃʼɛ́m+sɛ ̄ tʃʼɛ́m+sɛ ̄ tʃʼɛ́m+sɛ ̄ (be) come free PKmn *pʰal pɪ ̌ pɪ ̌ - - - - pʰál pʰál pʰál pʰál and fall off come, PKmn *ha hɔ̌ hɔ̌ hà + ʊ́ - - - - - - - come_SG copulate PCtrl *lud̪(a) - - lùd - - lùd̪á lūdá lūdá lūdá lūdá (animal)_2 copulate PKmn *ʃʊ́ɓ ʃʊ́pʼ ʃʊ́pʼ - ʃúb ʃúɓ - - - - - (animal)_3 PKmn corner *rOk⁽ʰ⁾Oɲ rɔ́kɔ̀n rɔ́kɔ̀n rɔ́kɔ̀n rúgùɲ à+rúkūɲ rɔ́kʰɔ̀n rʊ́kʊ̄n rʊ́kʊ̄n rʊ́kʊ̄n rʊ́kʊ̄n 642 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi cotton, thread, PKmn *tU̪s dʊ̄ʃ dʊ̄ʃ tūʃ tūs tū̪s - - - - - spider web_1 cotton, thread, PDaOp *latɔ̪n - - - - - lát ̪h ɔ̄n lǎtɔ́n lǎtɔ́n lǎtɔ́n lǎtɔ́n spider web_2 PKmn cough *kʼʊt ̪*kʼut ̪ kʼʊ́kʼʊ́t kʼʊ́kʼʊ́t kʼùt kʼútʰ kʼútʰ kʼút ̪h kʼútʼù kʼútʼù kʼútʼù kʼútʼù PCtrl count_1 *dɛ̀ŋ - - dɛ̀n dɛ̀ŋ+ɛ ̄ dɛ̀ŋ+ɛ ̄ - - - dɛ̀n dɛ̀n PGw count_2 - ǎt - - - - - - - - POp count_3 *kɔ́ndíl - - - - - - kɔ́ndíl kɔ́ndíl - - Dana count_4 - - - - - kʰɛ̀nɛ́ - - - - court (v.), PCtrl *pàl(Í) - - pàlí - - pàlɪ ́ pàlɪ ́ pàlɪ ́ pàlɪ ́ pàlɪ ́ flirt with_1 court (v.), PGw *sā sā sā - - - - - - - - flirt with_2 court (v.), PUd *dɔ̌ʃ - - - dɔ̌ʃ dɔ̌ʃ - - - - - flirt with_3 PKmn cover (v.) *kum(bi) kùmbì kùmbì kúm kūm kūm kúmā kűmá kűmá kűmá kűmá 643 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PCtrl cow_1 *bìb ~ *bìb - - bɪ̀b bìʔ bìpʰ pìʔ pìb pì pì pì PGw cow_2 *ɪḿɪ ́ ɪḿɪ ́ ɪḿɪ ́ - - - - - - - - PDaOp crawl_2 *lala - - - - - lálá lālá lālá lālá lālá PCtrl criticize_1 *wàg - - wàg wàkʰ wàkʰ wàg wàg wàg wàg wàg PGw criticize_2 *mǎjā mǎjā mǎjā - - - - - - - - PCtrl crocodile_1 *ɓɪs̄à - - ɓɪs̄à - - - ɓīsā ɓɪs̄ā ɓɪs̄ā - PGw crocodile_2 *sīzì sīzì sīzì - - - - - - - - GwKo cross legs_1 *gjá(j)má gjájá gjámá gɛ́m - - - - - - - *kāná+wɔ̀n kāná+w kāná+w kāná+w kāná+w POp cross legs_2 - - - - - - ɛ̀ ɔ̀nɛ̀ ɔ̀n ɔ̀n ɔ̀n crow PCtrl *kʼʊkʼʊl - - kʼʊ̀kʼʊ̀l - - kʼʊ́kʼʊ́l kʼʊ́kʼʊ́lʊ́ kʼʊ́kʼʊ́lʊ́ kʼʊ́kʼʊ́lʊ́ kʼʊ́kʼʊ́lʊ́ (verb)_1 crow GwKo *kʼɪnkʼɪlɪ kʼɪn̄kʼɪĺɪ ̄ kʼɪn̄kʼɪĺɪ ̄ kʼɪ̀ngɪĺ - - - - - - - (verb)_2 PKmn cry *kO(j) kū kū kɔ̀ kɔ̄ kɔ̄ kɔ̀j kwɛ ̄ kwɛ ̄ kwɛ ̄ kwɛ ̄ PDaOp curse_1 *cɛŋ - - - - - cɛ́ŋ tʃɛŋ̄ tʃɛŋ̄ tʃɛŋ̄ tʃɛŋ̄ PKoUd curse_2 *sʼɛ́ - - sʼɛ́ - tʼɛ́ - - - - - PGw curse_3 *kɛʃ̄ kɛʃ̄ kɛʃ̄ - - - - - - - - PUd curse_4 - - - zīlí - - - - - - 644 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi cut (meat into one PCtrl *ti̪til ~ titil - - tìtíl - - - - - tītíl tītíl long piece)_1 cut (meat into one PKmn *sɪĺ(ɪ) ʃɪĺɪ̀ ʃɪ ́ - - - - sɪ ́ - - - long piece)_2 cut (meat into one POp *làwà - - - - - - làwà làwà - - long piece)_3 cut (split in half *tʼwaŋ(k)a PKmn tʼāŋà tʼāŋà - tʼwákʰ tʼwákʰ - - - - - lengthwise) ~ t ̪̓waŋ(k)a _1 cut (split in half PCtrl *kwak - - kwàk - - kwɪ ́ kwāk kwāk kwāk kwāk lengthwise) _2 645 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi cut (split in half PGw - pɛ́ - - - - - - - - lengthwise) _3 *kʼísʼ ~ PCtrl cut_1 - - - cʼísʼ cʼít ̪h kʰísʼ kʰıt̋ʃʼ kʰıt̋ʃʼ kʰıt̋ʃʼ kʰıt̋ʃʼ *kʰísʼ PGw cut_2 *kʼʊ́t kʼʊ́t kʼʊ́t - - - - - - - - Dana d̪ɛ̀+kɛt̄ ̪h ā Dana (ethnonym) - - - - - - - - - +wád̪ _1 Dana PUd (ethnonym) *pùr - - - pùr pùr - - - - - _2 Dana PKmn (ethnonym) *d̪ana dànā dànā dāná - - d̪àná dànā dànā dànā dànā _3 PKoUd dance_1 *s(w)à - - ʃwà sā sā - - - - - PGw dance_2 *pʼá pʼá pʼá - - - - - - - - PDaOp dance_3 *ʃìjà - - - - - ʃìjà sījā sījā sījā ʃījā PCtrl daughter *ɓakʼum - - ɓā ɓwāʔ ɓāʔúm - ɓākʼ ɓākʼ ɓākʼ ɓākʼ 646 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi day (24 PCtrl *kɪm(V) - - kɪḿɪ ́ cím à+cím kɪ̀mà - - - - hours)_1 day (24 PGw *kājā kājā kājā - - - - - - - - hours)_2 Dazu (S. PKmn *dáɟV dázɔ̀ dázɔ̀ dázʊ̀ dǎɟɔ̀ dǎɟɔ̀ dáɟɔ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ dádʒʊ̀ Sudan) PKmn deep (be)_1 *gɔ̀ɗɔkʼ kwì kwǐ gʊ̀r - - kɔ̀ɗɔ̀kʼ kɔ̄rɔ́ kɔ̄rɔ́ kɔ̄rɔ́ kɔ̄rɔ́ PUd deep (be)_2 *kʼúl - - - kʼúl kʼúl - - - - - defecate_1, PKmn *sʼUDI úsʼ úsʼ sʼɪ ́ - t ̪̓í sʼʊ̀dʊ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ tʃʼʊ̄jɪ ́ diarrea Komo defecate_2 - - páp - - - - - - - PUd defecate_3 *kūcʼ - - - kūʃʼ kūcʼ - - - - - PGw defecate_4 *túʃ túʃ túʃ - - - - - - - - descend PKmn *T̪Ol(a) - ʃʊ̀ - sūl tū̪l - tʃɔ́já tʃɔ́já tʃɔ́já tʃɔ́já (land) PCtrl dew_1 *gjapaj - - - ɟápɛʔ̄ ɟápɛʔ̄ kʰìpʰɔ̂j gīpàj gīpàj gīpàj gīpàj Komo dew_2 - - zàpʊ́ - - - - - - - PGw dew_3 *kɪňtʼ kɪňtʼ kɪňtʼ - - - - - - - - PCtrl die_1 *hwOj - - wʊ́ wú wú - hwɔ̀ɪ ̄ hwɔ̀ɪ ̄ hɔ̌j hwɔ̀ɪ ̄ PGw die_2 *sʼí sʼí sʼí - - - - - - - - 647 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Dana die_3 - - - - - jɪd̪́ - - - - dig (for PKmn *kʼwàd̪(ɪ)́ kʼwī kʼwī kʼɔ̀ kʼwā - kʼwàd̪ɪ ́ kʼɔ̄j wārɪ ́ wādɪ ́ kʼwārɪ ́ water)_1 dig (for PUd - - - - gɔ̌pʰ - - - - - water)_2 PKmn dig_1 *ɟà ʃà ʃà - - - cà tʃà tʃà tʃà tʃà PUd dig_2 *kʰúɲ - - - kʰúɲ kʰǔɲ - - - - - Komo dig_3 - - mʊ̀tʼ - - - - - - - dip food in PKmn sauce with *sʼUB(V)(n) - sʼʊ́pʊ́n sʼūb sʼúpʰ t ̪̓úpʰ sʼúbá - - - - fingers_1 dip food in POp sauce with *sūtʼā - - - - - - sūtʼā sūtʼā sūtʼā sūtʼā fingers_2 disabled PKmn (be), angry *nap(a) nǎpá nǎpá nǎpá - nápʰɛ ̄ nápʰ nāp nāp nāp nāp (be) PKmn disregard *piʃ pīs pīs píʃ - píʃ píʃ pıs̋ pıs̋ pıs̋ pıs̋ PGw dive_1 *pʊ́gʊ́n pʊ́gʊ́n pʊ́gʊ́n - - - - - - - - PUd dive_2 *pɔ́ʃ - - - pɔ́ʃ pɔ́ʃ - - - - - 648 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PDaOp dive_3 *sír - - - - - sír sır̋ sır̋ sır̋ sır̋ Komo dive_4 - - kʼʊ̀lʊ̀m - - - - - - - PGw dog_1 *kānā kānā kānā - - - - - - - - *kʼáw ~ PCtrl dog_2 - - kʼáw kʼwáʔ à+kʼá - ʔwáj - - - *kʼwá PKmn doze off_1 *D̪i sɪśɪ ̄ sɪśɪ ̄ zìzí - - - dìsì dìsì zìz sìs Dana doze off_2 - - - - - hʊ́lʊ̀lá - - - - PUd doze off_3 - - - - ìʃ+kì+mís - - - - - PUd doze off_4 - - - kʰālā+ɛ ̄ - - - - - - PCtrl dream_1 *bɛs - - bɛš - - - bɛs̄ bɛs̄ bɛs̄ bɛs̄ Dana dream_2 - - - - - kʰɛ́t ̪h ɛ̀n - - - - PGw dream_3 *āmʊ́n āmʊ́n āmʊ́n - - - - - - - - PUd dream_4 *ɟàn - - - zàn ɟàn - - - - - PCtrl dregs *cʼVmaj - - sʼímá ʃʼɔ́má à+cʼúmá cʼɪḿáj tʃʼım̋áj tʃʼım̋áj tʃʼım̋áj tʃʼım̋áj dregs, PUd residue of - - - - t ̪̓ímát ̪̓ - - - - - fat PCtrl dress up_1 *hísʼ ~ *jísʼ - - jíz jísʼ hít ̪̓ hízá hıd̋ʒá hıd̋ʒá hıd̋ʒá hıd̋ʒá PGw dress up_2 *wɛ́ wɛ́ wɛ́ - - - - - - - - PGw drink *tʼɔ̄pʼ tʼɔ̄pʼ tʼɔ̄pʼ - - - - - - - - 649 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PCtrl drink_PL *ɪ̀pʰ(ɪ) - - ɪ̀p - - ɪ̀pʰɪ ́ ɪp̄ʰɪ ̄ ɪp̄ʰɪ ̄ ɪp̄ʰ ɪp̄ʰ PCtrl drink_SG *(ɪ)pʰɪ ́ - - ɪṕ pʰí pʰí ɪṕʰ pʰɪ ́ pʰɪ ́ ɪṕʰ ɪṕʰ drip (fall in PKmn *cʼɔ(tʼɔ)l sʼɔ́tʼɔ́ sʼɔ́tʼɔ́ sʼʊ̀lɪl̄ - cʼɔ̄cʼɔ̄lɔ̄cʼ cʼɔ̀ʔ tʃʼɔ̄ tʃʼɔ̄ tʃʼɔ̄ tʃʼɔ̄ globules)_1 drip (fall in PUd - - - ʃúkʼ - - - - - - globules)_2 PCtrl drum_A *bùl - - bùl - - bùl bùl bùl bùl bùl PKmn drum_B *bàmbá pàmbà pàmbà - bàmbá - - - - - - PKmn drum_C *tà̪rá - tàrá tàrá - - tà̪rá tàrá tàrá tàrá tàrá PKmn dry (be) *kʰʊ́sʼ kúsʼ kúsʼ kʊ́sʼ kʰúsʼ kʰút ̪̓ kʰʊ́sʼ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ kʰʊ́tʃʼʊ́ - dry out_1, roast next PCtrl *t ̪h ùd̪ - - tùd tʰūd tʰūr t ̪h ùd̪ - - - - to fire, aim at PGw dry out_2 *sàzà sàzà sàzà - - - - - - - - PCtrl dry out_3 *ɗAs - - ɗàʃ ɗās ɗās ɗɛ̀s - - - - dust, PKmn *burbuɗ bùrbūt bùrbūt - bǔtʰ bǔɗ būrkʼùs būrkʼùs būrkʼùs būrkʼùs būrkʼùs sand_1 dust, Komo - - pùlàní - - - - - - - sand_2 650 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi dwell (live, PGw *tùl tùl tùl - - - - - - - - reside)_1 dwell (live, PUd *dìd - - - dìd dì - - - - - reside)_2 dwell (live, Komo - - dʊ̀ - - - - - - - reside)_3 dwell (live, Dana *pʰɪ̀ - - - - - pʰɪ̀ - - - - reside)_4 dwell (live, POp *wà - - - - - - wà wà wà wà reside)_5 PKmn ear *cʼɛ sʼɛ ̄ sʼɛ ̄ sʼɛ ̄ ʃʼɛ́ cʼɛ́ kʼɛ ̄ tʃʼɛ̀ tʃʼɛ̀ tʃʼɛ̀ tʃʼɛ̀ earth, soil, PKmn ground, *k(j)asʼVN kʼjáŋásʼ kʼɛs̄ʼɛ́n kʼāsʼɪ̀ - - - - - - - floor_1 earth, soil, POp ground, *kʼɔ́tʃʼɔ́ - - - - - - kʼɔ́tʃʼɔ́ kʼɔ́tʃʼɔ́ kʼɔ́tʃʼɔ́ kʼɔ́tʃʼɔ́ floor_2 earth, soil, PUd ground, *cʼɛ́ʃ - - - cʼɛ́ʃ à+cʼɛ́ʃ - - - - - floor_3 651 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi earth, soil, Dana ground, - - - - - tɪ̪̀n - - - - floor_4 eat (hard PKmn *kʼama kʼā kʼā kʼá kʼáʔ kʼá kʼámá kʼámá kʼámá kʼámá kʼámá food) eat (soft PCtrl *uʃa - - ùʃ - - ùʃá ùsà ùsà ùsà ùʃà food)_PL eat (soft PKmn *ʃa ʃā ʃā ʃá ʃwá ʃwá úʃā sá sá sá ʃá food)_SG PKoUd egg_1 *ɔ̀m - - ɔ̀m ɔ̀m à+ʔɔ̄m - - - - - PGw egg_2 *símpʼ símpʼ símpʼ - - - - - - - - PDaOp egg_3 *kúmú - - - - - kúmú kűmű kűmű kűmű kűmű PKmn elephant_1 *gwàj kwɪ̀ kwɪ̀ gwà gwàj - - - - - - PUd elephant_2 *ɟɛ̀ - - - - ɟɛ̀ - - - - - POp elephant_3 *bàj - - - - - - bàj bàj bàj bàj Dana elephant_4 - - - - - dúkʼàn - - - - empty PCtrl *t ̪̓at ̪̓ - - tʼɪ ̌ - tʼíʔ t ̪̓àt ̪̓ tʼáj tʼáj tʼáj tʼáj (be)_1, dull empty dí+bw PUd - - - - - - - - - (be)_2 à 652 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi empty PGw *ʃákáná ʃákáná ʃákáná - - - - - - - - (be)_3 PKoUd enset_1 *wúsʼ - - wúsʼ wúsʼ à+wút ̪̓ - - - - - PGw enset_2 *ɛ́ʃɛ́ ɛ́ʃɛ́ ɛ́ʃɛ́ - - - - - - - - Dana enset_3 - - - - - bāɗà - - - - PKmn enter_1 *gìcʼ kɪšʼ kɪšʼ gìz - cīcʼ - kìtú kìtú kìtú kìtú enter_2, PKmn *t ̪̓wI tʼwɪ ́ tʼwɪ ́ - - - t ̪̓wī - - - - sprout PUd enter_3 - - - síʔ - - - - - - extinguish_ PCtrl *tú̪s - - túʃ ūs ūs tú̪s tűs tűs tűs tűs 1 extinguish_ PGw *ʃɪʃ́ ʃɪʃ́ ʃɪʃ́ - - - - - - - - 2 extract tooth_1, PKmn *bʊkʰ pʊ̌ pʊ̌ - - bǔkʰ - - - - - barking (of dog) extract PKoUd *jɛkʰ - - jɛ̀ʔ ɛ́kʰ - - - - - - tooth_2 653 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi extract POp *lɪ ̄ - - - - - - lɪ ̄ lɪ ̄ lɪ ̄ lɪ ̄ tooth_3 PKmn eye_1 *D̪E zì zì - ī ɛ ̄ d̪ɛ̀ dʒɛ̀ dʒɛ̀ zɛ̀ ʃɛ̀ Komo eye_2 - - ɓî - - - - - - - PKmn far (be)_1 *sɪ̀t ̪̓ - ʃɪt̄ʼ ʃɪ̀tʼ sīd sīɗ sɪ̀t ̪̓ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ sɪt̄ʼ PGw far (be)_2 ʃúl - - - - - - - - - PGw far (be)_3 - tàʔ - - - - - - - - GwKo farm (n.)_1 *kwar kwàlà kwɪ̂ kwàr - - - - - - - mɔ̀nzɛ̀ PUd farm (n.)_2 *mɔ̀nd̪ɛ̀ɗ - - - mɔ̀nd̪ɛ̀ɗ - - - - - d PDaOp farm (n.)