Trans-Himalayan Linguistics: Historical and Descriptive Linguistics of the Himalayan Area 9783110310832, 9783110310740

The Himalaya and surrounding regions are amongst the world's most linguistically diverse places. Of an estimated 60

275 95 8MB

English Pages 452 Year 2013

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Introduction
Trans-Himalayan
Creolization in the Divergence of the Tibeto-Burman Languages
Rethinking Sino-Tibetan phylogeny from the perspective of North East Indian languages
The Tibetic languages and their classification
Internal diversity in the Tamangic lexicon
A preliminary reconstruction of East Bodish
Burushaski kinship terminology of Indo-European origin
Subject and object agreement in Shumcho
The tone patterns of numeral-plus-classifier phrases in Yongning Na: a synchronic description and analysis
Rengmitca: the most endangered Kuki-Chin language of Bangladesh
Initial Grammatical Sketch of Tilung
Tshangla Phonology and a Standard Tshangla Orthography
Index
Recommend Papers

Trans-Himalayan Linguistics: Historical and Descriptive Linguistics of the Himalayan Area
 9783110310832, 9783110310740

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Thomas Owen-Smith, Nathan W. Hill (Eds.) Trans-Himalayan Linguistics

Trends in Linguistics Studies and Monographs

Editor Volker Gast Editorial Board Walter Bisang Jan Terje Faarlund Hans Henrich Hock Natalia Levshina Heiko Narrog Matthias Schlesewsky Amir Zeldes Niina Ning Zhang Editor responsible for this volume Walter Bisang

Volume 266

Trans-Himalayan Linguistics edited by Thomas Owen-Smith and Nathan W. Hill

ISBN 978-3-11-031074-0 e-ISBN 978-3-11-031083-2 ISSN 1861-4302 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.dnb.de. © 2014 Walter de Gruyter GmbH & Co. KG, Berlin/Boston Typesetting: PTP-Berlin Protago-TEX-Production GmbH, Berlin Printing: Hubert & Co. GmbH & Co. KG, Göttingen ♾ Printed on acid-free paper Printed in Germany www.degruyter.com

List of contributors Roger Blench Kay Williamson Educational ­Foundation 8 Guest Road, Cambridge, CB1 2AL, UK [email protected]

Isao Honda Department of English and ­Communication Nagoya College, Japan [email protected]

Tim Bodt Institut für Sprachwissenschaft Universität Bern, Switzerland [email protected]

Christian Huber Phonogrammarchiv Austrian Academy of Sciences Vienna, Austria [email protected]

Ilija Čašule Department of Linguistics Macquarie University, Sydney, ­Australia [email protected] Scott DeLancey Department of Linguistics University of Oregon, USA [email protected] George van Driem Institut für Sprachwissenschaft Universität Bern, Switzerland [email protected] Nathan Hill Department of Linguistics School of Oriental and African Studies (SOAS), University of London, UK [email protected]

Gwendolyn Hyslop ANU College of Asia and the Pacific, Australia [email protected] Alexis Michaud International Research Institute MICA (HUST-CNRS-Grenoble INP), Hanoi, Vietnam, and LACITO-CNRS, Paris, France [email protected] Jean Robert Opgenort www.opgenort.nl [email protected] Thomas Owen-Smith Department of Linguistics School of Oriental and African Studies (SOAS), University of London, UK [email protected]

vi   

   List of contributors

David Peterson Program in Linguistics and Cognitive Science Dartmouth College, USA [email protected] Mark W. Post Institut für Sprachwissenschaft Universität Bern, Switzerland ­[email protected]

Nicolas Tournadre Université de Provence, and CNRS (LACITO), France [email protected]

Content Thomas Owen-Smith and Nathan W. Hill Introduction  1 George van Driem Trans-Himalayan  11 Scott DeLancey Creolization in the Divergence of the Tibeto-Burman Languages  41 Roger Blench and Mark W. Post Rethinking Sino-Tibetan phylogeny from the perspective of North East Indian languages  71 Nicolas Tournadre The Tibetic languages and their classification  105 Isao Honda Internal diversity in the Tamangic lexicon  131 Gwendolyn Hyslop A preliminary reconstruction of East Bodish  155 Ilija Čašule Burushaski kinship terminology of ­Indo-European origin  181 Christian Huber Subject and object agreement in Shumcho  221 Alexis Michaud The tone patterns of numeral-plus-classifier phrases in Yongning Na: a synchronic description and analysis  275 David A. Peterson Rengmitca: the most endangered Kuki-Chin language of Bangladesh  313 Jean Robert Opgenort Initial Grammatical Sketch of Tilung  329

viii   

   Content

Tim Bodt Tshangla Phonology and a Standard Tshangla Orthography  393 Index  437

Thomas Owen-Smith and Nathan W. Hill

Introduction

The Himalayan Region is home to around 600 languages (Turin 2007: foreword), making it one of the most linguistically diverse regions of the world.1 But it is not only this diversity which constitutes the region’s great importance for linguistics. The typological profiles of Himalayan languages are also striking, with many displaying features which are rare in Eurasia and in the Old World in general. A large number of language families are present in the region, and complex histories of ethnic mixing and language contact have created a contemporary linguistic landscape which is as challenging as it is exhilarating to research. The mountains, though not an absolute barrier, hinder transport and communications. Even today travel in the hills of Nepal or Northeast India is more difficult than on the Indo-Gangetic plain, or for that matter the Tibetan Plateau (see footnote 2). This rugged topography has throughout history weakened the assimilating and centralizing tendencies of large states, and allowed for the existence of small, isolated communities living with a high degree of economic selfsufficiency and political autonomy. Such a pattern is reflected in other mountains regions of the world (Braudel [1949] 1976: 30–43; Gellner 1969: 1–2; Scott 2001: 13–22 inter alia), and not surprisingly fosters linguistic and cultural diversity. Bàrtoli points out that mountainous areas tend to preserve archaic forms of speech: nella storia del linguaggio…le isole sono di norma più conservative che i continenti…, e più le montagne che le pianure e le marine…, e più certe aree laterali che le aree di mezzo…, e più i centri minori che i maggiori [in the history of language…islands are generally more conservative than continents…, mountains more so than plains or the seaside…, certain marginal areas more so than central areas…, and minor centres more so than major ones] (Bàrtoli 1925: 4)

Historical linguistics handbooks refer to these isolated enclaves of linguistic conservatism as “relic areas” (Campbell 2004: 216; Hock 1991: 440; Dimmendaal 2011: 159). Nichols uses the term “accretion zone,” and notes that rather than

1 Our definition of the Himalayan Region is a wide one, including the Himalaya proper, the Tibetan Plateau, and the various mountain ranges which flare off from these in all directions, including the Hindu Kush, Pamir and Tian Shan, the highlands of Sichuan and Yunnan, and the hilly tracts of Northeast India, Bangladesh and Burma.

2   

   Thomas Owen-Smith and Nathan W. Hill

simply conservatism in such areas, the influx of languages from other places also affects their character: An accretion zone…is an area where genetic and structural diversity of languages are high and increase over time through immigration. Examples are the Caucasus, the Himalayas, the Ethiopian highlands and the northern Rift Valley, California, the Pacific Northwest of North America, Amazonia, northern Australia, and of course New Guinea… Accretion zones generally contain representatives of major stocks in the vicinity as well as some languages with no outside kin. (Nichols 1997: 369)

Mountains differ from lowlands in that lowland areas have historically been prone to large-scale spreads of particular languages or groups of languages (on a family or group scale examples include Iranian, Turkic, Indo-Aryan, Sinitic; on a language scale examples include Mandarin Chinese, Persian, Hindi-Urdu) which submerge earlier languages spoken in the same space; while when new languages move into the mountains they tend to add to the linguistic diversity of the region rather than obliterate it (see Nichols 1992).2 Linguistic diversity is manifested in the Himalayan Region in a number of ways. Firstly there is the density of discrete languages spoken there. There is then the diversity of linguistic stocks represented, which totals roughly nine depending on where the region’s boundary is drawn: Sino-Tibetan (or “Trans-Himalayan” – see below), Indo-European, Austroasiatic, Hmong-Mien, Tai, Turkic and perhaps Mongolic, as well as the presumed isolates Burushaski and Kusunda. Within several of these stocks, multiple branches are represented in the Himalayan Region. For instance, in Indo-European all three subbranches of IndoIranian are present: both Eastern (Pashto, the Pamir languages) and Western Iranian languages (Dari, Tajik, Parachi, Ormuri) are spoken, as well as several branches of Indo-Aryan,3 and the Nuristani languages, which are proposed to stand at the same phylogenetic level as Iranian and Indo-Aryan (see Nelson 1986). Worthy of special mention is the heterogeneous Dardic group spoken at the far Northwestern reach of the Himalaya. The Dardic languages are generally considered Indo-Aryan but have not participated in a number of developments 2 By this definition, the Tibetan Plateau may share more features with lowland areas and prototypical spread zones: although situated at a high altitude it is relatively flat and easy to traverse, and has been dominated throughout recorded history by the spread of a single language, Old Tibetan. However, as Zeisler (2005) and Tournadre (this volume) point out, the degree of diversity between the modern varieties collectively referred to as “Tibetan” is more akin to that of a family of languages than dialects of one language. The languages of the Tibetan Plateau also display some of the unusual linguistic features which appear to delineate the Himalayan Region as a typological enclave. 3 The details of subgrouping within Indo-Aryan itself are disputed (see Masica 1991: 446–463).

Introduction   

   3

which are general throughout the rest of Indo-Aryan, preserving some extremely conservative features while innovating others which are quite untypical of IndoAryan and Indo-European in general (see Bashir 2003).4 The Himalaya’s diversity is even better exemplified by the profusion of SinoTibetan/Trans-Himalayan language groups spoken on both sides of the range. In contrast to the Indo-Iranian languages, which moved into the area from elsewhere (originally from Central Asia and later from the Indo-Gangetic plain), the preponderance of high level Sino-Tibetan/Trans-Himalayan taxa within the Himalayan Region may indicate that it has always been the centre of gravity of this family (see van Driem this volume; Blench and Post this volume). Aside from their phylogenetic diversity, Himalayan languages display typological features rarely encountered in the Old World. From a survey of some 350 languages from around the world, Bickel and Nichols (2003) propose that the Himalaya and the Caucasus can be considered “typological enclaves,” due to the high frequency in these two regions (both mountainous relic areas/accretion zones) of certain features which are otherwise rare in Eurasia, and generally in most parts of the world. The unusual features which characterise the Himalayan enclave in particular are very high levels of morphological synthesis, polypersonal agreement on verbs,5 complex systems of evidentiality,6 bipartite verb stems, radical double marking (including agreement for more than one argument on the predicate and as well as case marking on the arguments themselves, and marking of possession on both the possessor and possessee), and multiple classes for possessive marking in the noun phrase. The multitude of languages in the Himalayan region and their typological diversity therefore more than warrant the steady stream of research on these languages, of which the current volume presents a recent selection. The volume arises from papers given and topics discussed at the 16th Himalayan Languages Symposium, held in September 2010 at the School of Oriental and African Studies, University of London; and it includes chapters on historical linguistics, and new descriptive work and original data on under-researched Himalayan languages. The majority of papers in the volume deal primarily with Sino-Tibetan/TransHimalayan languages. Since the earliest days of research on languages of this 4 Strand (1973) and Bashir (2003) indicate that the group “Dardic” is heuristic rather than genetic, arguing that the Dardic languages are highly diverse, and do not display shared innovations which delineate them as a subgroup. 5 Defined as “unconditionally obligatory verb agreement with more than one argument” in Bickel and Nichols (2003: 10). 6 Thse include patterns which have been analysed as conjunct/disjunct in some languages (see e.g. Hale 1980; Watters 2006a), although questions have also been raised about the applicability of the terms “conjunct” and “disjunct” to certain languages (see e.g. Tournadre 2008).

4   

   Thomas Owen-Smith and Nathan W. Hill

family, different classifications have vied for adherents. Numerous scholars have proposed classification and subgrouping schemes, including the early work of Hodgson ([1828] 1972, [1848] 1972) and Konow (1909), and more elaborate proposals from the mid-20th century onwards (Shafer 1955; DeLancey 1987; Bradley 1997, van Driem 2001 inter alia). Few of these taxonomic proposals have weathered the years free of controversy. The disputed names for the family, Indo-Chinese, SinoTibetan, Tibeto-Burman, Trans-Himalayan – all of which appear at some point in this volume – are one indicator of the controversial status of the Stammbaum. As van Driem relates in the present volume, “Indo-Chinese” originally grouped together all languages of Asia and Oceania. Over the decades various stocks were respectively distinguished and then removed from this presumed genetic unity. The term “Sino-Tibetan,” which has a close association with the University of California at Berkeley, inherited its reference from “Indo-Chinese.” Although “Sino-Tibetan” has varied in meaning, it is now normally understood to reflect a family with two primary branches, “Sinitic” and “Tibeto-Burman,” i.e. a binary split at the highest level between Chinese and everything else (see Matisoff 2003: 5–6). But as no decisive evidence has been put forward to suggest that all non-Sinitic languages of the family share a common innovation (or indeed that Sinitic languages share a common innovation to set them apart from all the rest of the family), some scholars are now searching for terms which reflect an agnostic approach to subgrouping. Although von Klaproth coined “Tibeto-Burman” in 1823 to serve as a phylogenetically neutral term for the family of which Chinese, Tibetan, and Burmese are members, the now common use of the same term for all non-Sinitic members of the family makes it difficult to use it with Klaproth’s original designation (i.e. the whole family) without causing confusion. Consequently, other terms have been proposed, including “Sino-Tibeto-Burman” (see Tournadre this volume; Matisoff 2011: i), and “Trans-Himalayan” (see van Driem 2007, this volume). The term “Sino-Tibeto-Burman” acknowledges the three most prominent literate civilizations, although as Blench and Post (this volume) point out, most languages in the family do not have literate histories, and there is no a priori reason to think that these three languages are of any special significance in the structure of the family or in the reconstruction of its ancestor. The term “Trans-Himalayan” has the advantage of being more geographically neutral and less culturally biased, and as a geographic term it parallels the names of other language families such as Indo-European and Austroasiatic. To instantiate the advantages of this neutral geographic term, Trans-Himalayan is an apt title for this volume. The typological divergence of Trans-Himalayan languages leads not only to disagreements over nomenclature and subgrouping, but also to differences of opinion about the morphosyntactic profile to be reconstructed for the proto-

Introduction   

   5

language. Whereas Rgyalrongic and Kiranti languages show a high degree of morphological complexity (e.g. synthetic verb inflection including agreement for more than one argument, double marking of possession on possessor and possessee), Sinitic and Lolo-Burmese languages are more analytic in structure (e.g. zero agreement on verbs and simple juxtaposition of nominal elements in a possessive clause in some languages). Some scholars, most prominently LaPolla (1989, 1992, 1994), argue that verbal agreement systems are independent innovations in various branches of the family, arising from the grammaticalization of erstwhile independent pronominal forms. In contrast, van Driem (1993), DeLancey (1989, 2010a) and Jacques (2012) propose a highly synthetic proto-language which has experienced varying degrees of phonological erosion in different subgroups. Fully accounting for the extremely divergent typologies of languages across the family naturally presents difficulties for both sides of this debate. In the current volume DeLancey proposes that the simpler typology of the analytic languages can be explained by the fact that these languages underwent a process of creolization when they emerged as lingua francas. This proposal could account for the correlation between analytic typology and the settings of urbanization and state centralization associated with languages such as Chinese, Burmese, Tibetan, Newar, and Meithei. Work on a Stammbaum for Trans-Himalayan also stands in a certain tension to what appear to be extremely complex histories of migration, contact, acculturation, and ethnic and linguistic shift (see e.g. Thomason and Kaufmann 1988; Hickey 2010) across the regions where Trans-Himalayan languages are spoken. The survival of remnant language isolates like Kusunda (see Watters 2006b) reminds us that there may once have been many more linguistic stocks in the Himalayan Region which were ultimately submerged by acculturation to the larger families. Various authors have argued that large areas where Trans-Himalayan languages are now spoken are the result of intrusive migrations, or acculturation and shift from now lost languages to Trans-Himalayan forms of speech (e.g. Grierson 1909; LaPolla 2001, 2009; Zeisler 2005, 2009; DeLancey 2010b, 2011 inter alia), but it has not yet been possible to incorporate these observations systematically into the theore­tical models about the linguistic history of the area. In this volume Blench and Post consider subsistence and environmental vocabulary from languages of Northeast India, and reiterate their position earlier stated in Post and Blench (2011) that the high levels of divergence amongst the languages of Arunachal Pradesh and surrounding areas have not been taken into account in any of the classification schemes for Trans-Himalayan. They further argue that some of these languages are so aberrant that there is a distinct possibility they represent the outcome of a situation in which speakers of unknown language stocks or isolates acquired a TransHimalayan lexicon through varying degrees of contact and acculturation.

6   

   Thomas Owen-Smith and Nathan W. Hill

Terminological and subgrouping controversies also beset the Trans-Himalayan subbranches. Drawing from the Tibetan autonym Bod, Shafer posits four “Bodish” subgroups for languages spoken in the region of the Tibetan Plateau: West, Central, Southern, and East Bodish (1966: 78–123), of which he derives only Central Bodish and Southern Bodish from Old Tibetan (1966: 87). Although some researchers still see West Bodish as a meaningful grouping (see Bielmeier 2004), it is becoming increasingly clear that all of these forms of speech except for the East Bodish languages (see Hyslop this volume) derive directly from Old Tibetan (see Hill 2010). Many authors refer to the languages descending from Old Tibetan as “Tibetan dialects” (e.g. Denwood 1999: 21–36), but Tournadre (this volume) presents a forceful argument that these languages should be called “Tibetic” languages to acknowledge the degree of their divergence from one another. The key to a successful analysis of subgrouping is the identification of shared innovations. As LaPolla (2001: 245) notes, many of the proposed subgrouping schemes for Trans-Himalayan do not give the reasons for their groupings. In this process, the distinction of inherited features from borrowed features is paramount. Three papers in the current volume aim to sharpen the distinction between inherited and borrowed vocabulary in the languages they study. Honda examines patterns of lexical diversity in Tamangic languages, looking to differentiate inherited Tamangic words from later borrowings from Tibetan. Hyslop attempts a preliminary comparative reconstruction of the East Bodish languages. With this reconstruction in hand she is able to show that several grammatical morphemes which Kurtöp shares with Tibetan are not cognate, as might first appear to be the case, but as they are not reconstructible to proto-East Bodish they must be borrowings from Tibetan. This approach is a step forward from earlier work that compares Tibetan directly to words in East Bodish languages. Pursuing a line of research which he began in 1998, Čašule presents new evidence of an Indo-European affiliation for Burushaski; on this occasion he focuses on kinship terminology. Although many Indo-Europeanists will regard his contribution with scepticism, his engagement with the methods and literature of this field, the wellspring and the touchstone for the methodology of historical linguistics, reminds us that our knowledge of Trans-Himalayan historical phonology, in which regular correspondences across what is taken to be cognate vocabulary in the modern languages have still not been established, has far to go. Historical linguistics, as well as areal linguistics and typology, are absolutely dependent on descriptive work on contemporary languages. Unfortunately, few Himalayan languages have been researched in depth and many remain virtually undocumented. This situation is partly due to the remoteness of many mountain localities which makes fieldwork difficult, but also to the periodic imposition of travel restrictions to certain areas. Four contributions to the current volume

Introduction   

   7

provide fresh data from in situ fieldwork on under-described languages. Huber presents the strategies of person marking in Shumcho, where a fairly complicated verbal agreement system indexes participants of main and logophoric clauses. Michaud examines the tonal patterns in numeral plus classifier phrases in Yongning Na, whose complexity presents a number of challenges for a tonological analysis. Peterson describes the phonology and various aspects of the morphology of the highly endangered language Rengmitca, which with its conservative segmental phonology could contribute greatly to the reconstruction of Proto-Kuki-Chin, or indeed Trans-Himalayan itself. Opgenort provides an initial grammatical sketch of another highly endangered language, Tilung, which holds a unique position within the Kiranti group, though now appears to be falling into obsolescence due to the spread of Nepali. All four of these languages are small and vulnerable. In the Himalayan Region as in other parts of the world, modernization is eroding traditional economies and lifestyles, and ancestral cultures and minority languages are (understandably) being fast abandoned for national languages such as Hindi, Chinese and Nepali which facilitate employment in the mainstream of the national economy. In such a context, the work of linguistic description and documentation often goes hand in hand with efforts in language development and education working with the community in which a fieldworker does their research. The long-term survival of a language across multiple domains of use in the modern world essentially necessitates that the language be written. Consequently, orthography development is an area in which a linguist can assist in promoting a language’s long-term viability. Bodt discusses issues in the development of Roman and ’Ucen orthographies for Tshangla, in view of the phonological variation across the language’s dialects, and the influence of literacy in a more prestigious language, Tibetan. Despite such efforts, under the current rate of development, and rapid linguistic and cultural loss, it is highly probable that a great many languages used in the Himalayan Region today will no longer exist as spoken languages within a few generations. As many voices have already warned, thorough and sensitive research on the minority languages of the Himalayan Region is a matter of urgency.

8   

   Thomas Owen-Smith and Nathan W. Hill

References Bàrtoli, Matteo. 1925. Introduzione alla neolinguistica. Geneva: L. S. Olschki. Bashir, Elena. 2003. Dardic. In George Cardona and Dhanesh Jain (eds.). The Indo-Aryan Languages, 818–9. London: Routledge. Bradley, David. 1997. Tibeto-Burman languages and classification. In David Bradley (ed.). Tibeto-Burman Languages of the Himalayas (Papers in Southeast Asian Linguistics No. 14), 1–71. Canberra: Department of Linguistics, The Australian National University. Braudel, Fernand. [1949] 1976. The Mediterranean and the Mediterranean World in the Age of Philip II, Vol 1. New York: Harper & Row, Publishers. Bickel, Balthasar and Johanna Nichols. 2003. Typological enclaves. Paper presented at the 5th Biannual Conference of the Association for Linguistic Typology, Cagliari, September 18, 2003. Bielmeier, Roland. 2004. Shafer’s proto-West Bodish hypothesis and the formation of the Tibetan verb paradigms. In Anju Saxena (ed.). Himalayan Languages Past and Present, 395–412. Berlin: Mouton de Gruyter. Campbell, Lyle. 2004. Historical Linguistics: An Introduction. Edinburgh: Edinburgh University Press. DeLancey, Scott. 1987. The Sino-Tibetan Languages. In Bernard Comrie (ed.). The World’s Major Languages, 799–810. New York: Oxford University Press. DeLancey, Scott. 1989. Verb agreement in Proto-Tibeto-Burman. Bulletin of the School of Oriental and African Studies 52 (2): 315–33. DeLancey, Scott. 2010a. Towards a history of verb agreement in Tibeto-Burman. Himalayan Linguistics 9 (1): 1–39. DeLancey, Scott. 2010b. Language replacement and the spread of Tibeto-Burman. Journal of the Southeast Asian Linguistics Society 3(1): 40–55. DeLancey, Scott. 2011. The origins of Sinitic. In Zhuo Jing-Schmidt (ed.). Proceedings of the 23rd North American Conference on Chinese Lingusitics (NACCL-23) 2011, Vol. 1, 51–64. Eugene: University of Oregon. Denwood, Philip. 1999. Tibetan. Amsterdam: John Benjamins. Dimmendaal, Gerrit J. 2011. Historical Linguistics and the Comparative Study of African Languages. Amsterdam: John Benjamin’s Publishing Company. van Driem, George. 1993. The Proto-Tibeto-Burman verbal agreement system. Bulletin of the School of Oriental and African Studies 56 (2): 292–334. van Driem, George. 2001. Languages of the Himalayas: An Ethnolinguistic Handbook of the Greater Himalayan Region, Containing an Introduction to the Symbiotic Theory of Language. Leiden: Brill. van Driem, George. 2007. The diversity of the Tibeto-Burman language family and the linguistic ancestry of Chinese. Bulletin of Chinese Linguistics 1 (2): 211–270. Gellner, Ernest. 1969. Saints of the Atlas. London: Weidenfeld and Nicolson. George A. Grierson (ed.). 1909. Linguistic Survey of India. Volume I, Part 1: Introductory. Calcutta: Office of the Superintendent of Government Printing. Hale, Austin. 1980. Person markers: Finite conjunct and disjunct verb forms in Newari. In S.A. Wurm (ed.). Papers in South-East Asian linguistics 7: 95–106. Canberra: Department of Linguistics, The Australian National University. Hickey, Raymond (ed.). 2010. The Handbook of Language Contact. Singapore: Wiley-Blackwell.

Introduction   

   9

Hill, Nathan W. 2010. An overview of Old Tibetan synchronic phonology. Transactions of the Philological Society 108 (2): 110–125. Hock, Hans Henrich. 1991. Principles of historical linguistics. Berlin: Mouton de Gruyter. Hodgson, Bryan Houghton. [1828] 1972. Notices of the Languages, Literature and Religion of Nepál and Tibet. In Bryan Houghton Hodgson. Essays on Nepal and Tibet, Part I, 1–35. Amsterdam: Philo Press. Hodgson, Bryan Houghton. [1848] 1972. On the Aborigines of the Himálaya. In Bryan Houghton Hodgson, Essays on Nepal and Tibet, Part II, 29–37. Amsterdam: Philo Press. Jacques, Guillaume. 2012. Agreement morphology: the case of Rgyalrongic and Kiranti. Language and Linguistics 13: 83–116. Konow, Sten. 1909. The Tibeto-Burman family. In George A. Grierson (ed.), Linguistic survey of India. Volume III, 1, Tibeto-Burman family. Calcutta: Office of the Superintendent of Government Printing. LaPolla, Randy J. 1989. Verb agreement, head-marking vs. dependent-marking, and the ‘deconstruction’ of Tibeto-Burman morpho-syntax. In Kira Hall et al. (eds.). Proceedings of the fifteenth annual meeting of the Berkeley Linguistics Society, 356–67. Berkeley: Berkeley Linguistics Society. LaPolla, Randy J. 1992. On the Dating and Nature of Verb Agreement in Tibeto Burman. Bulletin of the School of Oriental and African Studies 55 (2): 298–315. LaPolla, Randy J. 1994. Parallel Grammaticalizations in Tibeto-Burman: Evidence of Sapir’s ‘Drift.’ Linguistics of the Tibeto-Burman Area 17 (1): 61–80. LaPolla, Randy J. 2001. The Role of Migration and Language Contact in the Development of the Sino-Tibetan Language Family. In R. M. W. Dixon and A. Y. Aikhenvald. Areal Diffusion and Genetic Inheritance: Case Studies in Language Change, 225–254. Oxford: Oxford University Press. LaPolla, Randy J. 2009. Causes and effects of substratum, superstratum and adstratum influence, with reference to Tibeto-Burman languages. In Yasuhiko Nagano (ed.). Issues in Tibeto-Burman historical linguistics (Senri Ethnological Studies 75), 227–237. Osaka: National Museum of Ethnology. Masica, Colin. 1991. The Indo-Aryan Languages. Cambridge: Cambridge University Press. Matisoff, James A. 2003. Handbook of Proto-Tibeto-Burman: system and philosophy of Sino-Tibetan reconstruction. Berkeley, Los Angeles, and London: University of California Press. Matisoff, James A. 2011. Series Editor’s Introduction. In Christopher Button. Proto-Northern Chin. Berkeley: University of California. Nelson, David Niles. 1986. The Historical Development of the Nuristani Languages. Minneapolis and St. Paul, MN: University of Minnesota dissertation. Nichols, Johanna. 1992. Linguistic Diversity in Space and Time. Chicago: University of Chicago Press. Nichols, Johanna. 1997. Modeling Ancient Population Structures and Movement in Linguistics. Annual Review of Anthropology 26: 359–384. Post, Mark W. and Roger Blench. 2011. Siangic: A new language phylum in North East India. Paper presented at the 6th International Conference of the North East Indian Linguistics Society, Tezpur University, Assam, India, 31 January – 2 February 2011. Scott, James C. 2009. The Art of Not Being Governed: An Anarchist History of Upland Southeast Asia. New Haven and London: Yale University Press.

10   

   Thomas Owen-Smith and Nathan W. Hill

Shafer, Robert. 1955. Classification of the Sino-Tibetan languages. Word, Journal of the Linguistic Circle of New York 11: 94–111. Shafer, Robert. 1966. An Introduction to Sino-Tibetan: Part 1. Wiesbaden: Harrassowitz. Strand, Richard F. 1973. Notes on the Nuristani and Dardic Languages. Journal of the American Oriental Society 93 (3): 297–305. Thomason, Sarah G. and Terrence Kaufman (1988). Language contact, creolization, and genetic linguistics. Berkeley: University of California Press. Tournadre, Nicolas. 2008. Arguments against the concept of ‘conjunct’/‘disjunct’ in Tibetan. In B. Huber, M. Volkart and P. Widmer (eds.). Chomolangma, Demawend und Kasbek, Festschrift für Roland Bielmeier, 281–308. Halle (Saale): International Institute for Tibetan and Buddhist Studies. Turin, Mark 2007. Linguistic Diversity and the Preservation of Endangered Languages: A Case Study from Nepal. Kathmandu: International Centre for Integrated Mountain Development (ICIMOD) Watters, David E. 2006a. The Conjunct-Disjunct Distinction in Kaike. Nepalese Linguistics 22: 300–310. Watters, David E. 2006b. Notes on Kusunda Grammar: A language isolate of Nepal. Himalayan Linguistics Archive 3: 1–182. Zeisler, Bettina. 2005. On the position of Ladakhi and Balti in the Tibetan language family. In John Bray (ed.). Ladakhi histories: Local and regional perspectives, 41–64. Leiden: Brill. Zeisler, Bettina. 2009. Reducing phonological complexity and grammatical opaqueness: Old Tibetan as a lingua franca and the development of the modern Tibetan dialects. In Enoch Aboh and Norval Smith (eds.). Complex processes in new languages, 75–95. Amsterdam and Philadelphia: Benjamins.

George van Driem

Trans-Himalayan 1 Astride the Himalayas This Trans-Himalayan tale unites two narratives, an historical account of scholarly thinking regarding linguistic phylogeny in eastern Eurasia alongside a reconstruction of the ethnolinguistic prehistory of eastern Eurasia based on linguistic and human population genetic phylogeography. The first story traces the tale of transformation in thought regarding language relationships in eastern Eurasia from Tibeto-Burman to Trans-Himalayan. The path is strewn with defunct family trees such as Indo-Chinese, Sino-Tibetan, Sino-Himalayan and Sino-Kiranti. In the heyday of racism in scholarship, Social Darwinism coloured both language typology and the phylogenetic models of language relationship in eastern Eurasia. Its influential role in the perpetuation of the Indo-Chinese model is generally left untold. The second narrative presents a conjectural reconstruction of the ethnolinguistic prehistory of eastern Eurasia based on possible correlations between genes and language communities. In so doing, biological ancestry and linguistic affinity are meticulously distinguished, a distinction which the language typologists of yore sought to blur, although the independence of language and race was stressed time and again by prominent historical linguists.

2 T  he teetering trail from Tibeto-Burman to Trans-Himalayan The Tibeto-Burman linguistic phylum was identified in 1823. However, the term “Tibeto-Burman” was subsequently used in two different meanings, one by scholars following Julius von Klaproth’s polyphyletic framework and another by scholars operating within the Indo-Chinese or Sino-Tibetan paradigm. The essential differences between the two lineages of thought are contrasted, and the evidence is weighed. The geographical distribution of major subgroups and the phylogeny of the language family provide clues to Tibeto-Burman ethnolinguistic population prehistory. Several alternative theories of linguistic relationship are discussed and the major subgroups are presented. In 1823, Julius von Klaproth identified the Tibeto-Burman phylum in Paris in his polyphyletic view of Asian linguistic stocks. Klaproth’s model of many distinct Asian linguistic phyla was initially controversial because many scholars in

12   

   George van Driem

the West at the time entertained an undifferentiated view of Asian languages as all belonging to some nebulous all-encompassing language family. His TibetoBurman comprised Burmese, Tibetan and Chinese and all of the languages which could be demonstrated to be related to these three. He explicitly excluded languages today known to be Kradai or Daic (e.g. Thai, Lao, Shan), Austroasiatic (e.g. Mon, Vietnamese, Nicobarese, Khmer) and Altaic (e.g. Japanese, Korean, Mongolic, Turkic). The name Tibeto-Burman gained currency in English for the language family recognised by Klaproth and was widely used by scholars in the British Isles, e.g. Hodgson (1857), Cust (1878), Forbes (1878), Houghton (1896). Some other scholars of the day followed the Indo-Chinese theory proposed by the Scots amateur John Casper Leyden, who died at the age of 35 after making a short but dazzling career in the British colonial administration in Asia during the Napoleonic wars. In 1807, Leyden proposed his exuberant but poorly informed Indo-Chinese theory to George Barlow, Governor General of India at Fort William, in which he claimed that all the languages in Asia and Oceania shared some “common mixed origin” (Leyden 1808). This murky view held appeal to adherents of Biblical mythology who had been inclined to lump Chinese together with numerous other Asian languages into a grand Japhetic family, on the assumption that Chinese was one of the languages spoken by the descendants of Noah’s son Japhet, whilst some alternatively attempted to explain Chinese as an antediluvian language or as one of the “confounded” forms of speech with which Yahweh had afflicted mankind after the fall of the Tower of Babel. Klaproth was the first scholar to assign Chinese to its proper language family. Tibeto-Burman

Tibetan

Chinese

Burmese

…and all languages which can be demonstrated to be genetically related to these three

Fig. 1: Julius von Klaproth’s Tibeto-Burman family

The Biblically inspired Japhetic was not the only pan-Asian catch-all. Wilhelm Schott wrote personally to the famous scholar of Himalayan languages Brian Houghton Hodgson to warn him against the “Turanian” theory then being propa-

Trans-Himalayan   

   13

gated from Oxford. In 1856, Schott likewise published an essay warning against “Indo-Chinese” (Schott 1856). Schott foresaw that scholars who used the label would continue to think in terms of the mistaken phylogenetic model which the label designated. Yet the Indo-Chinese model became the favourite of racist language typologists who believed that Asian languages were generally more rudimentary and that Asian peoples were more primitive than their Western counterparts. Grammatical typology inspired language typologists such as Heymann Steinthal (1850, 1860), Ernest Renan (1858), Arthur de Gobineau (1854–1855) and John Beames (1868) to rank Chinese and Thai together on the lowest rung of the evolutionary ladder of language development based on their “monosyllabicity” and lack of inflection. These scholars argued that Chinese and Thai must be closely related and that neither was part of Tibeto-Burman. James Byrne argued that “the causes which have determined the structure of language” lay in the varying “degrees of quickness of mental excitability possessed by different races of men” (1885: 45). Chinese and Siamese ostensibly mediated a rudimentary, less evolved way of thinking and so were assigned to the lowest rungs of Steinthal’s ladder of language evolution.1 The following quote typifies this once widespread genre of scholarly discourse. …la langue chinoise, avec sa structure inorganique et incomplète, n’est-elle pas l’image de la sècheresse d’esprit et de cœur qui caractérise la race chinoise? …Suffisante pour les besoins de la vie, pour la technique des arts manuels, pour une littérature légère de petit aloi, pour une philosophie qui n’est que l’expression souvent fine, mais jamais élevée, du bon sens pratique, la langue chinoise excluait toute philosophie, toute science, toute religion, dans le sens où nous entendons ces mots. (Renan 1858: 195–196).

1 Through the lense of historical hindsight, racist linguistic typology in the 19th century had its burlesque moments, as, for example, when some linguists contested Steinthal’s hierarchy on the basis of the argument that ‘Negeridiomen’ could not possibly be positioned on rungs that were higher on the typological tree of language evolution than Chinese or Siamese in view of the differences in the material cultures of the language communities concerned. Another ludicrous moment was the coinage of the term ‘analytic’ to characterise languages such as English and French, which were no longer flamboyantly flexional and must therefore have ostensibly evolved beyond the stage of perfection purportedly reflected by Sanskrit. To account for the contrast between the technological advancement of Chinese civilisation and the ostensibly low rung on the typological ladder of language evolution ascribed to the Chinese language, Comte de Gobineau invented a distinction between so-called male and female races, whereby “les races males” possessed “un langage plus précis, plus abondant, plus riche que les races femelles” (1854, i: 190). His explanation, therefore, was that the Chinese “race” was in some sense “male” despite the inferior status which he imputed to the typological traits of the Chinese language.

14   

   George van Driem

Such reasoning contrasted starkly with the older but more sophisticated tradition of linguistic relativity, developed by John Locke (1690), Étienne de Condillac (1746), Pierre de Maupertuis (1748, 1756) and Wilhelm von Humboldt (1822, 1825, 1836). Linguists following this scholarly tradition, notably Julius von Klaproth (1823), Jean Jacques Nicolas Huot (Malte-Brun 1832, i: 521), August Friedrich Pott (1856) and Friedrich Max Müller (1871, 1881), vehemently opposed the ideas of the racist language typologists, stressed that biological ancestry was independent of language, and argued that the relationship between language structure and human cognition was not at all so simplistic, but more subtle, more interesting and, then as today, still largely unexplored.2 Indo-Chinese or Sino-Tibetan

Sino-Daic

pinioned ʻTibeto-Burmanʼ

Sinitic

Daic or Kra-Dai

Fig. 2: The Indo-Chinese or Sino-Tibetan theory: Kradai or Daic has been excluded since the 1940s

At first, Indo-Chinese encompassed Asian languages from the Caspian Sea to Polynesia. This untenable construct embodied numerous misguided phylogenetic conjectures and so it came to be whittled down in successive stages. After Philipp von Siebold (1832) and Anton Boller (1857) presented their case for a distinct Altaic phylum, Ernst Kuhn (1883, 1889) attempted to remedy what was still wrong with the Indo-Chinese model by correcting the erroneous inclusion of Austroasiatic, but the resulting model still represented a false family tree. Yet 2 The historical linguistic tradition of linguistic relativity was antagonistic to the racist tradition of the language typologists. Yet in the wake of the Second World War, the rejection of racism in most scholarly circles often went hand in hand with an unrefined, undifferentiated view of the distinct strands in the history of linguistic thought. Against this background, the backlash against the shortcomings in the writings of Benjamin Lee Whorf, who died in 1941, led to the view, dogmatically propounded in many introductory courses in general linguistics worldwide, that all languages are created equal. This smug spirit of linguistic equivalence would have been music to the ears of Pierre Maine de Biran (1815), but fortunately scholars such as George Grace (1989) continued to contest this post-war orthodoxy.

Trans-Himalayan   

   15

some scholars and several notable sinologists adopted the Indo-Chinese name and the false Indo-Chinese phylogeny, e.g. von der Gabelentz (1881), Forchhammer (1882), Conrady (1896), Laufer (1916), Wulff (1934). In 1924, the French orientalist Jean Przyluski coined sino-tibétain as the French term for Indo-Chinese in the English and German sense (Przyluski 1924).3 This French term entered English in 1931 when Jean Przyluski and Gordon Luce co-authored an article on the root for the numeral hundred in “Sino-Tibetan” (Przyluski and Luce 1931). The new term did not catch on at once, but during the Great Depression in 1935 the American president Franklin Roosevelt instituted the employment scheme called the Works Progress Administration. Through WPA, the famous Berkeley anthropologist Alfred Kroeber, inspired by the enthusiasm of Robert Shafer, managed to raise funding for his Sino-Tibetan Philology project. Changing the name of the model of linguistic relationship to the new Gallic label helped to deflect the widespread criticism against Indo-Chinese. Shafer effectively ran the project for Kroeber, but saw two things fundamentally wrong with “Sino-Tibetan.” In 1938, Shafer proposed to remove Kradai or Daic from the language family, but in the end he was not allowed to do so (Shafer 1955: 97–98). Shafer also put Sinitic on par with other divisions in the family. The two operations would effectively have heralded a return to Julius von Klaproth’s original Tibeto-Burman model. After Paul Benedict came to Berkeley in the winter of 1938–1939 to join the project, he traded in the name Indo-Chinese for “SinoTibetan.” Moreover, after the conclusion of the project in 1940, he took credit for removing Daic (Benedict 1942). Benedict (1972) also restored Sino-Tibetan to its original Indo-Chinese shape, again isolating Chinese as the odd man out. Ironically, after the Cultural Revolution, Chinese scholars adopted as orthodoxy the Indo-Chinese model as repackaged in America. Sino-Tibetan became 漢藏 語系 Hàn-Zàng yǔxì, notwithstanding its empirically unsupported phylogeny and its racist legacy. Historically, Sino-Tibetan is rooted in the fact that morphosyntactic typology had perplexed less enlightened linguists of 19th century into believing that Chinese and Thai represented an inferior developmental stage on a Steinthal’s ladder of language evolution. This view relied on the assumption that Sinitic languages had never evolved and that Chinese had remained typologically unchanged and “without inflection, without agglutination” for millennia, e.g. Chalmers (1866). 3 The need to coin a proper French term had become pressing, since in French indochinois ­referred politically and geographically to the French colonial dominions on the Indo-Chinese peninsula and linguistically to the Mon-Khmer-Kolarian or Mon-Annam linguistic phylum which Wilhelm Schmidt had renamed Austroasiatic at the beginning of the 20th century. Some British writers fond of terminological gallicisims also used the term ‘Indo-Chinese’ in the meaning Austroasiatic, e.g. Sir Richard Temple (1903: iii, 251–284).

16   

   George van Driem

By contrast, the informed historical linguistic view represented quite a different understanding of Chinese. Carl Richard Lepsius (1861: 492–496) proposed that Chinese tones had arisen from the merger of initials and the loss of finals based on correspondences between Chinese and Tibetan. He argued that entire syllables had been lost in Chinese and that Chinese ideograms once represented words which may often have contained more than just the root syllables whose reflexes survive in the modern pronunciations. The view of Chinese promulgated by Lepsius later inspired Bernhard Karlgren (1920, 1957) to conceive of Old Chinese as a langue flexionelle and to undertake the reconstruction of Old Chinese in accordance with the principles of the comparative method. Two models of phylogenetic relationship sought to defy the Sino-Tibetan paradigm propagated at Berkeley, i.e. Sino-Himalayan (Bodman 1973, 1980) and Sino-Kiranti (Starostin 1994). Although neither proposal gained acceptance, these sallies made the crucial point that to date no evidence has ever been adduced in support of the Sino-Tibetan phylogenetic model, defined by its truncated “Tibeto-Burman” taxon encompassing all non-Sinitic languages. Methodologically, attempts to define all non-Sinitic languages negatively in terms of Sinitic innovations which other languages lack or to invoke the argument of gross word order for Karen and Sinitic, as Benedict (1976) once did, are known to be phylogenetically meaningless. All comparative evidence amassed to date supports Julius von Klaproth’s 1823 minimalist Tibeto-Burman tree, which epistemologically therefore continues to represent the default model. However, the history of the field has left us with an unfortunate nomenclatural legacy. Whereas Tibeto-Burmanists in Klaproth’s tradition used the name “Tibeto-Burman” for the family as a whole, Sino-Tibetanists continue to use the term “Tibeto-Burman” to denote all non-Sinitic languages as comprising a single taxon. In an attempt to escape this terminological morass, in 2004 the alternative name “Trans-Himalayan” was proposed for the linguistic phylum because the world’s second most populous language family straddles the great Himalayan range along both its northern and southern flanks (van Driem 2007a: 226). This neutral geographical term is analogous to “Indo-European” and “AfroAsiatic” in reflecting the geographical distribution of the language family. The term “Afro-Asiatic” was coined in 1914 and replaced the earlier “HamitoSemitic” for similar reasons. Hamitic was shown not to be a valid subgroup, just as Sino-Tibetan, defined by its unitary non-Sinitic taxon, likewise denotes a false tree. The linguistic phylum is, of course, literally Trans-Himalayan in distribution. By far most of the roughly 300 different Tibeto-Burman languages and three fourths of the major Trans-Himalayan subgroups are situated along the southern flanks of the Himalayas (Figure 3), whilst by far most speakers of

Trans-Himalayan   

   17

Trans-Himalayan languages live to the north and east of the great Himalayan divide (Figure 4). Kazakhstan

Mongolia

Usbekistan Kyrgystan

Chinese Turkestan

Tajikistan

Korea

Japan

Inner Mongolia

Afghanistan Tibet

Pakistan Nepal India

China Bhutan

Bangladesh

Burma

Laos Thailand

Bay of Bengal

South China Sea Vietnam

Philippines

Cambodia

Fig. 3: Geographical distribution of the major Trans-Himalayan subgroups. Each dot represents not just one language, but the putative historical geographical centre of each of 42 major linguistic subgroups.

3 Towards a linguistic phylogeography Much more is known about the Tibeto-Burman language family today than in the days of Klaproth. Today we can identify 42 subgroups for which there appears to be evidence and about which there is some degree of consensus. The 2012 version model of the Fallen Leaves model, shown in Figure 5, contains a number of groups not mentioned when this model was first presented (van Driem 2001). The Rgyalrongic subgroup was proposed and validated by Jackson Sun (2000a, 2000b). The Nàic subgroup, comprising Nàmùyì and Shǐxīng and the closely related Nàish languages, i.e. 1Na2khi [nɑ˩hi˧] (Nàxī), Moso (Mósuō a.k.a. [nɑ˩˧]) and Laze [lɑ˧ze˧], has been proposed by Jacques and Michaud (2011). Evidence for an Ěrsūish subgroup has been presented by Yu (2011). The validation of lowerorder groups not only enables the validation of correctly delineated higher-order groups, but will also give us a clear view of their internal phylogeny.

18   

   George van Driem

Fig. 4: Geographical distribution of Trans-Himalayan languages

Post and Blench (2011) presented evidence for Siangic, a group comprising Milang and Koro. At one level, Post and Blench envisage Siangic not as a Tibeto-Burman subgroup, but as an altogether non-Tibeto-Burman phylum which has left vestiges in Koro and Milang. A more conservative stance would be to treat Koro and Milang together as a Tibeto-Burman subgroup in their own right. In a similar vein, many scholars have recently publicly aired the view that Puroik, also known as Sulung, normally deemed to be a member of the Kho-Bwa cluster of languages, is not a Tibeto-Burman language at all. Despite the apparently aberrant nature of some of the lexicon, Puroik, Koro and Milang all exhibit a good share of TibetoBurman vocabulary. The history of Indo-European is instructive in this regard. French shows a smidgen of Celtic lexicon that can be viewed as substrate (Lambert 1994), whilst the language itself is indisputably a Romance dialect. Words borrowed from the substrate language do not determine the linguistic affinity of a language. Until Ritter von Xylander (1835), Albanian was held to be a language isolate in Europe just like Basque. It is sobering to reflect that less is known today about Tibeto-Burman historical grammar than was known in 1835 about Indo-European historical grammar. The Gongduk language in Bhutan is analogous to Albanian, or for that matter much like Koro, Milang and Puroik,

Trans-Himalayan   

   19

in exhibiting much vocabulary which appears outlandish from a Tibeto-Burman perspective. Yet our perspective on Tibeto-Burman has been changing rapidly in recent years, as more becomes known about the less well documented languages of the phylum. Our understanding of what Starostin called “Tibeto-Burman in the narrow sense” is broadening to encompass a more informed and fine-mesh view. The growing awareness in the field that the Tibeto-Burman analogues of Armenian, Hittite and Albanian all appear to be found within the eastern Himalayas highlights the fact that the language family’s centre of phylogenetic diversity lies squarely within the eastern Himalayas. The lexical diversity observed in many subgroups of the eastern Himalayas is just one residue of a complex and many-layered ethnolinguistic prehistory in a region of ancient human habitation. The whereabouts and the names of the languages in the 42 leaves that have fallen from the Trans-Himalayan tree are listed below. The most obvious disambiguations are indicated with the symbol ≠ with additional elucidation. Realities on the ground are far more complex than any short list can show. Related but entirely distinct and mutually unintelligible languages sometimes go by the same name, e.g. Magar, Limbu, Chinese. So the roughly 280 language labels in this non-exhaustive list obscure a great deal of dialectal and linguistic diversity.

West Himalayish Tamangic

Bodish

Newaric Magaric

Tshangla

Dura

Raji-Raute

Bái

Lhokpu

Gongduk

Tǔjiā

Ěrsūish

Lolo-Burmese

Digarish

Nàic

Midźuish

Nungish

Tani Kho-Bwa

Sinitic

Qiāngic

Chepangic

Hrusish Dhimalish

Lepcha

Kiranti

Siangic Black Mountain

Rgyalrongic

Ao Zeme Brahmaputran

Kachinic Kukish

Angami-Pochuri Karbí Mru

Karenic

Meithei

Pyu

Tangkhul

Fig. 5: The 2012 version of the agnostic Fallen Leaves model. Thirty out of 42 Tibeto-Burman subgroups lie south of the great Himalayan divide, seven to the north and east of the Himalayas, and five, i.e. Tshangla, Bodish, Nungish, Lolo-Burmese and Kachinic, are distributed on both sides of the Himalayas

20   

   George van Driem

Sometimes the ethnic designation and the mother tongue do not match, as when a community, for example, consider themselves Jǐngpō but speak the LoloBurmese language Zaiwa or when a community consider themselves Tibetan but speak a Rgyalrongic language. Some languages are extinct, e.g. Pyu, Dura, believed to be extinct, e.g. the Sak languages, or moribund, e.g. Barām. In fact, most Tibeto-Burman languages are endangered with imminent extinction. A more detailed account can be found in the handbook Languages of the Himalayas (van Driem 2001) and in the literature referenced therein. Angami-Pochuri (southern Nagaland, northern Manipur, neighbouring portions of Burma and Assam): Angami, Chokri a.k.a. Chakri, Kheza, Mao a.k.a. Sopvoma, Pochuri, Ntenyi, Maluri a.k.a. Meluri, Sema, Rengma, Kezhama, Senkadong. Ao (central Nagaland and neighbouring portions of Burma): Yacham, Ao Chungli, Ao Mongsen, Yimchungrü a.k.a. Yachumi, Sangtam a.k.a. Thukumi, Yacham and Tengsa, Lotha a.k.a. Lhota. Bái (the area around Dàlǐ in Yúnnán province): Bái. Black Mountain Mönpa (the Black Mountains of Bhutan): ’Olekha, Riti, Jangbi and ’Wangling. Bodish (Tibet, Pakistan, India, Nepal, Sikkim, Bhutan): Balti, Purik, Ladakh, Zanskar, Lahul, Central Tibetan (Dbus and Gtsang), Sherpa, Ölmo Sherpa, Lhomi, Jirel, Kagate, Mustang, Limirong, Mugu, Northern Kham, Eastern Kham, Amdo Tibetan, Brokpa, Dzongkha, Lakha, Dränjoke, Cho-ca-nga-ca-kha, Bumthang, Kheng, Mangde, Kurtöp, Chali, Dzala, Dakpa. Brahmaputran a.k.a. Bodo-Koch and Northern Naga (West Bengal, Assam, Meghalaya, Arunachal Pradesh, northern Nagaland and adjacent portions of Burma): Chutiya, Kokborok, Tiwa, Dimasa a.k.a. Hills Kachāḍī, Bodo, Plains Kachāḍī, Meche, Garo, Atong, Pani Koch, Ruga, Rabha, Tangsa, Nocte, Wancho, Kuwa, Haimi, Htangan, Konyak, Ponyo, Phom, Chang, Welam, Nokaw. Chepangic (central Nepal): Chepang, Bhujeli. Dhimalish (eastern Nepalese Terai, western Bhutanese duars): Dhimal, Toto. Digarish a.k.a. ‘Northern Mishmi’ (Dibang river valley, Lohit district, Arunachal Pradesh): Idu, Taraon a.k.a Digaro. Dura (central Nepal’s Lamjung district): Dura. Ěrsūish (southern Sìchuān, northern Yúnnán): Ěrsū, Tosu, Lizu. Gongduk (south central Bhutan): Gongduk. Rgyalrongic (southern Sìchuān): Situ, Japhug, Tsobdun, Zbu, Lavrung (inc. Thurje Chenmo and nDzorogs), Horpa (inc. rTau and Stod-sde). Hrusish (western Arunachal Pradesh): Hruso a.k.a. Aka, Dhímmai a.k.a. Miji, Levai a.k.a. Bangru.

Trans-Himalayan   

   21

Kachinic a.k.a. Jinghpaw (northeastern India, northern Burma, southern Yúnnán): The various Kachin, Singpho, Jǐngpō or Jinghpaw languages and the Sak a.k.a. Luish languages Sak, Kadu, Andro, Sengmai, Chairel. Karbí a.k.a. Mikir (Mikir Hills or Karbí Anglóng, neighbouring districts of Assam): Karbí a.k.a. Mikir. Karenic (lower Burma, the Tenasserim and adjacent Thailand coastal regions): Pa’o, Pwo, Sgaw, Kayah, Brek a.k.a. Bwe, Bghai. Kho-Bwa (western Arunachal Pradesh): Khowa a.k.a. Bugun, Sherdukpen, Puroik a.k.a. Sulung, Lishpa. Kiranti (eastern Nepal): Pāñcthare Limbu, Phedāppe Limbu, Tamarkhole Limbu, Chathare Limbu, Yakkha, Chɨlɨng, Āṭhpahariyā (inc. Belhare), Lohorung, Yamphu, Mewahang, Kulung, Nachiring, Sampang, Sam, Chamling, Puma, Bantawa, Chintang, Dungmali, Thulung, Jero, Wambule, Tilung, Dumi, Khaling, Kohi, Bahing, Sunwar, Hayu. Kukish a.k.a. Mizo-Kuki-Chin (Mizoram and Indo-Burmese borderlands): Mizo a.k.a. Lushai, Lai, Sizang a.k.a Siyin, Thado, Tiddim Chin a.k.a. Paite a.k.a. Sokte a.k.a. Kamhau, Haka, Chinbok, Laizo, Lakher, Ashö, Khumi Chin, Hmar, Anal, Lakher a.k.a. Mara, Falam, Vaiphei, Lamgang, Simte. Lepcha (Sikkim, Darjeeling, Kalimpong): Lepcha. Lhokpu (southwestern Bhutan): Lhokpu a.k.a Doya. Lolo-Burmese (southwestern China, Burma, Southeast Asia): Burmese, Zaiwa (≠ Midźuish Zaiwa) a.k.a. Atsi, Lăshi, Măru (≠ Mru in the Chittagong), Maingtha a.k.a. Achang a.k.a. Ngachang, Hpon a.k.a Hpun, Dănu, Taungyo a.k.a. Tăru (≠ Danaw), Phunoi, Akha, Lahu, Lisu, mBisu, Ahsi, and various Yí languages. Magaric (central Nepal): Syāṅgjā Magar, Tanahũ Magar, Pālpā Magar, Khām Magar a.k.a. Kham (≠ Tibetan Kham). Meithei (Manipur): Meithei a.k.a. Manipuri Pyu (extinct language of pre-Burmese epigraphy in Burma): Pyu. Midźuish a.k.a. ‘Southern Mishmi’ (Lohit drainage, Lohit district, Arunachal Pradesh): Kaman a.k.a. Miju a.k.a Mijhu, Zaiwa (spoken by the Meyöl clan near Walong ≠ Burmic Zaiwa). Mru (in the Chittagong of Bangladesh): Mru a.k.a. Măru (≠ the Shan State Măru in Burma). Nàic (southern Sìchuān, northern Yúnnán): 1Na2khi (Nàxī), Moso (Nà, Mósuō), Laze, Nàmùyì, Shǐxīng. Newaric (central Nepal): Kathmandu Newar, Pahari Newar, Badikhel Newar, Chitlang Newar, Dolakha Newar, Barām, Thangmi. Nungish (Yúnnán province, northern Burma): Trung, Ālóng, Răwang, Róuruò, Nung inc. Nùsū and Ānù (≠ the Daic Nung in northern Vietnam).

22   

   George van Driem

Qiāngic (southern Sìchuān, northern Yúnnán): Southern Qiāngic, Northern Qiāngic, Mi-ñag (Mùyǎ), Prinmi (Pǔmǐ), Choyo (Quèyù), Tangut (Xīxià), Zhābā, Ěrgōng, Guìqióng. Raji-Raute (western Nepal, Uttarakhand): Raji, Raute. Siangic (Arunachal Pradesh): Koro, Milang. Sinitic (China): Mandarin, Cantonese, Wú, Gàn, Xiāng, Hakka a.k.a. Kèjiā, Southern Mǐn (inc. Hokkien), Eastern Mǐn, Northern Mǐn, Central Mǐn, Càijiā, Wǎxiāng. Tamangic (central Nepal): Tamang, Gurung, Thakali, Chantyal, Ghale, Kaike, ’Narpa, Manangba. Tangkhul (northeastern Manipur, neighbouring parts of Burma): Tangkhul, Maring. Tani a.k.a. Abor-Miri-Dafla (Arunachal Pradesh, neighbouring portions of Assam): Apatani, Nyisu, Bengni, Nishing, Tagin, Yano, Sarak a.k.a. Hill Miri, Galo, Bokar, Ramo, Ashing, Pailibo a.k.a. Libo, Damu, Bori, Mishing a.k.a. Plains Miri, Padam, Shimong, Pasi, Panggi, Tangam, Karko, Minyong. Tshangla a.k.a Shâchop (eastern Bhutan, enclaves in Arunachal Pradesh and Tibet): Tshangla a.k.a Shâchop or loconyms. Tǔjiā (Húnán, Húběi and Guìzhōu provinces): Tǔjiā. West Himalayish (Himachal Pradesh, Uttarakhand): Manchad, Tinan, Bunan a.k.a. Gari, Kanashi, Rangpo, Darma, Byangsi, Rangkas, Zhangzhung. Zeme (southwestern Nagaland, northwestern Manipur, neighbouring portions of Assam): Mzieme, Liangmai a.k.a Kwoireng, Zeme a.k.a. Empeo Naga a.k.a. Kacha Naga, Maram, Khoirao, Puiron, Rongmai a.k.a. Kabui a.k.a. Nruanghmei. Some of the subgroups in the above list of 42 fallen leaves represent tentative subgrouping hypotheses that have yet to be subjected to closer scrutiny, e.g. Newaric, Qiāngic. By the same token, questions arise such as whether Bodish should include East Bodish as well as Bodish proper, and how East Bodish should otherwise be renamed, or whether Brahmaputran should encompass both the Bodo-Koch as well as the Northern Naga languages. In historical linguistics, it is preferable to work from the bottom up, i.e. starting with the tangible leaves that have fallen from the tips of the branches, and then moving upward to gain an understanding of the nodes in the tree. Yet many Tibeto-Burman languages are still poorly documented and scantily described. The Fallen Leaves model is no definitive phylogeny by definition. Though agnostic about higher-order subgrouping, the model does not deny that there is a family tree whose structure must be ascertained by historical linguistic methods. The continuing identification of subgroups presents a challenge to the current generation and to future generations of historical linguists to reconstruct the

Trans-Himalayan   

   23

internal phylogeny of Trans-Himalayan on the basis of reliable data and regular sound laws and not to accept false family trees that we inherit from our predecessors or find in the literature without the support of historical comparative evidence. Two of Shafer’s old “divisions” continue to lead robust lives of their own as higher-order albeit vaguely delineated subgrouping proposals, i.e. Bodic and Burmic. Recently, Jacques and Michaud (2011) have proposed a higher-order subgroup called Burmo-Qiāngic, comprising Lolo-Burmese and a subgroup newly christened Nà-Qiāngic. Nà-Qiāngic essentially represents the same catch-all that used to be called ‘Qiāngic’ sensu lato. This constellation of subgroups has now been rendered less nebulous, however, by Sun (2000a, 2000b), Yu (2011) and Jacques and Michaud (2011), who have validated the Rgyalrongic, Ěrsūish and Nàic subgroups respectively. In addition to these three subgroups, Nà-Qiāngic also contains Mi-ñag (Mùyǎ), Prinmi (Pǔmǐ), Choyo (Quèyù), Tangut (Xīxià), Zhābā, Qiāngic sensu stricto and perhaps Ěrgōng and Guìqióng. The internal phylogeny of the latter medley of subgroups still has to be worked out, and the higher-order subgrouping hypotheses Nà-Qiāngic and Burmo-Qiāngic likewise require validation. The Càijiā 蔡家 language was recently discovered in the northwestern corner of Guìzhōu (Bó 2004). Zhèngzhāng (2010) considers Càijiā to be a member of the same subgroup as Bái, whereas Sagart believes that both Càijiā as well as the Wǎxiāng 瓦鄉 dialect of western Húnán could represent the first sub-branches of the Sinitic subgroup to have split off from Proto-Sinitic, even before the Mǐn dialects (de Sousa 2012). Another higher-order subgrouping hypothesis, Sino-Bodic, has a long history. Julius von Klaproth (1823) observed that Tibetan and Chinese appeared to be more closely related to each other than either were to Burmese. Simon (1927, 1928, 1929) and Forrest (1956, 1962) adduced lexical evidence which suggested a closer relationship between Chinese and Tibetan within the family. Although Shafer criticised Simon’s work, Shafer (1955) too observed that a closer genetic affinity obtained between Sinitic and Bodic than between any other two divisions. Later Bodman (1973, 1980) too adduced evidence indicating a closer relationship between Sinitic and Bodic. The name “Sino-Bodic” was proposed for the hypothesis, and additional lexical evidence for this affinity was adduced (van Driem 1997). Matisoff (2000) protested, but most of the Sino-Bodic evidence still stands (van Driem 2005). Possible new evidence for Sino-Bodic has been adduced by Nathan W. Hill (2011) and Zhèngzhāng Shàngfāng (2011). Future research will determine whether any of these supergroups will survive the test of time.

24   

   George van Driem

4 Paternal patterns Despite valiant efforts by David Bradley (2012), Blench’s (2009) claim still appears to hold that no rice agricultural terminology can be confidently reconstructed for the Tibeto-Burman phylum. Instead the linguistic ancestors of the Austroasiatics and the Hmong-Mien appear to be the likeliest candidates behind the early cultivation and later the domestication of Asian rice (van Driem 2011, 2012). Rather, as has long been widely presumed, the ancient Trans-Himalayans probably cultivated foxtail millet Setaria italica and broomcorn millet Panicum mileaceum. Yet significant advances in linguistic palaeontology, supported by detailed desciptions and lexicographical documentation, in tandem with genetic work on these two cultigens may one day bring us closer to unravelling this portion of the Trans-Himalayan past. A more obvious approach to tackling our prehistory than studying the link between languages and millet genes is the study of possible correlations between genetic markers in modern language communities and the phylogeography of the languages which they speak. However, from the beginning of the 19th century, when Jean-Baptist Lamarck elaborated his theory of evolution, to the Second World War, interdisciplinary approaches tying linguistics and human biological ancestry have had a chequered history. Since genes are always inherited by offspring from their parents, whilst the languages spoken by people are not necessarily those that were spoken by their parents or grandparents, correlations between languages and genes could only be probabilistic at best, and there need not be any relationship whatsoever. Therefore, it is highly interesting that when geneticists began to look for correlations between genetic markers and the geographical distribution of language communities, they began to find statistically relevant correlations, not with genetic markers on the maternally inherited mitochondrial DNA but with genetic markers on the paternally inherited Y chromosome. Such a tendency, first recognised in the pioneering studies of Poloni et al. (1997, 2000), has repeatedly been observed that some correlation obtains between the most frequent Y chromosomal haplo­groups of a community and the language which the people happen to speak. This correlation between a community’s language and that community’s prevalent paternal ancestries is what I called the Father Tongue hypothesis (van Driem 2002). There are a number of reasons why we might expect this outcome. Initial human colonisation of any part of the planet must have involved both sexes in order for a population of progeny to establish itself. Once a population is in place, however, subsequent migrations could have been heavily gender-biased. Subsequently, male intruders could impose their language whilst availing themselves

Trans-Himalayan   

   25

of the womenfolk already in place. Presumably, tribes of Amazons could have spread in a similar fashion. If so, however, then the tell-tale correspondences between mitochondrial lineages and the distribution of linguistic phyla should certainly have been detected by now, but any correlation between maternal lineages and linguistic phylogeography discerned to date has been underwhelming. The Father Tongue hypothesis suggests that linguistic dispersals were, at least in most parts of the world, posterior to initial human colonisation and that many linguistic dispersals were predominantly later male-biased intrusions. If we infer that a mother teaching her children their father’s tongue has been a recurrent, ubiquitous and prevalent pattern throughout linguistic his­tory, then some of the mechanisms of language change over time are likely to be inherent to the dynamics of this pathway of transmission. Such correlations are observed worldwide. The correlation of Niger-Congo languages with Y chromosomal haplogroups is a striking example (Wood et al. 2005). Likewise, the martial and malebiased historical spread of Hàn Chinese during the sinification of southern China, recounted in painstaking detail in the Chinese chronicles, is clearly reflected in the genetic evidence (Wen et al. 2004). A recent common ancestry between native Americans and indigenous Altaians is also based preponderantly on the shared Y chromosomal heritage and is not quite as well reflected in the mitochondrial lineages (Dulik et al. 2012). Whilst father tongues may predominate globally, mother tongues certainly do exist in the sense that there are areas on the planet where the linguistic affinity of a community correponds more closely to the maternally transmitted mitochondrial lineage which the speakers share with other linguistically related communities. In this sense, in the north of today’s Pakistan, the Balti speak a Tibetic mother tongue but profess a paternal religion that was first propagated in this area as early in the 8th century by men who came from the Near East, although the wholesale conversion of Baltistan to Islam is held to have begun only in the 14th century. The most prevalent mitochondrial DNA lineages amongst the Baltis are shared with other Tibetan communities, whereas the prevalent Y chromosomal haplogroups probably entered Baltistan during the introduction of Islam (Zerjal et al. 1997, Quintana-Murci et al. 2001, Qamar et al. 2002).4 At the same time, a jarring disconnect is sometimes seen between the occurrence of a highly salient genetic marker and the linguistic af­finity of a community’s language. Hungarians lack the TatC deletion defining the Y chromosomal

4 Ironically, the Balti call their language phaskat ‘father tongue’ (Roland Bielemeier, personal communication, 10 September 2012), just as they call their homeland phayul ‘fatherrealm’ and birthplace phasa ‘father-land’ (Sprigg 2002: 127).

ཕ་སྐད་

ཕ་ས་

ཕ་ཡུལ་

26   

   George van Driem

haplogroup N1c,5 despite the sheer prevalence of this marker amongst all other Uralic language communities (Lì et al. 1999). So, it deserves to be repeated that the lin­guistic ancestors of a language com­munity were not nec­essarily the same people as the bio­logical ances­tors of that community. In fact, some of them could not have been the same people. It also merits repeating that the time depth accessible to population geneticists studying polymorphisms on the genome is vastly greater than the reach of the linguistically reconstructible past. The wave of anatomically modern humans who introduced the proto-languages that were later to give rise to today’s Asian linguistic phyla and language isolates can be dated to between 25,000 to 38,000 years ago (Rasmussen et al. 2011), and the antiquity of Y chromosomal haplogroups such as O1 or O2 has been calculated to be greater than 10,000 years (Yan et al. 2011). Historical linguists, on the other hand, generally estimate the linguistically reconstructible past to be shallower than 10,000 years. This temporal gap must temper and inform all speculations regarding correlations between linguistic and genetic affinity. With such caveats in place, how can we address the question formulated at the beginning of this section? On the 28th of June 2006, at a symposium held at the École Française d’Extrême-Orient at Siem Reap, I identified the Y chromosomal haplogroup O2a (M95) as the marker for the spread of Austroasiatic on the basis of the then available genetic data (later published in van Driem 2007b). This view has been corroborated by subsequent genetic studies, e.g. Kumar et al. (2007), Chaubey et al. (2010). In the latter article, we concluded that Austroasiatic speakers in India today are derived from a dispersal from Southeast Asia, followed by extensive sex-specifix admixture with local populations indigenous to the Subcontinent. The autosomal data also reflect the distinction between two components in the genome, one represented by the predominantly indigenous maternal lineages and the other by the intrusive paternal O2a lineage that correlates with the linguistic affinity of the Austroasiatic language communities in the Indian subcontinent. These findings go well beyond Robert von Heine-Geldern’s model of a Southeast Asian homeland and envisage a father tongue spread of Austroasiatic, borne to the Indian subcontinent by predominantly male speakers from mainland Southeast Asia, but also involving a complex sociolinguistic prehistory of bidirectional gene flow across the Bay of Bengal (Chaubey et al. 2010). In many parts of the world, the mitochondrial DNA lineages often appear preponderantly to reflect older resident maternal lineages.

5 The 2008 Y Chromosome Consortium haplogroup labels are used here.

Trans-Himalayan   

   27

Fig. 6: Portion of the Y chromosome phylogenetic tree relevant to the Father Tongue hypo­ thesis with regard to Austronesian, Austroasiatic, Hmong-Mien and Trans-Himalayan and the peopling of eastern Asia, reproduced from Karafet et al. (2008) with the kind permission of the Cold Spring Harbor Laboratory Press.

The argument for the Father Tongue interpretation of the spread of major linguistic phyla in eastern Eurasia, such as Austroasiatic, is therefore not based solely on the frequencies of particular Y chromosomal haplogroups. The Father Tongue hypothesis is originally based on the differential correlation of Y chromosomal and mitochondrial lineages with the modern geographical distribution of language communities, i.e. the presence or absence of a strong correlation between linguistic affinity and genetic markers in the non-recombinant portions of the genome. As one might expect, a distinct provenance for the maternal and paternal lineages appears to be reflected by studies of autosomal markers as well (Chaubey et al. 2010). More importantly, a rooted topology of the Y chromosomal tree in its entirety and of the Y chromosomal haplogroup O in particular is central

28   

   George van Driem

to the reconstruction of linguistic population prehistory in eastern Eurasia, operating on the assumption of the veracity of the Father Tongue hypothesis. The avail­able genetic data also enabled us to identify a correlation of the Y chromo­somal haplo­group O3a3b (M7) with the spread of Hmong-Mien, whilst our genetic samplings throughout the Himalayan region had established a correlation between Trans-Himalayan and the paternal lineage O3a3c (M134) (Parkin et al. 2006, 2007, Kraaijenbrink et al. 2007a, 2007b, 2009, van Driem 2011). The Y chromosomal haplogroup O is becoming ever more minutely mapped, and most recently the phylogenetic positions of mutations P164 and PK4 within the haplo­ group have been revised (Yan et al. 2011). Yet the antiquity calculated for many of these mutations is generally greater than the time depth that most historical linguists are willing to ascribe to the major language phyla. Let us venture into the twilight beyond the linguistically reconstructible past to a time just after the Last Glacial Maximum, when the Y chromosomal haplogroup O (M175) had split up into the subclades O1 (M119), O2 (M268) and O3 (M122). Based on what is known about linguistic phylo­geny and about the geographical distribution of modern linguistic communities today, the three subclades can putatively be assigned to three geographical loci along an east-west axis. For the sake of argument and sche­matic representation, and without any claim to geographical precision or veracity, I shall assign the haplogroup O1 (M119) to the drainage of the Pearl River and its tributaries in what today is the Chinese province of Guǎngdōng. I shall situate haplogroup O2 (M268) in southern Yúnnán and O3 (M122) to the area where today’s northeastern India, southeastern Tibet and northern Burma adjoin. Since we have associated O2a (M95), which is a derivative clade of haplogroup O2 (M268), with the Austroasiatic language phylum, we might conjecture that Asian rice, perhaps both japonica and indica rice, was first domesticated roughly in the general area hypothetically imputed to O2 (M268) here. Whilst the bearers of the O2a (M95) haplo­group became the Stammväter of the Austroasiatics, the other derivative paternal subclade O2b (M176) spread eastward, where they introduced rice agriculture to the areas south of the Yang­tze. Though the bearers of the O2b (M176) haplogroup continued to sow seed as they continued to move ever further eastward, they left little or no linguistic traces, except maybe an Austroasiatic name for the Yang­tze river, as proposed by Pulleyblank (1993), reflected as the toponym borrowed by Old Chinese as 江 *kˤroŋ (jiāng). Meanwhile, back in southern Yúnnán, the early Austroasiatics spread from this locus initially to the Salween drainage in northeastern Burma and to the area that today is northern Thailand and western Laos. In time, the Austroasiatics would spread as far as the Mekong delta, the Malay peninsula, the Nicobars and later even into eastern India, where they would introduce both their language and their paternal lineage to indigenous peoples of the Subcontinent.

Trans-Himalayan   

   29

At the locus putatively assigned to the haplogroup O3 (M122), the bearers of this marker gave rise to the paternal lineages O3a3c (M134) and O3a3b (M7). Whilst the bearers of the polymorphism O3a3c (M134) stayed behind in the area comprising northeastern India, southeastern Tibet and northern Burma, the bearers of the O3a3b (M7) paternal lineage migrated eastward to settle in the areas south of the Yangtze. On their way, the early Hmong-Mien encountered the ancient Austroasiatics, from whom they adopted rice agriculture. The intimate interaction between ancient Austroasiatics and the early Hmong-Mien not only involved the sharing of knowledge about rice agriculture technology, but also left a genetic trace in the high frequencies of haplogroup O2a (M95) in today’s Hmong-Mien and of haplogroup O3a3b (M7) in today’s Austroasiatic populations. On the basis of these Y chromosomal haplogroup frequencies, Cai et al. observed that Austroasiatics and Hmong-Mien “are closely related genetically” and ventured to spe­culate about “a Mon-Khmer origin of Hmong-Mien populations” (2011: 8). More precisely, the incidence of haplogroup O3a3b (M7) in Austroasiatic language communities of Southeast Asia appears to indicate a significant Hmong-Mien paternal contribution to the early Austroasiatic populations whose descendants settled in Southeast Asia, whereas the incidence of haplogroup O3a3b (M7) in Austroasiatic communities of the Indian subcontinent is undetectably low. The incidence of haplogroup O2a amongst the Hmong-Mien appears to indicate a slightly more modest Austroasiatic paternal contribution to Hmong-Mien populations than vice versa. As the Hmong-Mien moved eastward, the bearers of haplogroup O2b (M176) likewise continued to move east. Even further east, the O1 (M119) paternal lineage gave rise to the O1a (M119) subclade, which moved from the Pearl River drainage eastward to the Mǐn river drainage in the hill tracts of Fújiàn province and across the strait to Formosa, which consequently became the Urheimat of the Austronesians. Back west in the easternmost spurs of the Himalayas, the bearers of Y chromosomal haplogroup O3a3c (M134) expanded eastward into Sìchuān and Yúnnán, north and northwest across the Tibetan plateau as well as westward into the Himalayas and southward into the Indo-Burmese borderlands. In the west and south, the early Trans-Himalayans encountered Austroasiatics, who had preceded them. Linguistic research on Trans-Himalayan languages can inform a chronologically layered view of ethnolinguistic prehistory. Not only do historical linguistics and genetics present two distinct and independent windows on the past. Even on a logarithmically distorted time scale the time depth accessible to historical linguistics can be seen to be far shallower than the prehistorical depth accessible to human population genetics. The human population genetic data from beyond the linguistically reconstructible past embolden us to speculate that there

30   

   George van Driem

must have been an early eastward and northward spread into East Asia, possibly including the linguistic ancestors of modern Trans-Himalayan language communities, who may have been the first bearers of the Y chromosomal haplogroup O3a3c (M134). After this post-glacial colonisation, there must have been a number of discrete expansions in different directions at different times in the past. To recapitulate the chronology of possible movements: (1) a post-glacial northward wave of peopling at a time depth beyond what is generally held to be linguistically reconstructible by historical linguists, (2) a northeasterly spread of a subset of the ancient Trans-Himalayans to the putative early locus of Sino-Bodic, (3) incremental spread of diverse ancient Trans-Himalayan groups throughout the Himalayas, where there appears to be both linguistic and genetic evidence of preTrans-Himalayan speaking populations, (4) a southward spread of Sino-Bodic, suggested by archaeology, genes and language, bringing Sino-Bodic groups, including Sinitic, into contact with the ancient Hmong-Mien, the early Austroasiatics, the Austronesians and a number of other Trans-Himalayan groups, (5) a Bodic spread across the Tibetan plateau spilling over into the Himalayas, as evinced by the distribution of Bodish, East Bodish, Tamangic, West Himalayish and several other groups, and (6) the spread of Trans-Himalayan groups from Yúnnán into Southeast Asia, e.g. Karen, Pyu and later Lolo-Burmese. Following these tentatively reconstructed prehistoric stages of peopling, there were the historically attested ethnolinguistic dispersals: (7) the historically documented Hàn spread, clearly evinced in linguistics and genetics, probably assimilating non-Trans-Himalayan as well as other Trans-Himalayan groups, and (8) the historically documented spread of Bodish (i.e. Tibetic) across the Tibetan plateau. The relative frequencies of the Y chromosomal haplogroup O2a (M95) in various Trans-Himalayan populations of the Indian subcontinent (Sahoo et al. 2006, Reddy et al. 2007) suggest that a subset of the paternal ancestors of particular Trans-Himalayan populations in north­eastern India, e.g. certain BodoKoch communities, may originally have been Austroasiatic speakers who married into Trans-Himalayan communities or were linguistically assimilated by ancient Trans-Himalayans. At the same time, median-joining net­work analyses of haplogroup O2a (M95) microsatellites have suggested a division in the Indian subcontinent between Trans-Himalayans vs. Austroasiatic and Dravidian language communities. Austroasiatics and Dravidians show greater Y chromosomal micro­ satellite diversification than Trans-Himalayan language communities, and the highest frequency of the O2a haplogroup is found in tribal populations in Orissa, Chattisgarh and Jharkhand (Sengupta et al. 2006). We must bear in mind that Y haplogroups are subject to selection and that frequencies change over time. As stressed above, haplotype frequencies by them-

Trans-Himalayan   

   31

selves are not a sufficient criterion. A rooted topology of the Y chromomosal tree and its subsidiary clades provides key evidence. Moreover, the ethnolinguistic significance of paternal lineages becomes even more manifest when other portions of the genome are scoured for correlations with linguistic phylogeography. At the same time, our understanding of what constitutes neutral diversity has been tempered by mathematical modelling. Simulations have shown that a normally low-frequency allele could surf on a demic wave of advance and so attain high frequency across a vast area. Gene surfing during a spatial expansion is likely to result in distinct geographical sectors of low genetic diversity separated by sharp allele frequency gradients. The result of recurrent bottleneck effects during range expansion into newly colonised territories can mimic complex phylogeographical patterns of adaptation and segregation into clades in post-glacial niche refugia. Likewise, the massive introgression of resident genes into the incursive population can also be misinterpreted as the result of a selective process (Excoffier and Ray 2008, Excoffier et al. 2009). Surfing on the crest of a demic wave of expansion confers a selective advantage when compared to alleles left behind in the core area (Klopfstein et al. 2006, Moreau et al. 2011). Both the dynamics of sex-biased dispersals as well as the process of the sexually asymmetrical introgression of resident alleles into incursive populations can be modelled in terms of hybridisation during range expansions (Petit and Excoffier 2009, Currat and Excoffier 2011). An observed state of affairs for which a particular model of population prehistory has been advanced may in many cases very well be either the result of demography or of selection on genome diversity (Fagundes et al. 2007). However, we must keep in mind that a scenario that has been computed to be the statistically more likely scenario may not necessarily correspond to prehistorical reality. Though presumably paternal lineages may often preferentially enjoy the benefits of surfing, incursive Y chromosomal lineages can go entirely extinct, as the linguistic evidence6 would suggest may very well have happened with the Y chromosomal haplogroup N1c in Hungary. We must also not lose sight of the fact that these speculations are based on correlations between language and Y chromosomal haplogroups and that these too are interpreted in the light of the assumed veracity of the Father Tongue hypothesis over a vast stretch of time. This assumption may not hold true for all times in the past. Furthermore, correlations may be due to different kinds of cir6 The presence of the Hungarian language in the region that was once Pannonia represents incontrovertible linguistic evidence of the advent of Uralic linguistic ancestors, a fact which is historically attested at any rate, but the hypothetical correlation of the Y chromosomal haplogroup N1c with the Uralic linguistic phylum, of course, remains conjectural.

32   

   George van Driem

cumstances other than causation or direct relationship. So, whilst we are free cautiously to develop arguments which buttress a speculative model of ethnolinguistic prehistory, such as the one outlined here, we must not lose sight of the essential distinction between the facts and our assumptions and inferences as well as the precise nature and limitations of the empirical basis for our speculations. Confronted with the overwhelming growing body of evidence in support of the Father Tongue hypothesis, Forster and Renfrew impute the spread of language families to “emigrating agriculturalists” who “took local wives” (2011: 1391). This interpretation is a transparent attempt to succour Bell­wood and Renfrew’s embattled First Farmers hypothesis, which seeks to ascribe the founding dis­persals of language families to the spread of agriculture (Bellwood and Renfrew 2002). At the same time, in order to buttress Renfrew’s widely doubted hypothesis of an Indo-European homeland in Asia Minor, Forster and Renfrew also propose a correlation of Indo-European with the Y chromosomal haplogroup J2a. In fact, it remains moot whether any part of Y chromosomal phylogeography correlates well with the spread of the Neolithic horizon. Not every population movement led to the spread of a language phylum, and population movements are not uniform in nature. Whether during the exodus of anatomically modern humans out of Africa or at the shallow time depth of the colonisation of Oceania by Austronesian populations, the colonisation of previously uninhabited lands invariably involved both sexes and the introduction of a language phylum. During the Neolithic horizon, the spread of farming was necessarily a sedentary and incremental process, which likewise must mostly have involved both sexes. Early farmers might only have been able to spread their language at times of great surplus and concomitant population growth, perhaps sometimes involving the establishment of agricultural colonies elsewhere. By contrast, the modern ethnolinguistic composition of Asian populations must be understood, at least in part, as having resulted from male-biased linguistic intrusions, whether motivated by conquest, land grab or the urge to seek out new habitats. In my argument against the premises and the reasoning behind the hypothesis of the founding dispersals of language phyla having been mediated by the spread of farming, I proposed the telic and more complex Centripetal Migration theory (van Driem 2007b). I shall not repeat that exposition here, but, with reference to Forster and Renfrew’s wilful interpretation of the Y chromosomal haplogroup J2, I shall reiterate that, in the context of the Indian subcontinent, “the J2 haplo­group… appears to emanate from the Arabian Peninsula and, unlike haplogroups N and R1a, attains no high frequency in Ceylon” and “probably reflects the historically attested male-borne eastward spread of Islam,” whereas Y chro-

Trans-Himalayan   

   33

mosomal haplogroups of the R subclades spread to the Subcontinent “from the northwest along with Indo-Aryan lan­guage across northern India and to Ceylon” (van Driem 2007b: 5). The spread of various Y chromosomal R subclades is likely to be linked to the dispersal of Indo-European from an original homeland in the Pontic-Caspian steppe, whilst the current geographical distribution of the Y chromosomal lineage L provides the likeliest candidate for a vestige of an earlier patrilingual dispersal of Elamo-Dravidian emanating from a region which encompassed the Bactria and Margiana of later prehistory.

References Beames, John. 1868. Outlines of Indian Philology. London: Trübner and Company. Bellwood, Peter, and Colin Renfrew (eds.). 2002 Examing the Farming/Language Dispersal Hypothesis. Cambridge: McDonald Institute for Archaeological Research. Benedict, Paul King. 1942. Thai, Kadai, and Indonesia: A new alignment in southeastern Asia. American Anthropologist 44: 576–601. Benedict, Paul King. 1972. Sino-Tibetan: a Conspectus. Cambridge: Cambridge University Press. Benedict, Paul King. 1976. Sino-Tibetan: another look. Journal of the American Oriental Society 96 (2): 167–197. Blench, Roger Marsh. 2009. If agriculture cannot be reconstructed for Proto-Sino-Tibetan, what are the consequences? Paper presented at the 42nd Conference on Sino-Tibetan Languages and Linguistics at Chiang Mai, 3 November 2009. Bó Wénzé. 2004. Càijiāhuà gàikuàng. Mínzú Yǔwén 2004 (2): 68–81. Bodman, Nicholas Cleaveland. 1973. Some Chinese reflexes of Sino-Tibetan s- clusters. Journal of Chinese Linguistics 1 (3): 383–396. Bodman, Nicholas Cleaveland. 1980. Proto-Chinese and Sino-Tibetan: data towards establishing the nature of the relationship. In Frans van Coetsem and Linda R. Waugh (eds.). Contributions to Historical Linguistics: Issues and Materials, 34–199. Leiden: E.J. Brill. Boller, Anton. 1857. Nachweis, dass das Japanische zum ural-altaischen Stamme gehört. Sitzungsberichte der philosophisch-historischen Classe der kaiserlichen Akademie der Wissenschaften, Wien, 23: 393–481. Bradley, David. 2012. Proto-Tibeto-Burman grain crops. Rice 4 (3–4): 134–141. Byrne, James. 1885. General Principles of the Structure of Language (2 volumes). London: Trübner and Company. Cai Xiaoyun, Zhendong Qin, Bo Wen, Shuhua Xu, Yi Wang, Yan Lu, Lanhai Wei, Chuanchao Wang, Shilin Li, Xingqiu Huang, Huí Lǐ and the Genographic Consortium. 2011. Human migration through bottlenecks from Southeast Asia into East Asia during Last Glacial Maxiumum revealed by Y chromosomes. Public Library of Science 6 (8): e24282. Chalmers, John. 1866. The Origin of the Chinese: An Attempt to Trace the Connection of the Chinese with Western Nations in their Religion, Superstitions, Arts, Language and Traditions. Hong Kong: De Souza & Co.

34   

   George van Driem

Chaubey, Gyaneshwer, Mait Metspalu, Ying Choi, Reedik Mägi, Irene Gallego Romero, Siiri Rootsi, Pedro Soares, Mannis van Oven, Doron M. Behar, Siiri Rootsi, Georgi Hudjashov, Chandana Basu Mallick, Monika Karmin, Mari Nelis, Jüri Parik, Alla Goverdhana Reddy, Ene Metspalu, George van Driem, Yali Xue, Chris Tyler-Smith, Kumarasamy Thangaraj, Lalji Singh, Maido Remm, Martin B. Richards, Marta Mirazon Lahr, Manfred Kayser, Richard Villems and Toomas Kivisild. 2010. Population genetic structure in Indian Austroasiatic speakers: The role of landscape barriers and sex-specific admixture. Molecular Biology and Evolution, 28(2): 1013–1024. de Condillac, Étienne Bonnot. 1746. Essai sur l’origine des connoissances humaines: Ouvrage où l’on réduit à un seul principe tout ce qui concerne l’Entendement Humain (2 volumes). Amsterdam: Pierre Mortier. Conrady, August. 1896. Eine indochinesische Causativ-Denominativ-Bildung und ihr Zusammenhang mit den Tonaccenten: Ein Beitrag zur vergleichenden Grammatik der indochinesischen Sprachen, insonderheit des Tibetischen, Barmanischen, Siamesischen und Chinesischen. Leipzig: Otto Harrassowitz. Currat, Mathias, and Laurent Excoffier. 2011. Strong reproductive isolation between humans and Neanderthals inferred from observed patterns of introgression. Proceedings of the National Academy of Sciences 108 (37): 15129–15134. Cust, Robert N. 1878. A Sketch of the Modern Languages of East India. London: Trübner and Company. van Driem, George. 1997. Sino-Bodic. Bulletin of the School of Oriental and African Studies 60 (3): 455–488. van Driem, George. 2001. Languages of the Himalayas: An Ethnolinguistic Handbook of the Greater Himalayan Region, containing an Introduction to the Symbiotic Theory of Language (2 volumes). Leiden: Brill. van Driem, George. 2002. The Father Tongue hypothesis: Sexually dimorphic dispersals in prehistory. Paper presented at the 17th conference of the Indo-Pacific Prehistory Association, held at the Academia Sinica, Taipei, 9–15 September 2002. van Driem, George. 2005. Sino-Austronesian vs. Sino-Caucasian, Sino-Bodic vs. Sino-Tibetan, and Tibeto-Burman as default theory. In Yogendra Prasada Yadava, Govinda Bhattarai, Ram Raj Lohani, Balaram Prasain and Krishna Parajuli (eds.). Contemporary Issues in Nepalese Linguistics, 285–338. Kathmandu: Linguistic Society of Nepal. van Driem, George. 2007a. The diversity of the Tibeto-Burman language family and the linguistic ancestry of Chinese. Bulletin of Chinese Linguistics 1 (2): 211–270. van Driem, George. 2007b. Austroasiatic phylogeny and the Austroasiatic homeland in light of recent population genetic studies. Mon-Khmer Studies 37: 1–14. van Driem, George. 2011. Rice and the Austroasiatic and Hmong-Mien homelands. In Nick J. Enfield (ed.). Dynamics of Human Diversity: The Case of Mainland Southeast Asia, 361–389. Canberra: Pacific Linguistics. van Driem, George. 2012. The ethnolinguistic identity of the domesticators of Asian rice. Comptes Rendus Palevol, 11 (2): 117–132. Dulik, Matthew C., Sergey I. Zhadanov, Ludmila P. Osipova, Ayken Askapuli, Lydia Gau, Omer Gokcumen, Samara Rubinstein and Theodore G. Schurr. 2012. Mitochondrial DNA and Y chromosome variation provides evidence for a recent common ancestry between native Americans and indigenous Altaians. American Journal of Human Genetics 90: 1–18. Excoffier, Laurent, Matthieu Foll and Rémy J. Petit. 2009. Genetic consequences of range expansions. Annual Review of Ecology, Evolution and Systematics 40: 481–501.

Trans-Himalayan   

   35

Excoffier, Laurent and Nicolas Ray. 2008. Surfing during population expansions promotes genetic revolutions and structuration. Trends in Ecology and Evolution 23 (7): 347–351. Fagundes, Nelson J. R., Nicolas Ray, Mark Beaumont, Samuel Neuenschwander, Francisco M. Salzano, Sandro L. Bonatto and Laurent Excoffier. 2007. Statistical evaluation of alternative models of human evolution. Proceedings of the National Academy of Sciences 104 (45): 17614–17619. Forbes, Charles James F. S. 1878. On Tibeto-Burman languages. Journal of the Royal Asiatic Society of Great Britain and Ireland 10: 210–227. Forchhammer, Emile. 1882. Indo-Chinese languages. Indian Antiquary 11: 177–189. Forrest, R. A. D. 1956. Towards common Tibeto-Burman. Wennti 10: 1–18. Forrest, R. A. D. 1962. The linguistic position of Róng (Lepcha). Journal of the American Oriental Society 82 (3): 331–335. Forster, Peter, and Colin Renfrew. 2011. Mother tongue and Y chromosomes. Science 333: 1390–1391. von der Gabelentz, Hans Georg Conon. 1881. Chinesische Grammatik mit Ausschluss des niederen Stiles und der heutigen Umgangssprache. Leipzig: T.O. Weigel. de Gobineau, Joseph Arthur. 1854–5. Essai sur l’inégalité des races humaines (4 volumes). Paris: Firmin-Didiot et Compagnie. Grace, George. 1989. The Linguistic Construction of Reality. London: Croom Helm. Hill, Nathan W. 2011. Multiple origins of Tibetan o. Language and Linguistics 12 (3): 707–721. Hodgson, Brian Houghton. 1857. Comparative Vocabulary of the Languages of the broken Tribes of Népál. Journal of the Asiatic Society of Bengal 26: 317–371. Houghton, Bernard. 1896. Outlines of Tibeto-Burman linguistic palæontology. Journal of the Royal Asiatic Society 1896: 23–55. von Humboldt, Wilhelm. 1822. Ueber das vergleichende Sprachstudium in Beziehung auf die verschiedenen Epochen der Sprachentwicklung. Abhandlungen der Königlichen Preussischen Akademie der Wissenschaften, historisch-philosophische Klasse, aus den Jahren 1820 und 1821: 239–260. von Humboldt, Wilhelm. 1825. Über das Entstehen der grammatischen Formen und ihren Einfluß auf die Ideenentwicklung. Abhandlungen der Königlichen Preussischen Akademie der Wissenschaften, historisch-philosophische Klasse, aus den Jahren 1822 und 1823: 401–430 von Humboldt, Wilhelm. 1836. Über die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluß auf die geistige Entwickelung des Menschengeschlechts. Berlin: Königliche Akademie der Wissenschaften. Jacques, Guillaume and Alexis Michaud. 2011. Approaching the historical phonology of three highly eroded Sino-Tibetan languages: Naxi, Na and Laze. Diachronica 28 (4): 468–499. Karefet, Tatiana M., Fernando L. Mendez, Monica B. Meilerman, Peter Underhill, Stephen L. Zegura and Michael F. Hammer. 2008. New binary polymorphisms reshape and increase resolution of the Y chromosomal haplogroup tree. Genome Research 18: 830–838. Karlgren, Bernhard. 1920. Le proto-chinois, langue flexionelle. Journal Asiatique (XIe Série) 40: 205–232. Karlgren, Bernhard. 1957. Grammata Serica Recensa (i.e. Bulletin of the Museum of Far Eastern Antiquities 29: 1–332). von Klaproth, Julius Heinrich. 1823. Asia Polyglotta. Paris: A. Schubart. Klopfstein, Seraina, Mathias Currat and Laurent Excoffier. 2006. The fate of mutations surfing on the wave of a range expansion. Molecular Biology and Evolution 23 (3): 482–490.

36   

   George van Driem

Kraaijenbrink, Thirsa, Emma J. Parkin, Denise R. Carvalho-Silva, George van Driem, Guido Barbujani, Chris Tyler-Smith, Mark A. Jobling and Peter de Knijff. 2009. Genetic and linguistic borders in the Himalayan region. In Francesco d’Errico and Jean-Marie Hombert (eds.). Becoming Eloquent: Advances in the Emergence of Language, Human Cognition and Modern Cultures, 181–201. Amsterdam: John Benjamins. Kraaijenbrink, Thirsa, Jean Robert M.L. Opgenort, Nirmal Man Tuladhar, George van Driem and Peter de Knijff. 2007. Allele frequency distribution for 21 autosomal STR loci in Nepal. Forensic Science International 168 (2–3): 227–231. Kraaijenbrink, Thirsa, Karma Tshering of Gaselô and Peter de Knijff. 2007. Allele frequency distribution for 21 autosomal STR loci in Bhutan. Forensic Science International 170: 68–72. Kuhn, Ernst. 1883. Ueber Herkunft und Sprache der transgangetischen Völker. Festrede zur Vorfeier des Allerhöchsten Geburts‑ und Namensfestes Seiner Majestät des Königs Ludwig II., gehalten in der öffentlichen Sitzung der Königlichen Akademie der Wissenschaften zu München am 25. Juli 1881. München: Verlag der Königlichen Bayerischen Akademie. Kuhn, Ernst. 1889. Beiträge zur Sprachenkunde Hinterindiens. Sitzungsberichte der Königlichen Bayerischen Akademie der Wissenschaften (München), Philosophisch-philo­ logische Classe II: 189–236. Kumar, Vikrant, Arimanda N.S. Reddy, Jagedeesh P. Babu, Tiprisetti N. Rao, Banrida T. Langstieh, Kumarasamy Thangaraj, Alla G. Reddy, Lalji Singh and Battini M. Reddy. 2007. Y-chromosome evidence suggests a common paternal heritage of Austro-Asiatic populations. BioMed Central Evolutionary Biology, 7: 47 (14 pp.) www.biomedcentral. com/1471–2148/7/47 Lambert, Pierre-Yves. 1994. La langue gauloise. Paris: Éditions Errance. Laufer, Berthold. 1916. The Si-hia Language, a Study in Indo-Chinese Philology. T‘oung Pao 17: 1–126. Lepsius, Carl Richard. 1861. Über die Umschrift und Lautverhältnisse einiger hinterasiatischer Sprachen, namentlich der Chinesischen und der Tibetischen. Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin aus dem Jahre 1860, 449–496. Leyden, John Casper. 1808. On the languages and literature of the Indo-Chinese nations. Asiatic Researches 10: 158–289. Lì Jīn, and Peter A. Underhill, Vishal Doctor, Ronald W. Davis, Peidong Shen, L. Luca CavalliSforza and Peter J. Oefner. 1999. Distribution of haplotypes from a chromosome 21 region distinguishes multiple prehistoric human migrations. Proceedings of the National Academy of Sciences of the United States of America 96: 3796–3800. Locke, John. 1690. An Essay concerning Humane Understanding, in Four Books. London: Tho. Basset. Maine de Biran, François-Pierre-Gonthier. [1815] 1932. Note sur les réflexions de Maupertuis et de Turgot au sujet de l’origine des langues. In Pierre Tisserand (ed.). Œuvres de Maine de Biran (14 volumes), vol. 10, 315–338. Paris: Librairie Félix Alcan. Malte-Brun, Conrad [posthumous]. 1832. Précis de la géographie univer­selle, ou description de toutes les parties du monde sur un plan nouveau, d’après les grandes divisions naturelles du globe: Nouvelle édition, revue, corrigée, mise dans un nouvel ordre et augmentée de toutes les nouvelles découvertes par J.J.N. Huot (6 volumes) Bruxelles: Th. Lejeune. Matisoff, James Alan. 2000. On “Sino-Bodic” and other symptoms of neosubgroupitis. Bulletin of the School of Oriental and African Studies 63 (3): 356–369.

Trans-Himalayan   

   37

de Maupertuis, Pierre Louis Moreau. 1756a [1748]. Réflexions philosophiques sur l’origine des langues et la signification des mots. In Œuvres (nouvelle édition corrigée et augmentée, 4 volumes), vol. 1, 258–309. Lyon: Jean-Marie Bruyset. de Maupertuis, Pierre Louis Moreau. 1756b. Dissertation sur les différents moyens dont les hommes se sont servis pour exprimer leurs idées. In Œuvres (nouvelle édition corrigée et augmentée, 4 volumes), vol. 3, 435–468. Lyon: Jean-Marie Bruyset. Moreau, Claudia, Claude Bhérer, Hélène Vézina, Michèle Jomphe, Damian Labuda and Laurent Excoffier. 2011. Deep human genalogies reveal a selective advantage to being on an expanding wave front. Science 334: 1148–1150. Müller, Friedrich Max. 1871. Lectures on the Science of Language (2 volumes). London: Longmans, Green and Company. Müller, Friedrich Max. 1881. Selected Essays on Language, Mythology and Religion (2 volumes). London: Longmans, Green and Company. Parkin, Emma J., Thirsa Kraaijenbrink, Jean Robert M.L. Opgenort, Nirmal Man Tuladhar, George van Driem, Peter de Knijff and Mark Jobling. 2007. Diversity of 26 Y-STR haplotypes in a Nepalese population sample: Isolation and drift in the Himalayas. Forensic Science International 166 (2–3): 176–181. Parkin, Emma J., Thirsa Kraaijenbrink, Karma Tshering of Gaselô, Peter de Knijff and Mark Jobling. 2006. 26-locus Y-STR typing in a Bhutanese population sample. Forensic Science International 161 (1): 1–7. Petit, Rémy J., and Laurent Excoffier. 2009. Gene flow and species delimitation. Trends in Ecology and Evolution 24 (7): 386–393. Poloni, Estella Simone, Ornella Semino, Giuseppe Passarino, A.S. Santachiara- Benerecetti, I. Dupanloup, André Langaney and Laurent Excoffier. 1997. Human genetic affinities for Y chromosome P49a,f/TaqI haploptypes show strong correspondence with linguistics. American Journal of Human Genetics 61: 1015–1035 (cf. the erratum published in 1998 in the American Journal of Human Genetics 62: 1267). Poloni, Estella Simone, Nicolas Ray, Stefan Schneider and André Langaney. 2000. Languages and genes: Modes of transmission observed through the analysis of male-specific and female-specific genes. In Jean-Louis Dessalles and Laleh Ghadakpour (eds.). Proceedings: Evolution of Language, 3rd International Conference 3–6 April 2000, 185–186. Paris: École Nationale Supérieure des Télécommunications. Post, Mark William, and Roger Marsh Blench. 2011. Siangic: A new language phylum in North East India. Paper presented at the 6th International Conference of the North East Indian Linguistics Society, Tezpur University, 31 January to 2 February 2011. Pott, August Friedrich. 1856. Die Ungleichheit menschlicher Rassen hauptsächlich vom sprachwissenschaftlichen Standpunkte. Lemgo und Detmold: Meyer’sche Hofbuchhandlung. Przyluski, Jean. 1924. Le sino-tibétain [and] Les langues austroasiatiques. In Antoine Meillet and Marcel Cohen (eds.) Les Langues du Monde, 361–384 [and] 385–403. Paris: Librairie Ancienne Édouard Champion. Przyluski, Jean and Gordon Hannington Luce. 1931. The number “a hundred” in Sino-Tibetan. Bulletin of the School of Oriental Studies 6 (3): 667–668. Pulleyblank, Edwin George. 1983. The Chinese and their neighbours in prehistoric and early historic times. In David N. Keightley (ed). The Origins of Chinese Civilization, 411–466. Berkeley: University of California Press.

38   

   George van Driem

Qamar, Raheel, Qasim Ayub, Aisha Mohyuddin, Agnar Helgason, Keh­kashan Mazhar, Atika Manoor, Tatiana Zerjal, Chris Tyler-Smith and S. Qasim Mehdi. 2002. Y-chromo­somal variation in Pakistan. Ameri­can Journal of Human Genetics 70: 1107–1124. Quintana-Murci, Lluís, C. Krausz, Tatiana Zerjal, S. Hamid Sayar, Michael F. Hammer, S. Qasim Mehdi, Qasim Ayub, Raheel Qamar, Aisha Mohyuddin, U. Radha­krishna, Mark A. Jobling Chris Tyler-Smith and Ken McElreavey. 2001. Y-chro­mosome lineages trace diffusion of people and languages in southwestern Asia. American Journal of Human Genetics 68: 537–542. Rasmussen, Morten, Xiaosen Guo, Yong Wang, Kirk E. Lohmueller, Simon Rasmussen, Anders Albrechtsen, Line Skotte, Stinus Lindgreen, Mait Metspalu, Thibaut Jombart, Toomas Kivisild, Weiwei Zhai, Anders Eriksson, Andrea Manica, Ludovic Orlando, Francisco de la Vega, Silvano Tridico, Ene Metspalu, Kasper Nielsen, María C. Ávila-Arcos, J. Víctor MorenoMayar, Craig Muller, Joe Dortch, M. Thomas P. Gilbert, Ole Lund, Agata Wesolowska, Monika Karmin, Lucy A. Weinert, Bo Wang, Jun Li, Shuaishuai Tai, Fei Xiao, Tsunehiko Hanihara, George van Driem, Aashish R. Jha, François-Xavier Ricaut, Peter de Knijff, Andrea B. Migliano, Irene Gallego-Romero, Karsten Kristiansen, David Lambert, Søren Brunak, Peter Forster, Bernd Brinkmann, Olaf Nehlich, Michael Bunce, Michael Richards, Ramneek Gupta, Carlos Bustamante, Anders Krogh, Robert A. Foley, Marta Mirazón Lahr, François Balloux, Thomas Sicheritz-Pontén, Richard Villems, Rasmus Nielsen, Wang Jun, and Eske Willerslev. 2011. An aboriginal Australian genome reveals separate human dispersals into Asia. Science 334 (6052): 94–98. Reddy, B. Mohan, B.T. Langstieh, Vikrant Kumar, T. Nagaraja, A.N.S. Reddy, M. Aruna, K. Thangaraj, A.G. Reddy and Lalji Singh. 2007. Austro-Asiatic tribes of Northeast India provide hitherto missing genetic link between South and Southeast Asia. Public Library of Science One 2 (11): e1141. doi:10.1371/journal.pone.0001141 Renan, Ernest. 1858. De l’origine du langage (deuxième édition, revue et considérablement augmentée). Paris: Michel Lévy, Frères, Librairies-Éditeurs. Ritter von Xylander, Joseph. 1835. Die Sprache der Albanesen oder Schkipetaren. Frankfurt am Main: die Andreäische Buchhandlung. Sahoo, Sanghamitra, Anamika Singh, G. Himabindu, Jheeman Banerjee, T. Sitalaximi, Sonali Gaikwad, R. Trivedi, Phillip Endicott, Toomas Kivisild, Mait Metspalu, Richard Villems and V. K. Kashyap. 2006. A prehistory of Indian Y chromosomes: Evaluating demic diffusion scenarios. Proceedings of the National Academy of Sciences of the United States of America 103 (4): 843–848. Schott, Wilhelm. 1856. Über die sogenanten indo-chinesischen Sprachen insonderheit das Siamische. Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin aus dem Jahre 1856, Philosophisch-historische Klasse: 161–179. Sengupta, Sanghamitra, Lev A. Zhivotovsky, Roy King, S.Q. Mehdi, Christopher A. Edmonds, Cheryl-Emiliane T. Chow, Alice A. Lin, Mitashree Mitra, Samir K. Sil, A. Ramesh, M.V. Usha Rani, Chitra M. Thakur, L. Luca Cavalli-Sforza, Partha P. Majumder and Peter A. Underhill. 2006. Polarity and temporality of high-resolution Y-chromosome distributions in India identify both indigenous and exogenous expansions and reveal minor genetic influence of Central Asian pastoralists. The American Journal of Human Genetics 78: 202–221. Shafer, Robert. 1955. Classification of the Sino-Tibetan languages. Word, Journal of the Linguistic Circle of New York 11: 94–111. von Siebold, Philipp Franz Balthazar. 1832. Verhandeling over de afkomst der Japanners. Verhandelingen van het Bataviaasch Genootschap van Kunsten en Wetenschappen 13: 183–275.

Trans-Himalayan   

   39

Simon, Walter. 1927. Zur Rekonstruktion der altchinesischen Endkonsonanten. Mitteilungen des Seminars für Orientalische Sprachen an der Friedrich-Wilhelms-Universität zu Berlin 30 (1): 147–167. Simon, Walter. 1928. Zur Rekonstruktion der altchinesischen Endkonsonanten, II. Teil. Mitteilungen des Seminars für Orientalische Sprachen an der Friedrich-WilhelmsUniversität zu Berlin 31 (1): 157–204. Simon, Walter. 1929. Tibetisch-chinesische Wortgleichungen, ein Versuch. Mitteilungen des Seminars für Orientalische Sprachen an der Friedrich-Wilhelms-Universität zu Berlin 32 (1): 157–228. de Sousa, Hilário. 2012. The southern Sinitic languages at the edge of mainland Southeast Asia. Paper read for the Berner Zirkel für Linguistik, University of Bern, 21 November 2012. Sprigg, Richard Keith. 2002. Balti-English English-Balti Dictionary. London: Routledge Curzon. Starostin, Sergej Anatol’evič. 1994. The reconstruction of Proto-Kiranti. Paper presented at the 27ème Congrès International sur les Langues et la Linguistique Sino-Tibétaines, Centre International d’Études Pédagogiques à Sèvres, 14 octobre 1994. Steinthal, Heymann. 1850. Die Classification der Sprachen dargestellt als die Entwickelung der Sprachidee. Berlin: Ferdinand Dümmler’s Buchhandlung. Steinthal, Heymann. 1860. Charakteristik der hauptsächlichsten Typen des Sprachbaues. Berlin: Ferdinand Dümmler’s Verlagsbuchhandlung. Sun, Jackson (= Sūn Tiānxīn). 2000a. Parallelisms in the verb morphology of Sidaba rGyalrong and Lavrung in rGgyalrongic. Language and Linguistics 1 (1): 161–190. Sun, Jackson (= Sūn Tiānxīn). 2000b. Stem alternations in Puxi verb inflection: Toward validating the Rgyalrongic subgroup in Qiangic. Language and Linguistics 1 (2): 211–232. Temple, Sir Richard C. 1903. Census of India, 1901, Vol. III: The Andaman and Nicobar Islands: Report on the Census. Calcutta: Office of the Superintendent of Government Printing, India. Wen Bo, Li Hui, Lu Daru, Song Xiufeng, Zhang Feng, He Yungang, Li Feng, Gao Yang, Mao Xianyun, Zhang Liang, Qian Ji, Tan Jingze, Jin Jianzhong, Huang Wei, Ranjan Deka, Sù Bīng, Ranajit Chakroborty and Jīn Lì. 2004. Genetic evidence supports demic diffusion of Han culture. Nature 431: 302–305. Wood, Elizabeth T., Daryn A. Stover, Christopher Ehret, Giovanni Destro-Bisol, Gabriella Spedini, Howard McLeiod, Leslie Louie, Mike Bamshad, Beverly I. Strassmann, Himla Soodyall and Michael F. Hammer. 2005. Contrasting pattersn of Y chromosome and mtDNA variation in Africa: Evidence for sex-biased demographic processes. European Journal of Human Genetics 13: 867–876. Wulff, Kurt. 1934. Chinesisch und Tai: Sprachvergleichende Untersuchungen. Copenhagen: Levin & Munksgaard. Yan Si, Chuan-Chao Wang, Huí Lǐ, Shi-Lin Li, Jīn Lì and the Genographic Consortium. 2011. An updated tree of Y-chromosome haplogroup O and revised phylogenetic positions of mutations P164 and PK4. European Journal of Human Genetics 19: 1013–1015. Yu, Dominic. 2011. Reconstructing Proto-Ersuic. Paper at the 17th Himalayan Languages Symposium, Kōbe City University of Foreign Studies, 6 September 2011. Zerjal, Tatjana, B. Dashnyam, A. Pandya, Manfred Kayser, Lutz Roewer, F.R. Santos, W. Schiefenhovel, N. Fretwell, M.A. Jobling, S. Harihara, K. Shimuzu, D. Semjidmaa, Antti Sajantila, P. Salo, M.H. Crawford, E.K. Ginter, Oleg V. Efgrafov and Chris Tyler-Smith. 1997. Genetic relationships of Asian and Northern Europeans, revealed by Y chromo­somal DNA analysis. American Journal of Human Genetics 62: 1174–1183.