_3 *(pɪ)tʊ̪j - - - - - tʊ̪̂j pɪ̀tɪ̀ pɪ̀tʊ̀j pɪ̀tʊ̀j pàtɔ̀j wū+d̪ùrù PCtrl fart_1 *d̪U(ru)s(E) - - tùʃ - tʊ̪̂s tīsí tīsí tʊ̄sɛ́ tʊ̄ʃɪ ́ s PGw fart_2 *bǔʃú bǔʃú bǔʃú - - - - - - - - fast (from PGw drinking or *baja bājá bājā - - - - - - - - eating) fast (from PKoUd drinking or *gakʼ - - gáʔ gǎ gǎkʼ - - - - - eating) 654 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi fast PCtrl *waC(a) - - ɪ+́wàʃà ŋwácʰà wácʰà wāt ̪h wāt wāt wāt wāt (velocity)_1 fast PGw - kāmākɔ́ - - - - - - - - (velocity)_2 fat (from PCtrl *mɔra - - mɔ̌rá - - mɔ́rá mɔ̄rá mɔ̄rá mɔ̄rá mɔ̄rá animals)_1 fat (from kʰwālà PUd *kʰwālàɲ - - - kʰwālàɲ - - - - - animals)_2 ɲ fat (from PGw *dà dà dà - - - - - - - - animals)_3 PKmn father_1 *bàbá - bàbá bǎ à+bàbá à+bàbá - àbá àbá àbá àbá PCtrl father_2 *cɔ̀m - - sɔ̀m à+cɔ̄m cɔ̄m sɔ̀m tʃɔ̀mɔ̀ tʃɔ̀mɔ̀ tʃɔ̀m tʃɔ̀m fear (be kʰɔ́k ~ PKmn *kʰwaGʼ kwāgà kwāgà kɔ̄g kʰɔ̄ʔ kʰɔ̄kʼ kʰɔ̄gɔ́ kʰɔ̄gɔ́ kʰɔ̄gɔ́ kʰɔ̄gɔ́ afraid) kʰɔ̀gɔ́ PCtrl fig_1 *pʰuku - - - pʰūʔ à+pʰūʔ pʰúj - pʰűkū pűj pűj Komo fig_2 - - sʼúŋá - - - - - - - PGw fig_3 *ɪśɪ̀ ɪśɪ̀ ɪśɪ̀ - - - - - - - - POp fig_4 - - - - - - dùrű - - - POp fight_2 *tā+dʒàj - - - - - - tā+dʒàj tā+dʒàj tā+zàj tā+ʃàj PGw fight_3 *hǐmì hǐmì hǐmì - - - - - - - - 655 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PKmn find, meet *gàm kàm kàm gàm gàm gàm kàm kàm kàm kàm kàm fingernail, toenail, PUd *gwǎpʰī - - - gwǎpʰī à+gwǎpʰī - - - - - claw, hoof_1 fingernail, toenail, PGw *kʼʊ̄mpʼ kʼʊ̄mpʼ kʼʊ̄mpʼ - - - - - - - - claw, hoof_2 fingernail, toenail, PCtrl *cʼɪḱʼɪŕ - - sʼɪḱʼɪĺ - - kʼɪḱʼɪŕ tʃʼɪḱʼɪŕ tʃʼɪḱʼɪŕ tʃʼɪḱʼɪŕ tʃʼɪḱʼɪŕ claw, hoof_3 PKmn finish_1 *dak - dak dàg dàkʰ dàkʰ dâk dàk - - - PKoUd finish_2 *kʼúɓ - - kúpʼ - kʼúɓ - - - - - POp finish_3 *mùtʰa̋ - - - - - - mùtʰa̋ mùtʰa̋ mùtʰá mùtʰá PCtrl finish_4 - - - - - dʊ̀kʰá - - - - fire or PKmn *wa(n)t ̪̓(Iʃ) āntʼ āntʼ wàtʼíʃ ɔ̄d ɔ̄ɗ ɔ̀t ̪̓ wʊ̄tʼɪ ́ wʊ̄tʼɪ ́ wʊ̄tʼɪ ́ ʊ̄tʼɪ ́ firewood PGw firefly_1 *bīsʼàn bīsʼàn bīsʼàn - - - - - - - - 656 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PUd firefly_2 *cúl - - - - à+cúl - - - - - fish (n.) PKmn general *wàcʼà wàsʼ wàsʼ wàsʼ wàʃʼ wàcʼ wàcʼà wàtʃʼà wàtʃʼà wàtʃʼà wàtʃʼà term PCtrl fish (v.)_1 *mal(i) - - māl mālí - màl màl màl màl màl PGw fish (v.)_2 - hɪṕɪ ́ - - - - - - - - PUd fish (v.)_3 - - - - rɔ̀kʰ - - - - - fish_(small, PKmn small *jàhút ̪ jàhú jàhú jàhú - - àhút ̪h àhű àhű àhűwı ̋ àhűj scales) fish_sp (big PCtrl and fat *gùr - - gùr - - gùr gùr gùr gùr gùr sized fish) fish_sp ʊ̄+wàsʼ+d PKmn *dɪŋ́kā - dɪŋ́kā - - - dıŋ̋ā dıŋ̋ā dıŋ̋ā dıŋ̋ā (electric)_1 ɪŋ́kā fish_sp Dana - - - - - à+d̪ìjɛ̀s - - - - (electric)_2 fish_sp PUd - - - - à+t ̪̓ ɛw̄ - - - - - (electric)_3 657 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi fish_sp (very small, scaled fish PKmn with a *dɔlɔ - dɔ̌lɔ́ dɔ̀lɔ́ - - dɔ́lɪʔ̄ dɔ̀lɔ́ dɔ̀lɔ́ dɔ̀lɔ́ dɔ́lɪ ̄ small rounded mouth) PKmn fishhook *Bɪncʼ bɪn̄sʼ bɪn̄sʼ bɪńsʼ bǐʃʼ à+bǐcʼ bɪćʼ ɓītʃʼ ɓītʃʼ ɓītʃʼ ɓītʃʼ tʼúm+ tʼúm+mɛ̀ PCtrl fist_1 *tʼum(á) - - tʼùmá tʼùmá tʼūmá tʼūmá tʼūmá tʼūmá mɛ̀ɗ ɗ PUd fist_2 - - - - - - - - - - PGw fist_3 *dʊ̄jʊ̄ dʊ̄jʊ̄ dʊ̄jʊ̄ - - - - - - - - PCtrl flea_1 *tá̪pʼí - - tápí - - tá̪pʼí - - - - Komo flour_1 - - bùdí - - - - - - - PGw flour_2 *pʼʊ̄mʊ̀tʼ pʼʊ̄mʊ̀tʼ pʼʊ̄mʊ̀tʼ - - - - - - - - PUd flour_3 *dǐpʰáɲ - - - dípʰɛ́ɲ à+dǐpʰáɲ - - - - - *pitʼɔn ~ PDaOp flour_4 - - - - - pìtʰɔ́n pìtʼɔ̀n pìtʼɔ̀n pìtʼɔ̀n pìtʼɔ̀n *pit⁽ʰ⁾ɔn PKmn fly (v.) *pʰàd̪ pāl pāj pàj pʰāj pʰɛ ̄ pʰàd̪ pʰāj pʰāj pʰāj pʰāj PCtrl follow_1 *basʼ - - - bǎsʼ bǎt ̪̓ pàt ̪h ɪ ́ - - - - 658 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PKmn follow_2 *uD ū ū út úr úr - - - - - POp follow_3 *kʰɔ̀m - - - - - - kʰɔ̀m kʰɔ̀m kʰɔ̀m kʰɔ̀m PGw follow_4 *tǎpʼ tǎpʼ tǎpʼ - - - - - - - - PUd follow_5 - - - súkʰ - - - - - - PCtrl food_1 *mà(ʔV) - - mɛ̀ʔí mà mà màʔá mǎ mǎ mǎ mǎ PGw food_2 *pwàʃ pwàʃ pwāʃ - - - - - - - - foot or PKmn *ʃɔnkʼ sɔ̄ŋkʼ sɔ̄ntʼ ʃɔ̀g ʃɔ̄kʼ ʃɔ̄ʔ ʃɔ̀g - - - - leg_1 foot or POp *wɔ̀nɛ̀ - - - - - - wɔ̀nɛ̀ wɔ̀nɛ̀ wɔ̀n wɔ̀n leg_2 PCtrl forbid_1 *tʼVʃ ~ t ̪̓Vʃ - - tʼáʃ tʼáʃ tʼáʃ - tʼɪśɪ ̄ tʼɪśɪ ̄ - - PGw forbid_2 *kàs kàs kɪ̀s - - - - - - - - PUd forbid_3 - - - - - - - - - - PDaOp forbid_4 *ɟíkʰì - - - - - ɟíkʰì - - zɪk̄ʰɪ ́ ʃɪk̄ʰɪ ́ PKmn forehead_1 *t⁽ʰ⁾(w)àg twã twã ̄ tàg - - tàg bī+tʰāg - pī+tʰāg - PUd forehead_2 *buɲɛ - - - bwìɲ bùjɛ̀ - - - - - PUd forehead_3 - - - - - - - - - - PKoUd four_1 *dɔ̀ŋgɔ̀n - - dɔ̀gɔ̀n dɔ̀ŋɔ̀n dɔ̀ŋɔ̀n - - - - - PGw four_2 *bɪśʼɪńɪ ̄ bɪśʼɪńɪ ̄ bɪśʼɪńɪ ̄ - - - - - - - - PDaOp four_3 *ŋwan - - - - - ŋwān hwàn hwàn hwàn hwàn 659 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi à+wāt ̪h ál à+wātál à+wātál à+wātál à+wātál PCtrl fox_1 *wata̪la - - wātálà - - ā à à à à PGw fox_2 *wǎŋʊ́ wǎŋʊ́ wǎŋʊ́ - - - - - - - - PUd fox_3 *mákʰ - - - mákʰ à+mákʰ - - - - - PGw frighten_1 *hɛ́gɛ̀ hɛ́gɛ̀ hɛ́gɛ̀ - - - - - - - - Komo frighten_2 - - málá - - - - - - - PUd frighten_3 *rúm - - - rúm rúm - - - - - *hɔ̀sà ~ PDaOp frighten_4 - - - - - wɔ̀sà hɔ̄sā hɔ̄sā hɔ̄sā hɔ̄sā wɔ̀sà bā+ɓɔ̌ŋkʼ à+ɓɔ̄ŋkʼɔ́ à+ɓɔ̄ŋkʼ à+bɔ̄ŋkʼ à+bɔ̄ŋkʼ à+bɔ̄ŋkʼ PKmn frog_1 *bɔŋkʼɔ́ pʼà+bɔ̌ŋgɔ́ pʼà+bɔ̌ŋgɔ́ - - ɔ́ ʔ ɔ́ ɔ́ ɔ́ ɔ́ PUd frog_2 *ɗwá - - - ɗwá à+ɗwá - - - - - frog_sp bā+pʼɛ̂nk GwKo *ɓapɛnkʼɛ pʼā+pɛ̀ŋgɛ̀ pʼā+pɛ̀ŋgɛ̀ - - - - - - - (long) ʼ *tā̪gāj ~ PCtrl giraffe_1 - - tāgɪ ̄ - - - tāgāj tāgāj tāgāj - tā̪gāj PKmn giraffe_2 *kʰOG kɔ́k kɔ́k - kʰúʔ kʰúʔ kʰɛ́g - - - kɛ̂k PDaOp girl_1 *ɓānápʰà - - - - - ɓānápʰà ɓānàpà ɓānàpà ɓānàpà ɓānàpà PCtrl girl_2 *ɓāʔɔ̄ - - bāʔɔ̄ - - - ɓāɔ́ ɓāɔ́ ɓāɔ́ ɓāɔ́ PGw girl_3 *dwā dwà dwā - - - - - - - - 660 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PUd girl_4 *ɲàrá - - - ɲàrá ɲàrá - - - - - PCtrl give *kʰɪ ́ - - kɪ+̌k cʰíʔ cʰí kʰɪ+́wā kʰɪʔ́ kʰɪʔ́ kʰɪʔ́ kʰɪʔ́ PGw give *tɪ ́ tɪ ́ tɪ ́ - - - - - - - - PCtrl give birth_1 *tʰú(i) - - tú - tʰú tʰwí tʰű tʰű tʰű tʰű PGw give birth_2 *pɪ ̄ pɪ ̄ pɪ ̄ - - - - - - - - PGw go *hɔ̄ hɔ̄ hɔ̄ hà - - - - - - - PCtrl go_PL *ʔja - - ɪ̀ ī ī - ʔjá ʔjá ʔjá ʔjá PCtrl go_SG *D̪a - - jà jà jà d̪ā dʒà dʒà zà ʃà PCtrl goat_1 *mɛ - - mɛ́ mɛ̀ mì mɛ ̄ mɛ̀ mɛ̀ mɛ̀ mɛ̀ PKmn goat_2 *ɲa(n) njã ̄ njã ̄ - - - - - - - - PKoUd good (be)_1 *ɓɔ́r - - ɓɔ́l ɓɔ́r ɓɔ́r - - - - - POp good (be)_2 *kʼáj - - - - - - kʼáj kʼáj kʼáj kʼáj PGw good (be)_3 *nɔ́kɔ́ nɔ́kɔ́ nɔ́kɔ́ - - - - - - - - Dana good (be)_4 - - - - - pɪd̪́á - - - - PCtrl gourd_1 *tʰul - - túl - tʰúl - tʰɔ̀j tʰɔ̀j tʰɔ̀j - PDaOp gourd_2 *gwàtʰá - - - - - gwàtʰá - - - gwàtʰá PUd gourd_3 *kʼwà - - - kʼwà kʼwà - - - - - PGw gourd_4 *dɔ̌lɔ́ dɔ̌lɔ́ dɔ̌lɔ́ - - - - - - - - 661 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi grab with PCtrl fingers, *t ̪̓O - - - - tʼú t ̪̓ ɔ́wá tʼʊ́ tʼʊ́ tʼʊ́ tʼʊ́ pinch_1 grab with PGw fingers, *kʼājà kʼājà kʼājà - - - - - - - - pinch_2 grab with Komo fingers, - - pʼɛ̀ - - - - - - - pinch_3 grandfather PKmn *càkʰO - sàkʊ́ sàkʊ́ - - - tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ tʃàkʰɔ́ _1 grandfather zàn+cɔ̂ à+d̪àn+cɔ̄ PUd *d̪àn+cɔm - - - - - - - - _2 m m grandmoth PKmn *kʰàkʰá - kàkáʔ kàká - - kʰàkʰá kàká kàká kàká kàká er_1 grandmoth tʰíŋ+k à+tʰíŋ+kū PUd *tʰíŋ+kùm - - - - - - - - er_2 ùm m PKmn grass *ʃO ʃʊ́ʃʊ̄ ʃʊ́ʃʊ̄ʃʊ̄ ʃɔ̀ʔí ʃɔ̄ à+ʃɔ̄ ʃɔ̄ʔɔ́ sɔ̀ sɔ̀ sɔ̀ sɔ̀ PKmn green_1 *D̪ir ? zı̃ ̀ zĩ ̂ zì zì d̪ì sīsī tʃìr tʃìr tʃìr tʃìr PUd green_2 - - - - - - - - - - 662 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi grind PDaOp *D̪ùkʼà - - - - - dùkʼà dʒùkʼà dʒùkʼà zùkʼà ʃùkʼà (dry)_1 grind Komo - - ɗɪḿ - - - - - - - (dry)_2 grind PGw *ʔù ʔù ʔù - - - - - - - - (dry)_3 grind PUd *jɔ̀r - - - jɔ̀r jɔ̀r - - - - - (dry)_4 grind PCtrl *tʼɪʃ́ ~ tʼíʃ - - tʼɪʃ́ tʼíʃ tʼíʃ tʼìʃíʔ tʼıs̋ tʼıs̋ tʼıs̋ - (fine)_1 grind PGw *mítʼī mítʼī mítʼī - - - - - - - - (fine)_2 grind PKmn (second *(ɔ)t⁽̪ʰ⁾ɔɗ - twɛ́j tɔ́ - tɔ̪́ɗ ɔ̀t ̪h ɔ́ ɔ́tɔ́ ɔ́tʰ ɔ́tʰ ɔ́tʰ grind) grind wet PKmn *Dàsʼ dàsʼ dàsʼ nàsʼ nàsʼ nàt ̪̓ - - - - - (first grind) guinea PKmn *T̪ʊ(n)G ʃʊ̌nkʼ ʃʊ̌nkʼ zʊ̀g - tū̪kʰ sʊ̄kʰ tʃʊ̄gʊ́ tʃʊ̄gʊ́ tʃʊ̄g tʃʊ̄g fowl_1 663 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi guinea PUd - - - dwàjá - - - - - - fowl_2 GwKo gums_1 *gùrɛ̀tʼ gùlɛ̀tʼ gùrɛ̀tʼ gùlɛ̀tʼ - - - - - - - kʼɔ̀d+ʃɛ̀ PUd gums_2 *kʼɔ̀d+ʃɛ̀ʔ - - - kʼɔ̀ɗ+ʃɛ̀ʔ - - - - - ʔ Dana gums_3 - - - - - ʊ́náʃɛ́ - - - - Gwama PKmn *gwama gwàmá kwāmà gwàmá - - gɔ̀má gɔ̀má gɔ̀má gɔ̀má gɔ̀má (ethnonym) PKmn hail, ice_1 *wasakʼ wàsà wàsà wàʃākʼ wàsáʔ à+wàsáʔ - - - - - hail, ice_2, PCtrl *tɛ̪n(ɛ) - - - - tɛ̪n̄ tɛ̪́n tɛ́nɛ́ - tɛ́n tɛ́n melt PKoUd hair_1 *mùr - - mùl mùr à+mùr - - - - - PGw hair_2 *bàk bàk bàk - - - - - - - - PDaOp hair_3 *ʃudaj - - - - - ʃùdáj sūjɛ́ sūjɛ́ sūjɛ́ sūjɛ́ PCtrl half_1 *tá̪kʼāláj - - tákálá - - tá̪kʼāláj tálá tálá tálá tákàlāj PGw half_2 - dāgà - - - - - - - - kʼúpʰ+kūt PUd half_3 - - - - - - - - - ʰ PKmn hand_1 *mɛ̀t ̪̓ bɪ̀tʼ mɪ̀tʼ - mɛ̀d mɛ̀ɗ mɛ̀t ̪̓ mɪ̀tʼɪ ́ mɪ̀tʼɪ ́ - mɛ̀tʼ 664 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PCtrl hand_2 *kʼɔlɔ - - kʼɔ̀lɔ̀ - - - kʼɔ̄j kʼɔ̄j kʼwī - PCtrl hang up_1 *sɪɗ́(a) - - ʃɪŕ ʃír ʃír sɪɗ́ā sɪŕ sɪŕ sɪŕ sɪŕ PGw hang up_2 *sūtʼ sūtʼ sūtʼ - - - - - - - - PDaOp have *kɔ̀tʰ - - - - - kɔ̀tʰ kɔ̀tʰ - kɔ̀tʰ kɔ̀tʰ PKmn have sex_1 *hag(a) hã́ʔ hã́ʔ hág - háʔ - hágá hágá hágá hágá PUd have sex_2 - - - màʃ - - - - - - Dana have sex_3 - - - - - bɔ̀d̪ɔ́ - - - - PKmn head *kʼʊ́p kʼʊ́p ʔʊ́p kʼʊ́p kʼúpʰ kʼúpʰ kʼʊ́pʰ kʼʊ́p kʼʊ́p kʼʊ́p kʼʊ́p head pad PCtrl (for head *kOta̪n - - - - - kʊ̀t ̪h áj kɔ̄tìn kɔ̄tìn kɔ̄tìn - carrying)_1 head pad PKoUd (for head *gwatʰV - - gwàtʊ́ gwǎtʰɛ ̄ gwǎtʰɛ ̄ - - - - - carrying)_2 head pad PGw (for head *kʼándí kʼándí kʼándí - - - - - - - - carrying)_3 heavy PKoUd *dìd - - dìd dìtʰ dìtʰ - - - - - (be)_1 665 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi heavy PGw *sìl sìl sìʔ - - - - - - - - (be)_2 heavy PDaOp *tʼɪḱʼá - - - - - tʼɪḱʼá tʼɪk̋ʼá tʼɪk̋ʼá tʼɪk̋ʼá tʼɪk̋ʼá (be)_3 PKoUd help_1 *wɔ̀d̪ - - wɔ̀l wɔ̀s wɔ̀t ̪h - - - - - PGw help_2 *pʼātʼà pʼātʼà pʼātʼà - - - - - - - - PDaOp help_3 *kúnā - - - - - kúnā kűná kűná kűná kűná PKmn herd (v.) *kʰaj kɛ̃ʔ̄ kɛ̃ʔ̄ kàʔí - - kʰàjɪʔ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ kʰājɪ ́ PCtrl hiccough_1 *háɗìkʼ - - - - àhǎɗkīʔ háɗìkʼ ha̋rīkʼ hárìkʼ hárìkʼ - Komo hiccough_2 - - zɔ̀r - - - - - - - PGw hiccough_3 *hīgìm hīgìm hīgìm - - - - - - - - PUd hiccough_4 - - - híʔtāʔ - - - - - - hide PCtrl *bwàkʼ - - bɔ̀ʔ bàʔ bàkʼ bàkʼ - - - bɔ̀kʼ (oneself) hide PCtrl *lus ~ *rus - - lùs - - rùs lùs lùs rùs - (sth.)_1 hide PGw *nɪn̄ɪ ̄ nɪ̃ ̄ nɪn̄ɪ ̄ - - - - - - - - (sth.)_2 PKmn hide, skulk *bɔ̀b ~ bɔ̀ɓ pɔ̀pʼ pɔ̀pʼ bɔ̀b - - - pɔ̀j pɔ̀j pɔ̀j - PCtrl hit_1 *dɔr - - - dɔ̌r dɔ̌r tɔ̀dɔ́r tɔ̀r tɔ̀r tɔ̀r tɔ̀r 666 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi GwKo hit_2 *ʃàpʼ ʃāpʼ ʃāb ʃàpʼ - - - - - - - PCtrl hoe (n.)_1 *k⁽ʰ⁾Orɛcʼ - - - kūrɛ́ kʰɔ̄rɛ̂cʼ kɔ̄rɛ́ kɔ̄rɛ́ kɔ̄rɛ́ kɔ̄rɛ́ kɔ̄rɛ́ GwKo hoe (n.)