40   

   George van Driem

Zhèngzhāng Shàngfāng. 2010. Càijiāhuà Báiyǔ kuānxì jí cígēn bǐjiào. In Pān Wǔyún and Shěn Zhōngwěi (eds.). Yánjūzhī Lè, The Joy of Research, II, 389–400. Shanghai: Shanghai Educational Publishing House. Zhèngzhāng Shàngfāng. 2011. Wǔshí shēntǐde Zàng-hàn duìyìng [On Tibeto-Chinese correspondence of fifty body part terms]. Mínzú Yǔwén 2011 (4): 16–18.

Scott DeLancey

Creolization in the Divergence of the Tibeto-Burman Languages One dimension along which language families differ considerably is in the range of typological variation across languages of the same stock.1 Indo-European, for example, shows dramatic divergence in word order patterning (contrast Irish, English, and Bengali) and inflectional complexity (contrast Russian and English). We have long been aware that typological divergence is often a result of contact with other languages: the typological consistency of Indic with the OV patterns of its South Asian neighbors, or the fact that within Finno-Ugric we find tendencies toward VO typology in Europe rather than Siberia, are hardly coincidental. “Sino-Tibetan”2 is an obvious example, with the isolating SVO typology of Sinitic contrasting with the SOV agglutinating patterns of the rest of the family. My principal interest in this paper is in a subtler but historically very important divergence among the other Tibeto-Burman languages between what I will refer to as archaic and creoloid types. There is widespread recognition that contact with neighboring languages, and among Tibeto-Burman languages, has played a significant role in the history of many branches and languages of the family (LaPolla 2001, 2009; van Driem 2001; Jacquesson 2009; Sun and Liu 2009; DeLancey 2010a, b, 2012, to appear inter alia). In this paper I will show that the “creoloid” pattern has arisen, repeatedly and independently, in situations of intense language contact, either through direct dominance of one group over others or through the adoption of Tibeto-Burman lingua francas in various points in the network of growing urban states throughout the Tibeto-Burman area from the 1st millennium BCE.

1 I am grateful to Robbins Burling, Nathan W. Hill, Guillaume Jacques, and Tom Owen-Smith for very useful comments on earlier drafts of this paper. 2 The hypothesis of a Sino-Tibetan family bifurcated into Sinitic and Tibeto-Burman branches has never been substantiated, and should be considered obsolete (see van Driem 1997, 2005). Henceforth I will use Tibeto-Burman to refer to the entire family, of which Sinitic is simply one branch among others. However, this paper is primarily concerned with the other branches; for the application of the argument to Sinitic see DeLancey (2013).

42   

   Scott DeLancey

1 The problem Across the Tibeto-Burman family we find two substantially different morphosyntactic profiles. In one, we find elaborate, hierarchical argument indexation systems in the verb, in the other a complete absence of agreement morphology. Languages of the second type are very reminscent of the minimal grammar which we find in creole languages (Burling 2007; DeLancey 2010a, 2012). The relative antiquity of the two patterns has been the subject of considerable controversy, which I will not rehash here (see DeLancey 2010b). The present paper contributes to this debate by pointing out the connection between the creoloid type and wellknown tendencies of linguistic change under contact.

1.1 Archaic and Creoloid verbs As an introduction to the difference, consider the first sentence of traditional narratives from several Tibeto-Burman languages: archaic Limbu (Kiranti) and Japhug (Rgyalrongic), and creoloid Lhasa Tibetan (Bodish) and Boro (Bodo-Garo):3 Limbu (van Driem 1987: 351) (1) nɛpphu hɛndzaʔ thik yaˑn-le kɛ-boˑŋ-ba way-ɛ-tchi two boys one day-gen nmz-be.born-nmz be-past-du.abs ‘There were two boys who had been born on the same day.’ Japhug Rgyalrong (Jacques 2004: 483) (2) kɯɕɯŋgɯ kɯɕɯŋgɯ tɕe tɤ-ru ʁnɯs pjɤ-tú-ndʑi in.the.old.days conj headman two inferential-3du-exist ‘Once upon a time there were two village headmen.’ Lhasa Tibetan (3) snga.mo snga.mo gcig la spo.bo cig dang rmo.bo cig long.ago old.man a and old.woman a yod-pa red exist-nmz be ‘Long, long ago, there were an old man and an old woman.’

3 Lhasa data were collected by the author in Eugene, Oregon, during the 1980s. Boro data were collected in Assam in 2009, by Prafulla Basumatary, Krishna Boro, Bihung Brahma, and the author.



Creolization in the Divergence of the Tibeto-Burman Languages   

   43

Boro (4) dhonpur gami hwn-nai-au bwrai burwi sa-nwi P.N. village call-nmz-loc old.man old.woman cls-two doŋ-mwn za-hwi-dwŋ exist-past become-elsewhere-real ‘It happened that in Dhonpur village there lived an old couple.’ The Limbu and Japhug inflected verbs each consist of two morphemes beside the copula stem: a TAME4 form which roughly grounds the event in the world of discourse, and a 3rd person dual index by which the verb sets out in skeletal form the event described by the clause. The verb is a tightly-structured schematic recapitulation of the clause: ‘there were two [protagonists].’ Typically in archaic languages we find “portmanteau” morphs with complex meanings, such as the combined person-number index in Japhug, and the role information attached to the dual suffix in Limbu. The Lhasa verb is equally complex, in the simple sense of the number of morphemes attached to the verb, but it is more transparent: the form is a recently grammaticalized nominalization construction (DeLancey 2010b, 2011a), and each morpheme – the lexical verb, the nominalizer pa, and the equational copula red – is synchronically recognizable, and has a specific, unitary contribution to the meaning of the construction (whether or not that unitary meaning is unitarily describable in English).5 This is characteristic of creoloid verbal constructions. More fundamentally, the Lhasa verb does nothing to present the scene – the number and identity of participants is recoverable only from the noun phrase arguments themselves. All of the inflection of the verb – and all of the copious paradigm of nominalizations, suffixes, serial verbs and nested copular constructions in the modern verbal system – is about grounding the event in the world of discourse. The same is true of the Boro example. The Boro construction shares with the Lhasa the quality of transparency. The only purely inflectional morpheme, -mwn, is a tense/mood suffix and nothing else, but many speakers can identify every other grammatical morpheme with its synchronic source. The Boro verbal expression involves an inflected lexical verb plus an inflected copula, and thus is in one sense the most complex of all. But, as in Lhasa, the complexity is all in the 4 Tense/aspect/modality/evidentiality. The acronym is particularly useful for Tibeto-Burman languages, where the categories marked in the verbal system are often difficult to sort neatly into these categories. 5 At least for literate speakers; I have met younger speakers, who were literate in Chinese but not Tibetan, who have no sense of the internal structure of the verb endings.

44   

   Scott DeLancey

service of grounding. The inflected final verb za ‘become, happen, come to be’ has two suffixes, both very shallowly-grammaticalized verbs. As an independent verb, hwi means ‘go (someplace else) and give [someone something];’ here it is a member of a motley category of serialized verbs and other postverbal adverbial elements which I am calling lexical suffixes (we will return to these later), and indicates that what the sentence describes occurred somewhere other than the present location of the speaker and hearer. The final suffix -dwŋ is a grammaticalization of the existential verb doŋ; its sense can be roughly described as nonfuture. This word then is an auxiliary to the lexical verb doŋ-mwn, with doŋ in its lexical sense of ‘live, dwell’ and -mwn indicating past time. The overall import is something like ‘so it happened that in the past, far from here …’ Nothing in the entire complex is about the event or state described in the sentence itself, everything is about the relation of that event or state to the present world of the speaker and hearer. The Boro verb sequence has only to do with situating the proposition in the mental world of the speech act participants; it makes no reference to the internal structure of the clause or to any of the arguments. The nature of the archaic verb is more strikingly evident in a clause with more arguments, as in this Limbu example (van Driem 1987: 274): (5) nɛtchi məna-re andeˑ staʔl-ille haʔr-aŋ two man-erg earlier jackal-erg bite-and yaˑg -ɛ-lle nis-ɛ-tch-u be-past-sub see-past-du.ag-3pat ‘Earlier two men had seen that the jackal was on the prowl.’ The final verb form, roughly ‘they two saw him/it,’ is a schematic summary of the transitivity, argument structure, and actants of the clause. Nothing like this can occur in a creoloid grammar.

1.2 Creoloid typology Recent work in creole studies has seen the development of the concept of a “creoloid” language type (Ansaldo and Matthews 2001), which arises in “heavy contact situations involving typologically distant varieties” (2001: 311). This is not the classic scenario of a creole crystallizing out of an inchoate pidgin, which by some definitions is still considered criterial to the identification of a “true” creoles, but rather a very common diachronic phenomenon resulting from “nonnormal” (Thomason and Kaufman 1988), “suboptimal” (Dahl 2004) or “interrupted” (McWhorter 2007) transmission, i.e. one or more historical episodes in



Creolization in the Divergence of the Tibeto-Burman Languages   

   45

which a significant number of adult members of the speech community were nonnative speakers using the language as a lingua franca or supralect. For Ansaldo and Matthews and a number of other scholars, this terminological distinction between “creoloid” and “true creole” may usefully label different historical origins, but not any actual typological distinction; on this view classic creole typology can arise in situations of heavy linguistic contact without there necessarily ever having been the pidgin stage which used to be taken as the defining event in the history of a creole: Language contact has this consequence [reduction in complexity] because of pidginization. The most extreme outcome of pidginization is the development of a pidgin language, but this is a very rare occurrence. It is only pidiginization at its most extreme, together with a number of other unusual factors, which combine to lead to the development of pidgin and, even more rarely, creole languages. Pidignization can be said to occur whenever adults and post-adolescents learn a new language. (Trudgill 2009: 99, emphasis original)

Trudgill’s “pidginization” is more generally, and more appropriately, referred to as creolization. Since I am not proposing a pidgin stage in the development of any of the languages discussed here, these are not true creoles, in the sense of McWhorter 2001. We do not need to be concerned with the debate about whether this is a typologically meaningful distinction; it is sufficient for our purposes that certain creole-like patterns can develop through intense language contact involving suboptimal transmission. What I am suggesting is that Proto-Bodo-Garo, Proto-Lolo-Burmese, Proto-Bodish, and probably others such as Proto-Tani, took on their grammatical shape in circumstances in which they were widely spoken by non-native speakers, as trade languages, languages of administration, soldier’s argot, or by mixed populations. Definitions of the creoloid pattern tend to be simplistic and of limited usefulness. For example, some scholars (e.g. McWhorter 2001) claim lack of phonemic tone to be a characteristic of creole languages, a notion which is not particularly useful in an East/South Asian context (Ansaldo and Matthews 2001: 315–17). Much of the recent discussion of these issues has been framed in terms of one or another notion of “complexity.” For our purposes a simpler and more easily defined value is what we may call transparency. A characteristically creoloid morpheme has a unitary, coherent meaning, which is inherent to the morpheme itself, not dependent on paradigmatic or syntagmatic relations to other morphemes. Archaic Tibeto-Burman languages, in contrast, have elaborate interconnected paradigms of forms, with formal rather than substantive semantic meanings, and the meaning of a form is inherent not so much in the morpheme itself, as in the overall paradigm of which it is part.

46   

   Scott DeLancey

1.3 The creoloid tendency in Tibeto-Burman languages What I am calling the archaic vs. creoloid split has always been a major problem in comparative Tibeto-Burman. While the family, with the exception of Karen and Sinitic, is consistently verb-final, and most of the modern languages show many of the gross features associated with that broad type, we find a wide divergence in morphological structure and typology across the family. In the languages which tended to serve as “typical” Tibeto-Burman exemplars – Classical and Modern Central Tibetan, Modern Standard Burmese, more recently Lahu and one or two others – we find grammatical typology very reminiscent of that of the creoloid languages of mainland Southeast Asia, except that it is organized along agglutinative verb-final rather than isolating verb-medial lines. At the other typological extreme, in Kiranti, Kham-Chepang, Rgyalrong, and Nung, we find highly-structured paradigmatic organization involving highly grammatical forms and tightlygrammaticalized constructions. We see throughout Tibeto-Burman examples of drastic, and often rapid, creolization affecting one variety of a language, or one subbranch, which has come into intense contact with other languages; compare, for example, Jinghpaw with Singpho, Bodo-Garo with Jinghpaw and Konyak, Dolakha with Kathmandu Newar (Genetti 1988, DeLancey 2010b). Sun and Liu (2009) present a detailed case study of this process in Anong. Consider a few more illustrative verbs, (6–7) from Japhug and (8) from Boro: (6) ɣɯ-ja-sɯ-ɣe cislocative-aor/3→3-caus-come ‘s/he invited her’

(Jacques 2010: 142)

(7) a-ɣɯ-thɯ-tɯ-khɤm irrealis-cislocative-perfective:downstream-2-give ‘You will give it to me.’

(Jacques 2004: 502)

(8) dán-so-hwi-zwb-phin-lia-mwn cut-bisecting-at.a.distance-exhaustive-again-no.longer-past ‘no longer intend to cut all into two pieces horizontally someplace else again’ In Rgyalrongic languages like Japhug we see a strongly paradigmatic system: each affix occupies a specfic slot in which it alternates with other affixes of the same category. The ordering is fixed and grammatical, easily described in terms of morphological position classes. Many morphemes are fusional, e.g. in example (6) one prefix, ja-, indicates both aspect and person agreement, and participates in the inverse system; in example (7) one prefix, thɯ-, marks both aspect and



Creolization in the Divergence of the Tibeto-Burman Languages   

   47

directional orientation. Substantively, the Japhug paradigm includes argument indexation and other affixes whose meaning is purely syntactic, such as inverse, 3→3, etc. Finally, it is worth noting that the paradigms of Rgyalrongic, like those of Nungish, Kham, and many Kiranti languages, involve both prefixes and suffixes. In the Boro example (which is a bit artificial; we would seldom find that many lexical suffixes in one verb, though speakers are happy to concoct such forms on demand, and occasionally do so for effect), we find strictly agglutinative morphology, almost exclusively suffixal, and completely transparent and regular. The verbal system consists of a large, ill-defined and semi-open set of postverbal elements, many easily identifiable as recently grammaticalized verbs, which express a broad and very loosely-structured set of aspectual, modal, and adverbial concepts, with no verb agreement, and transitivity essentially irrelevant. In the Boro example the morphemes are not strictly paradigmatic, and the ordering is more semantic than grammatical; while ‑­ so in the first example would always immediately follow the lexical verb, and -lia-mwn, in that order, could only be final, the other morphemes could occur in any order. There are no fusional or portmanteau forms in the Boro verb: each morpheme carries a discrete meaning. In contrast to Japhug, Boro has no verbal suffixes of purely syntactic function. Finally, in BodoGaro the productive morphology is almost exclusively suffixal. (The Boro verb has a single productive prefix, the prohibitive da-, which is also, not coincidentally, the only element of the productive verbal morphology of Proto-Tibeto-Burman provenance). This seems to be a strong tendency in the creoloid Tibeto-Burman languages; as old morphology is shed and new structure is built up, these languages innovate suffixes, almost never prefixes.

1.4 Archaic and creoloid patterns in historical context Higher-order classification of the Tibeto-Burman branches is still problematic (Benedict 1972, van Driem 2001), but the archaic-creoloid division does not correlate with any plausible genetic classification. To give an idea of the relation between genetic classification and creoloid typology, we can look at the distribution of the patterns across Bradley’s (1997) classification (Table 1). Languages in ordinary typeface represent the most archaic patterns. Languages showing significant evidence of creolization are in italics; extreme creoloid languages and branches are in boldface:

48   

   Scott DeLancey

Tab. 1: Distribution of Archaic and Creoloid Languages Western

NE India

Eastern NE

SE

Kiranti Bodo-Garo-Konyak Rgyalrong Lolo-Burmese Kham-Chepang Luish-Jinghpaw Tangut WHimalayan Kuki-Chin-Naga Qiangic Tshangla Naxi-Moso Bodish Unclassified (including Bradley’s “Central”): Nungish, Keman, Tani, Digarish, Lepcha

In this and almost all other classification schemes there is no correlation between archaic vs. creolized typology at any but the lowest-level genetic classification. In general there is no correlation between creoloid typology and gross geographical location, except that the languages spoken farthest to the south are the most creolized. However, it has long been noted that there is a strong association between Tibeto-Burman languages and upland/mountain locations. This ecological distribution has a strong cultural component as well. As we will discuss further below (Section 3.1), Southeast Asian communities can be categorized as “Hill” vs. “Valley” cultures, the latter characterized by cities, state-level political organization, literacy, and official Buddhist, Hindu or Confucian religion, while Hill cultures are “tribal,” with little political organization above the community level. Although we find some creoloid languages among Hill peoples, wherever we find Tibeto-Burman languages spoken by dominant, Valley cultures (Chinese, Tibetan, Burmese, Boro, Meithei, Newar), they are always creoloid.

2 Boro as a typical creoloid I will not present a detailed account of creoloid SOV syntax in this paper, but will survey some facts of Boro which will exemplify the type. The primary characteristic of the creoloid type Tibeto-Burman language is loosely-organized verb structure, only marginally paradigmatic, for the most part transparently based on nominalization or serial verb constructions. The Bodo-Garo languages are the most extreme instances of the creoloid pattern that I am aware of in the family, but students of Lolo-Burmese or Tani languages will instantly recognize much of what we see in Bodo-Garo, and students of a modern Bodish or Naga language will find most of it quite familiar. Classical Tibetan is less so, since the verb manifests substantial affixal and stem-alternating morphology. But much of this is



Creolization in the Divergence of the Tibeto-Burman Languages   

   49

secondary rather than inherited structure, and Classical Tibetan is definitely of the creoloid rather than the archaic type, in that there is no argument indexation in the verb. The modern Bodish languages are emphatically creoloid, and most of what I have to say about Boro is applicable in broad strokes to Lhasa, Kurtöp, or Tamang.

2.1 The Boro verb Boro (DeLancey and Boro, in preparation) has a motley set of elements which attach after the verb, all with some sort of tense/aspect/modality function. These combine quite freely. They can be arranged into position classes in terms of permissible and impermissible orderings of possible combinations, but only one of these has more than one member, and even that one is only partially paradigmatically organized: Tab. 2: The Boro verb ˗1

Stem

da-

1 0 -w -bai -thwŋ -a -akhwi -lia

2

3

4

5

6

-dw

-ni

-swi

-gwn

-mwn

The first position suffixes themselves do not form a structured paradigm, but a miscellaneous set of elements including two tense suffixes (-w and -bai), a polite imperative (-thwŋ), and an innovative three-member negative paradigm (-a/-akhwi/-lí-a): (9) a. mẃn-a

‘not get’

b. mẃn-akhwi ‘hasn’t gotten’ cp. gwi-a ‘negative existential’ c. mẃn-lía ‘not get anymore’ The negative suffix and the generic affirmative ‑w date back to Proto-Bodo-Garo; the other forms are more recent developments. Many Tibeto-Burman languages make few or no tense/aspect distinctions in the negative construction. Boro has

50   

   Scott DeLancey

developed an innovative three-way distinction out of Bodo-Garo materials. The negative perfect is transparently the negative suffix combined with the negative existential gwi. The ‘not anymore’ form is often pronounced simply as -la, but in careful pronunciation it carries lexical tone, showing that it is a stem element, presumably originally a lexical suffix. The essential position 1 paradigm, then, has three members: -w, -bai, and -a, and this is the extent of paradigmatic structure in the verb. The verb suffixes can be put together quite freely: (10) a. tháŋ-a ‘didn’t go’ b. tháŋ-a-khwi ‘hasn’t gone’ c. tháŋ-a-khwi-gwn ‘maybe hasn’t gone’ d. tháŋ-a-khwi-si-gwn ‘unexpectedly maybe hasn’t gone’ e. tháŋ-a-khwi-mwn ‘hadn’t gone’ f. tháŋ-gwn-mwn ‘would have gone’ Note in particular the combination of “Past” (actually a sort of past/counterfactual) and -mwn “Future” (perhaps better irrealis) -gwn to describe a complex modal sense: (11) assa khokhi aŋ uwa-khw dán-si-gwn-mwn OK friend 1sg bamboo-obj cut-immediate-future-past ‘OK, friend, I was going to cut the bamboo.’ (Treasure) Boro has grammaticalized a number of tense/aspect-like notions, but a language like Boro (or English, to take another creoloid language where sentences like (11) are quite normal) in which the “future” and “past” markers can co-occur in the same verb form, has not progressed very far toward a paradigmatically organized verb. Most of the work which in the archaic languages is done with more tightlybound tense/aspect morphology, in the creoloid pattern is accomplished through serial verb constructions, often, as in Boro, more or less attached to the lexical verb stem: (12) iniphrai dokan-au gaugau-ni dokan-au za-phai-phin-bai after.that shop-loc self-gen shop-loc sit-come-again-perf ‘Then [they] came back to his shop and sat.’ (Mirror)



Creolization in the Divergence of the Tibeto-Burman Languages   

   51

(13) hor-hor-na buŋ-nai-khai nou thrwb send-send-nf say-nmz-because homeward straight hab-phai-nanwi bakos-khw bwr-khaŋ-bai enter-come-nf box-obj dig.out-finish-perf ‘Because [he] asked [her], [she] went straight to [her] room and dug out the box.’ (Treasure) These are transparently recognizable as modestly-grammaticalized verbs which still occur lexically: phai ‘come,’ phin ‘return,’ khaŋ ‘put down’ (Boro 2012). Several of these, such as khaŋ, seem to fill in for some of the basic categories which seem to be missing in the verbal system (e.g. a straightforward perfective), but show no particular phonological or morphosyntactic features that would distinguish them from the huge (we have counted about 150, and the count is not complete) set of postverbal serialized verbs and lexical suffixes which can occur in the same position.

2.2 Reflexive/reciprocal constructions There is a striking difference in the archaic and the creoloid reflexive construction, neatly exemplified by the descriptions of Nungish and Lolo-Burmese systems in Bradley 1995 and LaPolla 2005. The archaic Tibeto-Burman reflexive is the verb suffix *-(n)si (van Driem 1993; LaPolla 2003), as in Tarong/Dulong (LaPolla 2005: 3): (14) ŋà sat-ɕī-ŋ 1sg hit-refl-1sg ‘I hit myself.’ Its reflex in Limbu is -siŋ, realized in these examples as -chiŋ (van Driem 1987: 86): (15) kɛ-niˑt-chiŋ-iˑ niˑt-chiŋ-aŋ 2-count-refl-q count-refl-1sg ‘Are you counting yourself ?’ ‘Yes, I’ve counted myself.’ (16) khɛʔoˑ məna-haʔ mɛ-n-chɛt-chiŋ-nɛn there man-pl pl-neg-kill-refl-neg ‘People don’t kill each other there.’

52   

   Scott DeLancey

In the archaic languages this morpheme functions as a middle construction, similar to Slavic or Romance: (17) àŋ ǎlɯp-ɕ̌ɯ  3sg happy-refl ‘He is happy.’ (18) warum-siŋ-aŋ  bathe-refl-1sg ‘I bathed.’

(Tarong, LaPolla 2005: 2)

(Limbu, van Driem 1987: 87)

This construction is lost in all of the creoloid languages, each of which has replaced it with a new free pronominal form (Tibetan rang, Boro gau, Burmese ko). These represent independent developments in each branch. The nominal sources of the Bodo-Garo and Burmese forms are cognate, both apparently deriving from Proto-Tibeto-Burman *kəw ‘body’ (Matisoff 2003), but neither reflexive construction is reconstructible to Proto-Bodo-Garo or Proto-Lolo-Burmese, so this must represent independent parallel innovation. The most obvious evidence for the antiquity of the morphological reflexivemiddle relative to the pronominal reflexive construction is that the morphological forms, are evidently cognate across many branches and languages, while the pronominal reflexive forms are not. Another, more deductive, argument is that the pronominal construction is what we would expect to see develop under creolization, while the morphological construction can only be the result of an extended period of development. Imagine the context of creolization, where speakers are trying to express themselves in a language which they know only imperfectly. For someone who does not know the native reflexive construction in the language, the simple and obvious strategy is simply to use the ordinary transitive clause construction, and fill the object slot with some noun which can easily be interpreted as anaphoric and coreferential. It is impossible to imagine a speaker in such a situation somehow concocting an affixal middle construction out of the linguistic material which he has at hand. We would expect the morphological middle/reflexive to be abandoned along with verbal argument indexation. This is not simply a matter of abandoning inflectional morphology in some typological drift pattern, but of abandoning a system in which the verb expresses its transitivity, argument structure and participants for one in which it expresses only the situational categories of tense/ aspect/modality/evidentiality, and the listener can recreate the scenario being described only by carefully tracking characters through their mention in independent noun phrases. Since pronouns are generally used only at first mention



Creolization in the Divergence of the Tibeto-Burman Languages   

   53

or contrastively,6 this can be quite complex – note the number of clauses in (13) with no mention whatever of the highly topical subject. (To this end the creoloid verb typically does express spatial deixis through the use of serialized motion verbs which help to keep track of the movements of various referents, as in both examples 12–13).

3 T  owards a history of creoloid languages in the Tibeto-Burman world Scholars have long noted the dramatically creoloid nature of Sinitic syntax, and the obvious connection to the origins of Chinese as a language of conquerors imposed on a non-Tibeto-Burman-speaking population (Terrien de la Couperie 1887; Benedict 1972; Hashimoto 1976a, b; Ballard 1984; van Driem 2001; Ansaldo and Matthews 2001; Blench 2008 inter alia; see DeLancey 2012, 2013). The prehistory of the Tibeto-Burman languages is even more obscure than that of Sinitic, and while there has been some recent attention to the role of language contact in the divergence of the family (van Driem 2001; LaPolla 2001, 2009; Jacquesson 2009; Sun and Liu 2009), the importance of creolization in the development of the modern languages deserves more attention than it has yet received.

3.1 Geography, history, and prehistory Southeast Asian populations, as well as those of Northeast India and the Himalayas, are dramatically split into what are generally called “Hill” and “Valley” cultures (Leach 1954; Cœdès 1962, 1964; Burling 1965; Scott 2009 inter alia). The essential difference for our present purpose is evident to everyone involved: I recall asking what the difference is, and being told again and again that ‘Vaai [Valley people] have towns and writings and kings while Zo [Hill people] do not’ … (Chit Hlaing 2007: 108)

Towns and cities, and kings with their attendant armies and conquests, are crucibles of creolization. Ethnic identities are notoriously fluid and opportunistic (Leach 1954; Barth 1981; Scott 2009 inter alia), but the Hill-Valley distinction is a 6 I.e. Boro is a “pro-drop” language, although to apply a special label to such languages and treat the English type as unmarked turns typological naturalness on its head.

54   

   Scott DeLancey

constant. The archaic vs. creoloid language type has a strong, though only a oneway, correlation with the Hill-Valley distinction. Highly archaic languages like Kiranti and Nung are found only among Hill communities, never in the Valley kingdoms. But though all dominant Valley cultures speak very creoloid languages, there are also many creoloid languages spoken by Hill peoples (Loloish, Tani, Mishmi, Naga). While the multiethnic and multilingual roots of Valley cultures like Thailand, Burma, and Assam are evident, we have a tendency to imagine the smaller, more fragmented Hill peoples as autochthonous bearers of long-distinct cultural lineages. Scott (2009) demonstrates at length that this is a gross oversimplification, and that the history of “Zomia,” his term for the upland area that we are concerned with here, has for the last two or even three millennia involved constant interchange with the Chinese, Tai, Burmese, and Assamese Valley cultures, including cultural as well as population exchange of many different kinds. In this perspective there is no surprise in seeing many creoloid languages spoken by Hill populations. And we can understand, from direct observation, something about the creolizing forces that are constantly in operation in the interaction of Hill communities. Even within “Zomia” there are dominant expansionist groups like the Jinghpaw (Leach 1954; Dai 1999; Maran 2007) and Nuosu (Harrell 2001), and dominated fugitive groups like the Akha (Geusau 2000) and Nung (Sun and Liu 2009; Gros 2010; Barnard 1934: ix-x, who describes them as perpetual victims of Shan to the west and Lisu to the east). In these contexts we see the way that dominance and institutionalized slavery drive creolization. The conservative Jinghpaw spoken in Burma and Yunnan has some notable creoloid tendencies, but retains a complex archaic verbal system (Hanson 1917; Dai and Xu 1992; DeLancey 2010b, 2011a). Singpho is a creoloid version of Jinghpaw, spoken in Assam as a result of Jinghpaw invaders enslaving local Ahom, Assamese and other speakers (Leach 1954; Maran 2007). Sun and Liu (2009) document the rapid creolization of Anong (Nung), one of the most archaic groups in the family, under pressure from more dominant Lisu. When we are trying to explain the creoloid structure of whole branches, rather than individual languages, our explanations must be deeper and more ambitious. For several substantial but not very internally divergent branches – Bodish, Bodo-Garo, and Lolo-Burmese – it appears that we must reconstruct an essentially creoloid morphosyntax for the protolanguage. Bodish appears to be the least so; the elaborate morphology of the Classical verb is not very creoloid.7 But except for the relict prefixes in the Eastern Tibetan languages and some sec7 Most of the rest of the notorious irregularity of the Classical verb, including the verbal ablaut, can be accounted for as secondary developments (Coblin 1976).



Creolization in the Divergence of the Tibeto-Burman Languages   

   55

ondary irregular stem ablaut and tone shift in Tibetan languages, the modern Tibetan, East Bodish and Tamangic languages, and Tshangla are all thoroughly creoloid. The situation is clearer for Proto-Lolo-Burmese and Proto-Bodo-Garo; although they certainly retained some old derivational morphology on the verb, the most economical reconstruction schemes for both posit a proto-language with the same creoloid syntax as its daughters. The linguistic histories of all three branches look quite parallel – starting about 2,500± years ago as languages holding on to a bit of ancient verbal morphology but innovating creoloid syntax on the basis of nominalization constructions (DeLancey 2010b, 2011b), and ending up thoroughly creoloid.

3.2 The origin of Bodo-Garo Bodo-Garo probably represents the most thoroughly creolized and mixed subbranch in the family, so its history provides an illustration of how this typology may come about. Bodo-Garo is a small, close-knit subbranch, consisting of about a dozen languages whose degree of divergence is roughly comparable to Tamangic, Tibetic, or Loloish. As we have seen in Boro, there is very little paradigmatic morphology, but parts of the TAME system seen to be reconstructible for ProtoBodo-Garo, which can thus be inferred to have had a similar morphosyntactic profile to the modern languages. There are two very striking things about Bodo-Garo, which make it stand out from other units of comparable size and divergence. One is its extreme creoloid grammar. The other is its considerable geographical spread: from the Meche language of Nepal in the west to Dimasa in eastern Assam is over 1200 kilometers. In modern times the Bodo-Garo languages form a patchwork across the Brahmaputra plain and beyond; prior to the Ahom conquests of the 13th-14th centuries and the later Aryanization of the Assam plain, Bodo-Garo probably covered the entire Brahmaputra Valley and modern Bangladesh, and much of modern West Bengal. We have no direct attestation of any Bodo-Garo language before modern times, but the dynasties which ruled Kāmarūpa (modern Assam) in the late 1st and early 2nd millennium CE have always been identified as Boro or “Kachari” (Gait [1926] 1983). Prāgjyotisha, the predecessor to Kāmarūpa in Assam, and its king participate in the Kurukshetra War in the Mahabharata, with armies of non-Aryan Kirāta and Mleccha, whom tradition and Assamese historians interpret as yellow-skinned Tibeto-Burman speakers (Barua [1933] 1960, S. Baruah 1985). The historicity and dating of the Mahabharata iself and the events described in it are uncertain

56   

   Scott DeLancey

and controversial (Sircar 1990a), but this reference constitutes evidence of nonIndic people subject to an urbanized kingdom as early as the mid-1st millennium BCE, and probably as far back as 1,000 BCE or more, some of them presumably Tibeto-Burman-speaking. The appearance of Prāgjyotisha in the Mahabharata is often taken, particularly by Indian scholars, as implying that by that point it was already a Hindu kingdom, but this evidence is not compelling (Sircar 1990b). It is improbable that Indic speech could have had any but the barest foothold in Assam prior to the Common Era. The first attestation of Indic in Assam is some Sanskrit rock inscriptions, and slightly later copperplate land grants, from the 5th century CE (Sircar 1990c; N. Boruah 2007). In the 7th century a Chinese Buddhist pilgrim, Xuanzang, who visited Kāmarūpa on his way back to China from a visit to Buddhist centers in India, described the language as “slightly different from that of Central India” (Xuanzang 1996: 299), so by that time Indic was established at least as a court language. But even if Tibeto-Burman speakers arrived in the Brahmaputra Valley 3,000 years ago (or more, see van Driem 2001), and even assuming that Bodo-Garo languages were spoken throughout the length and breadth of the valley 1,000 and more years ago, they would never have been the only languages spoken in the area. There is general agreement, among Tibeto-Burman communities in Assam and their traditions as well as among scholars, that Tibeto-Burman languages came into the Assam plain from the north and probably east, though opinions differ about how far north and/or east the original center of dispersal for the family is. There can be no question that there were speakers of other languages – certainly including Austroasiatic, but very possibly others as well – in the valley when Tibeto-Burman languages first entered (Kakati [1941] 1962; Choudhury 1987 inter alia). Thus the earliest kingdoms must have been already multiethnic and multilingual, regardless of whether they were founded by locals or immigrants from India. What this means is that increasingly from about 1,000 BCE on, the area which we can infer to have been substantially Bodo-Garo-speaking a millennium ago has been a multilingual zone subject to the waxing and waning of one or more states centered in Assam (and very likely influenced by the waxing and waning of other state-level polities in Bhutan, Tibet, and perhaps eastern kingdoms in northern Burma and Yunnan), which at some times was officially Indic-speaking. This is the kind of situation in which we can expect some language to develop into a lingua franca; this kind of economic-political system depends upon a vehicular language (Calvet 1981, 1987). So, we have a situation in which we would expect a vehicular language to arise, at a time when there were some substantial numbers of Tibeto-Burman speakers coinhabiting the plains of Assam, where in historical times we find



Creolization in the Divergence of the Tibeto-Burman Languages   

   57

a widespread but shallow Tibeto-Burman language group with dramatically stripped-down grammatical structure which immediately provokes images of creolization (Burling 2007). Although there is no actual historical attestation, it seems reasonable to identify Bodo-Garo, which in its structure and distribution looks very much like a former creolized lingua franca, with the lingua franca which we can infer to have been in use at the time and in the place where ProtoBodo-Garo must have formed. The exact mechanisms by which this may have occurred can only be inferred; common functions which could have been relevant include a trade language, the lingua franca of the army, or the language of administration and communication between an Indic or Tibeto-Burman-speaking court and a Tibeto-Burman and/or Austroasiatic-speaking population.

3.3 State formation and creolization Thus there is ample reason to link the formation of Proto-Bodo-Garo to the rise of the Prāgjyotisha-Kāmarūpa state. While the small-scale mechanisms which we have discussed in the introduction to this section as giving rise to Singpho were undoubtedly operating already in the times when branches like Bodo-Garo and Lolo-Burmese and Bodish were forming, these three in particular seem to have gotten off to very vigorous starts. These starts all were in times and places when growth and consolidation of early urban-centered states was accelerating. In the following sections I will suggest that this is not simply coincidence, but that these branches began as highly creolized varieties which spread presumably through state formation and consolidation and imperial conquest and administration.

3.3.1 Speculations on Lolo-Burmese The Hòu Hàn Shū 後漢書, a 5th century history of the Eastern Han kingdom, which contains extensive material on the western and southwestern frontiers, records a set of songs in the language of a non-Chinese people from the west of Sichuan called “Báiláng” 白狼, from the 1st century CE. This language is clearly Tibeto-Burman, with close lexical and phonological ties to Lolo-Burmese (Coblin 1974; Ma and Dai 1982; Beckwith 2008). Coblin (1974) notes that the phonology of Báiláng is more conservative than attested Lolo-Burmese languages, as we would expect. Ma and Dai (1982) note connections to both Burmese and Yi (as well as Naxi), and suggest that Báiláng represents a link between Burmish and Loloish. Both Coblin’s and Ma and Dai’s observations imply that the language is not far removed from Proto-Lolo-Burmese.

58   

   Scott DeLancey

If the Báiláng king and his people, as described in the Hòu Hàn Shū, spoke late Proto-Lolo-Burmese or an early offshoot of it, this lets us place the early development of Proto-Lolo-Burmese in Yunnan/Sichuan in the first millennium BCE, and place its early divergence in the second half, more or less contemporaneous with the Han Dynasty. We know from Chinese historical sources of a major state, the Shǔ 蜀 kingdom, in this place and time; with the discovery of the archaeological site of Sānxīngduī 三星堆, evidence for state-level organization and urbanization in Sichuan dates back to the late 2nd millennium BCE (Sage 1992). Thus, as in the case of Bodo-Garo, in precisely the time and place where we have historical evidence that Proto-Lolo-Burmese, one of the most highly creolized Tibeto-Burman branches, was formed, we also find the conditions which regularly lead to creolization. Chang (2005:126) cites historical assertions that in the realm of early China, 10,000 states (國) at the beginning of the Xia dynasty (2100 BCE) were reduced by 500 BCE by two orders of magnitude to 100. Of course this consolidation was not achieved peacefully: The wars and annexations of one state by another described by Gu Zuyu evidently started at the end of the Longshan period and resulted in the concentration of wealth and power—the core issues in the development of ancient China … The archaeological evidence for violent institutionalized warfare was quite extensive during the Longshan period [龍山, ca 3,000– 2,000 BCE]. (Chang 2005: 126)

The same applies no doubt to the history of Prāgjyotisha, and certainly to the growth of the urbanized state in Sichuan, from Sānxīngduī to Shǔ, and in Yunnan, from Dian to Nanzhao (Pirazzoli-t’Serstevens 1974). Thus once again, just as in Assam at about the same time, we have an extended historical phase involving urban centers and kingdoms, with multiethnic armies roaming around multiethnic empires, pulling and pushing anybody and everybody in the region into motion, in exactly the place and time where we have to place the formation of Proto-Lolo-Burmese, a branch for which we have every reason to reconstruct a creoloid proto-language.

3.3.2 Bodish We do not know enough about the state of urbanization in the east-central Himalayas in the first millennium BCE, or about the origin of Bodish as a distinct unit, to be able to speculate as concretely about this branch as we can about Bodo-Garo and Lolo-Burmese. But we know that when Tibetan is first recorded, it has the basic creoloid structure – the elaborate (for a creoloid Tibeto-Burman language)



Creolization in the Divergence of the Tibeto-Burman Languages   

   59

verbal morphology is still and all a nominalized construction, and in its essential structure the verb functions as in a creoloid language, providing grounding information but no recapitulation of the event and the arguments. Zeisler (2009, in preparation) presents a very clear case that Tibetan as we know it has undergone creolization effects, particularly in phonology: With the conquests of oasis towns in Eastern Turkistan, Tibet gained control over essential parts of the Silk Route, and this apparently led to the development of a particular *Archaic Tibetan variety into a lingua franca, or rather to the adoption of an already existing lingua franca as the language of administration and commerce in the new empire. (Zeisler 2009: 78 emphasis in original)

The idea of an “already existing lingua franca” is important here. Most of Zeisler’s argument is concerned with contact effects dating from the establishment of the historic Tibetan empire. But she also presents a tentative but very intriguing speculation that the language which spread with the empire was itself an older lingua franca which moved into what became the center of Tibetan imperial expansion two or three centuries earlier (Zeisler 2009: 87–8). This is important because the other Bodish languages – the Tamangic group and the East Bodish languages of Bhutan – are also resolutely creoloid in structure. There are various scenarios which could explain this. One which cries out for further research is the possibility that the Tamangic and/or East Bodish languages do not represent sister developments from a common Proto-Bodish language, but are the result of the imposition of an early Tibetan language on nonBodish speaking populations (see below, and Hyslop’s [2011] discussion of the ancestry of Kurtöp). It is also possible that Proto-Bodish had a more archaic structure and that the creoloid structure which we see in all the attested languages represents parallel secondary developments after the breakup of the proto-language. But a third possibility, which I want to put forward here, is that Proto-Bodish itself was already creoloid, so that the grammatical typology which is shared by the Bodish languages represents shared inheritance. It is impossible to say very much at present about the relative merits of these hypotheses (but see Zeisler in preparation for essential background to pursuing the question), but let us briefly explore the third. We have no concrete knowledge (but some recorded traditions concerning earlier dynasties) of Tibet or of Tibetan before it is recorded as a written language in the 7th century, but by the time the Tibetan language appears in history for the first time it is already a language of empire, which is why it appears in history for the first time. The emergence of Tibet as an imperial power must have been preceded by a period of state building, but beyond that we can say nothing definite. We know that imperial Tibet, like all other state-level polities of those times,

60   

   Scott DeLancey

impressed conquered peoples into its army (Richardson [1962] 1984: 31; Dotson 2009: 55). Undoubtedly this was the practice from the beginning, so that we can easily imagine the formation of a soldier’s lingua franca long before we have direct documentation of the Tibetan empire or the Tibetan language. At the time we first have documentation, Tibet is clearly a “Valley” culture: By the seventh century they already had walled towns and small castles surrounded by farmland; they were also skilled workers in metal, making highly serviceable armour and weapons as well as fine decorative gold mail for ceremonial use, and elaborate golden utensils. Such crafts suggest several centuries of development. (Richardson 1984: 29)

“Several” centuries back from the 7th century is probably getting beyond ProtoTibetic to the next node higher on the tree, i.e. common Tibetan-East Bodish, and perhaps well towards common Bodish, and thus to the days when we must imagine the conditions under which Proto-Bodish might have formed as a creoloid. To the west and north the polity (or one or another of them) which we refer to as “Zhang-zhung” can be traced back as far as 500 BCE (Aldenderfer and Zhang 2004, Aldenderfer 2007). The influence of Zhang-zhung in the Tibetan Plateau and the Himalayas provides a context for creolization effects in Tibeto-Burman languages of the region well before the Tibetan Empire emerges on the historical scene. This is important; although it is imaginable, as we have noted, that the spread of creoloid typology through the various Bodish languages could be a secondary effect of Tibetan political and later cultural imperialism, the simplest and most economical explanation for the stark typological split between Bodish and most of the rest of Western Tibeto-Burman is that Proto-Bodish, like Proto-BodoGaro and Proto-Lolo-Burmese, was already highly creolized when it first began to diverge into the daughter languages. While it is permissible to imagine a period of several centuries of urbanization and state-building leading up to the first emergence of Tibet on the larger regional stage, it is hard to imagine the divergence of Proto-Bodish any later than the beginning of the Common Era; the overall degree of divergence of the branch is comparable to, and arguably somewhat greater than, that of Bodo-Garo and Lolo-Burmese. Of course creolization pressures have continued since the time of Proto-Bodish and Old Tibetan. It is well-known that the Central Tibetan languages are much more progressive, phonologically as well as grammatically, than many of the geographically marginal varieties. In this connection Zeisler notes that: The innovation of the southern variety [i.e. Central Tibetan] might have had less to do with its centrality than with a possible contact or even mixing with other linguistic communities (Tibeto-Burman, Austro-Asiatic, and/or Indo-Arian) … (2005: 55)



Creolization in the Divergence of the Tibeto-Burman Languages   

   61

But “centrality” and “contact” are not competing alternatives, they are two perspectives on the same thing. The reason why language change is often accelerated in central, urban portions of a kingdom or empire is precisely because that is where dialect and language contact and mixture takes place (see Zeisler 2009). Certainly the reason why phonological and morphological change have proceeded more rapidly in Dbu-Gtsang than elsewhere is because within historic times this has been the most cosmopolitan part of the Tibetan-speaking world. The history of Tibetan, like that of any expansionist language, must have involved substantial and sustained contact influence. It is evident that when Tibetan culture expanded across the Tibetan Plateau, there were populations there already. Without historical records we cannot identify these pre-Tibetan inhabitants as to ethnicity or language, but there is historical, archaeological, and linguistic evidence to link them with a widespread Central Eurasian cultural complex which must have included Indo-European and Altaic elements (Roerich 1930; Eberhard [1942] 1979; Beckwith 2002, 2009; Aldenderfer 2007; Walter 2009, Zeisler 2009, in preparation). Tibetan sources refer to a pre-Tibetan stratum in the southern valleys where the Yarlung state was born (Eberhard 1979; Haarh 1969). While these accounts cannot be taken at face value, with or without them we can infer the same predatory process of state-building and expansion which we can now see in the archaeological record in Sichuan, Assam and Burma, and thus infer the same kinds of creolization process in the formation of Proto- or PreProto-Tibetan.8 We can only speculate about the chronological distance which separates Proto-Tibetan from Proto-Bodish. With the tremendous expansion of Tibetan cultural influence over the last millennium, other Bodish languages have been considerably affected by Tibetan, and in their attested forms clearly are more similar to Tibetan than they once were. At present we have no way to distinguish between creoloid features of modern Bodish languages which represent inheritance from a hypothetically creoloid Proto-Bodish and the effects of later creolization under Tibetan (or other) influence.

8 Some readers of an earlier version of this paper object to this inference, and argue (if I understand them correctly) that it is illegitimate to infer the existence of any kind of organized state-level polity much earlier than the beginning of reliable historical records. This strikes me as implausible; urban civilizations with state-level political organization do not simply spring up like mushrooms. Similar assumptions about the rise of kingdoms in other parts of the SinoTibetan area have had to be abandoned in the light of archaeological research (Stargardt 1990; Sage 1992; Aldenderfer 2007; N. Boruah 2007), and at least until we have better archaeological information, we should not suppose that Tibet is unique in this regard.

62   

   Scott DeLancey

The genetic classification of the Western or Bodic branch of Tibeto-Burman is approximately as follows: Tab. 3: Relationships within Western Tibeto-Burman Western TB Kiranti Kham-Chepang

1 Western Himalayan

Bodish Tshangla

2 Tamangic

3 Tibetan East Bodish

Since Kiranti and Kham-Chepang retain most of the original Tibeto-Burman verbal system, and West Himalayan a great deal of it, we cannot reconstruct creoloid typology farther back than the Bodish node on the tree. The most economical hypothesis would reconstruct Old Tibetan-like morphosyntax back as far as Proto-Bodish, but it is certainly possible that Proto-Bodish shared archaic traits with its sister Western Himalayan, and that the radical creolization which we see in modern Tshangla, Tamangic, and even East Bodish may represent changes which have occurred more recently, even during historical times. Local myths of ethnogenesis in this region often invoke ethnic mixing, often with a military theme. For example, a common folk etymology, in both Tibetan and Tamang, for the ethnic name Tamang, derives it from Tibetan rta-dmag (Lhasa /támaʔ/ ‘horse-soldier’ with an accompanying story about an army of Srong.btsan Sgam.po (circa 620–649) or Khri.srong Lde.btsan (755–797), sent to fight in Nepal and deciding to defect and not return to Tibet. (An internet search for “Tamang” will turn up numerous references to this account). This account cannot be accepted as literally true – if the origin of Tamang were this simple and this late, the Tamangic languages would be Tibetic, which they are not. (Moreover the necessary sound change would be quite irregular). But in recognizing the multiethnic origins of modern groups, such myths speak to the palpably creoloid nature of the languages in question. A more interesting and plausible tradition makes the Tamang the predecessors to the Newars in the Kathmandu Valley, placing speakers of pre-Tamang in a sister city-state to Yarlung in the latter days of the legendary pre-Tibetan kings, whose people, or at least armies, would have



Creolization in the Divergence of the Tibeto-Burman Languages   

   63

spoken Proto-Tibetan. This chronology makes reasonable sense in terms of the linguistic distance of Tamang from Tibetan, and places the formation of something like Proto-Tamangic in an early phase of urban state formation in Nepal, where it would have finished off its creolization process as the language of the urban center for and a lingua franca in Central and Eastern Nepal.

3.4 Language mixture in Zomia So we see that three of the largest and most conspicuous creoloid units have histories that involve massive, intense, long-term language contact, and probably long-term status as a lingua franca, connected with the development of state-level political structures. To these we may add the highly creoloid Meithei, the lingua franca and in recent times the state language of the valley of Manipur (see Singh 2002). But this type of explanation is not available for the other major pockets of creoloid structure – Arunachal Pradesh, Nagaland, and adjacent areas in China and Burma. Many of these peoples have traditions of having migrated to their present locations from the north and/or east, and some tell just-so stories along the lines of the Tamang horse-soldier account. The Jinghpaw, for example, are reported as relating that they moved west to their modern locations to escape conscription in the wars of the Three Kingdoms period (Wang 1997), and Denwood (2005) reports similar ethnogenetic traditions concerning the eastern Wa-skad dialects of Tibetan. The area where we find the Tani, Mishmi and Naga languages falls just outside the edge of greatest influence of major imperial expansion zones in all directions: Zhang-zhung and then Tibet in the west and north, Dian-Nanzhao, Sanxingdui-Shu, and China to the north and east, Pyu-Pagán to the east (Pyu, or some predecessor kingdom, has been an important factor in the region for a thousand years or more longer than historians have thought; see Stargardt 1990), Prāgjyotisha-Kāmarūpa and Manipur to the south, and steady Aryan pressure from the southwest.9 All of these were in contact with one another (Pelliot 1904) and sometimes exerted military force toward one another, so that our upland refuge zone could be expected to have collected the detritus of two or three millennia’s worth of military campaigns which have oozed through and around it. In each direction we have imperial urbanizing polities demonstrably dating at least to the mid-1st millennium BCE.

9 For example, local history records a long-term pattern of raiding into the Boro chiefdoms of northern Assam by Tibetic-speaking Bhutiyas from the north (A. Boruah 2011).

64   

   Scott DeLancey

In general it seems that what we have in these languages is the linguistic manifestation of the long-term interchange between Hill and Valley cultures recently described in detail by Scott (2009): Valley runaways have been replenishing the hill population for as long as we can tell. Hill people have been assimilating into valley-state societies also for as long as we can tell. The essentialized “line” between hill and valley peoples remains in place despite quite massive traffic back and forth in each direction. (Scott 2009: 332)

The languages which we have to account for – Tani, Naga, and small groups like Mishmi in between them – are all in the crunch zone where in pre-modern times the zone of influence of the major imperial polities of Assam, Tibet, China, and Burma fade out into the hills. This is exactly the zone which Scott is referring to, and his interpretation of the relations between Hill and Valley cultures in this area provides ample explanation for the existence of highly Creoloid languages in the mountainous refuge zone. This could very well be the explanation for the difficulty of establishing the wider genetic relations within Tibeto-Burman for small groups like Tani and Mishmi; their linguistic history may have involved episodes of interchange with various valley cultures, Tibeto-Burman speaking or otherwise. For example, Shafer’s speculations about Bodish affinities for Tani, Mishmi and Miju (his Miśingish, Digarish, and Midźuish; 1966: 3–4) could have very different interpretations, involving not direct lineal transmission of lexical items from an ancestral Bodish stock, but lateral transmission from Bodish languages to the north and west, through complex processes involving military contact, colonization, trade, and assimilation of Bodic-speaking refugees into Hill ethnicities (see also Blench and Post this volume).

4 Conclusion We have abundant attested examples of the process of creolization in TibetoBurman (see DeLancey 2010b for additional examples and discussion). In every case this involves a radical simplification of the verbal system, replacing a tightly structured paradigm of argument indexation and specification of transitivity. Thus this must be taken as a likely hypothesis to explain similar changes which have occurred in times and places where we have no direct knowledge of the external situation. When we look at what we do know or can plausibly infer about the context in which the major creoloid branches (Sinitic, Bodo-Garo, Bodish, and Lolo-Burmese) of the family developed, in every instance a good case can be



Creolization in the Divergence of the Tibeto-Burman Languages   

   65

made that these branches were born in the kinds of situations where creolization is to be expected. Thus in absence of some other equally plausible account, this must be taken as the default explanation for the dramatic typological divergence of the branches and languages of Tibeto-Burman. One important implication of this account is that our estimates of the time depth of the family are probably greatly exaggerated. If we think of the diversity of morphosyntactic structure which we see in the family as having evolved through “normal” language-internal processes,10 we will infer a long process of development. But under creolization a language can undergo very radical changes in phonology, lexicon and grammar in a very short time, measurable in centuries (the split of creoloid Singpho from archaic Jinghpaw is only a few centuries back; see S. Baruah 1985: 376, T. Baruah 1977) or even generations (again see Sun and Liu 2009 for a detailed case study of extremely rapid typological shift). It is quite plausible, and in my opinion probable, that the time of the breakup of Proto-Tibeto-Burman (including Sinitic) could be not that far – conceivably a millennium or less – prior to the time, 2,500–3,500 years ago, when we see the first creoloid branches, Sinitic, Bodo-Garo, and Lolo-Burmese, emerging into the light of history.

Abbreviations abs absolutive; ag agent; aor aorist; caus causative; cls classifier; conj conjunction; du dual; erg ergative; gen genitive; loc locative; neg negative; nf nonfinal; nmz nominalizer; obj object; pat patient; pl plural; q interrogative; real realized; refl reflexive; sg singular

References Aldenderfer, Mark. 2007. Defining Zhang zhung ethnicity: an archaeological perspective from far Western Tibet. In Amy Heller and Giacomella Orofino (eds.). Discoveries in Western Tibet and the Western Himalayas (Brill’s Tibetan Studies Library 10/8), 1–21. Leiden: Brill. Aldenderfer, Mark, and Yinong Zhang. 2004. The prehistory of the Tibetan Plateau to the Seventh Century A.D.: Perspectives and research from China and the West since 1950. Journal of World Prehistory 18 (1): 1–55.

10 Which, as we are increasingly realizing, are not normal at all. More and more we are seeing that purely internal language change is the exceptional rather than the typical case.

66   

   Scott DeLancey

Ansaldo, Umberto, and Stephen Matthews. 2001. Typical creoles and simple languages: The case of Sinitic. Linguistic Typology 5 (2–3): 311–326. Ballard, W. L. 1984. The mother soup: A South China recipe for tonometamorphogenesis. Computational Analysis of African and Asian Languages 22: 43–70. Barnard, J. T. O. 1934. A handbook of the Rawang Dialect of the Nung Language. Rangoon: Supt. of Government Printing and Stationery. Barth, Fredrik. 1981. Features of person and society in Swat: Collected essays on Pathans. (Selected essays of Fredrik Barth, Volume II). London: Routledge and Kegan Paul. Barua, K. L. [1933] 1960. Early history of Kāmarupa, From the earliest times to the end of the Sixteenth Century. Shillong: K.L. Barua Bahadur. Repr. Guwahati: Lawyers Book Stall. Baruah, S. L. 1985. A comprehensive history of Assam. New Delhi: Munshiram Manoharlal. Baruah, Tapan Kumar. 1977. The Singphos and their religion. Shillong: Govt. of Arunachal Pradesh. Beckwith, Christopher. 2002. The Sino-Tibetan problem. In Christopher Beckwith (ed.). Medieval Tibeto-Burman languages, 113–157. Leiden: Brill. Beckwith, Christopher. 2008. The Pai-lang songs: The earliest texts in a Tibeto-Burman language and their Late Old Chinese transcriptions. In Christopher Beckwith (ed.). Medieval Tibeto-Burman languages III (Proceedings of the 11th Seminar of the International Association for Tibetan Studies), 87–110. Halle: International Institute for Tibetan and Buddhist Studies GmbH. Beckwith, Christopher. 2009. Empires of the Silk Road: A history of Central Eurasia from the Bronze Age to the present. Princeton: Princeton University Press. Benedict, Paul. 1972. Sino-Tibetan: A conspectus. Cambridge: Cambridge University Press. Blench, Roger. 2008. Stratification in the peopling of China: How far does the linguistic evidence match genetics and archaeology? In Alicia Sanchez-Mazas, Roger Blench, Malcolm Ross, Ilia Peros and Marie Lin (eds.). Past human migrations in East Asia: Matching archaeology, linguistics and genetics, 105–132. London: Routledge. Boruah, Ajit Kumar. 2011. The Koch kingdom. Guwahati (India): DVS Publishers. Boruah, Nirode. 2007. Early Assam: State formation, political centres, cultural zones. Guwahati and Delhi: Spectrum Publications. Boro, Krishna. 2012. Serialized verbs in Boro. In Gwendolyn Hyslop, Stephen Morey and Mark Post (eds.). North East Indian linguistics 4, 83–103. New Delhi: Foundation/Cambridge University Press India. Bradley, David. 1995. Reflexives in Burmese. In David Bradley (ed.). Studies in Burmese languages. (Papers in Southeast Asian linguistics 13), 139–172. Canberra: Australian National University. Bradley, David. 1997. Tibeto-Burman languages and classification. In David Bradley (ed.). Tibeto-Burman languages of the Himalayas (Pacific Linguistics Series A, no.86; Papers in Southeast Asian Linguistics 14), 1–72. Canberra: Australian National University. Burling, Robbins. 1965. Hill farms and padi fields: Life in Mainland Southeast Asia. Englewood Clffs, NJ: Prentice-Hall. Burling, Robbins. 2007. The lingua franca cycle: Implications for language shift, language change, and language classification. Anthropological Linguistics 49 (3–4): 207–236. Calvet, Louis-Jean. 1981. Les langues véhiculaires. Paris: Presses Universitaires de France. Calvet, Louis-Jean. 1987. La guerre des langues et les politiques linguistiques. Paris: Payot.



Creolization in the Divergence of the Tibeto-Burman Languages   

   67

Chang, Kwang-chih. 2005. The rise of kings and the formation of city-states. In Sarah Allan (ed.). The formation of Chinese Civilization: An archaeological perspective, 125–139. New Haven: Yale University Press. Choudhury, Pratap Chandra. 1987. The history of civilisation of the people of Assam to the twelfth century A.D. Delhi: Spectrum. Chit Hlaing, F. K. L. (= F. K. Lehman). 2007. Some remarks upon ethnicity theory and Southeast Asia, with special reference to the Kayah and the Kachin. In Mikael Gravers (ed.). Exploring Ethnic Diversity in Burma (NIAS Studies in Asian Topics 39), 107–22. Copenhagen: Nordic Institute of Asian Studies. Cœdès, George. 1962. Les peoples de la péninsule indochinoise. Paris: Dunod. Cœdès, George. 1964. Les états hindouisés d’indochine et d’indonésie. Paris: F. de Boccard. Coblin, W. South. 1974. New study of the Pai-lang songs. Tsing-hua Journal of Chinese Studies (Qīnghuá xuébào) 12: 179–211. Coblin, W. South. 1976. Notes on Tibetan verbal morphology. T’oung Pao 52 (1–3): 45–60. Dahl, Östen. 2004. The growth and maintenance of linguistic complexity. Amsterdam and Philadelphia: Benjamins. Dai, Qingxia and Xu Xijian.1992. Jǐngpō yǔ yǔfǎ [Jingpo grammar]. Beijing: Mínzú chūbǎnshè. Dai, Qingxia. 1999. On the languages of the Jingpo nationality. Linguistics of the Tibeto-Burman Area 22 (2): 1–59. DeLancey, Scott. 2010a. Language replacement and the spread of Tibeto-Burman. Journal of the Southeast Asian Linguistics Society 3 (1): 40–55. http://www.jseals.org/JSEALS-3–1.pdf (accessed 4 December 2012). DeLancey, Scott. 2010b. Towards a history of verb agreement in Tibeto-Burman. Himalayan Linguistics 9 (1): 1–38. http://www.linguistics.ucsb.edu/HimalayanLinguistics/ articles/2010/PDF/HLJ0901A.pdf (accessed 4 December 2012). DeLancey, Scott. 2011a. Nocte and Jinghpaw: Morphological correspondences. In Gwendolyn Hyslop, Stephen Morey and Mark Post (eds.). North East Indian linguistics 3, 61–75. New Delhi: Foundation/Cambridge University Press India. DeLancey, Scott. 2011b. Finite structures from clausal nominalization in Tibeto-Burman languages. In Foong Ha Yap, Karen Grunow-Hårsta and Janick Wrona (eds.). Nominalization in Asian languages: Diachronic and typological perspectives, 343–362. Amsterdam and Philadelphia: John Benjamins. DeLancey, Scott. 2012. On the origins of Bodo-Garo. In Gwendolyn Hyslop, Stephen Morey and Mark Post, (eds.). North East Indian linguistics 4, 3–20. New Delhi: Foundation/Cambridge University Press India. DeLancey, Scott. 2013. The origins of Sinitic. In Zhou Jing-Schmidt (ed.), Increased empiricism: New advances in Chinese linguistics. 73–100. Amsterdam and Philadelphia: Benjamins. DeLancey, Scott, and Krishna Boro. in preparation. A grammar of spoken Boro. Denwood, Philip. 2005. Early connections between Ladakh/Baltistan and Amdo/Kham. In John Bray (ed.). Ladakhi histories: Local and regional perspectives, 31–39. Leiden: Brill. Dotson, Brandon. 2009. The Old Tibetan Annals: An annotated translation of Tibet’s first history (Philosophisch-Historische Klasse Denkschriften 381). Vienna: Österreichische Akademie der Wissenschaften. van Driem, George. 1987. A grammar of Limbu. Berlin and New York: Mouton de Gruyter. van Driem, George. 1993. The Proto-Tibeto-Burman verbal agreement system. Bulletin of the School of Oriental and African Studies 56 (2): 292–334.

68   

   Scott DeLancey

van Driem, George. 1997. Sino-Bodic. Bulletin of the School of Oriental and African Studies 60 (3): 455–488. van Driem, George. 2001. Languages of the Himalayas. Leiden: Brill. van Driem, George. 2005. Tibeto-Burman vs. Indo-Chinese: Implications for population geneticists, archaeologists and prehistorians. In Laurent Sagart, Roger Blench, and Alicia Sanchez-Mazas (eds.). The peopling of East Asia: Putting together archaeology, linguistics and genetics. London and New York: RoutledgeCurzon. Eberhard, Wolfram. [1942] 1979. Kultur und Siedlung Randvölker Chinas. (Supplement to T’oung Pao 36, repr. 1979). Leiden: Brill. Gait, Edward. [1926] 1983. A history of Assam. Calcutta: Thacker, Spink & Co. Repr. Guwahati (India): LBS Publications. Genetti, Carol. 1988. A contrastive study of the Dolakhali and Kathmandu Newari dialects. Cahiers de Linguistique – Asie Orientale 17 (2): 161–191. Geusau, Leo Alting von. 2000. Akha internal history: Marginalization and the ethnic alliance system. In Andrew Turton (ed.). Civility and savagery: Social identity in Tai states, 122–158. Surrey: Curzon. Gros, Stéphane. 2007. The missing share: The ritual language of sharing as a ‘total social fact’ in the Eastern Himalayas (Northwest Yunnan, China). In François Robinne and Mandy Sadan (eds.). Social dynamics in the highlands of Southeast Asia, 257–282. Leiden: Brill. Haarh, Erik. 1969. The Yar-Luṅ Dynasty. København: G.E.C. Gad’s Forlag. Hanson, Ola. 1917. A hand-book of the Kachin or Jinghpaw language. Rangoon: American Baptist Mission Press. Harrell, Stevan. 2001. Ways of being ethnic in Southwest China. Seattle: University of Washington Press. Hashimoto, Mantaro. 1976a. The agrarian and pastoral diffusion of languages. In Genetic relationship, diffusion, and typological similarities of East & Southeast Asian languages: Papers for the 1st Japan-US Joint Seminar on East & Southeast Asian Linguistics, 1–14. Tokyo: Japan Society for the Promotion of Science. Hashimoto, Mantaro. 1976b. Language diffusion on the Asian continent – Problems of typological diversity in Sino-Tibetan. Computational Analyses of Asian and African Languages 3: 49–65. Hyslop, Gwendolyn. 2011. A grammar of Kurtöp. Eugene, OR: University of Oregon dissertation. Jacques, Guillaume. 2004. Phonologie et morphologie du japhug (rGyalrong). Paris: Université Paris VII dissertation. Jacques, Guillaume. 2010. The inverse in Japhung Rgyalrong. Language and Linguistics 11(1): 127–157. Jacquesson, François. 2009. East meets West: The Assam corridor. In Francesco d’Errico and Jean-Marie Homber (eds.). Becoming eloquent: Advances in the emergence of language, human cognition, and modern cultures, 147–61. Amsterdam and Philadelphia: Benjamins. Kakati, Banikanta. [1941] 1962. Assamese, its formation and development: A scientific treatise on the history and philology of the Assamese language. Gauhati, India: Govt. of Assam Dept. of Historical and Antiquarian Studies, Narayani Handiqui Historical Institute. Repr. Guwahati: Lawyer’s Book Stall. LaPolla, Randy. 2001. The role of migration and language contact in the development of the Sino-Tibetan language family. In R. M. W. Dixon and A. Aikhenvald (eds.). Areal diffusion and genetic inheritance: Case studies in language change, 225–54. Oxford: Oxford University Press.



Creolization in the Divergence of the Tibeto-Burman Languages   

   69

LaPolla, Randy. 2003. Overview of Sino-Tibetan morphosyntax. In Graham Thurgood and LaPolla, Randy (eds.). The Sino-Tibetan languages, 22–42. London: Routledge. LaPolla, Randy. 2005. Reflexive and middle marking in Dulong-Rawang. Himalayan Linguistics 2: 1–13. http://www.linguistics.ucsb.edu/HimalayanLinguistics/articles/2004/PDF/ HLJ02_LaPolla.pdf (accessed December 4, 2012). LaPolla, Randy. 2009. Causes and effects of substratum, superstratum and adstratum influence, with reference to Tibeto-Burman languages. In Yasuhiko Nagano (ed.). Issues in Tibeto-Burman historical linguistics (Senri Ethnological Studies 75), 227–237. Osaka: National Museum of Ethnology. Leach, E. R. 1954. Political systems of highland Burma: A study of Kachin social structure. London: Bell & Sons. Ma Xueliang and Dai Qingxia. Báiláng gē yánjiū [Study of the Bailang songs]. Mínzú yǔwén 1982 (5): 16–26. Maran, LaRaw. 2007. On the continuing relevance of E. R. Leach’s Political Systems of Highland Burma to Kachin studies. In François Robinne and Mandy Sadan (eds.). Social dynamics in the highlands of Southeast Asia, 31–66. Leiden: Brill. Matisoff, James. 2003. Handbook of Tibeto-Burman: System and philosophy of Sino-Tibetan reconstruction. Berkeley and Los Angeles: University of California Press. McWhorter, John. 2001. The world’s simplest grammars are creole grammars. Linguistic Typology 5: 125–66. McWhorter, John. 2007. Language interrupted: Signs of non-native acquisition in standard language grammars. Oxford: Oxford University Press. Pirazzoli-t’Serstevens, Michèle. 1974. La civilsation du royaume de Dian a l’époque Han. Paris: École Française d’Extr̂eme-Orient. Pelliot, Paul. 1904. Deux itinéraires de Chine en Inde à la fin du VIIe siècle. Bulletin de l’Ecole française d’Extrême-Orient 4: 321–48. Richardson, Hugh. [1962] 1984. Tibet and its history. Boulder, CO & London: Shambala. (First edition London: Oxford University Press). Roerich, George Nicholas. (Юрий Николаевич Рерихъ). 1930. The animal style among the nomad tribes of Northern Tibet. Prague: Seminarium Kondakovianum. Sage, Steven. 1992. Ancient Sichuan and the unification of China. Albany: State University of New York Press. Scott, James. 2009. The Art of Not Being Governed: An Anarchist History of Upland Southeast Asia. New Haven: Yale University Press. Shafer, Robert. 1966. Introduction to Sino-Tibetan, vol. 1. Wiesbaden: Harrasowitz. Singh, Chungkham Yashawanta. 2002. The impact of historical events on Manipuri language. Indian Linguistics 63: 77–85. Sircar, D. D. 1990a. Prāgjyotisha-Kāmarūpa. In H. K. Barpujari (ed.). The comprehensive history of Assam: From the pre-historic times to the twelfth century A.D., 59–78. Guwahati (India): Publication Board Assam. Sircar, D. D. 1990b. Epico-Purānic myths and allied legends. In H. K. Barpujari (ed.), The comprehensive history of Assam: From the pre-historic times to the twelfth century A.D., 79–93. Guwahati (India): Publication Board Assam. Sircar, D. D. 1990c. Sources of early history. In H. K. Barpujari (ed.), The comprehensive history of Assam: From the pre-historic times to the twelfth century A.D., 44–58. Guwahati (India): Publication Board Assam.

70   

   Scott DeLancey

Stargardt, Janice. 1990. The Ancient Pyu of Burma, Volume 1: Early Pyu cities in a man-made landscape. Cambridge: PACSEA. Sun, Hongkai, and Liu Guangkun. 2009. A Grammar of Anong: Language death under intense contact. Leiden: Brill. Terrien de la Couperie, Albert. 1887. The languages of China before the Chinese. London: David Nutt. Thomason, Sarah and Terrence Kaufman. 1988. Language contact, creolization, and genetic linguistics. Berkeley and Los Angeles: University of California Press. Trudgill, Peter. 2009. Sociolinguistic typology and complexification. In Geoffrey Sampson, David Gil and Peter Trudgill (eds.). Language complexity as an evolving variable, 98–109. Oxford: Oxford University Press. Walter, Michael. 2009. Buddhism and Empire: The political and religious culture of early Tibet. Leiden: Brill. Wang, Zhusheng. 1997. The Jingpo Kachin of the Yunnan Plateau. Tempe, AZ: Program for Southeast Asian Studies, Arizona State University. Xuanzang. 1996. The great Tang Dynasty record of the Western Regions. (BDK English Tripitaka 79). Berkeley: Numata Center for Buddhist Translation & Research. Zeisler, Bettina. 2005. On the position of Ladakhi and Balti in the Tibetan language family. In John Bray (ed.). Ladakhi histories: Local and regional perspectives, 41–64. Leiden: Brill. Zeisler, Bettina. 2009. Reducing phonological complexity and grammatical opaqueness: Old Tibetan as a lingua franca and the development of the modern Tibetan dialects. In Enoch Aboh and Norval Smith (eds.). Complex processes in new languages, 75–95. Amsterdam and Philadelphia: Benjamins. Zeisler, Bettina. in preparation. Ethnic diversity, language contact, and the old Tibetan lingua franca: early Tibetan history and the development of the modern Tibetan languages.