_2 *gaʃa gàʃà gàʃà gâʃ - - - - - - - PCtrl hole_1 *gVs - - gìʃ ɟìs ɟìs kùs kùs kùs kùs kùs PGw hole_2 *bʊ̀ bʊ̀ bʊ̀ - - - - - - - - home, PCtrl *pa - - - pā pā - pà pà pà - place_1 home, PDaOp *wad̪ - - - - - wád̪ - - - wǎtʰ place_2 home, PGw *tʼʊ̀mʊ̀ tʼʊ̀mʊ̀ tʼʊ́mʊ̄ - - - - - - - - place_3 home, Komo - - làū - - - - - - - place_4 honey PKmn *njansʼɛsʼ jānsʼɛ́sʼ ɲɛn̄zɛ́sʼ nɛ̀z nɛ̀sʼ à+nɛ̀t ̪̓ nɛ̀s nɛ̀dʒɛ̀ - nɛ̀dʒɛ̀ - badger PKoUd hoof *cʊ̄k⁽ʰ⁾(ɪ) - - sʊ̄kɪ ́ - à+cūkʰ - - - - - horn PCtrl (anatomy)_ *k⁽ʰ⁾ɪd̪ - - kɪ ̄ cú cɛ́ kɪd̪̄ kʰɪw̄ kʰɪw̄ kʰɪw̄ kʰɪw̄ 1 667 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi horn PGw (anatomy)_ *kwàp kwàp kwàp - - - - - - - - 2 *parʃa ~ PCtrl horse_1 - - pǎrʃá - - pàrd̪á pàrsá pàrsá pàrsá pàrsá *pard̪a PGw horse_2 *gǎŋú gǎŋú gǎŋú - - - - - - - - PUd horse_3 *bàgál - - - bàgál bàgál - - - - - hot (be)_1, PKoUd *ɓàsʼ - - ɓàsʼ ɓàsʼ ɓāt ̪̓ - - - - - ill (be) PDaOp hot (be)_2 *ʃàdɪ ́ - - - - - ʃàdɪ ́ - - - ʃàrɪ ́ PGw hot (be)_3 *tàwàn tàwàn tàwàn - - - - - - - - POp hot (be)_4 *kʰāgá - - - - - - kʰāgá kʰāgá kʰāgá - PCtrl house_1 *gùɓ(V) - - gùbí gùb gùɓ kùʔú kù kù kù kù PGw house_2 *swāl swāl swɪ ̄ - - - - - - - - PCtrl hunger_1 *t ̪h (w)ád̪ - - twáj tʰɔ́j tʰɛ́ʔ t ̪h ád̪ tʰáj tʰáj tʰáj tʰáj PGw hunger_2 *pɪťɪ̀ pɪťɪ̀ pɪďɪ̀ - - - - - - - - hunt (in PCtrl *dwáɗ - - - wátʰ wáɗ - dwàr dwàr dwàr - group) PCtrl hunt_1 *hɔ̄j - - hɔ̄j - - - hɔ̄j hɔ̄j hɔ̄j hɔ̄j Dana hunt_2 - - - - - màt ̪h á - - - - 668 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PGw hunt_3 *ìrìŋ ìrìŋ ìrìŋ - - - - - - - - PUd hunt_4 *páD - - - pátʰ pár - - - - - PCtrl husband_1 *gàsʼ - - gàz à+kàsʼ kàt ̪h kàzūpʼ kàdʒʊ́m kàdʒʊ́m kàdʒʊ́m kàdʒʊ́m PGw husband_2 - jītī - - - - - - - - PGw hyena_1 *kàsʼmǎná kàsʼmǎná kàsʼmǎná - - - - - - - - PCtrl hyena_1 *tá̪pàj - - tápàjí - - à+tá̪pàjī à+tápàjɪ ́ à+tápàjɪ ́ à+tápàjɪ ́ à+tápàjɪ ́ ŋūríŋsʼ PUd hyena_2 *ɲùrúɲ - - - à+ɲùrúɲ - - - - - íʔ PUd hyena_3 - - - - - - - - - - POp hyrax_1 *ɓáŋgɔ̄ - - - - - - ɓáŋgɔ̄ ɓáŋgɔ̄ ɓáŋgɔ̄ ɓáŋgɔ̄ PGw hyrax_2 *dwákɪ̀ dwákɪ̀ dwákɪ̀ - - - - - - - - Komo hyrax_3 - - kʼwà - - - - - - - dùdù+kʼá dùdú+kʼ GwKo joint_1 *dùdù - - - - - - - - +mɪ̀tʼ ɔ̀lɔ̀ PUd joint_2 - - - ʃúʔ - - - - - - POp joint_3 *kʊ̋ká - - - - - - kʊ̋ká kʊ̋ká kʊ̋ká kʊ̋ká Dana joint_4 - - - - - t ̪̓ ā+súkʼ - - - - PKmn kick *t ̪h áɓ tápʼ tápʼ táb tʰáb tʰáɓ t ̪h áp tʰáp tʰáp tʰáp tʰáp d̪ɛ̀kʰwād̪ dɪ̀kwād dɪ̀kwād PDaOp kidney_1 *d̪Ekwād̪à - - - - - dɪ̀kwādà dɪ̀kwādà à à à 669 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi ɗwàkàɗ Komo kidney_2 - - - - - - - - - wàkà PKmn kill, fight *kʼOʃ kʼʊ́ʃ kʼʊ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́ʃ kʼɔ́sɔ́ kʼɔ́sɔ́ kʼɔ́s kʼɔ́ʃ PCtrl kiss_1 *sʼuɗ - - sʼùr - - sʼúɗ tʃʼűr tʃʼűr tʃʼűr tʃʼűr PGw kiss_2 *sʼʊ́pʊ́n sʼʊ́pʊ́n sʼʊ́pʊ́n - - - - - - - - pʰí+tʼw PUd kiss_3 *pʰí+tʼwā - - - pʰítʼwā - - - - - ā PCtrl knife *ʃìpʰá - - ʃìpá - - ʃìpʰá - - - - know or be PCtrl *tÉ̪ - - - - tɛ̪́ tɪ̪ ́ - - - - able_1 know or be PKoUd *mɪʃ - - mɪ̀ʃ míʃ míʃ - - - - - able_2 know or be GwKo *ar(V) ǎlā ǎjá àrí - - - - - - - able_3 Komo GwKo (ethnonym) *kwàmà kwàmà kwàmà kɔ̀m - - - - - - - _1 Komo PUd (ethnonym) *pʰɛk̄ʼɛ́ʃ - - - pʰɛk̄ʼɛ́ʃ pʰɛk̄ʼɛ́ʃ - - - - - _2 670 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Komo *bì+pāj+dʒ bì+pāj+ bì+pāj+ bì+pāj+ bì+pāj+ʃ POp (ethnonym) - - - - - - àw dʒàw dʒàw zàw àwà _3 Komo Dana (ethnonym) - - - - - màd̪ìn - - - - _4 lack (not PCtrl *kʼaʃ - - kʼáʃà kʼáʃ kʼáʃ kʼàʃ kʼās kʼās kʼās kʼāʃ have)_1 lack (not PGw *pǎŋgɔ̄ pǎŋ pǎŋgɔ̄ - - - - - - - - have)_2 tʃʼákʼűm tʃʼákʼűm tʃʼákʼűm tʃʼákʼűm PDaOp ladle_1 *cʼákʼúmú - - - - - cʼákʼúmú ű ű ű ű PKoUd ladle_2 *ɗAm - - ɗɔ́m ɗám ɗám - - - - - PGw large bird *bīt bīt bīt - - - - - - - - POp large bird *lìtʰ - - - - - - lìtʰ lìtʰ lìtʰ lìtʰ Komo large bird - - sʼʊ́p - - - - - - - PKmn laugh *pʰ(j)asʼ ɪs̄ʼ pāsʼ pɛ̀sʼ pʰɛs̄ʼ pʰɛt̄ ̪̓ pʰàsʼ pʰātʃʼ pʰātʃʼ pʰātʃʼ pʰātʃʼ PCtrl lay (v.)_1 *pʰáɗ(a) - - pár pʰád pʰáɗ pʰʊ́ɗā pʰárá pʰárá pʰárá pʰárá PGw lay (v.)_2 *sí sí sí - - - - - - - - PUd lay (v.)_3 - - - - t ̪̓í - - - - - 671 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi lead PKmn *sʊs ʃʊ́ʃ ʃʊ́ʃ ʃʊ́ʃ sús sús sʊ́ʔ - - - - (guide)_1 lead POp - - - - - - títā - - - (guide)_2 PGw leave_1 *jɪ ̌ jɪ ̌ jɪ ̌ - - - - - - - - PUd leave_2 *pí - - - pí pí - - - - - kʰāgɪ+́dʒ PDaOp leave_3 *kʰāg - - - - - kʰāgɪ+́d̪ɛ̀ - kʰāgɪ+́zɪ̀ kʰāgɪ+́ʃɪ̀ ɪ̀ PCtrl lick_1 *t ̪̓ ɛɗ - - tʼɛr̄ tʼɛd̄ tʼɛɗ̄ t ̪̓ ɛ́ɗ tʼɛ́rɛ́ tʼɛ́r tʼɛ́r tʼɛ́r PGw lick_2 *tʼápán tʼápán tʼápán - - - - - - - - lie down, PKmn *iʃ ìʃ ìʃ ìʃ īʃ īʃ īʃá - - - - sleep_1 lie down, POp *ɓā+tın̋ɛ ̄ - - - - - - ɓā+tın̋ɛ ̄ ɓā+tín ɓā+tín ɓā+tɪ̂n sleep_2 light PKmn *sʼa sʼā sʼā sʼà sʼā t ̪̓ ā sʼʊ́wà tʃʼǎ tʃʼǎ tʃʼǎ tʃʼǎ (ignite) light (the way with PKmn *kʰàɲ kɪn̄ kɪn̄ - kʰāɲ kʰāɲ - - - - - torch/flashl ight)_1 672 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi light (the way with PCtrl *sʼOpʰ - - sʼɔ̀p - - sʼʊ̄pʰ tʃʼʊ̄ tʃʼʊ̄ tʃʼʊ̄ tʃʼʊ̄ torch/flashl ight)_2 light PCtrl weight *pʰɔt ̪̓(ɔ) - - pɔ̀tʼ pʰɔ̄d pʰɔ̄ɗ pʰɔ́t ̪̓ pʰɔ́tʼɔ́ - pʰɔ́tʼ - (be)_1 light PGw weight *ʃɛ́ŋgɛ́l ʃɛ́ŋgɛ́l ʃɛ́ŋɪʔ́ - - - - - - - - (be)_2 PCtrl lion_1 *d̪ɔ̀bɔ - - dɔ̀b zɔ̀pʰ à+d̪ɔ̀pʰ - tɔ̀bɔ̀ tɔ̀bɔ̀ tɔ̀b tɔ̀b PGw lion_2 *tìʃàr tìʃàl tìʃàr - - - - - - - - Dana lion_3 - - - - - ɟɛ̂l - - - - *cikʼa ~ PCtrl listen_1 - - sɪ̀g ʃīkʼ cīkʰ sʼìkʼà - - - - cɪkʼa POp listen_2 *kàrà - - - - - - kàrà - kàrà kàrà PGw listen_3 *tíbí tíbí tíbí - - - - - - - - LOC, BE PKmn *ɪ ́~ *í - ɪ ́ ɪ ́ - í - ı ̋ - - - LOC 673 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi long or tall PKmn *tUr tǔ tǔ tʊ́l túr túr - - - - - (be)_1 long or tall POp *ʃɪḱʼ - - - - - - sɪḱʼ sɪḱʼ sɪḱʼ ʃɪḱʼ (be)_2 long or tall Dana - - - - - búd - - - - (be)_3 PCtrl lost (be)_1 *T̪is - - - tís tí̪s sís sísī sísī sísī sísī PGw lost (be)_2 *hɪ̀ntʼ hɪ̀ntʼ hɪ̀ntʼ - - - - - - - - Komo lost (be)_3 - - wʊ̀k - - - - - - - PKmn louse_1 *ʃOkʼ(VN) - ʃʊ̄gʊ̀n ʃùwɛ̀n ʃɔ̄kʼɔ̄m à+ʃɔ̄kʼɔ̄m ʃùkʼnáj sūkʼɛ́n sūkʼɛ́n sūkʼɛ́n ʃūkʼɛ́n PGw louse_2 sʼɔ̄nzɔ̀ - - - - - - - - - PCtrl love_1 *hwaj - - hwāj - - ɔ́j ɔ́já ɔ́já ɔ́já ɔ́já GwKo love_2 *ʃun(V) ʃūnù ʃǔn ʃùnà - - - - - - - PUd love_3 *ɛ́n - - - ɛ́n ɛ́n - - - - - PGw love_4 *ʃám ʃám ʃám - - - - - - - - PDaOp maggot_1 *D̪Vʃ - - - - - dìʃ dùsù dʒùs zùs ʃùʃ GwKo maggot_2 *sʼʊntʼa sʼʊ̄ntʼ sʼʊ̄ntʼ sʼwàndá - - - - - - - PUd maggot_3 *jìʃ - - - jìʃ à+jìʃ - - - - - PCtrl maize_1 *kʰɔba - - - kʰɔ́bà à+kʰɔ́bà kʰɔ́bā kʰɔ̀bà kʰɔ̀bà kʰɔ̀bà kʰɔ̀bà Komo maize_2 - - kāmā - - - - - - - 674 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PGw maize_3 *sàmún sàmún sàmún - - - - - - - - make go PKmn *tʰaʃ ~ t ̪h aʃ tāʃ tāʃ tàʃ tʰáʃ tʰáʃ - - - - - away meat, PKmn *ʃum(a) sūm sūm ʃùm ʃūm ʃūm ʃùmà sūmā sūmā sūmā ʃūmā animal PCtrl mediate_1 *t ̪h at ̪̓ - - - - t ̪h āt ̪̓ t ̪h ājd̪á - - - tʰájá POp mediate_2 *tʰágá - - - - - - tʰágá tʰágá tʰágá - Komo mediate_3 - - tìʃ - - - - - - - PGw mediate_4 *tābūk tābūk tābūk - - - - - - - - mediate_5, PUd *wɔ̀t ̪h - - - wòs wɔ̀t ̪h - - - - - help PGw milk_1 *bàsʼ bàsʼ bàsʼ - - - - - - - - PDaOp milk_2 *tɪʃ́ - - - - - tɪʃ́ tɪś tɪś tɪś - moon or PDaOp *àdɔ̀j - - - - - à+dɔ̀j à+dɔ̀j à+dɔ̀j à+dɔ̀j à+dɔ̀j month_1 moon or PKoUd *páj - - páj páj à+pɛ́ʔ - - - - - month_2 moon or *sʼɛw̄àn ~ PGw sʼɛw̄àn sʼjāwàn - - - - - - - - month_3 sʼjāwàn PCtrl mosquito_1 *taʃ - - - tāʃ à+tāʃ tāʃ tās tās tās tǎʃ 675 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi mosquito_2 PKmn *mɪmɪ mɪm̌ɪ ́ mɪm̌ɪ ́ mɪm̌ɪ ́ mǐmī - - - - - - , firefly PKoUd mother_1 *táɗā - - dâ à+tádā táɗā - - - - - POp mother_2 *àdıj̋ɛ ̄ - - - - - - àdıj̋ɛ ̄ àdıj̋ɛ ̄ àdıj̋ɛ ̄ àdıj̋ɛ ̄ mother, kūm ~ kʊ̀m ~ PKmn *kʊman - kūm kʊ̄mán kūmán kʊ̄mán kʊ̄mán kʊ̄mán kʊ̄mán female kūmán kwàn PKmn mouth *t ̪̓wa tʼwā tʼwā tʼā tʼwā tʼwā t ̪̓ āʔá tʼā tʼā tʼā tʼā mushroom_ PCtrl *dìʃ(a) - - dìʃ dìʃìʔ à+dìʃ tìʃà tìsà tìsà tìsà tìʃà 1 mushroom_ PGw *tʼáʃá tʼáʃá tʼáʃá - - - - - - - - 2 PCtrl name_1 *D̪uga - - zāgà - - - dʒùgà - - ʃùgà PDaOp name_2 *jɪ̂nt ̪h ɪ ́ - - - - - jɪ̂nt ̪h ɪ ́ - dʒɪn̄tʰɪ ̄ zɪn̄tʰɪ ̄ ʃɪn̄tʰɪ ̄ PUd name_3 *gwàj - - - gwàj gwàj - - - - - PGw name_4 - twī - - - - - - - - navel, PCtrl umbilical *kʼúmú - - kʼúmú - - kʼúmú kʼűmű kʼűmű kʼűmű kʼűmű cord PCtrl near_1 *ɗiʃa - - ɗìʃ ɗīʃ ɗīʃ ɗìʃà īsā īsí ɗīsā ɗīʃā PGw near_2 *kɪs̄ɪ ̄ kɪs̄ɪ ̄ kɪs̄ɪ ̄ - - - - - - - - 676 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PKmn neck *ɓId̪a pʼɪl̄ pʼɪ ̄ ɓàʔ ɓāʔ ɓāʔ ɓìd̪à ɓījā ɓījā ɓījā ɓījā PKmn new_1 *d̪is ~ d̪ɪs dīʃ dīʃ zɪ̀ʃ tīs tī̪s - - - - - PDaOp new_2 *k⁽ʰ⁾ɪ̀s - - - - - kɪ̀s kʰɪ̀s kʰɪ̀s kʰɪ̀s kʰɪ̀s PKmn nose *ʃʊnʃ ʃʊ̃ʃ̄ ʃʊ̃ʃ̄ ʃʊ̀nʃ ʃūʃ ʃūʃ ʃʊ̀ʃ sʊ̀sʊ̀ sʊ̀s sʊ̀s ʃʊ̀ʃ nosebleed_ PCtrl *(a)T̪ina ? - - zīnà - - - àtʃwìnà àtʃwìnà àtʃwìnà àtʃwìnà 1 nosebleed_ PKmn *SʼUn(tʼ)a - tʼʊ̀ntʼ - - - sʼùnáʔ - - - - 2 not know *mal(at) ~ PCtrl - - màlàt mɔ̀l mɔ̀l - tā+mál tā+mál tā+mál tā+mál (how)_1 *mɔl not know PGw *dāŋā dāŋā dāŋā - - - - - - - - (how)_2 Nuer PKmn (ethnonym) *ɟaŋg(ɔ)aj zǎgʊ́ zǎgʊ́ zǎgʊ́ ʒwǎŋgì - ɟàŋwɛ̀j dʒāŋɔ́ - zāŋwɛ́ ʃàŋwɛ̀ _1 Nuer POp (ethnonym) - - - - - - - nɔ̀r - - _2 PDaOp offend_1 *ʃalɪ - - - - - ʃàlɪ ́ sálɪ ̄ sálɪ ̄ sálɪ ̄ ʃál PGw offend_2 *kɛ́nzɛ́ kɛ́nzɛ́ kɛ́nzɛ́ - - - - - - - - 677 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PKoUd offend_3 *sɔ́D - - ʃɔ́t sɔ́d sɔ́r - - - - - oil (organic PCtrl substance)_ *jEn(ɪ) - - jɛ̀n jìn à+jìn jɪń dʒɪn̄ɪ ́ dʒɪn̄ɪ ́ zɪn̄ ʃɪn̄ 1 oil (organic PGw substance)_ *dà dà dà - - - - - - - - 2 PUd old (be)_1 *d̪àm(V) - - - zàmì d̪àmɔ̀ - - - - - one (1)_1, PCtrl *ɗɛ - - ɗɛ́ ɗɛ́ ɗɛ́ ɗɛ̀dɛ́ ɗjān ɗjān ɗɛd̄ɛ ̄ ɗɛd̄ɛ ̄ alone_2 PGw one (1)_2 *sɛ́nɛʔ̄ sɛ́nɛʔ̄ sɛ́nɛʔ̄ - - - - - - - - PKmn open *kʰáɗ(a) kálá kájá kár kʰǎd kʰǎɗ kʰátá kʰátá kʰátá kʰátá kʰátá Opo PKmn (ethnonym) *k⁽ʰ⁾ɪńáj kɪńá kɪńá kɪńá cʰínáj - kɪńáj - - - - _1 Opo POp (ethnonym) - - - - - - ʊ̀pɔ̀ - - - _2 678 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi pack in, PKmn stuff into *gɪ̀m gɪ̀m gɪ̀m - - ɟìm - - - - - container_1 pack in, PDaOp stuff into *ʃɔŋ(V) - - - - - ʃɔ̀ŋɔ́ sɔ̄ŋɪ ́ sɔ̄ŋɪ ́ sɔ̄ŋɪ ́ ʃɔ̄ŋɔ́ container_2 pack in, Komo stuff into - - wūb - - - - - - - container_3 pack in, PUd stuff into - - - dìl - - - - - - container_4 pack in, PUd stuff into - - - - - - - - - - container_5 pain, be GwKo *mɪ̀tʼ(V)n - mɪ̀tʼɪ ̄ mɪ̀tʼán - - - - - - - hurt_1 pain, be PCtrl *ʃáw - - ʃáw - - ʃáw sáw sáw sáw ʃáw hurt_2 679 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi pain, be PGw *tʼɔ̄ʃ tʼɔ̄ʃ tʼɔ̄ʃ - - - - - - - - hurt_3 pain, be PUd *ɲɔ̀r - - - ɲɔ̀r ɲɔ̀r - - - - - hurt_4 PKmn palate_1 *ɓalilaj pʼàlílì pʼàlílí ɓàlílá - - - - - - - *bwàhāɗgī( bwàgíd PUd palate_2 - - - bwàhāɗgī - - - - - da) à Dana palate_3 - - - - - kàltʼá - - - - path, PKmn road_1, *bwaŋ(a) pwǎŋà pwǎŋà - - bwàj - - - - - towards path, Dana - - - - - jàrábūd - - - - road_2 path, PUd - - - ŋɔ̂ŋ - - - - - - road_3 path, tʼā+ʊ̀+p tʼā+ʊ̀+p POp *̀pʊ́r - - - - - - dʒɛ̀pʊ́r dʒɛ̀pʊ́r road_4 ʊ́r ʊ́r path, Komo - - kʊ̀mà - - - - - - - road_5 PCtrl peel_1 *kʼɔj - - kʼɪ ́ cʼɛ́ cʼɛ́ kʼɔ̀j kʼɛ́ - - - 680 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PDaOp peel_2 *jɛkʼ ~ hɛkʼ - - - - - hɛ́kʼ - jɛk̄ʼ jɛk̄ʼ jɛk̄ʼ peel, PKmn *pʼi(n)Cʼ? pʼɪńsʼ pʼɪńsʼ - - pʼíɗ pʼɪt́ʼ - - - - husk_1 peel, skin (with PKoUd *pʰɛɗ - - pɛ̀t pʰítʰ - - - - - - knife), peel, husk, peel pelt_1, drive in, PCtrl pound, *tɛ̪́l - - tɛ́l - tɛ̪́l tɛ̪́l tɛ́l tɛ́l tɛ́l tɛ́l hammer (something) PGw pelt_2 *kānsʼ kānsʼ kānsʼ - - - - - - - - pelt_3, PCtrl *ʃin - - ʃìn ʃīn ʃīn ʃìnà sīnā sīnā - - throw PCtrl penis_1 *D̪Is - - jìʃ jīs jīs d̪ɪ̀s dʊ̄s dʒʊ̄s zʊ̄s sʊ̄s PGw penis_2 *sjã sjã sjã - - - - - - - - person_1, PCtrl *g(w)àsʼ - - jī+gwàz wàzí wàtí̪ jɛ+̄kàz ʊ̀+kàdʒ ʊ̀+kàdʒ ʊ̀+kàdʒ ʊ̀+kàdʒ man PKmn person_2 *D̪ita̪ sīt sīt - - - d̪ítà̪ - - ʊ̀+zìtà ʊ̀+ʃìtà 681 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi person_2, PCtrl *g(w)àsʼ - - gwàz gwàsʼ gwàt ̪h kwàz - - - - man pick up (small PCtrl *kwáɲ - - kwán kwáɲ kwáɲ kwán kwán kwán kwán - things), peck at_1 pick up (small PGw *nǎ nǎ nǎ - - - - - - - - things), peck at_2 PKmn pick_1 *pʼɔt ̪̓(a) pʊ́t pʊ́t pʼɔ̀tʼ - - pʼɔ̀t ̪h à pʼɔ̄tʼā pʼɔ̄tʼā pʼɔ̄tʼā pʼɔ̄tʼā PUd pick_2 - - - wɔ́ʔ hɔ́ - - - - - PKmn pierce_1 *tU̪b(a) sʊ́ sʊ́ - tūpʰ tūpʰ tú̪bá - - tʰūbá tʰūbá POp pierce_2 *tʃɔ̄t - - - - - - tʃɔ̄t tʃɔ̄t - - pierce, cut PCtrl (grass with *kʰɛm - - kɛ́m - - - kʰɛ̀m kʰɛ̀m kʰɛ̀m kʰɛ̀m sickle) PKmn pig_1 *gUd̪Um kʊ̀rʊ́m kʊ̀rʊ́m gùdúm - - gùd̪ùm kùdùmà kùdùmà kùdùmà kùdùmà pipe (for PKmn *dʊ́ɟɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ dʊ́zɛ̀ - - dʊ́ɟɛ̀ - - - - smoking)_1 682 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi pipe (for POp *kʊ̀dʊ́s - - - - - - kʊ̀dʊ́s kʊ̀dʊ́s kʊ̀dʊ́s kʊ̀dʊ́ʃ smoking)_2 plait or POp braid or *pʰákʼ - - - - - - pʰákʼ pʰákʼ pʰákʼ pʰákʼ weave_1 plait or Komo braid or - - ʃít - - - - - - - weave_2 plait or PGw braid or *sʼú sʼú sʼú - - - - - - - - weave_3 plait or Dana braid or - - - - - tâkʰ - - - - weave_4 plait or PUd braid or *tʰɔ́ɗ - - - tʰɔ́d tʰɔ́ɗ - - - - - weave_5 plaster (v.), Komo - - lɔ̀k - - - - - - - adhere_1 683 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi plaster (v.), PDaOp *tʰáp - - - - - tʰápʰ tʰáp tʰáp tʰáp tʰáp adhere_3 plaster (v.), PGw *pàntʼ pàntʼ pàntʼ - - - - - - - - adhere_4 PKmn platform *pʰárá párá párá párá - - pʰárá pʰárá pʰárá pʰárá pʰárá play Dana (instrument - - - - - ʃàj - - - - )_1 play PKmn (instrument *jɪ jɪ̀ jɪ̀ - - - - dʒɪ ̄ dʒɪ ̄ zɪ ̄ ʃɪ ̄ )_2 play Komo (instrument - - tíʃ - - - - - - - )_3 play PUd (instrument - - - pʰú - - - - - - )_4 PCtrl play_1 *bɔ̀g(ɔ̀) - - bɔ̀g - - pɔ̀gɔ̀ʔ pɔ̀gɔ̀ pɔ̀gɔ̀ pɔ̀gɔ̀ pɔ̀gɔ̀ PGw play_2 *jíl jíl jí - - - - - - - - PUd play_3 *lɔ̀b - - - lɔ̀b lɔ̀ɓ - - - - - 684 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi polygamou PCtrl *ɓ(w)áʃ - - ɓáʃ ɓwáʃ ɓáʃ ɓáʃ - - - - s (be)_1 polygamou POp *ʃʊ́ʃʊ́ - - - - - - sʊ́sʊ́ sʊ́sʊ́ sʊ́s ʃʊ́ʃ s (be)_2 polygamou PGw - pʼɔ̄sɔ́pʼ - - - - - - - - s (be)_3 PDaOp poor (be)_1 *cān - - - - - cān tʃān tʃān tʃān tʃān Komo poor (be)_2 - - dɛ̀gá - - - - - - - POp poor (be)_3 *sɔ̄sɔ́r - - - - - - sɔ̄sɔ́r sɔ̄sɔ́r sɔ̄sɔ́r sɔ̄sɔ́r PGw poor (be)_4 *gìrì gìrì gìrì - - - - - - - - hāsʼ+kʼ PUd poor (be)_5 *hāsʼ+kʼɔ̀d - - - hāt ̪̓+kʼɔ̀ɗ - - - - - ɔ̀d porcupine_ *k⁽ʰ⁾akʼas ~ PKmn kákʼàʃ kákʼàʃ káʃākʼ - - kásākʼ kʰásākʼ kʰásākʼ kʰásākʼ kʰáʃākʼ 1 k⁽ʰ⁾asakʼ porcupine_ tɔ̄nsúlɛ́ PUd - - - - - - - - - 2 ɲ porcupine_ PUd - - - - jápʰ - - - - - 3 pottery, PCtrl *tʰi(aj) - - sɪ ̄ - - tījā tʰɪ ̌ tʰɪ ̌ tʰɪ ̌ tʰɪ ̌ pot_1 685 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi pottery, PGw *pǎjá pǎjá pǎjá - - - - - - - - pot_2 pottery, PUd *nɔ́s - - - nɔ́s à+nɔ́s - - - - - pot_3 pound PDaOp *kʰɔm - - - - - kʰɔ̀m - - kʰɔ́m kʰɔ́m (v.)_1 pound PUd *t ̪h ásʼ - - - tʰásʼ t ̪h át ̪̓ - - - - - (v.)_2 pound PCtrl *Vl - - - ál ál ɔ̀l ɔ̄l ɔ̄l ɔ̄l ɔ̄l (v.)_3 pound PKmn *kɔ́pʰ kɔ́p kɔ́p - kúpʰ - - - - - - (v.)_4 pound PCtrl *dùm - - dùm - - dùm dùm dùm dùm dùm (v.)_5 pound POp *pàtʃʼɪ ́ - - - - - - pàtʃʼɪ ́ pàtʃʼɪ ́ - - (v.)_6 pound PGw *gʊ̌pʊ̀ gʊ̌pʊ̀ gʊ̌pʊ̀ - - - - - - - - (v.)_7 PCtrl pour_1 *kʰar - - kál - - kʰár kʰārɪ ́ kʰārɪ ́ kʰārɪ ́ kʰārɪ ́ POp pour_2 *tʰır̋ - - - - - - tʰır̋ tʰır̋ tʰır̋ tʰır̋ 686 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PGw pour_3 *swɪ ́ swɪ ́ swɪ ́ - - - - - - - - Dana pour_4 - - - - - tí̪sà - - - - PUd pour_5 *jú - - - jú jú - - - - - PCtrl pray, beg_1 *ta̪m - - dàm - tá̪m tá̪má tāmá tāmá tāmá tāmá PGw pray, beg_2 *jàn jàn jàn - - - - - - - - PUd pray, beg_3 *sʼɔ̄ - - - sʼɔ̄ t ̪̓ ɔ̄ - - - - - *tí̪t(á) ~ PCtrl precede_1 - - tít - - - tıt̋á tıt̋á tıt̋á tıt̋á tít(á) Dana precede_2 - - - - - kʊ̀d̪á - - - - PGw precede_3 *tʼɔ́zí tʼɔ́zí tʼɔ́zí - - - - - - - - PUd precede_4 *bāsʼ - - - bāsʼ bāt ̪h - - - - - pregnant PKmn *bUma pʊ̀m pʊ̀m ɓú pwá bwà pùmá pǔmá pǔmá pǔmá pǔmá (be) PDaOp protect_1 *kát⁽̪ʰ⁾ - - - - - kát ̪h - - kát kát PCtrl protect_2 *jigi - - jìgíʔ - - - dʒīgí dʒīgí - - PGw protect_3 *tɪm̌ tɪm̌ tɪm̌ - - - - - - - - PUd protect_4 *cɛʃ̄ - - - cɛʃ̄ cɛʃ̄ - - - - - pull, pull PCtrl *haɗ - - hàd - hǎɗ hàdí hādɪ ́ hādɪ ́ hādɪ ́ hādɪ ́ off, drag 687 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi *sūsá ~ PDaOp push_1 - - - - - sūsá sʊ̄sá sʊ̄sá sʊ̄sá sʊ̄sá *sʊ̄sá PKmn python_1 *sʊm ʃʊ̄ʃʊ́m ʃʊ̄ʃʊ́m ʃʊ̌m - súm sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ sʊ̄mʊ́ kàldɔ̄ŋ PUd python_2 - - - - - - - - - ā rain PDaOp (precipitati *d̪ìbà - - - - - d̪ìbà - - - dìbà on)_1 rain PKmn (precipitati *ʃɔkʼ - ʃʊ́ ɔ́ ʃɔ̄kʼ à+ʃɔ̄kʼ - hɔ́ hɔ̌ hɔ̌ - on)_2 rain PGw (precipitati hìnsʼ - - - - - - - - - on)_3 PCtrl rain (v.)_1 *ʃɪ ́ - - ʃɪ ́ - - ʃɪ ́ sɪ ́ sɪ ́ sɪ ́ sɪ ́ PGw rain (v.)_2 *ʃāpʼ ʃāpʼ ʃāpʼ - - - - - - - - PUd rain (v.)_3 - - - jɔ́sʼ - - - - - - PUd rain (v.)_4 - - - - hɛ́t ̪̓ - - - - - raise (a PCtrl *tù̪n ~ tùn - - tùn - - - tūn tūn tūn tún child)_1 688 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi raise (a PCtrl *t ̪h ipʰ - - - sīpʰ tī̪pʰ t ̪h ì - - - - child)_2 PKmn rat_1 *sʼÍk sʼɪ ́ sʼɪ ́ sʼík sʼíʔ à+t ̪̓íkʰ - tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ tʃʼıg̋ı ̋ PDaOp rat_2 *kʰīd̪í - - - - - kʰīd̪í kʰıʔ̋ kʰıʔ̋ kʰıʔ̋ kʰıʔ̋ PKoUd red (be)_1 *pʼɛ́r - - pʼɛ́l pʼɛ́r pʼɛ́r - - - - - PDaOp red (be)_2 *kʰāpʰā - - - - - kʰāpʰā kʰāpā kʰāpā kʰāpā kʰāpā PGw red (be)_3 *kʼāʃ kʼāʃ kʼāʃ - - - - - - - - PDaOp refuse_1 *k⁽ʰ⁾a - - - - - kâ kʰà kʰà kʰà kʰà PKoUd refuse_2 *ʊʃ - - ʊ̄ʃ úʃ úʃ - - - - - PKmn repair_1 *kʰaɓ áp - áb - - ʊ́bā kʰápʼ kʰápʼ kʰápʼ kʰápʼ PGw repair_2 - át - - - - - - - - PUd repair_3 - - - - tɔ̪̄sɔ̄n - - - - - PUd repair_4 - - - ɲɔ̀ŋ - - - - - - PKmn replace *(n(j)a)gaD nāgát njāgát gàdá gàs gàr gàtá kàrá kàrá kàrá kàrá resemble PCtrl *T̪ál - - sál sál tá̪l hálà tʃál tʃál tʃál tʃál (reflect)_1 resemble PGw zí zí - - - - - - - - (reflect)_2 689 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi right PDaOp (direction)_ *k⁽ʰ⁾ɪ̀t⁽̪ʰ⁾(à) - - - - - kɪ̀t ̪h à kʰɪt̄ kʰɪt̄ kʰɪt̄ kɪt̄ā 1 right GwKo (direction)_ *kʼana - kʼǎná kʼāná - - - - - - - 2 right mɛd̄+p PUd (direction)_ - - - - - - - - - à 3 right à+bǐmpɔ́r PUd (direction)_ - - - - - - - - - ɔ̄s 4 PCtrl rinse face_1 *t ̪̓àm - - tʼà+ɓî - - t ̪̓àm tʼām tʼām tʼām tʼām PUd rinse face_2 *lám+ɛ ̄ - - - lám+ɛ ̄ lám+ʔɛ ̄ - - - - - PGw rinse face_3 *mùsʼ mùsʼ mùsʼ - - - - - - - - rinse tʃākʼʊ́m PKmn *CVkʻUm ʃʊ̀kʊ́m ʃʊ̀kʊ́m zùkʼúm - - ɟákʼʊ́m tʃākʼʊ́má tʃākʼʊ́má sākʼʊ́má mouth_1 á rinse *kʼūcʰūr ~ PUd - - - kʼúʒūr kʼūcʰūr - - - - - mouth_2 kʼūcʼūr PKmn ripen *ís ~ ís ɪś ɪś íʃ ís ís ísá ıs̋á ıs̋á ıs̋á ıt̋ʃá 690 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi rise PCtrl *p⁽ʰ⁾ɪt́(̪á) - - - pí pí pɪt́á̪ pʰɪj́á pʰɪj́á - pʰɪʃ́á (oneself)_1 rise PGw *kʊ̀ɪ ̄ kʊ̀ɪ ̄ kʊ̀ɪ ̄ - - - - - - - - (oneself)_2 rise POp *ɗɪĺá - - - - - - - - ɗɪĺá ɗɪĺá (oneself)_3 rise PKoUd *Kɪ̀-mɪś - - gɪ̀mɪʃ́ cì+mís cì+mís - - - - - (oneself)_4 kʼʊ́sʊ́+d kʼʊ́sʊ́+d PDaOp river_1 *kʼʊ́s(V) - - - - - kʼʊ́sī kʼʊ́sɪ+́zì kʼʊ́sɪ+́sì ʒì ʒì Komo river_2 - - sʼʊ́ʔ - - - - - - - PGw river_3 *pɔ́gɔ́ pɔ́gɔ́ pɔ́gɔ́ - - - - - - - - PUd river_4 *wɔ̀r - - - wɔ̀r wɔ̀r - - - - - roast (burn PDaOp hairs off of *Hil - - - - - híl jǐl jǐl - sīl pig skin)_1 roast PKmn *t ̪̓ ɔ́s tʼɔ̄ʃ tʼɔ̄ʃ tʼɔ́ʃ tʼɔ́s t ̪̓ ɔ́s - - - - - (something) roast or PCtrl *kɛ́s(ɛ́) - - kɛ́ʃ cɛ́s cɛ́s kɛ́s kɛ́sɛ́ kɛ́sɛ́ kɛ́s kɛ́sɛ́ fry_1 691 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi roast or PGw *kɔ́ kɔ́ kɔ́ - - - - - - - - fry_2 PCtrl roll_1 *Dangal - - - ɗáŋgàl ɗáŋgál nàŋgàl làŋgàl làŋgàl làŋgàl làŋgàl GwKo roll_2 *tVndVl tìndìl tìndì tùndàl - - - - - - - PCtrl roof_1 *kʰáɗúm - - kárúm - - kʰáɗúm kʰa̋rűm kʰa̋rűm kʰa̋rűm kʰa̋rűm kʼɔ́ŋgù PUd roof_2 - - - - - - - - - b bàmbɔ̀r+ PUd roof_3 - - - - - - - - - gùɓ bìl+mà+c PCtrl root_1 *bVr(màn) - - bɪ̀l bǐr bǐrmàn - - - - á POp root_2 *kʰɪ̀l - - - - - - kʰɪ̀l kʰɪ̀l kʰɪ̀l kʰɪ̀l PGw root_3 *sʼānsʼ sʼānsʼ sʼānsʼ - - - - - - - - PKmn rope_1 *ʃutʼ ? ʃʊ̄dʊ̀l ʃwītʼìn ʃʊ̀ʔɪ ́ ʃí à+ʃí ʃʊ́l - - - - POp rope_2 *tʼɪŕá - - - - - - tʼɪŕá tʼɪŕá tʼɪŕá tʼɪŕá rough PCtrl *kʼVkʼVr - - - ʃʼɛ́gɛr̄ - kʼàkʼàr kʼɛk̄ʼɛr̄ kʼɛk̄ʼɛr̄ kʼɛk̄ʼɛr̄ kʼákʼár (be)_1 rough Komo - - kákán - - - - - - - (be)_2 692 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi rough PGw *gɔ́kʼɔ̀m gɔ́kʼɔ̀m gɔ́kʼɔ̀m - - - - - - - - (be)_3 rough PUd - - - - kʰɔ̄kɔ̄r - - - - - (be)_4 roughen PKmn (stone for *tɪ̪t́ ̪ sɪt́ sɪt́ tɪt́ títʰ tír tɪ̪t́ ̪h tɪt́ɪ ́ tɪt́ʰ tɪt́ʰ tɪt́ʰ grinding) rub hands together (e.g. to PGw *sʼú sʼú sʼú - - - - - - - - make fire using stick)_1 rub hands together (e.g. to PUd *sʼɛ́kʰ - - - sʼɛ́kʰ t ̪̓ ɛ́kʰ - - - - - make fire using stick)_2 693 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi rub hands together (e.g. to PCtrl *piɟ(i) - - pìz - - pʰìcʰ pīdʒí - pīdʒ - make fire using stick)_3 run (SG)_1, PKmn *gÙs gʊ̀s gʊ̀s gùʃ gùs gùs - - - - - flow, bleed PCtrl run_PL_1 *sɔ - - - sɔ̄ - - sɔ́ sɔ́ sɔ́ sɔ́ PKoUd saliva_1 *maɲa - - mālà màɲà màɲà - - - - - Dana saliva_2 - - - - - mʊ̀lt ̪̓à - - - - PGw saliva_3 *takal tākāl tāgɪ ́ - - - - - - - - POp saliva_4 *dʒì+tʼà - - - - - - dʒì+tʼà dʒìtʼà zì+tʼà ʃìtʼà salt_(from ash of a POp particular *kàrɔ̀ - - - - - - - kàrɔ̀ kàrɔ̀ kàrɔ̀ plant/tree)_ 1 694 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi salt_(made from ash of PKmn a particular *tʼaʃ tʼāʃ tʼāʃ tʼàʃ tʼāʃ tʼāʃ tʼàʃ tʼās - - - plant/tree)_ 2 PDaOp salt_1 *ʃàpɔ́ - - - - - à+ʃàbɔ́ à+sàpɔ́ à+sàpɔ́ à+sàpɔ́ à+ʃàbɔ́ʔ PGw salt_2 - nɔ̌dɔ́ - - - - - - - - ɗɔ̄ŋgɔ̄r à+ɗɔ̄ŋgɔ̄r PUd salt_3 *ɗɔ̄ŋgɔ̄rɔ̄ - - - - - - - - ɔ̄ ɔ̄ PUd sand_1 *sīɓ - - - sīb à+sīɓ - - - - - GwKo sand_2 *pUʃU puʃu puʃu pʊ̂ʃ - - - - - - - PKoUd sap_1 *jìman - - ìmàn+sá - jìmán - - - - - POp sap_2 *ɓūntʃʼú - - - - - - ɓūntʃʼú ɓūntʃʼú ɓūntʃʼú ɓūntʃʼú PGw sap_3 *hɔ̄ʃɔ́ʃ hɔ̄ʃɔ́ʃ hɔ̄ʃɔ́ʃ - - - - - - - - satiated PKoUd *pɔ́kʼ - - pɔ́g pɔ́kʼ - - - - - - (be)_1 satiated PGw *pīs pīs pīs - - - - - - - - (be)_2 satiated PUd - - - - kāpʰ - - - - - (be)_3 695 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi satiated PDaOp *put⁽ʰ⁾ - - - - - pútʰ pǔt pǔt pǔt pǔt (be)_4 PKmn say_1 *k⁽ʰ⁾O kɔ̄ kɔ̄ ʊ́ ɔ́ ɔ́ - - - - - scootch *hɔ̀kʰá PDaOp (move - - - - - hɔ̀kʰá - - wɔ̀kʰá wɔ̀kʰá ~wɔ̀kʰá over)_1 scootch PKoUd (move *ʃɔ̀d - - ʃɔ̀d ʃɔ̄r ʃɔ̄r - - - - - over)_2 scootch POp (move - - - - - - - jɛ̀kʰá - - over)_3 scootch POp (move - - - - - - tʃɛr̄á - - - over)_4 scootch PGw (move *tʊ̄rʊ̀m tʊ̄rʊ̀m tʊ̄lʊ̀m - - - - - - - - over)_5 tʼwānkʼɪt́ʼ tʼwānkʼɪt́ʼ PKmn scorpion *ɗ(w)ankʼI bāɗāgíʔ ɗwākʰ à+ɗākʰ ɗāgí ɗāgı ̋ ɗāgı ̋ ɗāgı ̋ ɗāgı ̋ wānkʼ wānkʼ 696 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PDaOp scratch_1 *gɔ̀n - - - - - gɔ̀n gɔ̀n gɔ̀n gɔ̀n gɔ̀n PKoUd scratch_2 *sʼʊ́sʼ - - sʼʊ́sʼ sʼúsʼ t ̪̓út ̪h - - - - - PGw scratch_3 *kʼʊ̃̄sʼ kʼʊ̃̄sʼ kʼʊ̃̄sʼ - - - - - - - - PKmn seed_1 *D̪E(se) zì zì - - - d̪ɛ̀sɛ̂ dʒɛ̀sɛ ̄ dʒɛ̀sɛ ̄ zɛ̀sɛ ̄ sɛ̀sɛ ̄ PKoUd seed_2 *ʔjamVn - - ʔjāmʊ́n ɛm̄ɛ́n ɛm̄ɛ́n - - - - - PGw sell_1 *dwì dwì dwì - - - - - - - - PUd sell_2 *jɔ́l - - - jɔ́l jɔ́l - - - - - send PKmn *ɗar(a) tʼálà tʼájà ɗàr ɗɛt̄ʰ - ɗɛ̀d - ɗɛr̄ - ɗɛr̄ someone_1 send POp - - - - - - tʃàŋɪ ́ - - - someone_2 send PUd - - - - hāʃ - - - - - someone_3 send POp - - - - - - - - kʰátá - someone_4 PKoUd sesame_1 *bū ~ *pū - - bū pū à+pū - - - - - dʒɪňɪ+́kɔ̄ dʒɪňɪ+́kɔ̄ zɪn̄ɪ+́kɔ̄ POp sesame_2 *dʒɪňɪ+́kɔ̄rɛ́ - - - - - - - rɛ́ rɛ́ rɛ́ PGw sesame_3 *ís ís ís - - - - - - - - PDaOp sesame_4 *tɪ̪̀p⁽ʰ⁾a - - - - - tɪ̪̀pʰà - - - tɪp̄á 697 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi set Komo (heavenly - - wùdɪd̄ - - - - - - - bodies)_1 set PGw (heavenly *kɪs̄ʼ kɪs̄ʼ kɪs̄ʼ - - - - - - - - bodies)_2 set PUd (heavenly *jVl - - - jíl júl - - - - - bodies)_3 set PDaOp (heavenly *T̪út⁽̪ʰ⁾(a) - - - - - sút ̪h ā - - - tʃútʰá bodies)_4 PKmn sew_1 *kɔ́s ɔ́s ɔ́s - - - kɔ́s kɔ́s kɔ́s kɔ́s kɔ́s PKoUd sew_2 *ʃɔ́ - - ʃɔ́ ʃɔ́ ʃɔ́ - - - - - PDaOp shadow_1 *jílɔ́ŋ - - - - - jílɔ́ŋ - - - jıl̋ɔ́ŋ PUd shadow_2 - - - cʰìzàrà - - - - - - PUd shadow_3 - - - - à+kʰāʃīrā - - - - - PGw shadow_4 *Kʊ̄ʃʊ̀n gʊ̄ʃʊ̀n kʊ̄ʃʊ̀n - - - - - - - - PCtrl shadow_5 *D̪ǐtʰ - - zǐt - - - dʒǐtʰ dʒǐtʰ zǐtʰ ʃǐtʰ 698 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi shake PKmn *pid(V) - pǐdí pídá pítʰ pír - - - - - (sth.)_1 shake PKmn *tEŋ(g)(E) tɪḡɪ ̄ tɪḡɪ ̄ - - - tɛ́ŋ - tɪŋ́há tɪŋ́há tɪŋ́há (sth.)_2 shake POp - - - - - - kʰıh̋á - - - (sth.)_3 shake PUd *jikin - - - jígìn jǐkʰīn - - - - - (sth.)_4 shallow PCtrl *tɛ̪́ɗ - - - tɛd̄ tɛɗ̄ tɛ̪́ɗ tɛ́rɛ́ tɛ́r tɛ́r tɛ́r (be)_1 shallow Komo - - bàd - - - - - - - (be)_2 shallow PGw *wárkɪ̀n wárkɪ̀n wálkɪ̀n - - - - - - - - (be)_3 PCtrl sharpen_1 *kíl - - kíl cíl cíl - kıl̋ kıl̋ kıl̋ kıl̋ PGw sharpen_2 *mɛ́ʃɛ́ mɛ́ʃɛ́ mɛ́ʃɛ́ - - - - - - - - Dana sharpen_3 - - - - - síkʼ - - - - PKmn shave *sʼɛ̀d̪ ~ sʼɪ̀d̪ sʼɪl̄ sʼɪ ̄ sʼɛ̀ sʼī cʼɛ ̄ sʼɛ̀d̪ tʃʼɛ ̄ tʃʼɛ ̄ tʃʼɛ ̄ tʃʼɛ ̄ PCtrl sheep_1 *kʰālí - - kālí - - kʰālí kʰālı ̋ kʰālı ̋ kʰālı ̋ kʰālı ̋ PDaOp shiver_1 *ʃij(a) - - - - - ɛ̀s+ʃìjà ɪs̄+sıj̋á ɪs̄+sıj̋á ɪs̄+sıj̋á ɪs̄+sıj̋á 699 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi Komo shiver_2 - - zùg - - - - - - - PGw shiver_3 *bíʃ bíʃ bíʃ - - - - - - - - PUd shiver_4 *dV - - - dɔ̄ dɛ ̌ - - - - - PKmn shoe *pʰákʼá pák pák páʔ pʰāʔ à+pʰāʔ - pʰákʼá pʰákʼ pʰákʼ - PKmn short (be)_1 *KUtʰ gʊ̄t gʊ̄t kùt kūtʰ kūtʰ - - - - - POp short (be)_2 *tūl - - - - - - tūl tūl tūl tūl Dana short (be)_3 - - - - - tùm - - - - PKmn shout_1 *dɔl(ɔ) dɔ́l dɔ́l dɔ́l - - ɔ̀lɔ́ - - - - PCtrl shout_2 *kíw - - - cú cú - kıw̋ kıw̋ kıw̋ kıw̋ PDaOp show_1 *t ̪h ùbá - - - - - t ̪h ùbá tʰūbá tʰūbá tʰūbá tʰūbá PUd show_2 *tʰɔ́r - - - tʰɔ́r tʰɔ́r - - - - - PGw show_3 - dɛp̄ɛ ̄ - - - - - - - - Komo show_4 - - sùn - - - - - - - shut_1, PKmn *mVsʼ(a) mɪ̀sʼ mɪ̀sʼ mʊ̀sʼ músʼ+ɛ ̄ mút ̪̓+ɛ ̄ mìsʼà mʊ̀tʃʼà mʊ̀tʃʼà mʊ̀tʃʼà mʊ̀tʃʼà close eyes PKmn shut_2 *kʰacʼ kǎʃ kǎʃ - kʰáʃʼ kʰácʼ - - - - - *húsʼ(a) ~ PCtrl shut_3 - - wúsʼ wúsʼ - húsʼā hɪt̋ʃʼá hʊ̋tʃʼá hʊ̋tʃʼá hʊ̋tʃʼá wúsʼ(a) side of PDaOp *pājā - - - - - pājā - - zɪ+̄pāj - body, rib_1 700 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi side of Komo - - gɛm̌ - - - - - - - body, rib_2 side of POp *gàŋgārà - - - - - - gàŋgārà gàŋgārà - - body, rib_3 side of PUd *gwǎr - - - gwǎr gwǎr - - - - - body, rib_4 side of PGw *sāmpʼ sāmpʼ sāmpʼ - - - - - - - - body, rib_5 PCtrl sing_1 *kʼɛm - - - cʼɛm̄ - kʼɛ̀m - - - - Komo sing_2 - - ɗɔ̄d - - - - - - - POp sing_3 *ɔ̄lá - - - - - - ɔ̄lá ɔ̄lá ɔ̄lá ɔ̄lá sink PKmn *lili līlí līlí lìlí - - līl līlí līlí līlı ̋ līlí (descend) *wɔpʼ ~ PKmn sip (liquid) hɔ̌bɔ̀sʼhɔ̀pʼ hɔ̌bɔ̀sʼ wɔ́pʼ kʰɔ̄bɔ̄s - hɔ̀pʼ hɔ̄pʼɔ̄ hɔ̄pʼ hɔ̄pʼ hɔ̄pʼ hɔpʼ sit_1, dwell PCtrl (live, *kʼɔ́ - - - kʼɔ́ kʼɔ́ kʼɔ́ʔ kʼɔ́ kʼɔ́ kʼɔ́ kʼɔ́ reside)_PL PGw sit_2 *zàl zàl zɛ̀ - - - - - - - - PKoUd sit_3 *cɔkʰ - - sɔ̀k - cɔ́kʰ - - - - - 701 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PCtrl skin *dìncʼà - - dìnsʼā - - dìncʼà dìntʃʼà dìntʃʼà dìntʃʼà dìntʃʼà skin, hide (of animal), PKmn *gɔnkʼ(ɔʃ) gɔ̀kʼɔ́ʃ gɔ̀kʼɔ́ʃ gɔ̀nkʼí gɔ̀kʰ - kɔ̀gɔ̀ kɔ̀gɔ̀ kɔ̀gɔ̀ kɔ̀gɔ̀ gwàŋgɪ ́ bark of tree_1 skin, hide (of animal), jɪ̀ɗ+mà+ʃ PCtrl *jiɗ - - - jīd à+jǐɗ - - - - bark of úmà tree_2 skin, hide (of animal), PKoUd *jìcʼ - - jìsʼ jìnʒà jìcʼ - - - - - bark of tree_3 skin, hide (of animal), PGw *kɔ̀ʃì kɔ̀ʃì kɔ̀ʃì - - - - - - - - bark of tree_4 skin, hide PUd *kʰūr - - - - kʰūr - - - - - (of animal), 702 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi bark of tree_5 PKmn sky_1 *wUs wús wús - - - - wʊ̀s wʊ̀s wʊ̀s wʊ̀s PKoUd sky_2 *mɪś - - mɪʃ́ mís mís - - - - - slaughter, cross PCtrl *t ̪̓ ɔ̀ɗ - - tʼɔ̀r - - t ̪̓ ɔ̀ɗ - - tʼɔ̄r tʼɔ̄r boundary, chop PKmn slip (v.)_1 *d̪ɛrkʼɛsʼ dɛr̄gɛ́sʼ dɛr̄gɛ́sʼ - dɛ̀rɛ̀s d̪ɛ̀rɛ̀s d̪ɛ́rkʼɛ́s - - - - PCtrl slip (v.)_2 *bɛsɛr - - bɛ̀sɛ̀r - - - bɛšɛ́r bɛšɛ́r bɛšɛ́r bɛšɛ́r slippery, PDaOp smooth *p⁽ʰ⁾ɔ̀t ̪ - - - - - pɔ̀t ̪h pʰɔ̄t pʰɔ̄t pʰɔ̄t pʰɔ̄t (be)_1 slippery, PGw smooth *ʃāwā ʃāwā ʃāwā - - - - - - - - (be)_1 PCtrl slurp_1 *ɟìbí - - zìbí - - ɟìbí - - - tʃibi POp slurp_2 *tɔ̀dɔ́ - - - - - - tɔ̀dɔ́ tɔ̀dɔ́ tɔ̀dɔ́ - PKmn smell (v.) *gàŋ(a) kɛ̃ ̄ kɛ̃ ̄ gàg - - kàŋà kàŋà kàŋà kàŋà kàŋà 703 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi smoke PCtrl *kuɗ - - kūr kúd à+kúɗ kūɗāʔ kűrā kűrā kűrā kűrā (exhaust)_1 smoke PGw *sīnkʼ sīnkʼ sīnkʼ - - - - - - - - (exhaust)_2 smoke out (e.g an PKmn *kʰuʃ kʊ̌ʃ kʊ̌ʃ ūs - - húʃ űs űs űs űʃ animal out of a hole)_1 PDaOp snake_1 *d̪ɔt ̪h ɛn - - - - - d̪ɔ̄t ̪h ɛ̀n dʒɔ̌ dʒɔ̌ zɔ̌ ʃɔ̌ *daʃV ~ PKoUd snake_2 - - dàʃʊ́ tānʒí tāʃá - - - - - ta(n)ʃV PGw snake_3 *bwǎʃà bwǎʃà bwǎʃà - - - - - - - - PKmn sneeze_1 *hat ̪̓is hǎtʼìʃ - - - - hát ̪̓ìs ha̋tʼīs ha̋tʼīs ha̋tʼīs ha̋tʼīs Komo sneeze_2 - - gɛ̀t - - - - - - - PGw sneeze_3 - tɪ̂nsʼɪ̀ - - - - - - - - PUd sneeze_4 *cʼīsân - - - ɟìsân cʼīsân - - - - - PKmn snore *kʰO(r)nOn kɔ̀nɔ̀n kɔ̀nɔ̀n kɔ̀nɔ̀n kʰūnūn àkʰɔ̄rnɛʔ̄ kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n kʰʊ̀nʊ̀n PKmn soak_1 *pʰàcʼ pāsʼ pāsʼ pàsʼ pʰāʃʼ pʰācʼ - - - - - PDaOp soak_2 *pʰɔpʰ - - - - - pʰɔ̀pʰ pʰɔ́pʰɔ́ pʰɔ́pʰ pʰɔ́pʰ pʰɔ́pʰ PKmn soft (be)_1 *kʼáɗ kʼátʼ kʼátʼ kʼátʼ kʼád kʼáɗ - - - - - 704 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PDaOp soft (be)_2 *ɓʊ́t ̪̓ - - - - - ɓʊ́t ̪̓ ɓʊ̋tʼ ɓʊ̋tʼ ɓʊ̋tʼ ɓʊ̋tʼ POp soft (be)_3 *pɪd̄ʒɪ ́ - - - - - - pɪd̄ʒɪ ́ pɪd̄ʒɪ ́ pɪd̄ʒɪ ́ pɪd̄ʒɪ ́ PCtrl son *jàkʼ - - jà jàʔ jàʔ jàkʼ dʒàkʼ dʒàkʼ zàkʼ sàkʼ sorghum, PKmn *ɟana sjànà sjànà zɛ̀nā - - - dʒɛ̀ná - dʒɛ̀ná - millet dʒì+hwā PCtrl soup_1 *hwaja - - wɪ̀ wì à+wìj hɔ̀jàʔ - - - jā PGw soup_2 *mùsʼ mùsʼ mùsʼ - - - - - - - - POp soup_3 *(dʒì+)kārɔ̄ - - - - - - - dʒì+kārɔ̄ zì+kārɔ̄ ʃì+kārɔ̄ tʃʼɪk̄ʼ+dʒ tʃʼɪk̄ʼ+dʒ PCtrl sour (be)_1 *cʼɪkʼ - - sʼɪk̄ʼ - - cʼɪ̀kʼ tʃʼɪk̄ʼ+zɛ̀ tʃʼɪk̄ʼ+sɛ̀ ɛ̀ ɛ̀ PUd sour (be)_3 *pʼàcʰ - - - pʼàcʰ pʼàcʰ - - - - - sow seeds PCtrl (by *sE - - ʃɛ̀ʔ sī sī sɛ̀ sɛ ̄ sɛ ̄ sɛ ̄ - planting)_1 sow seeds PGw (by *àʃ àʃ àʃ - - - - - - - - planting)_2 sow seeds PKmn *jEkʰ jɪ̀ jɪ̀ jɛ̀k jɛ̀kʰ jɛ̀kʰ - - - - - (by 705 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi throwing)_ 1 sow seeds (by PDaOp *gàɟɪ ́ - - - - - gàɟɪ ́ gàdʒɪ ́ gàdʒɪ ́ gàdʒɪ ́ gàdʒɪ ́ throwing)_ 2 PKmn spicy (be)_1 *pʰUl? pʊ̀l pwì - - - pʰîl pʰıl̋ī pʰıl̋ī pʰıl̋ī pʰıl̋ī Komo spicy (be)_2 - - pʼìpʼ - - - - - - - PKmn spit (v.)_1 *t ̪h u tū tū - - - t ̪h úwà tʰűjhá tʰűjhá tʰűjhá tʰűjhá PUd spit (v.)_2 *t ̪̓ ākʰ - - - tʼākʰ t ̪̓ ākʰ - - - - - Komo spit (v.)_3 - - tʼàp - - - - - - - spoil PDaOp (become *nokʼa - - - - - nɔ́kʼà nɔ̌kʼá nɔ̌kʼá nɔ̌kʼá nɔ̌kʼá spoiled)_1 spoil Komo (become - - kʼúʃ - - - - - - - spoiled)_2 spoil PGw (become *kɪʃ̄ kɪʃ̄ kɪʃ̄ - - - - - - - - spoiled)_3 706 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi spoil, stink, PCtrl *pʰú ~ *pʰʊ́ - - pʊ́ - - - pʰűwú pʰűjɪ ́ pʰűrű pʰűrű smell bad_1 spoil, stink, PUd *bùl - - - bùl bùl - - - - - smell bad_2 sprout PCtrl *kʼáɗ - - kʼáʔ kʼáj kʼáɗ kʼájī kʼáʔ kʼáʔ kʼáʔ kʼáʔ (verb) PKmn stab_1 *sɔ́p sɔ́ sɔ́ - - - - sɔ́p - sɔ́p - PKoUd stab_2 *jɪp̄ʼ - - jɪp̄ʼ jípʰ jípʰ - - - - - PUd stab_3 *sūkʼ - - - sūkʼ sūkʼ - - - - - Dana stab_4 - - - - - mɔ̀cʼ - - - - POp stab_5 *pʰɪ ̄ - - - - - - pʰɪ ̄ pʰɪ ̄ pʰɪ ̄ pʰɪ ̄ PKoUd stand_1 *dɔ̀ʃ - - dɔ̀ʃ dɔ̀ʃ dɔ̀ʃ - - - - - PGw stand_2 *zūgū zūgū zūgū - - - - - - - - PKoUd star_1 *kɪl - - kɪľ cúl à+cúl - - - - - PGw star_2 *bīsʼàn bīsʼàn bīsʼàn - - - - - - - - star_3, PCtrl *písāk⁽ʼ⁾ɔ́ - - písākɔ́ - - písākʼɔ́ pıs̋ákʼɔ̄ pıs̋ákʼɔ̄ pıs̋ákʼɔ̄ pıs̋ákʼɔ̄ firefly PKmn steal_1 *kwabOʃ kɔ̄bɔ́ʃ kɔ̄bɔ́ʃ - - - - kǎbús kǎbús kǎbús - PCtrl steal_2 *kʰwal - - kwál kʰwāl kʰwāl kʰwàlà - - - kʰwālā 707 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi stomp, step PKmn *djalIsʼ dɛ̀lísʼ dàlísʼ dɪ̀l dìl dìl - tɪ̀lɪ ́ tɪ̀lɪ ́ tɪ̀lɪ ́ tɪ̀lɪ ́ on_1 stomp, step PUd - - - - - - - - - - on_2 stone or PKoUd *wɔ̀ʃ - - ɔ̀ʃ wɔ̀ʃ wɔ̀ʃ - - - - - rock_1 stone or PGw *pɪd̄ɪ̀N pɪd̄ɪ̀l pīdìn - - - - - - - - rock_2 stone or rock_3, PCtrl *jawa ? - - wàʔí jɔ̀ʔ à+jɔ̀ jàwà dʒàw dʒàw zàw ʃàwà grindstone (bottom) strain PKmn (solids from *d̪ìm - zìŋā zìm zǐm d̪ìm tì̪m tìm tìm tìm tìm liquid)_1 strain PGw (solids from *táʃ táʃ táʃ - - - - - - - - liquid)_2 PGw stretch *bīrītʼ bīrītʼ bīrītʼ - - - - - - - - 708 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi tʼīnīn+ī PCtrl stretch *t ̪̓in - - tʼìn tʼīnīn+īs t ̪̓ìn tʼín tʼín tʼín tʼín s string (v.) PCtrl (e.g. *D̪ɔm - - ɔ̀m wɔ̌m hɔ́m d̪ɔ̀m dʒɔ̀m dʒɔ̀m dʒɔ̀m ʃɔ̀m beads)_1 string (v.) PGw (e.g. *ún ún ún - - - - - - - - beads)+2 strong PKmn *ɓɪśʼ ~ ɓísʼ pʼɪ ́ pʼɪ ́ ɓísʼ ɓísʼ ɓít ̪̓ - - - - - (be)_1 strong Dana - - - - - tʰɛ̀ʔ - - - - (be)_2 strong POp *kāw - - - - - - kāw kāw kāw kāw (be)_3 strong PGw *tʼwālā tʼwālā tʼwājā - - - - - - - - (be)_4 PUd stuff into_1 *ʃúm+īs - - - ʃúm+īs ʃúm+īs - - - - - PGw stuff into_2 *gɪ̀m gɪ̀m gɪ̀m - - - - - - - - PDaOp stuff into_3 *ʃOl(V) - - - - - ʃɔ̀lɔ́ sʊ̄lɪ ́ sʊ̄lɪ ́ sʊ̄lɪ ́ ʃʊ̄lɪ ́ Komo stuff into_4 - - ʃútʼ - - - - - - - 709 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PCtrl stumble_1 *kʰɔr - - kɔ̀r+ɔ̀ʃ - - kʰɔ̀kʰɔ̀r kʰɔ̄kʰɔ̄r kʰɔ̄kʰɔ̄r kʰɔ̄kʰɔ̄r kʰɔ̄kʰɔ̄r PGw stumble_2 *kɛ́ntɛ́ʃ kɛ́tɛ́ʃ kɛ́ntɛ́ʃ - - - - - - - - PUd stumble_3 *gɔ̀l - - - gɔ̀l gɔ̀l - - - - - PCtrl suck_1 *d̪ɔ̀sʼ - - dɔ̀sʼ dɔ̀sʼ d̪ɔ̀t ̪̓ tɔ̪̀sʼ tɔ̀tʃʼɔ̀ tɔ̀tʃʼ tɔ̀tʃʼ tɔ̀tʃʼ PGw suck_2 *sʼúpʼ sʼúpʼ sʼúpʼ - - - - - - - - PUd suck_3 *cʼúm~cʼúm - - - ʃʼúʒùm cʼúɲcʼúm - - - - - PCtrl sun_1 *tá̪j - - tɛ́ tíndí à+tɛ́ntɛ́ tá̪j táj táj táj táj PGw sun_2 *kájá kájá kájá - - - - - - - - swagger, *(gɔ)kwar( PKmn arrogant - kwárá gɔ̀kɔ̀l kār+īs kār+īs kʰɔ̄r kʰɔ̄r kʰɔ̄r kʰɔ̄r kʰɔ̄r a) (be) PKmn swallow_1 *gUsʼ gùsʼ kùsʼ gʊ̀sʼ - - kʊ̄sʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá kʊ̀tʃʼá PUd swallow_2 *lɔ́s - - - lɔ́s lɔ́s - - - - - sweat PKoUd (substance) *ɟVtʰid - - zàtít ɟìtʰíd ɟìtʰí - - - - - _1 sweat PKmn (substance) *jàgàl jàgàl jàgɪ̀ - - - - dʒīkʼāj dʒīkʼāj zīkʼāj ʃīkʼāj _2 710 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi sweat Dana (substance) - - - - - hīpʰ - - - - _3 PUd sweep_1 *d̪ìsʼ - - - dìsʼ d̪ìt ̪̓ - - - - - PCtrl sweep_2 *ʃɛ̀ɗ - - ʃɛ̀r - - ʃɛ̀ɗ - - - - POp sweep_3 *bàr - - - - - - bàr bàr bàr bàr PGw sweep_4 *kɛ ̌ kɛ ̌ kɛ ̌ - - - - - - - - sweet PUd *d̪ɛl - - - zɛl̄ d̪ɛľɛl̄ - - - - - (be)_1 sweet PCtrl *sʼ(j)am - - sʼām - - sʼɛ́m tʃʼɛ́m tʃʼɛ́m tʃʼɛ́m tʃʼɛ́m (be)_2 sweet PGw - mǐzí - - - - - - - - (be)_3 sweet PGw tǎm - - - - - - - - - (be)_4 PDaOp swim_1 *tÉ̪ɗ(á) - - - - - tɛ̪́ɗá tɪ ́ tɪ ́ tɪ ́ tɪ ́ Komo swim_2 - - pàg - - - - - - - PGw swim_3 *sáŋkʼ sáŋkʼ sáŋkʼ - - - - - - - - PUd swim_4 - - - líd - - - - - - PUd swim_5 - - - - káŋ - - - - - 711 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PUd swim_6 - - - mɔ́l - - - - - - tail (of PDaOp *hùl ~ wùl - - - - - hùl wùl wùl wùl wùl animal)_1 tail (of PKoUd *ti̪n - - sìn sīn tī̪n - - - - - animal)_2 tail (of PGw *ʊ̃ŋʊ̃ ʊ̃ŋʊ̃ ʊ̃ŋʊ̃ - - - - - - - - animal)_3 PDaOp tasty (be)_1 *ɗúbá - - - - - ɗúbá ɗűbá ɗűbá ɗűbá ɗűbá Komo tasty (be)_2 - - sʼɪĺá - - - - - - - PGw tasty (be)_3 *íŋgíʃ íŋgíʃ íŋgíʃ - - - - - - - - PUd tasty (be)_4 *kʼūɲ - - - kʼūɲ kʼūɲ - - - - - PDaOp teach_1 *ŋEɟ - - - - - ŋɪ̀ɟ ŋɛ̀tʃ - - ŋitʃ PGw teach_2 *dɔ̀zɔ̀ dɔ̀zɔ̀ dɔ̀zɔ̀ - - - - - - - - tear PKoUd *d̪ɛ̀rɪń - - sɛ̀rɪń zɛř d̪ɛř - - - - - (shred)_1 tear PDaOp *k⁽ʰ⁾ír(a) - - - - - kír kʰírá kʰírá kʰírá kʰírá (shred)_2 tear PKmn *cʼ(w)ɛs̄ʼ sʼwɛ̃ ̌ sʼwɛ̃ ̌ - - cʼɛt̄ ̪̓ - - - - - (shred)_3 PKmn tell_1 *(w)ɔt(V) ɔ̌dɔ́ ɔ̌d - - - - ɔ́tá ɔ́tá ɔ́tá - 712 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PDaOp tell_2 *lɔ̀gɔ́ - - - - - lɔ̀gɔ́ - - - lɔ̀gɔ́ PKoUd tell_3 *wɔ - - wɔ̀ ɔ́ ɔ́ - - - - - tendon, PKmn *ʃʊ(n)kʼ ʃʊ̄nkʼ ʃʊ̄nkʼ ʃʊ̄ ʃúʔ à+ʃúʔ ʃùkʼ sʊ̄kʼ sʊ̄kʼ sʊ̄kʼ ʃūkʼ vein termite kʼɔ́pʰāg PCtrl *gàj - - gàʔ kʼúpʰāɟɛ̀ kʼʊ́kʼàj kàj kàj kàj kàj mound_1 àj termite PGw *úndījá úndījá úndījá - - - - - - - - mound_2 PCtrl termite_1 *sʼɛ̀sʼ(ɛ̀) - - sʼɛ̀sʼ - - sʼɛ̀sʼ tʃʼɛt̄ʃʼɛ ̄ tʃʼɛt̄ʃʼɛ ̄ tʃʼɛt̄ʃʼ tʃʼɛt̄ʃʼ PUd termite_2 *cʼɛk̄ʼ - - - cʼɛk̄ʼ à+cʼɛk̄ʼ - - - - - PGw termite_3 *kʼʊ́mkɛ̃ kʼʊ́mkɛ̃ kʼʊ́mkɛ̃ - - - - - - - - PKmn testicles_1 *lùt ̪̓ ~ lʊ̀t ̪̓ dūtʼ dūtʼ lʊ̀tʼ lùd à+lùɗ - - - - - PDaOp testicles_2 *lɔ̀ŋɔ̀ - - - - - lɔ̀ŋɔ̀ lɔ̀ŋɔ̀ lɔ̀ŋɔ̀ lɔ̀ŋɔ̀ lɔ̀ŋɔ̀ PKoUd thatch *sʼíl(ít⁽ʰ⁾) - - sʼílít sʼíl t ̪̓íl - - - - - PCtrl thigh_1 *ɓácʼ - - ɓásʼ - - ɓácʼ ɓátʃʼ ɓátʃʼ ɓátʃʼ ɓátʃʼ PGw thigh_2 *tjǎʃá tjǎʃá tjǎʃá - - - - - - - - PUd thigh_3 - - - ɗūhúm - - - - - - PUd thigh_4 - - - - - - - - - - PUd thigh_5 - - - - à+wúm - - - - - PCtrl thin (be)_1 *t ̪̓ ɪ(t ̪̓V) - - tʼɪ̀̀tʼá tʼī tʼī - tʼɪt̄ʼɪ ́ tʼɪt̄ʼɪ ́ tʼɪt̄ʼɪ ́ tʼɪt̄ʼɪ ́ 713 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PGw thin (be)_2 *zɪńzɪ̀ zɪńzɪ̀ zɪńzɪ̀ - - - - - - - - thorn, PKmn *k⁽ʰ⁾a(ŋ)k⁽ʼ⁾a kāŋà kāŋà kákā kāʔ kǎkā kʰákʰāʔ kʰákʼā kʰákʼā kʰákʼā kʰákʼā sharp (be) PKmn throat *kʼʊ́s k’úʃ k’úʃ kʼʊ́ʃ kʼús kʼús kʼʊ́s kʼʊ́sʊ́ kʼʊ́s kʼʊ́s kʼʊ́s kʼwāntʼɪ ́̓k kʼwāntʼɪ ́̓k PKmn tick *kʼwànt ̪̓ kʼwàtʼ kʼwātʰ à+kʼwāɗ kʼwàt ̪̓ kʼwātʼ kʼwātʼ kʼwātʼ kʼwātʼ wāntʼ wāntʼ tie PKmn *gì(n)sʼ kɪ̀nsʼ kɪ̀nsʼ gìsʼ - - - - - - - (bundle)_1 tie POp *mɪn̄ - - - - - - mɪn̄ mɪn̄ mɪn̄ mɪn̄ (bundle)_2 tie PUd *dɛǩʰ - - - dɛǩʰ dɛǩʰ - - - - - (bundle)_3 tie Dana - - - - - tɪj́á - - - - (bundle)_4 PCtrl tie (knot)_1 *rukʼ(i) - - - rūkʼ rūh rwí rwì rwì rwì rwì PGw tie (knot)_2 - kʼʊ́p - - - - - - - - Komo tie (knot)_3 - - lùdùm - - - - - - - tie up PDaOp *d̪ɔ̀t ̪h (á) - - - - - d̪ɔ̀t ̪h á dɔ̄tʰ dɔ̄tʰ dɔ̄tʰ dɔ̄tʰ (tether)_1 714 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi tie up PUd *ʃíɗ - - - ʃíd ʃīɗ - - - - - (tether)_2 to prepare PCtrl earth for *kʰɔ́T - - kɔ́t kʰɔ́d kʰɔ́r kʰɔ́tʰ kʰɔ̄tɔ́ kʰɔ̄tɔ́ kʰɔ́t kʰɔ́t farming_1 to stew PCtrl *ɗɛm - - ɗɛ̀m - - ɗɛ̀m ɗɛm̄ ɗɛm̄ ɗɛm̄ ɗɛm̄ (food)_1 to stew PGw - ʃʊ̄kʼʊ̀n - - - - - - - - (food)_2 to stew PGw wūʃ - - - - - - - - - (food)_3 to stew PUd *kʼúpʰ - - - kʼúpʰ kʼúpʰ - - - - - (food)_4 tomorrow_ PKmn *kjana - gɪ̀+kjānā - - - - dʒì+kɛ́n dʒì+kɛ́n zì+kɛ́n ʃì+kɛ́n 1 tomorrow_ Komo - - sʼáŋgà - - - - - - - 2 tomorrow_ Dana - - - - - lɛ̀sʼɛ́nɛ ̄ - - - - 3 715 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi tomorrow_ PUd - - - sʼámɛ́ - - - - - - 4 tomorrow_ PUd - - - - kā+t ̪̓ámɔ̄ - - - - - 5 PCtrl tongue_1 *lEt ̪̓a - - lɛ̀tʼ lɛ̀d à+lɛ̀ɗ lɪ̀t ̪̓á lītʼá lītʼá lītʼá lītʼá PGw tongue_2 *tʼákʼál tʼákʼál tʼákɪ ́ - - - - - - - - PCtrl tonsils_1 *kʼalala - - kʼálálá - - kʼàlálá kǎltʼā kǎltʼā kǎltʼā kǎltʼā PGw tonsils_2 *sjãkʼúʃ sjãkʼúʃ sjãkʼúʃ - - - - - - - - ārgám PUd tonsils_3 - - - - - - - - - +lɛ̀d lɛ̀ɗà+gwá PUd tonsils_4 - - - - - - - - - ɗ+gà PKmn tooth *ʃE ʃī ʃī ʃɛ̀ ʃɛ ̄ ʃɛ ̄ ʃɛ ̄ sɛ ̄ sɛ ̄ sɛ ̄ ʃɛ ̄ tooth PCtrl *sIrE - - ʃɪ̀l - - sírɛ́ sır̋á sır̋á sır̋á sır̋á (canine)_1 tooth PGw *ʃīkānā ʃī+kānā ʃīkānā - - - - - - - - (canine)_2 tooth PUd - - - - ʃɛŋ̄kʼá - - - - - (canine)_3 716 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi toss, throw PKmn away, fall *bit⁽ʰ⁾ pɪ̀t pɪ̀t bìt bǐtʰ bǐtʰ - - - - - over touch_1, PKmn *pàD pāt pāt pàt pātʰ pār - - - - - crawl PDaOp touch_2 *ɔl - - - - - ɔ̄lá ɔ́l ɔ́l ɔ́l ɔ́l trade or PCtrl *hwand̪a - - wādá wǎn+ɛ ̄ wǎn wànd̪á hwān hwān hwān hwān barter_1 trade or PGw *nāmā nāmā nāmā - - - - - - - - barter_2 trample, PKmn *hasʼ - hāsʼɪ̀ hásʼ hásʼ hát ̪̓ hásʼ hátʃʼ hátʃʼ hátʃʼ hátʃʼ ruminate transplant PCtrl (e.g. *tʰát ~ t ̪h át - - tát - - - tʰát tʰát tʰát tʰát plant)_1 transplant PUd (e.g. *cʰúsʼ - - - cʰúsʼ cʰút ̪̓ - - - - - plant)_2 717 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi transplant Dana (e.g. - - - - - t ̪h àgɪʔ́ - - - - plant)_3 transplant PGw (e.g. *pǎʃá pǎʃá pǎʃá - - - - - - - - plant)_4 PKmn tree *cwálá swálá swájá sá ʃwá cwá cá tʃá tʃá tʃá tʃá tree (sp.)(sausag PKmn e *dùmàj - ū+dùmɪ̀ dùmɛ̀ - - dùmáj dùmàj dùmàj dùmàj dùmàj tree_Kigelia africana) PKoUd tree_sp. *jɛ̀sʼ - - jɛ̀sʼ jɛ̀sʼ à+jɛ̀t ̪̓ - - - - - tree_sp. (mahogany, PKmn *k⁽ʰ⁾ɪs kɪśɪ̀ ʊ̄+kɪs̄ kɪs̄ cɛ́s cʰís kɪś kʰɪś kʰɪś kʰɪś kʰɪś Trichilia emetica) PKmn tree_sp._1 *ɓaʃa pʼàʃ - ɓàʃ ɓàʃà - ɓàʃà ɓāsā ɓāsā ɓāsā - trunk (of PCtrl *lut⁽̪ʰ⁾u - - lūd - - lùt ̪h ú - - - - elephant)_1 718 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi trunk (of PGw wūmbū - - - - - - - - - elephant)_2 PDaOp try (test)_1 *tɛ̪̀mɛ̀ - - - - - tɛ̪̀mɛ̀ tɛ̀mɛ̀ - - tɛmɛ GwKo try (test)_2 *tɔ́g tɔ́k tɔ́k tɔ́g - - - - - - - PUd try (test)_3 *ɲɔ̀kʰ - - - ɲɔ̀kʰ ɲɔ̀kʰ - - - - - *tʰánā ~ PCtrl turtle_1 - - bā+tánɪ ̄ - - tʰánā tʰánā tʰánā tʰánā tʰánā tʰánɪ ̄ *kʰwāʃkʰā kʰwāʃk à+kʰwāʃk PUd turtle_2 - - - - - - - - m ʰām ʰām PKmn two *sʊkʼa swījā swījā sʊ̄ sú sú sʊ́kʼà sʊ̄kʼá sʊ̄kʼá sʊ̋kʼá sʊ̋kʼá PKmn ululate_1 *ɪlɪl ~ ilil ɪĺɪ̀l ɪĺɪ̀l ílīl - - ílíl ıl̋ıl̋ ıl̋ıl̋ ıl̋ıl̋ ıl̋ıl̋ PUd ululate_2 - - - jísāɲ - - - - - - PUd ululate_3 - - - - kájī - - - - - untie, take out PCtrl (quickly, *pʰɛɗ̄(ɛ) - - pɛ̀l pʰɛd̄ pʰɛɗ̄ pʰɛ̀rɛ́ - - - - e.g. out of fire) untie, PCtrl *cʼúg - - sʼúk - cʼúkʰ - tʃʼűg tʃʼűg tʃʼűg tʃʼűg undress_1 719 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi untie, PGw *sʼɪl̄ sʼɪl̄ sʼɪʔ̄ - - - - - - - - undress_2 untie, Dana - - - - - gùd̪úʔ - - - - undress_3 PUd urinate_2 *tɛ̪r̄ - - - - tɛ̪r̄ - - - - - PCtrl urinate_3 *T̪isʼ(a) - - - - tī̪t ̪̓ - tʃītʃʼá tʃītʃʼá tʃītʃʼá tʃītʃʼá urine_1, PKmn *dùcʼá dùsʼ tùsʼ dʊ̀sʼ - - tùcʼáʔ - - - - urinate_1 PKoUd urine_2 *d̪(w)àràcʼ - - dàràsʼ dwàràʃ d̪àràcʼ - - - - - PKmn vagina_1 *pít ̪h ~ pɪt́ ̪h pɪt́ pɪt́ pít - - pít ̪h - - - - PUd vagina_2 *kúkʰ - - - kúʔ à+kúkʰ - - - - - POp vagina_3 *kʰɛb̄ɛ ̄ - - - - - - kʰɛb̄ɛ ̄ kʰɛb̄ɛ ̄ kʰɛb̄ɛ ̄ kʰɛb̄ɛ ̄ very, PCtrl *caja - - sà - cājá - sáj sáj sáj sáj many_1 very, Dana - - - - - bwɪ - - - - many_2 PKmn vomit_1 *(pa)jasʼ pājàsʼ pājàsʼ jàʔ ɟǎʔ jǎʔ - - - - - PDaOp vomit_2 *D̪akʼa - - - - - d̪ákʼà dʒākʼá dʒākʼá zākʼá ʃākʼá 720 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi waist, *baɗa ~ PCtrl - - bàr pād pāɗ pàɗáʔ - - pìpārá pìpāráʔ hip_1 paɗa waist, POp *tʃápūm - - - - - - tʃápūm tʃápūm - - hip_2 wake PKmn *ʃukʼ(i) sūgì sūgì ʃùg ʃūkʼ - ʃùg sūg sūg sūg ʃūg (trs.)_1 wake PUd - - - - mɛ̀r+ɛ ̄ - - - - - (trs.)_2 *k⁽ʰ⁾wal ~ PCtrl want_1 - - kʼwàl - - kʰwāj kwāj kʰwáj kʰwáj kʰwáj *kʼwal PGw want_2 *ʃám ʃám ʃám - - - - - - - - PUd want_3 *ŋápʰ - - - ŋápʰ ŋápʰ - - - - - warm PKmn *s(w)am sɔ́m sɔ́m ʃɔ́m sām sām sɔ̀m sɔ̄m sɔ̄m sɔ̄m ʃɔ̄m oneself warm up PKmn *Cisʼ ʃɪʃ̄ ʃɪʃ̄ jíz jísʼ jít ̪h hízá ıs̋á ıs̋á ıs̋á - (sth.) PKmn water *jiɗE ìjáʔ ìjáʔ jǐ wùdíʔ jìɗɛ́ʔ jìʔí dʒì dʒì zì ʃìʔ wet_1, PKmn *jEsI ɪʃ́ɪ̀ ɪʃ́ɪ̀ jɛ̀ʃ jɛ̀s jɛ̀s sìʔ - - - - slippery PKoUd wet_2 *jàsʼ - - jɛ̀sʼ jàsʼ jàt ̪̓ - - - - - 721 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PKmn whistle *cut - ʃút ʃɪt́ - cúwā cʊ́ɪ̀ tʃúwì tʃúwì tʃúwì tʃúwì white PCtrl *pʼátʼá - - pʼátʼá - - pʼátʼá pʼátʼá pʼátʼá pʼátʼá pʼátʼá (be)_1 white PGw - sɛ́lɛ́ - - - - - - - - (be)_2 white PUd *kúʃ - - - kúʃ kúʃ - - - - - (be)_3 PKmn wide (be) *bàj ~ ɓàj pàj pàj ɓàjá ɓàn bɛ̀ - pàj pàj pàj pàj wife_1, PKmn marry, *màʃ - pʼā+màʃ màʃ màʃ màʃ màʃ màs màs màs màs wedding PKoUd wife_2 *(w)àʃ - - wàʃ āʃ āʃ - - - - - PCtrl wind (n.)