Roger Blench and Mark W. Post

Rethinking Sino-Tibetan phylogeny from the perspective of North East Indian languages 1 Introduction Sino-Tibetan has more speakers than any other language phylum, and covers a major proportion of the land area of East Asia. Despite some two centuries of study and publication, the subclassification of Sino-Tibetan remains highly controversial, as does its external affiliation (van Driem 2008a; Blench 2008a,b; Handel 2008). Originating as “Indo-Chinese” in the middle of the nineteenth century, it originally carried racial connotations (van Driem 2002). The first recognition of the phylum probably dates to Julius von Klaproth (1823) who recognised three parallel branches: Chinese, Burmese and Tibetan. Von Klaproth explicitly excluded Austroasiatic and Daic, unlike many later classifications, which sequentially included all the regional phyla. Although such views still sometimes surface (primarily in Chinese publications), they have been fairly conclusively rejected by most scholars. Considering the importance of Sino-Tibetan and its history of scholarship, there is a striking lack of consensus as to its internal classification. Historically speaking, there have been two opposing camps: those who consider Sinitic (i.e. the several varieties of Chinese) as representing a primary branch of the family (Wolfenden 1927; Benedict 1972, 1976; Bodman 1980; Weidert 1987; Bradley 1997b, 2002; Matisoff 2003, 2008; Thurgood and LaPolla 2003; Handel 2008) and those who situate it within the remaining languages, consequently applying the name Tibeto-Burman to the whole phylum (Shafer 1955, 1966/67; van Driem 2002). In recent years, successful reconstructions of low-level groups have begun to appear (e.g. Sun 1993; Mortensen 2003; VanBik 2007; Wood 2008; Button 2009), raising hope that higher-level reconstructions may eventually be able to be placed on a stronger footing - or at least, that their validity will be able to be more rigorously tested. Nonetheless, many putatively Sino-Tibetan languages remain very sparsely documented, with accessible comparative lexical material of any significant scale being largely confined to Chinese and (to a lesser extent) Indian sources. The largest-scale comparative database of Sino-Tibetan languages compiled to date, the Sino-Tibetan Etymological Dictionary and Thesaurus (STEDT) project, was finally made generally available in “beta” form in October 2010, with some additional functionality added in 2012 (though still in “beta”). Consequently, historical linguists can now see the evidence for Proto-Tibeto-Burman

72   

   Roger Blench and Mark W. Post

forms given in Matisoff (2003), which did not explicitly present the data on which most reconstructions are based.1 Some potential problems with the linguistic data employed in Matisoff (2003) have been pointed out in various reviews (e.g. Sagart 2006; response by Matisoff 2007; further reply Sagart 2008; Hill 2009). One point that we will underscore here, however, is that there are also problematic disconnects with the archaeological evidence (Blench 2008b). For example, while ‘iron’ is reconstructed at the Proto-Tibeto-Burman level, it seems clear that terms for ‘iron’ are in fact a series of loanwords, reflecting the diffusion of ironsmelting technology (Chang 1972;2 Blench in press a). Perhaps most significantly, however, Proto-Sino-Tibetan (PST) reconstructions have traditionally relied very heavily on citations from “major” languages, and reference to epigraphic and written sources for earlier forms of certain such languages, i.e. Chinese, Tibetan and Burmese. It cannot be emphasised too strongly that these are, if not indeed irrelevant, of relatively very low significance for the reconstruction of proto-forms of a phylum the great majority of whose members have never been written and which must be far beyond the reach of epigraphy. This emphasis on “major” languages has had another consequence: “minor” and often poorly documented languages have generally been excluded from consideration. This is particularly true of the languages of North East India, where the way of life hardly matches the settled agricultural lifestyle depicted for Proto-Sino-Tibetan speakers.3 This paper,4 then, is intended to redress the balance in two ways. First, by suggesting what sorts of revisions to the image of Sino-Tibetan are required if 1 To be fair, Matisoff (2008) does present much supporting data, but generally only forms which contribute to the argument; complete accounts of all available data are not found in this work either. 2 Chang does not say this; indeed, his observations of roots that cross-phylic boundaries were adduced as evidence of a large-scale “inclusive” Sino-Tibetan. 3 We do not of course mean to suggest that the earliest Sino-Tibetan comparativists deliberately excluded “minor” languages which they could have included had they cared to. As we note in several places herein, and as was pointed out by Matisoff (1991), data for many North East Indian Tibeto-Burman languages were simply unavailable throughout much of the last century, and indeed remain sparse to this day. However, the result is nevertheless the same: undue prominence has been given and generally, without qualification – to potentially only marginally useful languages such as Chinese, and little mention has been made of potentially much more important languages such as Puroik (Sulung). These structures of unwarranted imbalance, having come to exist for whatever reason, are perpetuated almost without remark throughout the Sino-Tibetanist literature. 4 Paper presented at the 16th Himalayan Languages Symposium 2–5 September 2010, School of Oriental and African Studies, London. The authors would like to thank the many individuals who assisted with language data and offered hospitality in remote areas, and particularly Jiken



Rethinking Sino-Tibetan phylogeny   

   73

the full spectrum of minority languages are given equal weight, and second, to sketch the sort of model of its evolution that would be required to be congruent with the available archaeological and historical data. It cannot be emphasised too strongly that this approach is provisional; as new data appear, the model presented herein can and should be revised.

2 Assumptions about Proto-Sino-Tibetan The “standard model” of Sino-Tibetan is represented by the reference volume Thurgood and LaPolla (2003) and by the listing in Ethnologue (Lewis et al. [eds.] 2013). Convention dictates a primary Sinitic/Tibeto-Burman split, despite that there is no unambiguous published evidence to support such a view. Instead, it appears to us that Sinitic was likely to have been historically set apart for cultural and/or other non-linguistic reasons. Along similar lines, van Driem (1997) has long claimed that Sinitic is co-ordinate with Bodic and thus simply within TibetoBurman, a view strongly opposed by Matisoff (2000). However, even if Sinitic is simply part of Tibeto-Burman, its position remains far from obvious (Matisoff 2008; van Driem 2008a). Similar considerations apply to arguments concerning the homeland of ­Sino-Tibetan. Matisoff (STEDT website) proposes the Tibetan Plateau, which supposes the ancestral speakers of the different Sino-Tibetan branches to have migrated down various river valleys to reach their present locations. Van Driem (1998) challenges this view on the basis of archaeological evidence, with a claim that we should look instead to Sichuan. Northern China is favoured by other researchers, and the Yangshao culture (Yǎngsháo wénhuà 仰韶文化) which flourished on the Yellow River from around 6800 BP, has also been canvassed (e.g. LaPolla 2001). Related to this is the conventional reconstruction of Proto-TibetoBurman in Matisoff (2003) which implies that its speakers were fully settled agriculturalists, with a wide range of livestock and crop species, and using iron tools. However, where starred forms are not congruent with the known archaeology of the region, and the reconstruction methodology produces false positives, it becomes harder to have confidence in items with credible semantics. Is it then presently possible to have long lists of starred Proto-Tibeto-Burman reconstruc-

Bomjen, Serwa Dajusow, Rebecca Gnuechtel, Tia Toshi Jamir, Jummar Koyu, Meri Koyu, Jokut Modi, Adde Modi, Kindi Modi, Aamin Modi, Yankee Modi, and Oken Tayeng. Some of the results presented here derive from fieldwork by the authors in Arunachal Pradesh 2010–2012 subsequent to the initial conference presentation.

74   

   Roger Blench and Mark W. Post

tions? Only if they are treated with appropriate scepticism, as hypotheses for consideration. Attributing a putative reconstruction to a genetic node that is not congruent with external evidence, such as dates for iron or the horse, is unjustified and potentially misleading.5

3 Large groupings and spiky trees Using the density of individual branches to speculate on the homeland of a phylum is fraught with possible errors. The expansion of an individual branch may simply obscure a former diversity, and secondary diversification can also occur. Nonetheless, deep divisions among languages in close proximity are at least highly indicative; in the case of Sino-Tibetan, this is nowhere more the case than in the Eastern Himalaya.6 While relatively little has been known about the languages of this region up to and including the present time, this has not stopped scholars from proposing that these languages either constitute or fall within some other Tibeto-Burman subgroup. However, in absence of any sort of systematic comparison – whether the data are thought reliable or not – such “subgroupings” are essentially vacuous. The use of pseudo-genetic labels such as “Himalayish” and “Kamarupan” inevitably give an impression of coherence which is at best misleading. As is well known from voluminous research in contact linguistics, common features in a geographical area are far from proof of genetic affiliation; while it may well be the case that an armchair glance at, say, a 200-item Puroik (Sulung) wordlist yields greater-than-chance resemblances among certain forms and parallel items in well-known Tibeto-Burman languages (the usual suspects tend to be ‘fire,’ ‘sun/day,’ ‘person,’ ‘two’ and ‘three,’ the first and second person pronouns, and a handful of other common forms), it is wrong to discount the possibility that such forms could have come about via contact and borrowing. Although the reality of language contact in the Tibeto-Burman region and the difficulties it can pose for subgrouping and reconstruction have long been discussed 5 Not all linguists take this view, and Roger Blench has been taken to task at conference presentations for “missing the point” of historical linguistics by requiring reconstructed forms to match external evidence. The authors remain unrepentant and stand by the rather strong statement in the main text. 6 By “Eastern Himalaya,” we mean the area immediately South of the Tibetan plateau, to the East of Bhutan, and to the North of the Brahmaputra and Irrawaddy Valleys. In modern geopolitical terms, this region includes almost all of India’s state of Arunachal Pradesh, spilling slightly across the Tibet, Burma, and China border regions. This area has also been identified from a cultural perspective by Blackburn (2007), among others.



Rethinking Sino-Tibetan phylogeny   

   75

in the Tibeto-Burman literature (e.g. Hashimoto 1976; Matisoff 1991), caution is often thrown to the wind when a large-scale subgrouping scheme is in fact attempted.7 Robbins Burling (1999) has correctly attacked the use of geographic groupings such as “Kamarupan” as unhelpful. Language families do not necessarily develop in ways convenient for graphic designers, especially in montane regions, where communities may be several days’ walk from one another across difficult terrain. Two examples will serve to illustrate this point. The Gongduk language is spoken by around 1000 individuals in central Bhutan (van Driem 2001). Little has been published on it as yet, although a grammar is in preparation (van Driem p.c.). This language has complex verbal morphology characteristic of the Kiranti languages, although it is highly lexically divergent. On this basis, it is presumably Sino-Tibetan but does not seem to be part of any major subgroup. It therefore will probably have to be assigned to a single branch; either extremely high on the tree as part of the primary diversification or as a single node within the general grouping of Himalayan languages. Similar problems arise with ’Olekha (Black Mountain Monpa) which has highly conservative verbal morphology (van Driem 1995). Originally assumed to be part of East Bodish, it now appears that it has very little in common with this grouping (Hyslop, this volume). The initial explanation was that ’Olekha was conservative and the rest of the group had thus innovated; but the low level of cognacy with other East Bodish languages provides no evidence for this. Provisionally, ’Olekha may accordingly be treated as an isolate branch. In these two cases and more generally, high-level branching (essentially, the postulation of “within-family isolates”) should be practised until we have better evidence for the position of individual languages. If this results in an untidy “tree” which is hard to capture, so be it. In language classification, Ockham’s razor must sometimes be turned on its head; entities must be multiplied rather than assumed. Put differently, the postulation of differently-structured “family trees” in absence of adequate supporting evidence is nothing more than windowdressing – an exercise in comparative aesthetics – the results of which are then naively referenced in the literature as though they were true taxonomies. They are nothing of the kind.

7 Even when acknowledging the linguistic shaping effects of contact, there is a widespread tendency to understand the dynamics involved in terms of “major” languages influencing “minor” languages. Matisoff’s widely-adopted labels “Indosphere” and “Sinosphere” (Matisoff 1991) neatly encapsulate this tendency (Post 2011; see also DeLancey, this volume).

76   

   Roger Blench and Mark W. Post

4 T  he contribution of the languages of North East India A comparison of the various classificatory proposals for Sino-Tibetan makes clear that they have a common feature, a tendency to pass over the highly diverse languages of North-Eastern India. The most recent survey of North East Indian Sino-Tibetan languages is Robbins Burling (2003) which, although valuable in terms of bringing some order to the region, is far from comprehensive. Only the Ethnologue (Lewis et al. [eds.] 2013), whose agenda mandates completeness, includes all those so far recorded; however, its ethnonyms and classification are frequently inaccurate, apparently privileging the unpublished reports of SIL’s local operatives over established international research (as is the case with Tani, cf. the corrected table in Appendix 1). An intriguing source, Abraham et al. (2005) which provides the only available material on a variety of hard to classify lects of Arunachal Pradesh, including Sartang, Koro, Chug and Lish, is grossly misrepresented in the Ethnologue. The difficulties of making accurate maps are reflected in the Anthropological Map of Northeast India, re-issued in 2009 by the Archbishop’s House at Guwahati. This includes many ethnolinguistic groups not found in other publications. There is no accompanying text, so it is not possible to be sure of the source of the data, and some of the ethnonyms may not represent distinct linguistic entities. Nonetheless, it represents a major advance in linguistic geography of the region, and creates a series of problems that will keep linguists occupied for years to come.8 A revised and updated version of this map is given in Figure 1. Political difficulties for researchers have made this region inaccessible, but the situation is changing rapidly.9 There is, moreover, a wealth of local publications available concerning the languages of North East India, in particular those of Arunachal Pradesh. Often the only record of a language is a descriptive but thin and under-researched account published by local administrative officers in a QWERTYstyle Romanization. That is to say, there is generally no identification of non-standard consonants, non-cardinal vowel qualities, contrastive vowel length, nasalization, or prosodic features such as glottalization, breathiness and tone (all of which are highly endemic to the region). These “language guides” nonetheless generally 8 This map will not be reproduced here in the interest of space, but may be found in reproduction in Matisoff (1991). 9 Between drafts of this paper, Nagaland and Manipur were both opened to international researchers, joining Mizoram and Tripura. This leaves only Arunachal Pradesh among the North Eastern “seven sisters” to remain in principle closed to outside research, a highly regrettable situation which we can only hope will also eventually change.



Rethinking Sino-Tibetan phylogeny   

   77

provide a basic vocabulary and grammar, and often include a list of subsistence vocabulary; there is no excuse for not making use of them.

Fig. 1: Languages of North East India. Source: Redrawn and language names edited by Roger Blench from a map published by Bishop’s House, Guwahati

Once this material is taken into account, it becomes clear that these languages are very different from one another. There are large groups of related languages, such

78   

   Roger Blench and Mark W. Post

as the Tani and Tangsa clusters, whose Tibeto-Burman affiliation is not generally disputed (even if the genetic status of at least some putative members remains somewhat unclear). But for many more isolated languages of the region, assignment to Tibeto-Burman is highly questionable. In addition to languages for which there is some data, the Ethnologue has listed a number of “ghost” languages, such as Anu, Lui and Palu, as “unclassified Sino-Tibetan.” But if there is no data, it is hard to see how such languages – if they exist – can be classified as Sino-Tibetan (or allied to any other recognized phylum), and astonishing that such “classifications” are actually accepted by scholarship.10 What is even more striking, however, is the acceptance of received wisdom concerning languages for which data in fact exists. For example, Aka (Hruso) has been known since the early nineteenth century, and has been referenced as a “Sino-Tibetan” language ever since it was surveyed by Konow (in Grierson 1909) and Shafer (1947). However, evidence for this alignment is minimal. Aka not only shows few clear cognates with Sino-Tibetan, but there does not even appear to be a significant level of borrowing. Similar doubts must arise concerning Miji, Koro, Puroik (Sulung), Bugun and Mey (Sherdukpen), at a minimum. Moreover, histories of intensive contact and borrowing may well have undermined our ability to see clearly. Milang, which has been classified as an “aberrant” Tani language by Sun (1993) is distinguished by the strikingly irregular relationship of cognates with mainstream Tani as well as a highly-differentiated lexicon and morphology (Post and Modi 2011). Recent research links some of these items with the geographically remote Koro language (data for which is provided in Grewal 1997 and Abraham et al. 2005), pointing to a possible Koro-Milang small phylum which may further underlie some aspects of Tani (Post and Blench 2011). Konow (1909) was thus correct to frame the “North Assam” group as a heuristic, geographical classification rather than a defensible genetic subgroup. An even more appropriately cautious assumption would have been to suppose that some area languages may in fact be isolates with Sino-Tibetan loans, and that moreover, substrate lexicon and morphology from such isolates may indeed account for the extreme diversity of languages which nonetheless “are” Sino-Tibetan. If this is so, we need to change our perspective on the “Tibeto-Burman” languages of North East India, and to regard them as constituting a centre of lin10 The ISO, for example, whose coding system and nomenclatures are increasingly obligatory features of grant applications and archive deposits, accepts without question the lists of languages and non-languages put forth by SIL/Ethnologue (whether or not there is evidence for them) while requiring the rest of the world’s scholars to petition SIL/Ethnologue on the basis of published argumentation. This “standardization” regime, whatever its laudable intentions, is thus neither equitable, nor scientific.



Rethinking Sino-Tibetan phylogeny   

   79

guistic diversity more characteristic of North East Asia or parts of the Amazon, whose potential importance has been obscured only by unwarranted assumptions and assertions concerning the classification of its languages. Even where a Sino-Tibetan affiliation is well-accepted, groups such as Bodo-Garo only have about 25% of their basic lexicon drawn from inherited Sino-Tibetan lexemes (DeLancey p.c.). The rest must be presumed to derive from assimilated substrate languages. Given the massive grammatical restructuring characteristic of BodoGaro (DeLancey 2012), it would be just as logical to regard the family as an isolate with heavy relexification from Sino-Tibetan.

Fig. 2: Sago palm under cultivation, Riga (Minyong) Village

80   

   Roger Blench and Mark W. Post

5 Foragers and sago palms Linked directly with a pattern of extreme linguistic diversity is the subsistence strategy which remains dominant in Arunachal Pradesh and adjacent regions, which balances extensive hunting and foraging with swidden cultivation. A particularly striking element in foraging strategies is the traditional exploitation of the sago palm (Figure 2) (Blench in press d). As in New Guinea (Rhoads 1981) and Borneo (Morris 1991), several varieties of sago palm grow wild throughout central Arunachal Pradesh, although it has also been adapted for cultivation in some areas. The trunk of sago contains a starchy interior which can be processed as a staple food once the tree is cut down (Figure 3). In a region with a low human population density and dense forest, this is an attractive subsistence strategy, requires considerably less work than conventional agriculture, and has the added advantage of year-round availability. The ethnographic literature makes it clear that peoples such as the Puroik/Sulung (Stonor 1952; Deuri 1982), Milang (Yankee Modi, p.c.) Idu (Bhattacharjee 1983) and others have only adopted rice farming in recent times.11 Some groups, such as the Sulung and Bangni/Nyishi, continue to balance sago exploitation with other productive strategies to this day (Tassar ms), while other groups, such as the Upper Minyong and the Milang, retain sago cultivation only to provide animal fodder (authors’ field notes). Similar patterns are found in livestock production. Although livestock is highly characteristic of the high Himalayan way of life in general, with yaks and sheep being predominant species until recently, the mithun, or gayal (Bos frontalis) is the most prominent animal exploited by Eastern Himalayan groups (Figure 4). The mithun is a semi-domesticate, managed in fenced tracts of forests rather than being kept in or near villages. Outside North East India, mithun are primarily imported for the purpose of cross-breeding with other bovids, for example in Bhutan. It is very common among Eastern Himalayan languages to find lexical sets denoting fauna in which the mithun is lexicalized as a “prototypical” meat animal, with all other terms being derived. Table 1 illustrates such a set for Hruso (Aka); similar sets can be formed for many other area languages, including Proto-Tani.

11 The possibility that there was an older farming culture which was dropped in favour of sago exploitation, something which does happen with Austronesians arriving in Papua, seems less likely here, as pseudo-grain field crops such as Job’s tears (Coix lacrima-jobi) have been maintained in this region.



Fig. 3: Prepared sago trunk, Riga Village

Fig. 4: Mithun (Bos frontalis), Dali (Galo) Village

Rethinking Sino-Tibetan phylogeny   

   81

82   

   Roger Blench and Mark W. Post

Tab. 1: Hruso livestock terminology Hruso

Gloss

fú fǔ babu fú-glu f(ú)-gra fú ɦu fú-ɲ fǔ lhu imɲi fú msu fú fu bʃə

mithun donkey sheep horse wild pig cattle cow wild dog, wolf buffalo

Source: Fieldwork by Roger Blench

Terms for ‘mithun’ in other languages of Arunachal Pradesh are typically cognate with Aka fu (e.g. Miji ʃu, Koro sù, Puroik ʧa and Proto-Tani *ɕo), suggesting that this is probably not a case of semantic shift from a wild species. The implication is that the semi-wild mithun was seen as the core species, and the true domesticates such as cattle, which arrived subsequently, as marginal to the system. In the light of this, the earliest phases of Sino-Tibetan take on a wholly new appearance. Ethnolinguistic diversity is highly characteristic of hunter-gatherers in other regions of the world, for example among the Khoisan of Southern Africa, in Siberia, the Amazon and of course among the sago palm exploiters of Western Papua. The overlay of agriculture in recent times in North East India has disguised this pattern so that it is not usually seen as comparable to elsewhere in the world. But in reconstructing a language phylum surface appearances must be discarded. Logically, if languages show highly diverse structures, they are more likely to have a place at the top of the Sino-Tibetan phylogenetic tree. The better known groupings, and therefore the proposed reconstructed forms, must be downgraded to meso-languages. Apart from the diverse forager languages, North East India is also home to agricultural populations falling under the general labels of “Naga” and others such as the Kuki-Chin (Marrison 1967). Some of these form more coherent groupings than others; Naga in particular is a term that describes more a lifestyle than a linguistic unity (van Driem 2008b). As with the former foragers, it seems likely that cereal culture is a relatively recent superimposition on these populations. The millets (Setaria and Panicum) probably spread down from the Tibetan Plateau, whereas rice-growing (as well as sorghum, a highly marginal recent arrival) appears to be an innovation spreading up from the valley of the Brahmaputra. The basic subsistence strategy, as in much of island and humid South-



Rethinking Sino-Tibetan phylogeny   

   83

east Asia, seems to have been vegeculture, the cultivation of tubers and other non-seed cultigens such as bananas, palms and ferns (Blench in press b). Islands such as Borneo have switched from vegeculture to rice systems in the relatively recent past. Taylor (1953) mapped out these crops for North East India some time ago, but this approach has been little exploited, in part because of the difficulties of finding vegetative crops in the archaeological record. Nonetheless, there is every reason to consider that the basic cultigen repertoire of the Naga and related peoples may well have been yams (Dioscorea spp.), taro (Colocasia spp.) and the Musaceae. As it happens, there is evidence, provided by Matisoff (2003), that taro (both the plant and the word) was borrowed from Austroasiatic into Sino-Tibetan. Matisoff (2003:173) proffers *sr(y)a as proto-Tibeto-Burman for ‘yam/potato’ and *grwa for taro. The evidence for the former, according to the STEDT database, is as follows: Lushai [Mizo] Meithei Dimasa Garo

ba-hra  ha  tha  tha 

This number of attestations is restricted at best and these languages are geographically extremely close to one another; this suggests a regional loanword. Table 2 compiles vernacular names for ‘taro’ from Austroasiatic; these are the probable sources of Tibeto-Burman forms for ‘yam/potato:’ Tab. 2: Some reflexes of #trawʔ, ‘taro’ in Austroasiatic Branch

Language

Attestation

Monic Vietic Khmuic Katuic Palaungic Khasian Muṇḍā

Nyah Kur Proto-Vietic Khmu Bru Riang Khasi Santal

traw *sroʔ sroʔ ʔara̤w sroʔ shriew saru

Source: Blench (2012)

In terms of livestock, ‘cow’ and/or ‘buffalo’ appears to be a regional borrowing among Sino-Tibetan, Austroasiatic and Indo-Aryan. Cattle appear in the archaeological record on the Yellow River in China around 4500–4200 BP and must have been regionally diffusing around this time period (Yuan et al. 2008). Table 3 shows the reflexes of a widespread root which appears in all three phyla.

84   

   Roger Blench and Mark W. Post

Tab. 3: Attestations for a form #mVCV ‘cow’ in Sino-Tibetan and Austroasiatic Phylum

Branch

Language

Attestation

Gloss

Source

Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Austroasiatic Austroasiatic Austroasiatic

Loloish Luish Bodish Tangkhulic Bodo-Garo Jingpho-Konyak Puroik Mijuish Mishmi Palaungic Palaungic Bahnaric

hmyaŋ453 θa ˋmulʔ mʌyi si.muk ? < Wa maˑsu maan ce amwa mansöu macu moi2 *mɤk rəmɔɔ

cow cow buffalo cow cow cow cow cow cow cow cow cow

Edmondson (n.d.) Bernot (1966) Hildebrandt (2003) Mortensen (2003) Burling (2003) Bandyopadhyay (1989) Tayeng (1990) Boro (1978) Pulu (2002) Paulsen (1989) Diffloth (1980) Sidwell (2000)

Austroasiatic Austroasiatic Austroasiatic Austroasiatic Indo-Aryan

Khasic Khasic Munda Munda Eastern

Phu Noi Cak Manange P-Tangkhul Garo Tangsa Puroik Miju Idu Samtao Proto-Waic Bahnar [Pleiku] Khasi War Jaintia Santali Ho Assamese

masiː mut mĩhũ miu maŋso

cow buffalo calf calf meat

Indo-Aryan

Central

Hindi

mans

meat

Singh (1906) Brightbill et al. (2007) Stampe (ined.) Stampe (ined.) University of Gauhati (1962) Caturvedi (1970)

It is striking that the different consonants in C2 position, the nasal /n/, the fricative /s/ and the velar /k/ occur in both Tibeto-Burman and Austroasiatic phyla, suggesting a CV root with affixation that has undergone a complex history of borrowing, presumably with the animal itself. It is a long speculative leap to characterise a whole nexus of languages, but the broad picture may be that speakers of Naga-like languages gradually adopted vegeculture and livestock production thus slowly parting company with the foragers. If there was an arc of diverse foragers in the North East Indian region, then it is not difficult to imagine the slow exploration of higher latitudes in search of game. Not all of these foragers would have been Sino-Tibetan speaking, as the presence of a language isolate such as Kusunda (in Nepal) suggests. The archaeology of the Tibetan region is sketchy at best, but Middle Stone Age foragers were reaching the high altitudes as early as 20,000 BP (Zhang et al. 2003). A second phase of occupation, beginning by 7500 BP, is marked by the presence of microliths indicating seasonal exploitation by foragers (Huang 1994). It is reasonable to imagine that these represent the first forays by Sino-Tibetan speakers into the Tibetan Plateau in search of wild yak and antelope. However, without stored crops and other sources of nourishment, climatic conditions would have



Rethinking Sino-Tibetan phylogeny   

   85

been too forbidding to stay there all year round. Permanent human occupation of the Plateau begins by 5000–4000 BP (Aldenderfer and Zhang 2004) and this is most likely to reflect the domestication of the yak, which would make it possible to exploit the pastures of the Plateau and subsist in the inimical climate all year round (Xue-bin et al. 2008). Fu Daxiong (2001: 66) has reported carbonised Setaria grains in Eastern Tibet ca. 5000 BP, which would fit with the early dates reported for this millet in China. The other aspect of the structure of Sino-Tibetan that needs to be explained is the early eastward spread of isolated groups prior to the expansion of Sinitic. The whole region of China would presumably have been occupied by very different language phyla prior to the Sinitic expansion, Altaic and Koreanic in the North, Hmong-Mien, Austroasiatic and probably Austronesian in the centresouth. Archaeology suggests pig domestication by 8000 BP (Yuan and Flad 2002; Yuan et al. 2008), broomcorn millet by ~10,000 BP (Lu et al. 2009), foxtail millet by 6000 BP (Hiroo et al. 2007) and rice by 7700 BP in the Yangtse Valley (Zong et al. 2007).12 The shouldered celts which connect the Himalayas and Szechuan focused on by van Driem (1998) may well reflect this eastward flow of diverse early populations who were either foragers or had begun the management of pigs and high-altitude crops such as buckwheat (see Section 6.3). Populations such as the Bai (Wiersma 1990) and Tujia (Brassett and Brassett 2005) represent preSinitic migrations of Sino-Tibetan peoples. Although much of their lexicon has been replaced by deep-level Chinese loans, both languages retain non-Chinese names for both crops and livestock (Blench 2011). Their encounter with resident farming populations would have encouraged sedentarisation and the adoption of a wider range of crops. Starostin (2008) has argued that some key terms relating to subsistence in Old Chinese are of Altaic origin. The proto-Sinitic encounter with livestock-producing, millet-growing Altaic speakers could well have led to a subsistence and military revolution which in turn resulted in the overwhelming expansion of Sinitic and thus the dominance of this branch of Sino-Tibetan in East Asia today.

12 Although the dating of truly domestic rice has been the subject of much debate (Fuller, Qin and Harvey 2008).

86   

   Roger Blench and Mark W. Post

6 Inferences from ecological reconstruction 6.1 General Assumptions about homelands and dispersals can be tested using reconstructions of lexical items characteristic of particular ecologies. If, for example, the Sino-Tibetan phylum did originate in the mid-level Himalayas, then it should be possible to reconstruct the fauna, flora and meteorological phenomena typical of that ecozone. If the early phases of the expansion were indeed characterised by agriculture, crops typical of such high-altitude regions should be prominent and humid zone crops absent. This section discusses two examples of what can be deduced from this type of linguistic data, using the examples of ‘snow/ice’ and buckwheat.

6.2 ‘Snow’ and ‘ice’ and a Himalayan origin Dempsey (1995) may have been the first author to consider the terms for ‘snow’ and ‘ice’ as relevant to the quest for a Sino-Tibetan homeland, although he gives no actual data. If a language phylum originated in a region where these were common there should be a deep-level root. And indeed there is a claimed form *khyam for Proto-Tibeto-Burman (suspiciously similar to Burmese). Appendix Table 2 collates the words for ‘snow’ in Sino-Tibetan. But of 190 languages and dialects collated there are some 30% unidentifiable forms, the remainder assigned to some ten different roots, each of low frequency. In Sinitic, we find attestations of four of these roots suggesting that this may in fact represent a complex network of borrowing rather than reconstructions of great antiquity. Accordingly, the probability is low that ‘snow’ was part of the environment of early Sino-Tibetan speakers.

6.3 Buckwheat: A high altitude crop Buckwheat is the most important crop of the mountain regions above 1600m both for grain and greens and occupies about 90% of the cultivated land in the higher Himalayas. The domestication of buckwheat is described in Joshi and Rana (1995) and Ohnishi (1998). There are two species of domestic buckwheat: “bitter” buckwheat (Fagopyrum tartaricum), which is cold and high altitude tolerant and occurs wild throughout the Tibetan plateau, and “sweet” buckwheat (F. esculentum), which is restricted to the eastern Plateau and some hills in



Rethinking Sino-Tibetan phylogeny   

   87

Yunnan and Sichuan. Bradley (1997a) has compiled terms for buckwheat in LoloBurmese, and shows that it clearly reconstructs to Proto-Lolo-Burmese. Table 4 compiles terms for buckwheat in the Sino-Tibetan area. There appears to be a common root, something like #ʨiau, but also at least one other term in Qiangic of unknown etymology. Tab. 4: Terms for ‘buckwheat’ in Sino-Tibetan languages Language

#

Sino-Tibetan Sinitic

Chinese

Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan

Chinese Chinese SW Chinese Tibetan (Lhasa) Amdo Tibetan Tujia Jinghua Taoba Caodeng Queyu (Yajiang) Queyu (Yajiang) Ersu Muya Qiang (Mawo) Qiang (Mawo) Rgyalrong Achang (Longchuan) Atsi [Zaiwa] Bola (Luxi) Jinuo Kaman Taraon

qiáo mài (蕎麥) ku qiao tian qiao ʨiau31 tʂhau¹⁵ tʂu khu²¹tɕhiau²¹ tãu ʧə13 tō35 ʨi35 ʃɔ ʐõ³⁵qa⁵⁵ ʐõ³⁵tʂa⁵³

Phylum

Branch

Sinitic Sinitic Sinitic Tibetic Tibetic Tujia Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Rgyalrongic Burmish Burmish Burmish Loloish Miju Mishmic

ʨiau [?]

ʃok ʨhauʔ⁵⁵ khjau⁵⁵ khjau³¹ ʨhɔ³¹tsi⁴⁴ ʨi³¹kɑ⁵⁵ tɯ³¹kɑ⁵³

Others

Comment

bitter sweet < Sinitic? < Sinitic? bitter

ndzɿ³³ ȵɛ³⁵ndʑyɯ⁵³ dzə dzəʂ

sweet bitter sweet sweet sweet bitter < Sinitic? < Sinitic? < Sinitic? < Sinitic? bitter bitter

Source: Blench in press c

The importance of buckwheat among the Qiangic peoples and the phonological diversity of the names, does suggest its possible origin in this region. However, the SW Sinitic #ʨiau is clearly an important source of secondary loans, as suggested in the comment column. In particular it looks as if Tibetic forms are borrowings from Sinitic, suggesting that buckwheat travelled west from China once domesticated, as it was suitable for high-altitude cultivation. Table 5 shows the distribution of a second root, #bram-, for buckwheat:

88   

   Roger Blench and Mark W. Post

Tab. 5: The #bram- root for ‘buckwheat’ in SE Asian languages Phylum

Branch

Language

Attestation

Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan Sino-Tibetan

Qiangic Nungish Bodish Tibetic Kiranti Kiranti Mishmic Mishmic Tani

Horpa Trung Tshona (Wenlang) Written Tibetan Bahing Kulung Idu Taraon Damu

brɛ və ɉɑm⁵⁵bɹɑi⁵⁵ bre³⁵mo⁵⁵ bra bo bramtbham ɑ⁵⁵bɹɑ⁵⁵ xɑ³¹bɹɑ⁵⁵ pra-ɦu

Comment

bitter

bitter probable loan

It seems at least possible that this root was originally associated with the domestication of bitter buckwheat. Apart from these, there are many more low-frequency roots, especially in the Himalayas. We can therefore conclude that buckwheat domestication was important in the early period of the development of agriculture among Sino-Tibetan speakers, but that it was certainly not part of the protolanguage but was rather a later spread.

7 How old is Sino-Tibetan? Determining the age and homeland of a linguistic phylum depends on several types of evidence coming together. It is assumed here that the results from linguistic reconstruction should be congruent with known archaeological, ecoclimatic and genetic data; if they are not, then the reconstruction should be treated as problematic. It is possible to claim that reconstruction is an abstract historical exercise, such that if there is apparently a form for ‘trousers’ in proto-Tibeto-Burman then it is irrelevant whether that was indeed its original referent. But most linguists would be unhappy with this; they would rather there were some correspondence between their constructs and real-world situations. Alternatively, one can suppose that there might have been a systematic semantic shift; that a proto-referent has been discarded in favour of a modern item. Such shifts clearly occur, but usually they leave traces, semantic doubling in some languages or the original referent in conservative cultures. At any rate, historical reconstruction ought surely to show awareness that the semantics of proto-forms should also be credible, not merely their phonology and morphology. Without adhering to any strict version of glottochronology, it is reasonable to expect there to be some correlation between internal diversity and age. There



Rethinking Sino-Tibetan phylogeny   

   89

are now reasonable dates for the diversification of phyla or subgroups such as Polynesian, Bantu, Mayan or Turkic. These estimates are based on a combination of linguistic trees, reconstructible roots and archaeology in the presumed homeland. Furthermore, these are all branches of families where agriculture can be reconstructed without question. In other words, these allow us to estimate approximately the level of diversity there should be over a period of 3000–4000 years, the approximate age of Sinitic. If the arguments of this paper are accepted, then in its earliest phase SinoTibetan was a congeries of diverse foragers in the Eastern Himalayan region. Dates for systematic exploitation of the Tibetan Plateau by hunters go back to 7500 BP, and presumably some time must be allowed for the movement from thick forests to the more open montane regions. If this is so, it seems reasonable to place the origins of Sino-Tibetan at around 9000–8000 BP. The diversification of the Naga and related peoples through vegeculture can be placed at around 6000–5000 BP and the beginnings of livestock production in the Himalayas immediately after this. At the same time, the primary movement eastward towards China begins and the expansion of Sinitic proper can then be placed at around 4500–4000 BP. These dates remain approximate and further archaeological research may well provide a far more nuanced picture. But this model at least has the advantage of not contradicting the known parameters of prehistory and not requiring improbable reconstructions of subsistence lexemes at various levels of Sino-Tibetan.

8 Sino-Tibetan: an alternative model The evidence presented in this paper is marked by absences; lack of cognate reflexes in many of the smaller branches of Sino-Tibetan, lack of evidence for a coherent internal structure and a failure of congruence with archaeology and genetics. Given this, any hypothesis concerning the spread and diversification of the phylum must be speculative and subject to revision. However, we can do better than any of the claims presently on the table with an account which at least does not contradict the external evidence. With this in mind, the following summary is put forward as a model of the evolution of the phylum: –– The earliest speakers of Sino-Tibetan were highly diverse foragers living in an arc between the eastern slopes of the Himalayas and regional lowland jungles up to 9,000 years ago and practising arboriculture (sago) –– Some spoke early Sino-Tibetan languages, others unknown languages now present primarily as substrates and the rare case of a modern isolate such as Kusunda

90   

   Roger Blench and Mark W. Post

–– Seasonal foragers exploit the high Tibetan Plateau from 7500 BP –– Perhaps 6000–5000 BP a “livestock revolution” takes place in the mid-level Himalayas. Yak herders move up and settle the Tibetan Plateau permanently. –– Gathering of wild cereals (buckwheat etc.) and tubers (high-altitude taro) leads to proto-agriculture in the mid-level Himalayas –– Foragers who will become the Naga complex began to practise vegeculture (taro, plantains) (NE India) and animal management (mithun) by 6000 BP –– By 5000 BP diverse early Sino-Tibetan groups in the Himalayas begin spreading eastwards to China. Sinitic is not a primary branch, but simply the language of one of many migratory groups –– Proto-Tujia, Proto-Bai and probably others meet unknown populations (Hmong-Mienic? Austronesians?) with domestic pigs, millet, while also cultivating and beginning to domesticate rice –– Proto-Sinitic speakers encounter early Altaic speakers with foxtail millet and other crops –– The Sinitic languages expand southwards, assimilating or encapsulating many small groups. They encounter Hmong-Mien speakers with rice and switch millet terminology to rice –– Rice moves up from India but also westwards from China (hence hybridised types) and overlays older cereals where ecologically possible –– Ruminants (cows, sheep, goats) spread downwards into China from Central Asia 4400 BP (? Altaic for small ruminants but not cattle) –– Tibetic speakers undergo a major expansion (perhaps as late as 500 CE?) assimilating linguistic diversity on the Plateau –– Rice invades the lowland vegecultural zones rather later, pushing taro into residual systems, possibly as late as the spread of Assamese (10th century?) –– Groups such as early Burmic spread southwards, fragmenting resident Austroasiatic-speaking peoples Figure 5 shows a highly simplified map of the early phases of this proposed movement:



Rethinking Sino-Tibetan phylogeny   

   91

taic l A ly Early Sinitic Ear

Nucleus (livestock, millet)

China

Tibet Nepal

Gathered cereals and Diverse foragers hunting large animals Vegeculture and livestock

India

Ea

rly

o

Hm

n

ie

M

ng

Burma South China Sea Bay of Bengal

Fig. 5: Possible pathways of early Sino-Tibetan expansion

9 Rethinking Sino-Tibetan phylogeny The subtext of this paper is a critique of existing Sino-Tibetan phylogenies. Is it possible to put anything more nuanced in its place? If it is to be based on numerous low level reconstructions and regular sound correspondences then this will be impossible for a long time to come. Any “tree” should thus be treated as a speculation, a tool for thinking, a graphic that minimally shows consilience with lowlevel classifications and which is credible in the light of historical, epigraphic and archaeological evidence. Such a classification should not be afraid of single language branches. Especially in the arc of the Himalayas, where individual communities have been diverging from one another for millennia, it is entirely possible they will no longer have transparent relatives. With these caveats, Figure 6 puts forward a new proposal for Sino-Tibetan phylogeny. Languages backgrounded in grey are those for which there is apparently no data, so their position is simply a default. Also marked (with a box outline) are languages whose Sino-Tibetan

92   

   Roger Blench and Mark W. Post

affiliation has not in fact yet been demonstrated, and where the authors maintain doubts as to the likelihood of a genetic relationship in fact being demonstrable. Proto-Sino-Tibetan

Palu

Kamengic

Lui Ayi

Puroik

Bugun Mey Cluster

Mishmi

Miji

Idu Digaro Miju Hruso Koro Milang

ST Affiliation uncertain No Data

Tujia

N.Qiangic

Siangic

JingphoKarenic West Konyak-Bodo = Kachinic Gongduk Bai

Nagish East Tani

Lepcha Nungish Olekha Mahakiranti Greater

LoloSinitic Burmish- Tibetic Naxi (Bodish) S. Qiangic

Mikir (=Karbi) Mruish (Mru-AnuHkongsǝ)

KukiZeme Meithei Chin Ao Tangkhul AngamiPochuri Tamangic KhamMagaricChepang

Lhokpu

Fig. 6: A new proposal for Sino-Tibetan phylogeny

Future developments will quite likely involve the exclusion of some of these languages from the Sino-Tibetan phylum as they are treated as isolates with a SinoTibetan superstrate, as well as the inclusion of barely documented languages within larger groups as they become better known. But it seems likely that some languages will remain controversial, even when we know more about them.

10 W  hat’s in a name? “Trans-Himalayan” and other possibilities If these arguments are even partway accepted, then “Sino-Tibetan” becomes a highly inappropriate name for the phylum, privileging as it does two low-level subgroups. It has been proposed to use the term “Tibeto-Burman” to refer to the



Rethinking Sino-Tibetan phylogeny   

   93

whole phylum (van Driem 2002); but in fact, the same objection applies – Tibetan and Burmese simply being two culturally prominent groups, with no special classificatory significance. One proposal on the table, also endorsed by the authors, is to use the term “Trans-Himalayan;” this would capture the geographical locus of much of the phylum without emphasising individual subgroups. Of course, the academic community may well put forward other suggestions, and we should be open to these.

11 Conclusions The primary object of this paper has not been to put forward a definitive phylogenetic proposal, but instead to suggest that for too long a bundle of ideas and assumptions has been repeated in the literature without any serious evidential base. “Reconstructions” have been proposed which have failed to take many languages of high phyletic significance into account; these forms have been repeatedly quoted without remark in the literature, in the process gaining a lustre they hardly deserve. Sino-Tibetan has no agreed internal structure, and yet its advocates have been happy to propose dates for its origin, expansion and homeland in stark contradiction to the known archaeological evidence. A focus on “high cultures” (Chinese, Tibetan, Burmese) has led to an emphasis on these languages and their written records, something wholly inappropriate for a phylum where an overwhelming proportion of its members speak unwritten languages. Standard handbooks have ignored minority languages whose lexicon and grammar do not fit with prevailing stereotypes. This paper is intended as a contribution towards redressing this balance.

12 Abbreviations, acronyms and symbols * # BP C N PST PTB V

reconstruction regular based on established sound correspondences quasi-reconstruction based on rapid inspection of forms before present consonant nasal proto-Sino-Tibetan proto-Tibeto-Burman vowel

94   

   Roger Blench and Mark W. Post

Appendix: Languages of NE India and Sino-Tibetan terms for ‘snow’ Appendix Table 1 is a list of the languages of NE India and adjacent territories listed in Ethnologue, with correct names, classification and other information where this is available. Absence of an ISO number implies that it is likely that the lect has the status of a language but is not listed separately or at all in the Ethnologue. We also here identify “languages” with ISO codes, but whose status is either in clear error, or whose identification as languages is seemingly not based on any available data. Appendix Tab. 1: Corrected table of languages of NE India and adjacent territories Branch

Language

ISO

Country

Mikir

Amri Karbi Karbi Mru Anu Hkongsə Singpho Lower Adi

ajz mjw mro anl

India India Bangladesh Myanmar Bangladesh India India

Mruish

Jingpho Tani

sgp adi

Upper Adi

India, Tibet

Tangam Mising Pao

mrg ppa

India India India

Galo

adl

India

Central-Western Lower Nyishi Hills Miri Apatani Upper Nyishi

Siangic

Tagin Milang Koro

India, Tibet dap apt nbt

India India India India, Tibet India, Tibet India India

Also

Abor (pejorative), Padam, Pasi, Panggi, Minyong, Komkar Karko, Shimong, Bori, Aashing Miri (pejorative) (this is not a language, but rather a Mising clan name) (Adi) Gallong (exonym) Bokar, Ramo, Pailibo Dafla (pejorative) Hill Miri Na(h), Bengni, Bangni M(a)ra Mala, Holon, Dalbo (Koro) Aka



Rethinking Sino-Tibetan phylogeny   

   95

Branch

Language

ISO

Country

Also

East Bodish Tibetic?

Khamba Memba Meyor Bugun Mey Sartang

kbg mmc zkr bgg sdp onp

India India, Tibet India India India India

Lish

bqh

India

Chug

cvg

India

Mishmic

Idu

clk

India

Mijiic

Taraon Miji

mhu sjl

India India

Bangru Puroik Miju Hruso Baima Ayi

suv mxj hru lsh ayx

India India, Tibet India India China China

Khams Tibetan Tshangla? Zakhring Khowa Sherdukpen But Monpa, dialect of Sherdukpen forms a close dialect cluster with Chug forms a close dialect cluster with Lish Chulikata (exonym, pejorative) Digaru Sajalong, Dhammai Levai Sulung Kaman Aka

Lui

lba

Myanmar

Palu

pbz

Myanmar

Kamengic

Isolate Isolate Isolate Unclassified

Erroneous. Deleted from most recent Ethnologue Erroneous. Deleted from most recent Ethnologue Erroneous. Deleted from most recent Ethnologue

Appendix Tab. 2: Sino-Tibetan terms for ‘snow’ Branch

Language

Attestation

Root

*Tibeto-Burman Bai Bodo-Garo Bodo-Garo Bodish Bodish

Bai Atong Garo (Bangladesh) Cuona Menba Motuo Menba

*kyam suɛ² suri bo-rop kha⁵⁵ru⁵³ phom

#kyam #ʃü[ri] #ʃü[ri] ? #kyam #pham

96   

   Roger Blench and Mark W. Post

Branch

Language

Attestation

Root

Burmish Burmish Burmish Burmish Burmish Burmish Burmish Burmish Himalayish Himalayish Himalayish Himalayish Himalayish Jingpho-Konyak Jingpho-Konyak Jingpho-Konyak Jingpho-Konyak Jingpho-Konyak Jingpho-Konyak Karenic Kiranti Kiranti Kiranti Kiranti Kiranti Kiranti Kuki-Chin Kuki-Chin Kuki-Chin Kuki-Chin Kuki-Chin Kuki-Chin Kuki-Chin Kuki-Chin Loloish Loloish Loloish Loloish Loloish Loloish Loloish Loloish Loloish Loloish Loloish

Achang (Longchuan) Atsi [Zaiwa] Burmese (Yangon) Burmese (Written) Burmese (Written) Lalo Marma Sani [Nyi] Kanauri Kanauri Kanauri Kanauri Pattani [Manchati] Jingpho Jingpho Phom Tangsa (Moshang) Chang Konyak Pa-O Bahing Dumi Kulung Limbu Thulung Yamphu Daai Khumi Lushai [Mizo] Matu Tiddim Mkaang Wakung Sorbung Ahi Bola (Luxi) Gazhuo Hani (Gelanghe) Jinuo Jinuo (Youle) Lahu (Black) Lipho Lisu Lusu Nasu

xan³¹ʑai⁵⁵ kjoʔ²¹ shi⁵⁵n̥ĩ⁵⁵ khyàm hsi³ hnɑŋ³ va²¹ rəkhébraŋ va²¹ tshō rīsŭr stil(h) ṭhāṇöṅ mug khyen² kjo³¹naʔ⁵⁵ shü thikhek ninthu yi wàʔ phyrky rim noŋˍ naŋ phomu naŋ ɴaːŋ kʰu.tᵊmáiˀ vuːr xɔsɔŋ vuːk² ɓɔk.kɜm sʰənãŋ ʔəhúr ɣo²¹ ŋɛ̱̃⁵⁵ xoa⁵⁵ xa³¹ njɛ³³ji⁴⁴ se⁵⁵thɑ⁵⁵ vᴀ⁵³mei¹¹ ɣo²¹ wa⁵/wɑ³¹ ʑy³⁵ vo³³

#xan #kliN #ʃü[ri] #kyam #naŋ #[te] van ? #[te] van #ʃü[ri] #ʃü[ri] ? ? ? #kliN #kliN #ʃü[ri] ? #naŋ ? #[te] van ? ? #naŋ #naŋ #pham #naŋ #naŋ ? #[te] van ? #[te] van #kyam #naŋ #ʃü[ri] #kyam #naŋ #xan #xan ? ? #[te] van #kyam #[te] van #dʑyɛp #[te] van



Rethinking Sino-Tibetan phylogeny   

Branch

Language

Attestation

Root

Loloish Loloish Loloish Loloish Loloish Meithei Mikir Mikir Mijuic Mishmic Mishmic Naish Naga Angami Naga Angami Naga Angami Naga Angami Naga Angami Naga Angami Naga Ao Naga Ao Naga Ao Naga Ao Naga Ao Naga Zeme Naga Zeme Newaric Nungish Nungish Puroik Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic Qiangic

Neisu Nusu (Bijiang) Yi (Dafang) Yi (Nanhua) Yi (Nanjian) Meithei Mikir Mikir Miju Idu Darang [Taraon] Naxi Khonoma Ntenyi Mao Sumi Sumi Sumi Ao (Chungli) Ao (Mongsen) Lotha Naga Sangtam Yimchungrü Mzieme Rongmei Newar Anong Trung Puroik (Sulung) Caodeng Daofu Ergong (Danba) Ersu Guiqiong Muya Muya [Minyak] Namuyi Pumi (Jinghua) Pumi (Jiulong) Pumi (Lanping) Qiang (Mawo) Qiang (Mawo) Queyu (Xinlong) rGBenzhen Rgyalrong

ɣu³³ vɑ⁵⁵ vu³³ ɣo²¹ vɑ²¹ un arekelok ephik so⁵⁵ põ ɑ³¹ɹuɑi⁵⁵ be³³ pekri ghula ovumara mora sü apüghü kukhukite rürjep azu šérə̀ shurü aru moŋ zui gun chwã̄põ thi³¹vɛn⁵³ tɯ³¹wɑn⁵³ kə³³ʑɯh⁵³ tʰɐ-jpe kha va nkhɛ va zɿ⁵⁵ khə⁵⁵wu̵⁵⁵ βə²⁴ vɯ³⁵ ju̵³¹ spy⁵⁵ py⁵⁵ fpy⁵⁵ piɛ tɕi qəʴ βu kha⁵⁵wa⁵⁵ te va tɐi jpa

#kyam #[te] van #[te] van #kyam #[te] van ? ? ? #ʃü[ri] #pham #aru #pu[n] ? ? ? ? #ʃü[ri] ? ? ? #ʃü[ri] ? #ʃü[ri] #aru ? #kyam #pu[n] #te van #te van #ʃü[ri] #pu[n] #kyam #[te] van #dʑyɛp #kyam #pu[n] #pu[n] #pu[n] #pu[n] #pu[n] #pu[n] #pu[n] #pu[n] #kyam #te van #pu[n]

   97

98   

   Roger Blench and Mark W. Post

Branch

Language

Attestation

Root

Qiangic Qiangic Qiangic Qiangic Qiangic Siangic Sinitic Sinitic Sinitic Sinitic Sinitic Sinitic Tangkhulic Tani Tani Tani Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tamangic Tibetic Tibetic Tibetic Tibetic Tibetic Tibetic Tibetic Tibetic Tibetic Tujia

Rgyalrong Shixing Tangut [Xixia] Taoping Zhaba Milang Modern Chinese Old Chinese Modern Chinese Old Chinese Modern Chinese Old Chinese Tangkhul *Tani Apatani Galo *Tamang *Tamang Tamang (Risiangku) Tamang (Sahu) Tamang (Taglung) Gurung (Ghachok) Gurung (Ghachok) Manang (Gyaru) Manang (Gyaru) Manang (Ngawal) Manang (Prakaa) Thakali (Marpha) Thakali (Syang) Thakali (Tukche) Tibetan (Alike) Tibetan (Amdo) Tibetan (Batang) Tibetan (Khams:Dege) Tibetan (Lhasa) Tibetan (Spiti) Tibetan (Written) Tibetan (Written) Tibetan (Xiahe) Tujia

tey va dʑyɛ³⁵ wjị mə³¹pɑ⁵⁵ vʑi¹³ ta-pim bīng [ice] *prəŋ [ice] jiāo [frost] *krû xuě *sot pham *ta-pam pẽ́m-bè ta-pam¹ gliŋ⁴ gaŋᴬ kliŋ⁴ ‘sera ‘yu-pa khliŋ klĩq ɬĩ² kyin⁴ gᴇː⁴ khĩ⁴ kaŋ³ lin⁴ lim¹¹ kin⁴ kaŋ khaŋ khɑ⁵⁵ kha⁵⁵wa⁵³ khaŋ¹³² kʰá gaŋs [ice] khaba khaŋ su³⁵su⁵⁵

#te van #dʑyɛp ? #pu[n] #dʑyɛp #pham #pham #pham ? ? #kliN #ʃü[ri] #ʃü[ri] #pham #pham #pham #pham #kliN #kyam #kliN ? #kliN #kliN #kliN #kliN ? #kliN #kyam #kliN #kliN #kliN #kyam #kyam #kyam #kyam #kyam #kyam #kyam #kyam #kyam #ʃü[ri]

Sources: Compiled from the STEDT database, with additional inputs from other published and manuscript sources, as well as the authors’ field notes. The proposed quasi-reconstructions are by the authors



Rethinking Sino-Tibetan phylogeny   

   99

References Abraham, Binny, Kara Sako, Elina Kinny and Isapdaile Zeliang. 2005. A Sociolinguistic Research Among Selected Groups in Western Arunachal Pradesh Highlighting MONPA. Unpublished report. Aldenderfer, Mark, and Yinong Zhang. 2004. The prehistory of the Tibetan Plateau to the seventh century A.D: perspectives and research from China and the West since 1950. Journal of World Prehistory 18 (1): 1–55. Bandyopadhyay, Swapon Kumar. 1989. A Tangsa wordlist. Linguistics of the Tibeto-Burman Area 12 (2): 79–91. Benedict, Paul King. 1972. Sino-Tibetan: a conspectus (“STC”). Contributing editor, J. A. Matisoff. Princeton-Cambridge Series in Chinese Linguistics #2. New York: Cambridge University Press. Benedict, Paul King. 1976. Sino-Tibetan: another look. Journal of the American Oriental Society 96 (2): 167–197. Bernot, Lucien 1966. Eléments de vocabulaire Cak recueilli dans le Pakistan Oriental. Essays Offered to G. H. Luce by His Colleagues and Friends in Honour of His Seventy-Fifth Birthday. Volume 1: Papers on Asian History, Religion, Languages, Literature, Music Folklore, and Anthropology. Artibus Asiae, Supplementum 23: 67–91. Bhattacharjee, Tarun Kumar. 1983. The Idus of Mathun and Dri Valley. Itanagar: Government of Arunachal Pradesh. Blackburn, Stuart. 2007. Oral stories and culture areas: from Northeast India to Southwest China. South Asia: Journal of South Asian Studies 30 (3): 419–437. Blench, Roger M. 2008a. Stratification in the peopling of China: how far does the linguistic evidence match genetics and archaeology? In Alicia Sanchez-Mazas, Roger M. Blench, Malcolm D. Ross, I. Peiros and Marie Lin (eds.). Human migrations in continental East Asia and Taiwan. Matching archaeology, linguistics and genetics, 105–132. London: Routledge. Blench, Roger M. 2008b. Re-evaluating the linguistic prehistory of South Asia. In Toshiki Osada and Akinori Uesugi (eds.). Occasional Paper 3: Linguistics, Archaeology and the Human Past, 159–178. Kyoto: Indus Project, Research Institute for Humanity and Nature. Blench, Roger M. 2011. The role of agriculture in the evolution of Southeast Asian language phyla. In N. Enfield (ed.). Dynamics of Human Diversity in Mainland SE Asia, 125–152. Canberra: Pacific Linguistics. Blench, Roger M. 2012. Vernacular names for taro in the Indo-Pacific region and their possible implications for centres of diversification. In M. Spriggs, David Addison and Peter J. Matthews (eds.). Irrigated Taro Colocasia esculenta in the Indo-Pacific: Biological, Social and Historical Perspectives, 21–43. Osaka: Minpaku. Blench, Roger M. in press a. The contribution of linguistics to understanding the foraging/ farming transition in NE India. In Tia Toshi Jamir and Manjil Hazarika (eds.), 51 Years after Daojali Hading: Emerging perspectives in the Archaeology of Northeast India. Essays in Honour of T.C. Sharma. Blench, Roger M. in press b. Was there once an arc of vegeculture linking Melanesia with Northeast India? In Glenn Summerhayes, David Addison (eds.). Selected papers from the 2011 Lapita Pacific Archaeology Conference: ‘Pacific Archaeology: documenting the past 50,000 years to the present.’ Auckland: Auckland University Press.

100   

   Roger Blench and Mark W. Post

Blench, Roger M. in press c. Vernacular names for millets and other minor cereals in east and southeast Asia and their relevance to centres of domestication and spread. Archaeological and Anthropological Science. [Special issue on small millets ed. Dorian Fuller]. Blench, Roger M. in press d. The world turned upside down: sago-palm processors in Northeast India and the origins of Chinese civilisation. Paper presented at the panel ‘The sub-Himalayan Corridor: Just what is going on in North East India?’ Northeast India Panel, EurASEAA XIV, Dublin 2012. Submitted for a volume of selected papers, Helen Lewis ed. to be published by National University of Singapore Press. Bodman, Nicholas Cleaveland. 1980. Proto-Chinese and Sino-Tibetan: data towards estab­lishing the nature of the relationship. In Frans van Coet­sem and Linda R. Waugh (eds.). Con­tri­butions to Historical Linguistics: Issues and Materials, 34–199. Leiden: E.J. Brill. Boro, A. 1978. Miju dictionary. Shillong: Director of Research, Arunachal Pradesh Government. Bradley, David 1997a. What did they eat? Grain crops of the Burmic peoples. Mon-Khmer Studies 27: 161–170. Bradley, David. 1997b. Tibeto-Burman languages and classification. In David Bradley (ed.) Tibeto-Burman languages of the Himalayas, 1–72. Pacific Linguistics A-86. Canberra: Australian National University. Bradley, David. 2002. The subgrouping of Tibeto-Burman. In Christopher Beckwith (ed.). Medieval Tibeto-Burman Languages, 73–112. Leiden: Brill. Brassett, Philip and Cecilia Brassett. 2005. Imperial Tiger Hunters: An Introduction to the Tujia People of China. Antony Rowe Publishing Services. Brightbill, Jeremy, Amy Kim and Seung Kim. 2007. The War-Jaintia in Bangladesh: A Sociolinguistic Survey. Electronic Survey Report. SIL International. Burling, Robbins. 1999. On Kamarupan. Linguistics of the Tibeto-Burman Area 22 (2): 169–171. Burling, Robbins. 2003. The Tibeto-Burman languages of Northeastern India. In Graham Thurgood and Randy J. LaPolla (eds.). The Sino-Tibetan languages, 169–191. London and New York: Routledge. Button, Christopher T. J. 2009. A Reconstruction of Proto Northern Chin in Old Burmese and Old Chinese Perspective. London: School of Oriental and African Studies, Univeristy of London dissertation. Caturvedi, Mahendra 1970. A practical Hindi-English dictionary. Delhi: National Publishing House. Chang, Kun. 1972. Sino-Tibetan ‘Iron:’ *Qhleks. Journal of the American Oriental Society 92 (3): 436–446. DeLancey, Scott. 2012. On the origins of Bodo-Garo. In Gwendolyn Hyslop, Mark W. Post and Stephen Morey (eds.), North East Indian Linguistics Volume 4. 3–20. New Delhi, Cambridge University Press India DeLancey, Scott. (this volume). Creolization in the Divergence of the Tibeto-Burman Languages. Dempsey, Jacob M. 1995. A reconsideration of some phonological issues involved in reconstructing Sino-Tibetan numerals. Seattle, WA: University of Washington dissertation. Deuri, Ram Kumar 1982. The Sulungs. Shillong: Government of Arunachal Pradesh. Diffloth, Gérard 1980. The Wa Anaphors. Linguistics of the Tibeto-Burman Area 5 (2): 1–182. van Driem, George. 1995. Black Mountain verbal agreement morphology, Proto-Tibeto-Burman morphosyntax and the linguistic position of Chinese. In New Horizons in Tibeto-Burman Morphosyntax, 229–259. SENRI Ethnological Studies 41. Osaka: National Museum of Ethnology.



Rethinking Sino-Tibetan phylogeny   

   101

van Driem, George. 1997. Sino-Bodic. Bulletin of the School of Oriental and African Studies 60 (3): 455–488. van Driem, George. 1998. Neolithic correlates of ancient Tibeto-Burman migrations. In Roger M. Blench and M. Spriggs (eds.). Archaeology and Language II, 67–102. London: Routledge. van Driem, George. 2001. Languages of the Himalayas: An Ethnolinguistic Handbook of the Greater Himalayan Region containing an Introduction to the Symbiotic Theory of Language. (2 volumes). Leiden: Brill. van Driem, George. 2002. Tibeto-Burman replaces Indo-Chinese in the 1990s: Review of a decade of scholarship. Lingua 111: 79–102. van Driem, George. 2008a. To which language family does Chinese belong, or what’s in a name. In Alicia Sanchez-Mazas, Roger Blench, Malcolm D. Ross, Ilia Peiros and Marie Lin (eds.). Past Human Migrations in East Asia: Matching Archaeology, Linguistics and Genetics, 219–253. London and New York: Routledge. van Driem, George. 2008b. The Naga language groups within the Tibeto-Burman language family. In Michael Oppitz, Thomas Kaiser, Alban von Stockhausen and Marion Wettstein (eds.). Naga identities: changing local cultures in the Northeast of India, 311–321. Zurich: Snoeck Publishers. Edmondson, Jerold A. n.d. Phu Kha, Xá Phó, Mantsi, Coong, Sila, Lahu, Hani, and Proto-Loloish data list. Undated electronic ms. Fu Daxiong. 2001. Xizang Changguogou yizhi xin shiqi shidai nongzuowu yicun de faxian, jianding he yanjiu [Discovery, Identification and Study of the Remains of Neolithic Cereals from the Changguogou Site, Tibet]. Kaogu 3: 66–74. Fuller, Dorian Q., Lin Qin and Emma Harvey 2008. Evidence for the late onset of agriculture in the Lower Yangtze region and challenges for an archaeobotany of rice. In Alicia SanchezMazas, Roger Blench, M.D. Ross, I. Peiros and Marie Lin (eds.). Human migrations in continental East Asia and Taiwan. Matching archaeology, linguistics and genetics, 40–83. London: Routledge. Grewal, Dalvinder Singh 1997. The tribes of Arunachal Pradesh. (2 volumes). Delhi: South Asia Publications. Grierson, George A. (ed.) 1909. Linguistic Survey of India. Volume III, 1, Tibeto-Burman family. Calcutta: Superintendent of Government Printing. Handel, Zev. 2008. What is Sino-Tibetan? Snapshot of a Field and a Language Family in Flux. Language and Linguistics Compass 2 (3): 422–441. Hashimoto, Mantaro J. 1976. Language diffusion on the Asian continent: Problems of typological diversity in Sino-Tibetan. Computational Analyses of Asian and African Languages 3: 49–65. Hildebrandt, Kristine Ann. 2003. Manange Tone: Scenarios of Retention and Loss in Two Communities. Santa Barbara, CA: University of California at Santa Barbara dissertation. Hill, Nathan W. 2009. Review of Handbook of Proto-Tibeto-Burman: System and Philosophy of Sino-Tibetan Reconstruction. By James A. Matisoff. Berkeley: University of California Press, 2003. Language and Linguistics 10 (1): 173–195. Hiroo, Nasu Momohara Arata, Yasuda Yoshinori and Jiejun He. 2007. The occurrence and identification of Setaria italica (L.) P. Beauv. (foxtail millet) grains from the Chengtoushan site (ca. 5800 cal B.P.) in central China, with reference to the domestication centre in Asia. Vegetation history and archaeobotany 16 (6): 481–494. Huang, W. W. 1994. The prehistoric human occupation of the Qinghai-Xizang Plateau. Göttinger Geographische Abhandlungen 95: 201–219.

102   

   Roger Blench and Mark W. Post

Hyslop, Gwendolyn. (this volume). A preliminary reconstruction of East Bodish. Joshi, B. D. and R. S. Rana. 1995. Buckwheat (Fagopyrum esculentum). In J.T. Williams (ed.). Cereals and pseudo-cereals, 84–127. London: Chapman and Hall. von Klaproth, Julius Heinrich. 1823. Asia Polyglotta. Paris: A. Schubart. LaPolla, Randy J. 2001. The role of migration and language contact in the development of the Sino-Tibetan language family. In Alexandra Y. Aikhenvald and R. M. W. Dixon (eds.). Areal Diffusion and Genetic Inheritance, 225–254. Oxford: Oxford University Press. Lewis, M. Paul, Gary F. Simons, and Charles D. Fennig (eds.). 2013. Ethnologue: Languages of the World, Seventeenth edition. Dallas, Texas: SIL International. Online version: http://www.ethnologue.com. Lu, Tracey L-D. 2009. Prehistoric coexistence: the expansion of farming society from the Yangzi River Valley to western south China. In K. Ikeya, H. Ogawa, P. Mitchell (eds.). Interactions between Hunter-Gatherers and Farmers: from Prehistory to Present. Senri Ethnological Studies 73, 47–52. Osaka: National Museum of Ethnology. Marrison, Geoffrey E. 1967. The classification of the Naga Languages of north-east India. London: School of Oriental and African Studies, University of London dissertation. (2 volumes). Matisoff, James Alan. 1991. Sino-Tibetan linguistics: Present state and future prospects. Annual Review of Anthropology 20: 469–504. Matisoff, James Alan. 2000. On “Sino-Bodic” and Other Symptoms of Neosubgroupitis. Bulletin of the School of Oriental and African Studies 63 (3): 356–369. Matisoff, James Alan. 2003. Handbook of Proto-Tibeto-Burman: system and philosophy of Sino-Tibetan reconstruction. Berkeley, Los Angeles, and London: University of California Press. Matisoff, James Alan. 2007. Response to Laurent Sagart’s review of Handbook of Proto-TibetoBurman: System and philosophy of Sino-Tibetan reconstruction. Diachronica 24 (2): 435–444. Matisoff, James Alan. 2008. The Tibeto-Burman Reproductive System: Toward an Etymological Thesaurus. Berkeley: University of California Press. Morris, Stephen. 1991. The Oya Melanau. Kuching: Malaysian Historical Society, Sarawak Branch. Mortensen, David. 2003. Comparative Tangkhul. ms. Ohnishi, O. 1998. Search for the wild ancestor of the buckwheat. III. The wild ancestor of the cultivated common buckwheat, and of tartary buckwheat. Economic Botany 52: 123–133. Paulsen, D. 1989. A phonological reconstruction of proto-Plang. Mon-Khmer Studies 18–19: 160–222. Post, Mark W. 2011. Prosody and typological drift in Austroasiatic and Tibeto-Burman: Against “Sinosphere” and “Indosphere.” In Sophana Srichampa, Paul Sidwell and Kenneth J. Gregerson (eds.). Austroasiatic Studies: Papers from ICAAL4. Mon-Khmer Studies Special Issue No. 3, 198–221. Canberra: Pacific Linguistics. Post, Mark W. and Roger M. Blench. 2011. Siangic: A new language phylum in North East India. 6th International Conference of the North East Indian Linguistics Society. Tezpur University, Tezpur, Assam, India. Post, Mark W. and Yankee Modi. 2011. Language contact and the genetic position of Milang (Eastern Himalaya). Anthropological Linguistics 53 (3): 215–258. Pulu, Shri Jimi 2002. A handbook on Idu Mishmi language. Itanagar: Arunachal Pradesh State Government Directorate of Research.



Rethinking Sino-Tibetan phylogeny   

   103

Rhoads, J. W. 1981. Variation in land use strategies among Melanesian Sago Eaters. Canberra Anthropology 4 (2): 45–73. Sagart, Laurent. 2006. Review of Handbook of Proto-Tibeto-Burman: System and philosophy of Sino-Tibeto-Burman reconstruction. By James A. Matisoff. Berkeley: University of California Press, 2003. Diachronica 23 (1): 206–223. Sagart, Laurent. 2008. Reply to Matisoff on the Handbook of Proto-Tibeto-Burman: System and philosophy of Sino-Tibetan reconstruction. Diachronica 25 (1): 153–155. Shafer, Robert. 1947. Hruso. Bulletin of the School of Oriental and African Studies 12 (1): 184–196. Shafer, Robert. 1955. Classification of the Sino-Tibetan languages. Word 11 (1): 94–111. Shafer, Robert. 1966/1967. Introduction to Sino-Tibetan (Part I: 1966, Part II: 1967). Wiesbaden: Otto Harrassowitz. Sidwell, Paul. 2000. Proto South Bahnaric: a reconstruction of a Mon-Khmer language of Indo-China. PL 501. Canberra: Pacific Linguistics. Singh, U. Nissor. [1906] 1983. Khasi-English Dictionary. Shillong: Eastern Bengal and Assam Secretariat Press. Sino-Tibetan Etymological Dictionary and Thesaurus (STEDT). http://stedt.berkeley.edu/ Stampe, David. ined. Comparative Munda wordlists. Electronic ms. available from the author. Starostin, Sergei. 2008. Altaic loans in Old Chinese. In Alicia Sanchez-Mazas, Roger M. Blench, M.D. Ross, I. Peiros and Marie Lin (eds.). Human migrations in continental East Asia and Taiwan. Matching archaeology, linguistics and genetics, 254–262. London: Routledge. Stonor, C. R. 1952. The Sulung tribe of the Assam Himalayas. Anthropos 47: 947–962. Sun, Jackson Tian-Shin. 1993. A Historical-Comparative Study of the Tani Branch of TibetoBurman. Berkeley, CA: University of California dissertation. Tassar, L. (Unpublished manuscript). The tasse: A food plant of Arunachal Pradesh. Available at http://arunachalipr.gov.in/Latest_publication/The%20Tasse%20A%20Food%20Plant%20 of%20Arunachal.pdf. Accessed May 30, 2011. Tayeng, Aduk 1990. Sulung language guide. Itanagar: Government of Arunachal Pradesh. Taylor, Gordon D. 1953. Crop Distributions by Tribes in Upland Southeast Asia. Southwestern Journal of Anthropology 9 (3): 296–308. Thurgood, Graham and Randy J. LaPolla. (eds.) 2003. The Sino-Tibetan languages. London and New York: Routledge. University of Gauhati. 1962. Candrakanta abhidhana : Asamiyi sabdara butpatti aru udaharanere Asamiya-Ingraji dui bhashara artha thaka abhidhana. 2nd ed. Guwahati: Guwahati Bisbabidyalaya. VanBik, Kenneth. 2007. Proto-Kuki-Chin. Berkeley, CA: University of California dissertation. Weidert, Alfons. 1987. Tibeto-Burman Tonology: A comparative analysis. Amsterdam and Philadelphia: John Benjamins. Wiersma, Grace. 1990. Investigation of the Bai (Minjia) language along historical lines. Berkeley, CA: University of California dissertation. Wolfenden, Stuart N. 1927. Outlines of Tibeto-Burman Linguistic Morphology. London: The Royal Asiatic Society. Wood, Daniel C. 2008. An initial reconstruction of Proto-Boro-Garo. Eugene, OR: University of Oregon MA thesis Xue-bin Qi, Han Jianlin, Roger Blench, J. Edward O. Rege and Olivier Hanotte. 2008. Understanding yak pastoralism in Central Asian Highlands: mitochondrial DNA evidence for origin, domestication and dispersal of domestic yak. In Alicia Sanchez-Mazas, Roger

104   

   Roger Blench and Mark W. Post

M. Blench, M.D. Ross, I. Peiros and Marie Lin (eds.). Human migrations in continental East Asia and Taiwan. Matching archaeology, linguistics and genetics, 427–442. London: Routledge. Yuan, Jing, Han Jianlin and Roger M. Blench. 2008. Livestock in ancient China: an archaeozoological perspective. In Alicia Sanchez-Mazas, Roger M. Blench, M.D. Ross, I. Peiros and Marie Lin (eds.). Human migrations in continental East Asia and Taiwan. Matching archaeology, linguistics and genetics. 84–104. London: Routledge. Yuan, Jing and R. Flad. 2002. Pig domestication in ancient China. Antiquity 76: 724–732. Zhang, D. D., S. H. Li, Y.Q. He, and B.S. Li. 2003. Human settlement of the last glaciation on the Tibetan plateau. Current Science 84: 701–704. Zong, Y. 2007. Fire and flood management of coastal swamp enabled first rice paddy cultivation in east China. Nature 449: 459–462.