_1 *kʰab - - - - hāpāpʰ kʰàb kʰāb kʰāb kʰāb kʰāb Komo wind (n.)_2 - - pùt - - - - - - - PGw wind (n.)_3 - kās - - - - - - - - PDaOp woman *ɓāpʰā - - - - - ɓāpʰā ɓāpʰà ɓāpʰā ɓāpʰā ɓāpʰā PKoUd woman *ɓOm(it⁽ʰ⁾) - - ɓāmít ɓûm à+ɓɔ́m - - - - - PCtrl women_1 *Upʰ(U) - - ʊ̄p ūpʰ ūpʰ ɔ̄pʰ ɔ̄pʰɔ̄ ɔ̄pʰɔ̄ ɔ̄pʰ ɔ̄pʰ *mā+kɪḱjāt mā+kɪḱjāt mā+kɪḱjāt PGw women_2 - - - - - - - - à à à 722 NODE Meaning *proto GwHi GwLo Komo UdYa UdCh Dana OpBi OpMo OpPa OpKi PKmn wound_1 *gjama? kāmā kāmā zàmā ɟàmá ɟàmá - - - - - PDaOp wound_2 *ɓɪb́ɪ ̄ - - - - - ɓɪb́ɪ ̄ ɓɪb́ɪ ̄ ɓɪb́ɪ ̄ ɓɪb́ɪ ̄ ɓɪb́ɪ ̄ PKmn wrap *pʊʃ ~ puʃ pʊ̄ʃ pʊ̄ʃ - pūʃ pūʃ púʃá pūsá pūsá pūsá pūsá PCtrl wring_1 *ùm - - ùm ūm ūm - ùmà ùmà - - Dana wring_2 - - - - - nàɟɪ ́ - - - - PGw wring_3 *mʊ̀sʼ mʊ̀sʼ mʊ̀sʼ - - - - - - - - PUd wring_4 *ɲú - - - ɲú - - - - - - POp wring_5 - - - - - - - - tʼɔ́tʃʼ - PCtrl yawn_1 *ham - - hám hám à+hāmɛ ̄ hâm hām hām hām - PGw yawn_2 *hāwā hāwā hāwā - - - - - - - - young PCtrl *pʰāl - - pāl - - pʰāl pʰāl pʰāl pʰāl pʰāl people_1 young PGw *mǎn mǎn mǎn - - - - - - - - people_2 young PUd *dǔrVcʼ - - - dǔrīcʼ dǔrūcʼ - - - - - people_3 723 REFERENCES CITED 13 REFERENCES CITED Ahland, Colleen. 2010. The classification of Gumuz and Koman languages. Paper presented at the International Workshop on Language Isolates in Africa, Laboratoire Dynamique du Langage (DDL), 3–4 December, Lyon. Ahland, Colleen. 2012. A grammar of Northern and Southern Gumuz. Eugene, OR: University of Oregon Ph.D. dissertation. Ahland, Colleen. 2013. The status of Gumuz as a language isolate. Paper presented at the 87th annual meeting of the Linguistic Society of America, 3–6 January, Boston. Ahland, Michael. 2012. A grammar of Northern Mao (Màwés Aasʼè). Eugene, OR: University of Oregon Ph.D. dissertation. Alamin, Suzan, Gertrud Schneider-Blum, & Gerrit Dimmendaal. 2012. Finding your way in Tima. In Angelika Mietzner, & Ulrike Claudi (eds.), Directionality in grammar and discourse: Case studies from Africa, 9-33. Köln: Rüdriger Köppe Verlag. Amare Tsehay. 2013. Documentation and description of tense, aspect and mood in Gwama. Addis Ababa: Addis Ababa University MA thesis. Andersen, Torben. 1987a. The phonemic system of Agar Dinka. Journal of African Languages and Linguistics 9(1). 1-28. Andersen, Torben. 1987b. An outline of Lulubo phonology. Studies in African Linguistics 18(1). Andersen, Torben. 1988. Consonant alternation in the verbal morphology of Päri. Afrika und Übersee 71. 63-113. Andersen, Torben. 1999a. Vowel quality alternation in Mabaan and its Western Nilotic history. Journal of African Languages and Linguistics 20(2). 97-120. Andersen, Torben. 1999b. Vowel harmony and vowel alternation in Mayak (Western Nilotic). Studies in African Linguistics 28(1). 1-30. Andersen, Torben. 2004. Jumjum Phonology. Studies in African Linguistics 33(2). Andersen, Torben. 2007. Kurmuk phonology. Studies in African Linguistics 36. 29-90. 724 Andersen, Torben. 2012a. Verbal directionality and argument alternation in Dinka. In Angelika Mietzner, & Ulrike Claudi (eds.), Directionality in grammar and discourse: Case studies from Africa, 35-53. Köln: Rüdiger Köppe Verlag. Andersen, Torben. 2012b. Spatial roles and verbal directionality in Dinka. Journal of African Languages and Linguistics 33(2). 143-179. Aralova, Natalia. 2015. Vowel harmony in two Even dialects: Production and perception. Amsterdam: LOT. Baković, Eric. 2000. Harmony, dominance and control. New Brunswick, NJ: Rutgers University Ph.D. dissertation. Beam, Mary S., & A. Elizabeth Cridland. 1956. Uduk-English dictionary. Khartoum, Sudan: University of Khartoum. Belkadi, Aicha. 2015. Associated motion with deictic directionals: A comparative overview. SOAS Working Papers in Linguistics 17. 49–76. Belkadi, Aicha. 2016. Associated motion constructions in African languages. Africana Linguistica 22. 43–70. Bender, M. Lionel. 1971. The languages of Ethiopia: A new lexicostatistic classification and some problems of diffusion. Anthropological Linguistics 13(5). 165-288. Bender, M. Lionel. 1975. The Ethiopian Nilo-Saharans. Addis Ababa: Artistic Printers. Bender, M. Lionel. 1983. Proto-Koman phonology and lexicon. Afrika und Übersee 66(2). 259-297. Bender, M. Lionel. 1985. Gumuz, Koman, Mao and Omotic. Studies in African Linguistics (supplement) 9. 19-21. Bender, M. Lionel. 1989. Nilo-Saharan pronouns/demonstratives. In M. Lionel Bender (ed.), Topics in Nilo-Saharan Linguistics, 1-34. Hamburg: Helmut Buske Verlag. Bender, M. Lionel. 1991. Sub-classification of Nilo-Saharan. In M. Lionel Bender (ed.), Proceedings of the Fourth Nilo-Saharan Linguistics Colloquium, 1-36. Hamburg: Helmut Buske Verlag. Bender, M. Lionel. 1994. Comparative Komuz grammar. Afrika und Übersee 77(1). 31-54. Bender, M. Lionel. 1996. The Nilo-Saharan languages: A comparative essay. München: Lincom Europa. Bender, M. Lionel. 2008. Koman. In Siegbert Uhlig (ed.), Encyclopaedia Aethiopica, Vol. 4, 416-418. Wiesbaden: Otto Harrassowitz. 725 Bloomfield, Leonard. 1973 [1933]. Language. London: George Allen & Unwin Ltd. Bourdin, Philippe. 2006. The marking of directional deixis in Somali: How typologically idiosyncratic is it? In Erhard Voeltz (ed.), Studies in African linguistic typology, 13– 41. Amsterdam: John Benjamins. Burns, Samuel J. 1947. Notes toward a grammar of the Koma language. London: Sudan Interior Mission, MS. Bryan, Margaret. 1945. A linguistic no-man’s land. Africa: Journal of the International African Institute 15(4). 188-205. Campbell, Lyle. 1998. Historical linguistics: An introduction. Cambridge, MA: The MIT press. Casali, Roderic F. 2003. [ATR] value asymmetries and underlying vowel inventory structure in Niger-Congo and Nilo-Saharan. Linguistic Typology 7(3). 307-382. Casali, Roderic F. 2008. ATR harmony in African languages. Language and Linguistics Compass 2(3). 496-549. Cipriani, Lidio. 1942. Sui Berta, Coma e Mao dell’ovest etiopico. Rassegna di Studi Etiopici 2(3). 273-276. Clements, George. 2000. Phonology. In Bernd Heine, & Derek Nurse (eds.), African languages: An introduction, 123-160. Cambridge: Cambridge University Press. Comrie, Bernard. 1989. Language universals and linguistic typology: Syntax and morphology. Chicago: University of Chicago press. Corbett, Greville. 1991. Gender. Cambridge: Cambridge University Press. Corbett, Greville. 2000. Number. Cambridge: Cambridge University Press. Corfield, Frank D. 1938. The Koma. Sudan Notes and Records 21(1). 123-165. Davies, H. R. J. 1960. Some tribes of the Ethiopian borderland between the Blue Nile and Sobat rivers. Sudan Notes and Records 41. 21-34. Dimmendaal, Gerrit J. 1988. The lexical reconstruction of Proto-Nilotic: A first reconnaissance. Afrikanistische Arbeitspapiere 16. 5-67 Dimmendaal, Gerrit J. 2003. Locatives as core constituents. In Erin Shay, & Uwe Seibert (eds.), Motion, direction and location in languages, vol. 56, 91-109. Amsterdam: John Benjamins. 726 Dimmendaal, Gerrit J. 2011. Historical linguistics and the comparative study of African languages. Amsterdam: John Benjamins. Dimmendaal, Gerrit J. 2018. On stable and unstable features in Nilo-Saharan. In Helga Schröeder, & Prisca Jerono (eds.), Nilo-Saharan issues and perspectives, 9-24. Köln: Rüdiger Köppe Verlag. Dimmendaal, Gerrit. forthcoming. On word-final voicing contrasts in Nilotic and beyond. In Klaus Beyer et al. (eds.), Linguistics across Africa: Festschrift for Rainer Vossen, Koln: Rüdiger Köppe Verlag. Dimmendaal, Gerrit. to appear. Nilo-Saharan and its limits. In Rainer Vossen, & Gerrit Dimmendaal (eds.), Oxford handbook of African languages, Oxford: Oxford University Press. Ehret, Christopher. 2001. A historical-comparative reconstruction of Nilo-Saharan. Köln: Rüdiger Köppe Verlag. Evans-Pritchard, Edward E. 1932. Ethnological observations in Dar Fung. Sudan Notes and Records 15. 1-61. Federal Democratic Commission Republic of Ethiopia Population Census (FDCREPC). 2008. Summary and statistical report of the 2007 population and housing census. Addis Ababa, Ethiopia. Fernández, Víctor M. 2004. Prospección arqueológica y etnoarqueológica en el Nilo Azul (Sudán y Etiopía). Bienes Culturales: Revista del Instituto del Patrimonio Histórico Español 2. 121-129. Fernández, Víctor M. et al. 2007. A Late Stone Age sequence from west Ethiopia: The sites of K’aaba and Bel K’urk’umu (Assosa, Benishangul-Gumuz regional state). Journal of African Archaeology 5(1). 91-126. Goldberg, Joelle. 2018. Person marking in Gwama. In Helga Schröeder, & Prisca Jerono (eds.), Nilo-Saharan issues and perspectives, 57-72. Köln: Rüdiger Köppe Verlag. Goldberg, Justin. 2018. Obstruent neutralization in Gwama. In Helga Schröeder, & Prisca Jerono (eds.), Nilo-Saharan issues and perspectives, 57-72. Köln: Rüdiger Köppe Verlag. Goldberg, Justin, Joelle Goldberg, & Anne-Christie Hellenthal. 2017. Gwama write-up. Addis Ababa: SIL Ethiopia, MS. González-Ruibal, Alfredo. 2014. An archaeology of resistance: Materiality and time in an African borderland. Rowman & Littlefield. 727 Good, Jeff, & Michael Cysouw. 2013. Languoid, doculect, and glossonym: Formalizing the notion ‘language’. Journal of Language Documentation and Conservation 7. 331-359. Greenberg, Joseph H. 1963. The languages of Africa. Bloomington: Indiana University. Greenberg, Joseph H. 1981. Nilo-Saharan moveable-k as a stage III article:(with a Penutian typological parallel). Journal of African Languages and Linguistics 3(2). 105-112. Guillaume, Antoine. 2016. Associated motion in South America: Typological and areal perspectives. Linguistic Typology 20(1). 81–177. Guillaume, Antoine. 2017. Associated motion: Australia, South America and beyond. Paper presented at the 12th Meeting of the Association for Linguistic Typology, Workshop on Associated Motion, Canberra, Australia. Güldemann, Tom. 2018. Historical linguistics and genealogical language classification in Africa. In Tom Güldemann (ed.), The languages and linguistics of Africa, 58-444. Berlin: De Gruyter Mouton. Heath, Jeffrey. 2005. A grammar of Tamashek (Tuareg of Mali). Berlin: Mouton de Gruyter. Hellenthal, Anne-Christie. 2005. Notes on Mao. Addis Ababa: SIL Ethiopia, MS. Hellenthal, Anne-Christie. 2018. Semantics of directional verb morphology in Gwama. In Helga Schröeder, & Prisca Jerono (eds.), Nilo-Saharan issues and perspectives, 178-192. Köln: Rüdiger Köppe Verlag. Hellenthal, Anne-Christie, & Constance Kutsch-Lojenga. 2011. Phonology/Orthography statement for the Gwama language. Addis Ababa: SIL Ethiopia, MS. Hockett, Charles F. 1958. A course in modern linguistics. New York: Macmillan. Hombert, Jean-Marie. 1978. Consonant types, vowel quality and tone. In Victoria A. Fromkin (ed.), Tone: a linguistic survey, 77-111. New York: Academic Press. Huffman, Ray. 1929. Nuer-English dictionary. Berlin: Dietrich Reimer. Hyman, Larry M. 2001. Tone systems. In Martin Haspelmath et al. (eds.), Language typology and language universals, vol. II, 1367-1380. Berlin: Walter de Gruyter. James, Wendy R. 1968. A crisis in Uduk history. Sudan Notes and Records 49. 17-44. James, Wendy. 1975. Sister-exchange marriage. Scientific American 233(6). 84-94. 728 James, Wendy. 1978. Ephemeral names: The Uduk case. Aspects of Language in the Sudan 5. 114-144. James, Wendy. 1979. ’Kwanim Pa: The making of the Uduk people: An ethnographic study of survival in the Sudan-Ethiopian borderlands. Oxford: Clarendon Press. James, Wendy, Gerd Baumann, & Douglas Hamilton Johnson (eds.). 1996. Juan Maria Schuver’s travels in North-East Africa: 1880-1883. London: The Hakluyt Society. Jedrej, M. Charles. 2004. The Southern Funj of the Sudan as a frontier society, 1820- 1980. Comparative Studies in Society and History 46(4). 709-729. Johnson, Douglas H. 2016. South Sudan: a new history for a new nation. Athens: Ohio University Press. Kiessling, Roland. 2007. Space and reference in Datooga verbal morphosyntax. In Doris Payne, & Mechthild Reh (eds.), Advances in Nilo-Saharan Linguistics: Proceedings of the 8th Nilo-Saharan Linguistics Colloquium, Hamburg, August 22-25, 2001, 123- 142. Köln: Rüdiger Köppe. Kievit, Dirk, & Erika Robertson. 2012. Notes on Gwama grammar. Studies in African Linguistics 41(1). 39-97. Killian, Don. 2014. Uduk-English dictionary. Helsinki: MS. Killian, Don. 2015. Topics in Uduk phonology and morphosyntax. Helsinki: University of Helsinki Ph.D. dissertation. Killian, Don. to appear. Gender in Uduk. In Francesca Di Garbo, & Wälchili (eds.), Grammatical gender and linguistic complexity, 137-160. Berlin: Language Science Press. Kirchner, Robert. 1993. Turkish vowel harmony and disharmony: An Optimality Theoretic account. Paper presented at the Rutgers Optimality Workshop I (ROW- I), October 22, 1993, Rutgers University, New Jersey. Koch, Harold. 1984. The category of ‘associated motion’ in Kaytej. Language in central Australia 1. 23–34. Kutsch Lojenga, Constance, & Manuel A. Otero. 2011a. First Komo lexicon. Addis Ababa: SIL Ethiopia, MS. Kutsch Lojenga, Constance, & Manuel A. Otero. 2011b. Phonology/orthography statement for the Komo language. Addis Ababa: SIL Ethiopia, MS. Lemi Kebebw. 2010. A grammatical description of Opo. Addis Ababa: Addis Ababa University MA thesis. 729 Krell, Amy. 2011. A sociolinguistic survey of the Ganza, Komo, and ”Baruun be Magtole” language groups. SIL Electronic Survey Report 2011-039. Lejean, Guillaume. 1865. Note sur les Fougn et leur idiome. Bulletin de la Société de Géographie de Paris, 5e Sèrie, IX. 238-252. Lewis, Paul M., Gary F. Simons, & Charles Fennig (eds.). 2015. Ethnologue: Languages of the world, 18th edition. Dallas, TX: SIL international. List, Johann-Mattis, Simon Greenhill, Tiago Tresoldi & Robert Forkel. 2018. LingPy: A Python library for historical linguistics. Version 2.6.4. Online: http://lingpy.org. Marno, Ernst. 1874. Reisen im Gebiete des blauen und weissen Nil, im egyptischen Sudan und den angrenzenden Negerländern, in den Jahren 1869 bis 1873. Vienna: C. Gerold. Meckelburg, Alexander. 2016. From “subject to citizen”? History, identity and minority citizenship: The case of the Mao and Komo of Western Ethiopia. Hamburg: Universität Hamburg Ph.D. dissertation. Mellese Gelaneh Alemu. 2017. Documentation and grammatical description of Tapo. Addis Ababa: Addis Ababa University Ph.D. dissertation. Mietzner, Angelika. 2012. Spatial orientation in Nilotic languages and the forces of innovation. In Angelika Mietzner, & Ulrike Claudi (eds.), Directionality in grammar and discourse: Case studies from Africa, 165-175. Köln: Rüdiger Köppe Verlag. Negaso Gidada. 2001. History of the Sayyoo Oromoo of Southwestern Wallaga, Ethiopia: from about 1730 to 1886. Addis Ababa: Mega Printing Enterprise. Olejarczuk, Paul, Manuel A. Otero, & Melissa Baese-Berk. 2019. Acoustic correlates of anticipatory and progressive [ATR] harmony processes in Ethiopian Komo. Journal of Phonetics 74. 18-41. Otero, Manuel A. 2014. Notes from the Komo language discover your grammar workshop. Addis Ababa: SIL Ethiopia. Otero, Manuel A. 2015a. [+ATR] dominant vowel harmony except when it’s not? Evidence from Ethiopian Komo. In Ruth Kramer, Elizabeth C. Zsiga, & One Tlale Boyer (eds.), Selected Proceedings of the 44th Annual Conference on African Linguistics, 212-220. Somerville, MA: Cascadilla Proceedings Project. Otero, Manuel A. 2015b. Dual number in Ethiopian Komo. In Angelika Mietzner, & Anne Storch (eds.), Nilo-Saharan: Models and descriptions, 123-134. Köln: Rüdiger Köppe Verlag. 730 Otero, Manuel A. 2015c. Nominal morphology and ‘topic’ in Ethiopian Komo. In Osamu Hieda (ed.), Information Structure and Nilotic Languages, 19-35. Tokyo: Research Institute for Languages and Cultures of Asia and Africa. Otero, Manuel A. 2016. Reconstructing pronominal morphology in Koman languages. Paper presented at the 47th Annual Conference on African Linguistics, Berkeley, CA. Otero, Manuel A. 2017. (Deictic)direction/associated motion in Ethiopian Komo: A typological perspective. Paper presented at the 12th Conference of the Association for Linguistics Typology, Workshop on Associated Motion, Canberra, Australia. Otero, Manuel A. 2018a. Directional verb morphology in Ethiopian Komo. In Helga Schröeder, & Prisca Jerono (eds.), Nilo-Saharan issues and perspectives, 165-177. Köln: Rüdiger Köppe Verlag. Otero, Manuel A. 2018b. Aspects of Ethiopian Komo (morpho-)phonology. Linguistic Discovery 16(2). 136-178. Otero, Manuel A. accepted. Associated motion, direction and (exchoative) aspect in Ethiopian Komo. Studies in Language. Otero, Manuel A., Sumale Pogi, & Giregna Tesfaye. 2015. Tta Komo, Tta Gozome, Tta Inglizi dɨkshenerɨ [Komo, Amharic, English dictionary]. Addis Ababa: SIL Ethiopia. Payne, Doris L. 2013. The challenge of Maa ‘Away’. In Tim Thornes, Gwendolyn Hyslop, & Joana Jansen (eds.), Functional-historical approaches to explanation: In honor of Scott DeLancey, 259-282. Amsterdam: John Benjamins. Reh, Mechthild. 1996. Anywa language: Description and internal reconstructions. Köln: Rüdiger Köppe Verlag. Ribeiro, Eduardo Rivail. 2000. [ATR] vowel harmony and palatalization in Karajá. Santa Barbara Papers in Linguistics 10. 80-92. Ribeiro, Eduardo Rivail. 2002. Directionality in vowel harmony: The case of Karajá (Macro-Jê). Annual Meeting of the Berkeley Linguistics Society 28(1). 475-485. Seligmann, Brenda Z. 1912. Note on two languages spoken in the Sennar Province of the Anglo-Egyptian Sudan. Zeitschrift für Kolonialsprachen 2. 297-308. Serzisko, Fritz. 1988. On bounding in Ik. In B. Rudzka-Ostyn (ed.), Topics in cognitive linguistics, 429-445. John Benjamins: Amsterdam. Simons, Gary F., & Charles D. Fennig (eds.). 2018. Ethnologue: Languages of the world, 21st edition. Dallas, TX: SIL International. 731 Smolders, Joshua. 2017. A wordlist of T’apo (Opuuo): Version 0.1-28/02/2017. Addis Ababa: SIL International, MS. Smolders, Joshua. forthcoming. Nominal and verbal number in Bilugu Opo. Studies in African Linguistics. MS. Starostin, George (ed.). 2011-2016. The Global Lexicostatistical Database. Moscow: Russian State University for the Humanities, & Santa Fe: Santa Fe Institute. Available online at http://lexstat.tk/databases/, accessed on [April 14, 2018]. Stevenson, Roland C. 1942. Notes toward a grammar of the Uduk language. MS. Storch, Anne. 2014. A Grammar of Luwo: An anthropological approach. Philadelphia: John Benjamins. Tesfaye Negash Bayou. 2015. Documentation and grammatical description of Kwom. Addis Ababa: Addis Ababa University Ph.D. dissertation. Teshome Yehualashet. 2008. A brief phonological description and nomenclature of Kwama/ Komo (a Nilo-Saharan language). Addis Ababa: University of Addis Ababa, MS. Thelwall, Robin. 1983. Twampa phonology. In M Lionel Bender (ed.), Nilo-Saharan language studies, 323-335. East Lansing, MI: African Studies Center, Michigan State University. Trask, Robert Lawrence. 1996. Historical linguistics. Oxford: Oxford University Press. Triulzi, Alessandro. 1981. Salt, gold, and legitimacy: Prelude to the history of a no- man’s land, Belä Shangul, Wallaggä, Ethiopia (ca. 1800-1898). Napoli: Istituto Universitario Orientale. Tucker, Archibald N., & Margaret A. Bryan. 1966. Linguistic analysis: The non-Bantu languages of North-Eastern Africa. London: Oxford University Press. van Silfhout, Marijke. 2013. Opuo: Towards a phonology. Leiden: Leiden University MA Thesis. Voisin, Sylvie. 2010. Les morphèmes -i et -si en wolof. Sciences et Techniques du Langage 7. 25-34. Voisin, Sylvie. 2013. Expressions de trajectoire dans quelques langues atlantiques (groupe Nord). Faits de langues 42(2). 131-152. 732 Wedekind, Charlotte, & Charlotte Wedekind. 2002. Sociolinguistic survey report of the Asosa-Begi-Komosha area: part II. SIL Electronic Survey Reports 55. Wilkins, David P. 1989. Mparntwe Arrernte (Aranda): Studies in the structure and semantics of grammar. Canberra: The Australian National University Ph.D. dissertation. Wilkins, David P. 1991. The semantics, pragmatics and diachronic development of ‘associated motion’ in Mparntwe Arrernte. Buffalo Papers in Linguistics 1. 207– 257. Wolff, H. Ekkehard. 1987. Consonant-tone interference in Chadic and its implications for a theory of tonogenesis in Afroasiatic. In Daniel Barreteau (ed.), Langues et cultures dans le bassin du lac Tchad, 193-216. Paris: ORSTOM. Zelealem Leyew. 2005. Gwama, a little-known endangered language of Ethiopia: A sketch of its grammar and lexicon. Addis Ababa: Addis Ababa University, MS. 733