Nicolas Tournadre1

The Tibetic languages and their classification 1 Various meanings of “Tibetic” The term “Tibetic” has been used in the recent past by some authors in different ways to refer to various intermediate levels of classification within TibetoBurman (e.g. Matisoff 2003, Beckwith 2006, Lewis 2009, Blench 2011, de Haan 2011, DeLancey 2012, van Driem 2012, Noonan 2012, Chirkova in press inter alia).2 Depending on the author, “Tibetic” may be used to refer to a high level of subgrouping more or less equivalent to “Bodic” or “Himalayish” or to a lower level of sub-grouping such as “Bodish” or an even lower level as an alternative form for “Tibetan languages/dialects.” “Bodic,” also referred to as “Tibeto-Himalayan” (Michailovsky 2011), groups together many Eastern Himalayan languages (such as the Kiranti languages), Western Himalayan languages (such as Kannauri, Almora), the so-called “Tibetan dialects” and “Bodish languages” (such as Tamang, Gurung, Bumthang, Tshangla etc.). “Bodic” is a “heterogenous and impressionistic” group (van Driem 2011: 31). “Bodish,” sometimes referred to as Tibeto-Kannauri, although less problematic is also not well-defined. It groups together the “Tibetan dialects,” Tamangic languages and some other languages such as Tshangla, Bumthang, Kurtoep, Bake3 etc. The historical comparative methodology has so far not provided common innovations which would delimit clearly the Bodish subgroup. We can only agree with Matisoff’s (1989, 1990) note of caution to megaloreconstructionists, or van Driem who states that it is safer “to discover the structure of the [Tibeto-Burman] family tree by working up from the firmer group of lower-level subgroups to the higher levels of superordinate subgroups” (2011: 37).

1 Aix-Marseille University and CNRS (Lacito). I would like to warmly thank Nathan W. Hill, Guillaume Jacques, Hiroyuki Suzuki, Randy LaPolla, Lauren Gawne and two anonymous reviewers, who read an earlier version of this article for their relevant remarks and suggestions. 2 The Sino-Tibetan macrofamily groups together the “Tibeto-Burman” and “Sinitic” languages. I think it is more convenient and appropriate to call this macrofamily Sino-Tibeto-Burman (STB), because it allows us to clearly include the Sinitic languages (whatever their status may be) when discussing general issues about the inner and outer classification of the Tibeto-Burman subgroup. However, in this paper, I will use the traditional taxon ‘Tibeto-Burman’ (TB). 3 Bake or brag-skad is a language spoken near the Basum lake (brag-gsum mtsho) in the Kongpo area, less than 500 kms from Lhasa.

106   

   Nicolas Tournadre

Thus the use of “Tibetic” to replace already existing labels referring to subgroups which are not well-defined or to create new intermediate levels of classification without clear scientific criteria is not appropriate. Because of its various meanings, the term “Tibetic” could become useless, unless it receives a precise definition and is used in a consistent way.

2 Definition of “Tibetic” The term “Tibetic” could, however, become a useful replacement for the notion of “Tibetan dialects,” which is not appropriate for various reasons. First, the notion of “Tibetan dialects” implies the existence of a single “Tibetan language.” However, the so-called “Tibetan dialects” refer in fact to various languages which do not allow mutual intelligibility. Until the last decades of the 20th century, many scholars would still classify the “Tibetan dialects” into three major groups: Khams,4 Amdo, Ü-Tsang (thus taking into account only the “dialects” found in China) or into five major dialects: Central (Ü-Tsang), Southeastern (Khams), Western, Southern and Northeastern (Amdo).5 In recent years, some authors such as Sun (2003a, 2006, 2007), Suzuki (2009b, 2009c, 2009d, 2011a, 2011b), Zeisler (2011, 2012), Haller (2000, 2004, 2009), and Chirkova (2005, 2008b), to name a few, have provided new data on various languages and dialects of the area. These very significant contributions have resulted in new developments in the understanding of the Tibetan linguistic area and its diversity. Second, these “Tibetan dialects” are spoken not only by Tibetans per se but also by other ethnic groups such as Ladakhi, Balti, Lahuli, Sherpa, Bhutanese, Sikkimese Lhopo, etc. who do not consider themselves to be Tibetans.6 They do not call their language “Tibetan.” In a similar way, we do not talk of Latin Languages but of Romance languages and do not think of French, Portuguese, Italian, Catalan or Romanian as various dialects of Latin.7

4 I use here the orthograph Khams (which corresponds to the Wylie transliteration and not the pronunciation) because the word Kham also refers to a Tibeto-Burman language spoken in Nepal and not related to Tibetan. 5 See Gesang Jumian [1964] 2002; Qu Aitang 1996. For a critical approach, see Sun 2003a or Tournadre 2005. 6 In many areas outside Tibet, such as the southern Himalayas, people speaking Tibetan-related languages sometimes call themselves ‘Bhutia’ (བྷོ་ཊི་ bhoTi) or ‘Lamas’ (བླ་མ་ bla-ma) . 7 Nor would we think of English as a dialect of German just because German and English belong to the Germanic family. However the term “dialect” is sometimes used in a broader sense and means related varieties: so, for example, one of the definitions of “dialect” given in the Merriam-



The Tibetic languages and their classification   

   107

With the recent descriptions of many new “dialects” or “languages,” scholars of Tibetan linguistics have come to realize the incredible diversity of this linguistic area. The representation of a single language is no longer viable and we have to speak of a language family. In fact, the Tibetic linguistic family is comparable in size and diversity to the Romance or Germanic families. The term “Tibetic” is thus very convenient to denote a well-defined family of languages derived from Old Tibetan.8 The language called Old Tibetan was spoken at the time of the Tibetan empire (7th-9th centuries). According to Hill “Old Tibetan was originally spoken in the Yarlung valley, the cradle of the Tibetan empire” (2010: 111). While I tend to agree with this statement, it is hard to delimit the precise extension of the area where this language was spoken as well the degree of dialectal diversity at the time of the Empire. Much more controversial is Hill’s remark about the relationship between “Common Tibetan,” the ancestor of all the modern languages (see Section 4 below on Proto-Tibetic) and Old Tibetan: “Because Tibetan languages began to diverge from each other some centuries after Old Tibetan was committed to writing, the written system represents an état de langue (Old Tibetan) older than that reconstructible via the comparative method (Common Tibetan)” (2010: 112). Old Tibetan is very similar to the Classical Literary language which has preserved a very archaic orthography. And indeed, all the modern languages not only have regular reflexes of Classical Literary Tibetan (CLT), they also share a core vocabulary and grammar. The Tibetic language family belongs to a very small circle of language families throughout the world, derived from a common language which is identical or closely related to an old literary language.9 This small group includes for example the Romance languages with Latin, the Arabic languages (or “dialects”) with Classical Arabic, the Sinitic languages with Middle Chinese,10 the modern Indic languages with Vedic Sanskrit11 etc. There are only a few other examples of this Webster Online Dictionary (http://www.merriam-webster.com/) is “one of two or more cognate languages  : French and Italian are Romance dialects.” 8 Although in some rare cases such as Baima or Khalong, a Qiangic substratum is a very probable hypothesis. 9 The ideas elaborated in this paragraph were suggested by Guillaume Jacques, whom I thank. 10 Except maybe the Min dialects which bear more archaic features. 11 A few languages have old literary relatives but these literary forms can not be considered as (nearly) identical to the proto-languages: this is for example the case of the Runic inscriptions in relation to Proto-Germanic, Avestan and Old Persian in relation to Proto-Iranian, or Old Church Slavonic in relation to Proto-Slavic. In the case of Iranian or Slavonic languages, the Old Literary languages are precious to reconstruct the Proto-languages but their reconstruction is achieved by the comparative method.

108   

   Nicolas Tournadre

specific relation between a proto-language and an old literary language. These language groups or families have a great significance not only for the theory of language evolution but also for the typology of sound changes, for morphology and syntax. It is particularly true for Tibetan since some of the attested sound changes have no equivalent in European languages. As Jacques (p.c.) rightly points out, “a general model about the evolution of languages has to take into consideration the specific features of the Tibetic languages. Their interest goes well beyond the Sino-Tibetan macrofamily.”

3 Identification of Tibetic languages The identification of a Tibetic language should thus be based on a number of phonological, morphosyntactic and lexical criteria, which we will mention below. It is important to use such criteria also for the so-called “Bodish” languages, as well as the Rgyalrongic and Qiangic languages languages spoken in eastern Tibet (China). These Tibeto-Burman subgroups have been heavily influenced by Classical Literary Tibetan and various spoken forms of Tibetan because of political, cultural and liturgical factors, and together with the southern Himalayan Bodish languages of Nepal, India and Bhutan they can be considered to belong to the “Tibetosphere,”12 i.e. the sphere of influence of the Tibetan language, culture and religion. In the cases of some languages spoken in the Tibetosphere (such as Bake, Tshona Mönpa, Kheng, Kurtöp etc.), the existence of numerous loanwords from Tibetan may at first give the wrong impression that they are Tibetic languages. While all modern Tibetic developments can be directly derived from Classical Literary Tibetan (or forms very closely related to it) or indirectly derived from it (in the case of local innovations), this is not the case for Bodic or even Bodish developments. The examination of the phonological, lexical and morphosyntactic criteria allows us to sort out Tibetospheric languages from the actual Tibetic languages. These criteria reveal that, beyond the first impression, there is a real gap between the Tibetic languages and other closely related Bodish languages.

12 I coined this term by analogy with the terms “Sinosphere” and “Indosphere,” proposed by Matisoff (1990). The people speaking Rgyalrongic languages are all “Ethnic Tibetans” in that they identify with Tibetan ethnicity. A majority of speakers of Qiangic languages are considered as “Tibetans” and are officially recognized as such by the Chinese state. Other groups of Qiangic speakers are classified by the Chinese government as members of the Qiang Nationality and the Pumi Nationality.



The Tibetic languages and their classification   

   109

3.1 Phonology Tibetic languages exhibit a great variation on the phonological level. For example, some languages have suprasegmental distinctive features (pitch tone, phonation registers such as creaky voice, breathy voice) while other languages, which have preserved some consonant clusters (or traces of these clusters), do not have such distinctions. However, unlike Bodish languages, all the modern Tibetic languages have regular reflexes of Classical Literary Tibetan. This phonological criterion for the identification of a “Tibetic language” specifies that it is not enough for a given language to have a great number of cognates with Classical Literary Tibetan, these cognates should match regular phonological reflexes of the Literary forms. The regularity of sound changes is of course a fundamental rule of classical historical and comparative linguistics and applies very well to the Tibetic family. Let us exemplify the above statement with two examples. The reflexes of the initial consonant cluster LT found in Classical Literary Tibetan may be quite different in the various modern Tibetic languages, but they are absolutely regular. For example, we find the following reflexes: /lta/, /rta/, /hta/, /tā/, /lhā/ etc. as shown below:13 LTA ལྟ་ ‘look at’ > /lta/ (Ladakhi, Balti), /rta/ (“archaic Amdo”) or /hta/ (“innovative Amdo”),14 /tā/ (Ü,Tsang, Khams), /lhā/ (Sherpa). LTOGS ལྟོགས་ ‘be hungry’ > /ltoks/ (Balti), /rtox/ (archaic Amdo) or /htox/ (innovative Amdo), /tōʔ/ (Ü,Tsang, Khams), /lhōʔ/ (Sherpa) Thus in a given Tibetic language, any word derived from a Classical Literary Tibetan word that contains an initial cluster LT, should have the same derivation. In the above list of LT reflexes, all the modern Tibetic forms contain a /t/, except Sherpa which has a surprising form /lh/. In order to check the linguistic affiliation of Sherpa and confirm that it is a Tibetic language, we should find that all the Sherpa words derived from Literary forms that contain a LT sequence should yield /lh/ in this language. And this is actually the case. The cluster LT in words 13 These examples provide the main reflexes for LT and SR in a sample of Tibetic languages. However, the reflexes given for one language are not exhaustive. They also depend on the dialectal variety. For example, the main reflex of SR in Amdo is /ʂ/ but in Golok Amdo, the reflex is /ʈ/. The same is true for other dialect groups, particularly for Khams. 14 See Nishi 1986. The distinction between Archaic and Innovative Amdo could also be referred to as Pastoralists’ Amdo versus Cultivators’ Amdo.

110   

   Nicolas Tournadre

such as lta ‘look at,’ ltogs ‘be hungry’ (see examples above) or lte-ba ‘navel,’ ltas-mo ‘show’ etc. all yield the expected reflex /lh/. The regularity of sound changes in Tibetic may again be exemplified by the reflexes of the initial consonant cluster SR: /ʂ/, /str/, /ʈ/, /s/. SROG སྲོག་ ‘life’ > /ʂox/ (Amdo), /ʂok/ (Ladakhi), /stroq/ (Balti), /ʂōʔ/ (Tsang, Sherpa), /ʈōʔ/ or /sōʔ/ (Ü), /sōʔ/ (Khams) SRUNG སྲུང་ ‘guard’ (verb) > /ʂoŋ/ (Amdo), /ʂuŋ/ (Ladakhi), /struŋ / (Balti), /ʂūŋ / (Tsang, Sherpa), /ʈūŋ/ (Ü), /sūŋ/ (Khams)

As shown in the two examples above, the retroflex sound /ʈ/ is the normal reflex in Central Tibetan (Ü) and will normally appear in any word that contains a SR initial cluster such as sran-ma ‘pea,’ srang-lam ‘street,’ srab ‘horse bit,’ srab-mo ‘thin,’ sreg ‘burn,’ etc. There are some exceptions in the regularity but they may be explained as loanwords from other dialects or as prestigious pronunciations influenced by the Classical Literary language. For example, the word srog ‘life’ may be pronounced as /sōʔ/, which reflects the reading pronunciation. In some Tibetic languages, reflexes can be quite remote from the original but they remain regular. For example, the word KLAD-PA ‘brain’ becomes /xlatpa/ (Balti); /lǟpa/ (Ü); /lēta/ (Sherpa) and even /lēp/ (Dzongkha).15 In Baima (Zhang Jichuan 1997; Chirkova 2008a) and in many Southern Khams dialects (Suzuki 2009a), the phonetic evolution is even more spectacular. For example, in Baima LAG-PA ‘hand’ becomes /iɑ/; LUG ‘sheep’ > /y/. However, these changes are totally regular. The initial L yields a glide in Baima and the second syllable is reduced in many contexts (e.g. MCHIL-MA > /dʒuɛ |̷ / ‘spittle’). Tibetic languages have also developed a number of phonotactic restrictions which are not present in neighbouring Bodish languages (such as Khengkha, Tamang, etc.). This is the case for example with the onsets /ml/, /pl/ and /ŋr/, which are not allowed in Tibetic languages, but are found in non-Tibetic Bodish languages.

3.2 Lexicon In order to define “Tibetic,” a simple test based on the word for ‘seven’ was proposed by Nishi (1986: 849), Beyer (1992: 7), and by Michailovky and Mazaudon

15 The latter example is the result of syllable merger.



The Tibetic languages and their classification   

   111

(1994: 2). In the Tibetic languages, the word corresponding to ‘seven’ is regularly derived from Classical Tibetan bdun while it is not the case in other languages, even when in the case of Tibetospheric languages closely related to the Tibetic languages. The above test is generally valid and could be used as a first indication, but numbers can always be borrowed. Indeed in Japhug (a Rgyalrongic language) the ordinal number16 has been borrowed from Tibetan: βdɯnpa “seventh” < bdun-pa (see Jacques 2004b). Sometimes the lexical items have distinct etymons but they are all cognates with words attested in Classical Literary Tibetan, whose meanings are equivalent or similar. For example, the verb ‘fear’ in the various Tibetic languages is derived from at least 5 different etyma: ‘fear’ : ɕeʔ (Ü) < ZHED ་ཞེད་ ʈheʔ (Ts) < BRED བྲེད་, ʈāʔ, sāʔ (Kh), ʈāk (Ho), ɕʨax (Am) < SKRAG སྐྲག་; dʑiʔ (Tö), ʑiwa laŋ (Sh); dʑiks (La) < ‘jigs, ʑiks (Ba) < ‘JIGS འཇིགས་; ɖo: (Dz) < ‘DROG འདྲོག་

Tournadre (2005) provides a list of one hundred lexical items found across the Tibetic languages. The Comparative Dictionary of Tibetan Dialects (Bielmeier et al. in preparation) contains hundreds of roots found across the various Tibetic languages.17

3.3 Morphosyntax All the Tibetic languages share a number of fundamental morphosyntactic characteristics, some of which sharply differ from the neighbouring closely related languages. The “pronominalized languages” include many Tibeto-Burman languages belonging to various subgroups such as Qiangic, Rgyalrongic and Bodic (Kiranti, Kanauri, Dolakha Newar), but no modern Tibetic language has preserved a system of verb agreement.18 The ancient verb tense-aspect morphology inherited from 16 While the cardinal numeral ‘seven’ has remained a Rgyalrongic cognate. 17 See also “Etymological notes about widespread Tibetic roots” in Tournadre and Suzuki (in preparation). 18 The question of the existence of agreement markers in Proto-Sino-Tibetan is a very controversial issue. See e.g. LaPolla (1992, 1994) and Jacques who remarks: “Some scholars such as Bauman (1975), van Driem (1993), or DeLancey (2010) argue that such a system must be reconstructed for Proto-Sino-Tibetan, while others such as LaPolla (2003) propose that the agreement systems found in various Sino-Tibetan languages are independent innovations” (2012: 83). There is a general consensus that agreement systems are absent in Literary Tibetan and in the Tibetic

112   

   Nicolas Tournadre

Tibeto-Burman has been replaced in the Tibetic languages by a system of auxiliary verbs used with nominalized forms of the verb. Most (if not all) the Tibetic languages have developed a rich system of evidential and epistemic markers, which appear as verb suffixes. Classical Tibetan has a system of 10 nominal cases (see Tournadre 2010). Most of the modern Tibetic languages have preserved to some extent a nominal case system inherited from Old Tibetan, although most of the modern case systems are quite reduced and may use only four cases (ergative, absolutive, genitive and dative). Classifiers are not found in Literary Tibetan and none of the modern languages have developed a system of classifiers, although a few rare classifiers do exist in a marginal way. The grammatical morphemes that are cognate in the various Tibetic languages include the following: a) the negative ma- or m(y)i-19, b) the nominalizing suffix -pa, c) the auxiliary verbs yin and yod, and d) the genitive and ergative cases -gi and -gis.

4 Proto-Tibetic It is necessary to distinguish clearly between Pre-Tibetic and Proto-Tibetic20 (PT). The former is the ancestor not only of Tibetic languages but also of languages belonging to a higher grouping that may correspond partly to the Bodish branch (but possibly also higher groupings of Tibeto-Burman). The latter refers to the direct ancestor of the Tibetic languages. Many authors have used Proto-Tibetan to refer to forms that are indeed Pre-Tibetic.

languages. Jacques (2010) has mentioned what he believes is one single trace of agreement in Literary Tibetan, but the arguments which are based on the slim evidence of a single verb form zos ‘eat (past)’ are not convincing. Even if it were a trace of verb agreement, it already was reinterpreted in Old Tibetan as a tense phenomenon and did not function as a person agreement. 19 For example, Bake (brag-gsum skad) has a negation in a- which is evidence that it can not be a Tibetic language. As we have seen above, the negation is always derived from ma- or myiin the Tibetic languages. There is much other phonological evidence such as the lack of palatalization of the dental *ti or the fricative *si or lexical evidences such as the words for ‘meat’ and ‘red’ which are not derived from the pandialectal roots respectively sha and dmar. Despite these anomalies, Bake was however classified as a “Tibetan dialect” by some authors such as Qu ­Aitang (1996). 20 In some works, the term “Proto-Tibetan” is used. It corresponds to the term “Proto-Tibetic” used in the present paper.



The Tibetic languages and their classification   

   113

Proto-Tibetic reconstructed forms are often similar or identical to the orthography of Classical Literary Tibetan. The reality is of course somewhat more complex. As Sprigg points out “Shafer would have said that we already know what Proto-Tibetan looks like: it is embalmed in the orthographic forms of Written Tibetan,” but he adds that “none of the dictionaries gives a reliable picture of the phonological structure of Written Tibetan during a given Etat de langue” (1972: 556). Furthermore, as Hill (2011) has remarked, the phonology of Old Tibetan is not well researched. The main phonological features, which characterize Proto-Tibetic are: 1. The preservation of the prefixes inherited from Proto-Tibeto-Burman. 2. The palatalization of dental and alveolar before y. 3. The change from lateral to dental after m. 4. The emergence of distinctive aspirated initial plosives.

4.1 The preservation of prefixes The numerous morphological prefixes of Tibeto-Burman, are still clearly pronounced, most probably with an epenthetic vowel in Proto-Tibetic, (see Matisoff, 2003: 97). The main prefixes found in Proto-Tibetic are: *s(ǝ),*d(ǝ)/g(ǝ),*m(ǝ) and *b(ǝ). The numbers gives a good illustration of this phenomenon: *g(ǝ)-tɕik ‘one,’ *g(ǝ)-nyis ‘two,’ *g(ǝ)-su- ‘three,’ *b(ǝ)-ʑi ‘four,’ *l(ǝ)-ŋa ‘five,’ *d(ǝ)-ruk ‘six,’ *b(ǝ)-dun ‘seven,’ *b(ǝ)-rgyat ‘eight,’ *d(ǝ)-gu ‘nine,’ *b(ǝ)-tɕu ‘ten.’ The prefix ‘s’ used for animals and body parts: *s(ǝ)-dik-pa ‘scorpion,’ *s(ǝ)-bal ‘frog,’ *s(ǝ)-tak ‘tiger,’ *s(ǝ)-b-rul ‘snake,’21 *s(ǝ)-pra ‘monkey,’ *s(ǝ)-kra ‘hair,’ *s(ǝ)-nyiŋ ‘heart,’ s(ǝ)-na ‘nose.’ Other prefixes such as *d(ǝ)-/*g(ǝ)-, *m(ǝ)-/*r(ǝ)- are also used for body parts as well as some animals: *d(ǝ)-myik ‘eye,’ *m(ǝ)-go ‘head,’ *r(ǝ)na ‘ear.’ For a discussion about the status of the morphological prefix /d/, see Jacques (2008) and Hill (2011).

4.2 Palatalization of dentals and alveolars before y Palatalization is one of the main features of Proto-Tibetic. The combinations *ty *ly, *sy were not palatalized in Pre-Tibetic22 but all these combinations have undergone in Proto-Tibetic a palatalization, which is recorded in the orthography 21 Sagart and Jacques propose the reconstruction *s-m-rul for some ealier stage (p.c. also compare Hill 2011: 448). 22 See Jacques (2004a) and Gong (1977) for these reconstructions.

114   

   Nicolas Tournadre

of Literary Tibetan. About the phonemic status of the palatals, see Hill (2011). All the modern languages and dialects have now developed reflexes of these palatalized forms. *ty > tɕ *g(ǝ)-tyik ‘one’ > PT : *g(ǝ)- tɕ(h)ik > OT : gcig/gchig གཅིག་ /གཆིག *tye ‘big’ > PT : * tɕ(h)e > OT: che ཆེ་ *b(ǝ)-tyu ‘ten’ > PT : b(ǝ)-tɕu > OT: bcu / bchu བཅུ་/ བཆུ་ *tyi ‘what’ > PT : * tɕ(h)i > OT: ci / chi ཅི་ / ཆི་

In a lot of Bodic and even Bodish languages closely related to Tibetan, we do not find palatalized forms of * t+y. See for example Bake (Basum lake) /ti/ “what,” /tɨʔ/ ‘one’ which reflects a stage closer to Pre-Tibetic. *sy > ɕ *sya ‘flesh’ > PT : ɕa > CLT : sha ཤ་ *syes ‘know’ > PT : ɕes > CLT : shes ཤེས་ *sying ‘wood’ > PT : ɕiŋ > CLT : shing ཤིང་

A lot of Bodish languages such as Tamang, Kurtöp have not undergone this change. tsy > tɕ *b(ǝ)-tsyat ’to cut’ (past stem) > PT : *b(ǝ)-tɕat > CLT : bcad བཅད་ *m(ǝ)-tsyil-ma ‘spittle’ > PT : *m(ǝ)-tɕ(h)il-ma > CLT : mchil-ma མཆིལ་མ་ *m(ǝ)- tsin-pa ‘liver’ > PT : *m(ǝ)-tɕ(h)in-pa, CLT : mchin-pa མཆིན་པ

Among the important innovations of Proto-Tibetic is the palatalization of the lateral /l/ in front of y (see Jacques 2004a). This sound law has been dubbed ‘Benedict’s law’ by Hill (2011: 445). For example: *b(ǝ)-lyi ‘four’ > PT : * b(ǝ)ʑi > CLT : bzhi བཞི་ *lying ‘field’ > PT : *ʑiŋ > CLT zhing ཞིང་ *ldi “flea” > PT : *ldʑi > CLT: lji ལྗི་ ‘ji ་འཇི་23 *s(ǝ)-lak(s) “iron” > *l-sak(s) > *l-tsyak(s) > PT : *ltɕaks > CLT : lcags ལྕགས་

23 Both ‘ji-ba and lji-ba are found in Classical Literary Tibetan. The Tshig-mdzod chen-mo [Great Tibetan dictionary] gives two slightly different meanings for ‘ji-ba and lji-ba, but they have certainly a common etymology. See also Beyer (1992: 78) and Zhang Jichuan (2009).



The Tibetic languages and their classification   

   115

This last mentioned is found in all the Tibetic languages but not in the neighbouring Bodish languages such as Kurtöp /Hla:/ and Bumthap /lak/ (Michailovsky and Mazaudon 1994). The lateral of the sequence *bli is also preserved in many other Bodish or even Sino-Tibeto-Burman languages, such as Kurtöp, Tshona (mtshosna), and Old Chinese (see Jacques 2004a).

4.3 The change from lateral to dental after m The change from *ml to md occurred in proto-Tibetan and its reflex is found in all the modern Tibetic languages. Thus PTB *b/m-la ‘arrow’24 > PT : mda > CLT : mda’ མདའ་ Some Bodish languages closely related to Tibetic did not undergo this mutation (see Michailovsky and Mazaudon 1994). Hill calls this sound change ‘Bodman’s law’ (2011: 450 note 12).

4.4 The assimilation of b before nasal n The change from *bn to *mn is also a characteristic feature of proto-Tibetan. *bnans > PT : mnans > OT : mnand ‘to suppress’ *bnyan > PT : nyan > mnyand ‘to listen’ This sound law has been discovered by Chang (1971) and Hill (2011: 445–446) has proposed to call it Chang’s law.

4.5 Emergence of distinctive aspirated initial plosives. Another characteristic of proto-Tibetan is the emergence of distinctive aspirations for plosives as was shown by Li Fang Kuei (1933) and Jacques (2004b). According to Shafer (1950–1951: 772–773) aspiration in Tibetan [Proto-Tibetic] was originally non-distinctive. Hill calls the sub-phonemic status of aspiration “Shafer’s rule” (2011: 441–442). In Old Tibetan the emergence of aspirated sounds is attested, especially for initial plosives, but it is not yet organized as a system. This was already noticed by Shafer (1950–1951). This can easily be proven by the fluctua-

24 According to Matisoff’s reconstruction (2003: 50 et passim).

116   

   Nicolas Tournadre

tion found the orthography of Old Tibetan between aspirated and non aspirated consonants. Ex. : gcig/gchig གཅིག་ /གཆིག་ ‘one;’

phyin-chad/ phyin-cad ཕྱིན་ཆད་ / ཕྱིན་ཅད་ ‘from now on’

ci / chi ཅི་ / ཆི་ ‘what,’ cu/chu ཅུ་ /ཆུ་ ‘water,’ etc.

4.6 C  ases of non-coincidence between Proto-Tibetic and Classical Tibetan For some words, the classical orthography does not correspond to Proto-Tibetic. For example for the word mig ‘eye,’ many dialects of Amdo have a form such as /ɣnjǝx/ or /mnjǝx/. Some southern Khams dialects also exhibit an archaic reflex of a sound preceding the labial /m/ (Tournadre and Suzuki, in preparation). The Proto-Tibetic form should thus be reconstructed *d(ǝ)myik. Fortunately, an archaic orthography dmyig དམྱིག་ is attested in some old documents.25 On the basis of many dialects (Tö, Balti, Ladakhi, Sherpa, Gyalsumdo, Lhoke, etc.), we should reconstruct for ‘flower’ PT *mentok, while Classical Tibetan has me-tog. Fortunately, the form men-tog མེན་ ཏོག་ is also attested in Old Literary Tibetan (see Hill 2007: 480 note 8). Reconstructable Proto-Tibetic forms are not always attested, however. For the word sbom-po ‘big (for rope),’ the reconstruction based on some dialects of Tö and Amdo (Ngaba) which have rompo should be PT *s(ǝ)brompo. This form is not attested in Literary Tibetan. In some cases, as shown by Tournadre and Suzuki (in preparation), one could be tempted to reconstruct the proto-Tibetan form, however a phonological innovative rule may provide a better account for the phenomenon. For example, the word khang-pa ‘house’ is sometimes prenasalised in some dialects of Amdo, Khöpokhok, Minyag Khams and Baima (Zhang 1997) and thus we could have proposed to reconstruct PT *nkhangpa. But in this case, the phonological environment may provide a better explanation.26 A few words with this type of prenasalisation (e.g. tshang ‘nest,’ phreng ‘beads,’ tshang-ma ‘all’) attested in these dialects originally have an aspirated obstruent initial with -ng final, which was omitted and caused a prenasalisation instead (except Amdo Machu, mGolog). Thus, we have to be careful with the reconstruction of Proto25 This spelling does not appear in Classical Tibetan texts. 26 The phonological hypothesis was proposed by Suzuki (p.c.).



The Tibetic languages and their classification   

   117

Tibetic forms when they are based on a single region of the Tibetic area and never rule out the possibility of local areal innovations. Bielmeier et al. (in preparation) raises similar issues about the reconstruction of the common ancestor: In a number of cases the comparative evidence of the dialects does not lead back directly to the Written Tibetan etymological equivalent. Either the evidence leads to a form previous to the Written Tibetan etymological equivalent, making the Written Tibetan equivalent thus not the “ancestor” but simply an “older relative,” or else we have to accept that certain morphonological or grammatical changes took place within individual dialects. To give an example, we have gyang ‘wall’ in Written Tibetan with regular etymological correspondences in all dialect groups of Tibetan, but in Balti we find rgyang ‘wall.’ In such a case we have either to start from a Common Tibetan [PT] *rgyang of which Written Tibetan gyang is a later offshoot, comparable to Purik gyang, or else we may assume that there was an internal Balti development from Common Tibetan [PT] *gyang to Balti rgyang by prefixing r-. We would then have to explain the reason for this prefixing.27

5 T  he geographic distribution and status of Tibetic languages Tibetic languages are spoken above all in five countries: China, Pakistan, India, Nepal and Bhutan. Additionally, Sangdam, a Khams dialect is spoken in the Kachin state of Myanmar. For all the Tibetic languages taken together, the total number of speakers is roughly six million. However, this figure is approximate since there is no precise and reliable census. Here is a list of the Tibetic languages and where they are spoken: China: Ü-Tsang (དབུས་གཙང་སྐད་) , Khams (ཁམས་སྐད་), Hor ( ཧོར་སྐད་), Amdo (ཨ་མདོ འི་

སྐད་དམ་ཨམ་སྐད་), Kyirong (སྐིྱ ད་རོང་སྐད་), Zhongu (ཞོ་ངུ་སྐད་), Khalong (ཁ་ལོང་སྐད), gSerpa (གསེར་པ་སྐད་), Khöpokhok (ཁོད་པོ་ཁོག་), Palkyi [Pashi] (དཔལ་སྐིད) / Chos-rje ྱ (ཆོས་ᩞེ་),28 Sharkhok (ཤར་ཁོག་), Thewo (ཐེ་བོ འ་ི སྐད་), Chone (ཅོ་ནེའ་ི སྐད), Drugchu (འབྲུག་ཆུའི་སྐད་), Baima (བོད་དམག་སྐད་).

27 Website of the Comparative Dictionary of Tibetan Dialects (CTDT), now accessible at: http:// www.himalayanlanguages.org/cdtd 28 Sun (2003b) uses Chos-rje but according to Suzuki (p.c.) Dpal-skyid [Pashi] is better suited to refer to a group of four dialects which include Chos-rje.

118   

   Nicolas Tournadre

Pakistan: Balti (བལ་ཏིའི་སྐད་)29 (northern Pakistan).

India: Purik (པུ་རིག་སྐད་), Ladakhi (ལ་᩺གས་སྐད་), Zangskari (ཟངས་དཀར་སྐད་), Spiti (སྤི་ཏིའི་སྐད་), Lahuli or Gharsha (གར་ཞྭའི་སྐད་), Khunu (ᨯ་ནུའ་ི སྐད་), Jad or Dzad (འཇད་སྐད་), Drengjong (འབྲས་ལྗོངས་སྐད་) often locally called Lhoke (ལྷོ་སྐད་).

Nepal: Humla (ཧུམ་ལའི་སྐད་), Mugu (མུ་གུའི་སྐད་), Dolpo (དོལ་པོ འི་སྐད་), Lo-ke or

Mustang (གོ་ླ སྐད་དམ་གོ་ླ སྨོན་ཐང་གི་སྐད་), Nubri (ནུབ་རིའི་སྐད་), Tsum (ཙུམ་སྐད་), Langtang

(ལང་ཐང་སྐད་), Yolmo (ཡོལ་མོ འིསྐད་), Gyalsumdo (ᩂལ་གསུམ་མདོ་), Jirel (ཇི་རེལ་སྐད་),

Sherpa (ཤར་བའི་སྐད་) also locally called Sharwi Tamnye (ཤར་བའི་གཏམ་སྙད), Kagate also called Shupa (ཤོག་པའི་སྐད་), Lhomi (ལྷོ་མིའི་སྐད་ ), Walung (ཝ་ལུང་ཆུང་སྒོ་ལའི་སྐད་

དམ་ཝ་ལུང་སྐད) and Tokpe Gola (གྲོག་པོ་སྒོ་ལའི་སྐད་དམ་གྲོག་པའི་སྒོ་ལ་).

Bhutan: Dzongkha (ོ᫊ང་ཁ་), Tsamang (᫂་མང་ཁ་) or Chocha-ngacha (ཁྱོད་ཅ་ང་ཅ་ཁ་),

Lakha (ལ་ཁ་) also called Tshangkha (ཚང་ཁ་), Dur Brokkat (དུར་གི་འབྲོག་སྐད་) also



called Bjokha in Dzongkha, Mera Sakteng Brokpa-ke (མེ་རག་སག་སྟེང་སྐད་).

This list contains nearly fifty Tibetic languages, all of which are derived from Old Tibetan.30 However the total number of dialects and varieties certainly comes up to more than two hundred. This incredible diversity, which is largely due to the extension of the Tibetic linguistic area and to the geographic isolation of many localities, has been underestimated until quite recently.31 As mentioned earlier, some of these languages are tonal, while others are non-tonal. Although they do share a common basic vocabulary and grammar, they may largely differ in some aspects of phonology, morphology, syntax and lexicon. In general, the languages listed above do not allow mutual intelligibility. However some of them do allow a certain degree of intelligibility for example

29 Balti is traditionally written sbal-ti in Tibetan but Balti people write it and pronounce it bal-ti. 30 As mentioned earlier, a couple of languages in the list might originally have a distinct substratum. 31 This is partly due to the fact that many authors have continued to use the expression “Tibetan dialects” instead of “Tibetan languages” (although Zeisler 2004 uses this term) or “Tibetic languages.”



The Tibetic languages and their classification   

   119

Yolmo and Kagate (in Nepal) or Balti and Purik (on both sides of the India-Pakistan border).32 Some taxons in the list correspond to large groups of quite diversified dialects. This is the case for “Amdo,” “Ü-Tsang,” “Khams” or even “Ladakhi,” which include dozens of dialects (see e.g. Zeisler 2011; Tournadre and Konchok Jiatso 2001; Tournadre 2005, 2008). On the other hand, a number of taxons mentioned in the list above refer to languages with little or no dialectal diversity, e.g. Jirel or Sherpa.33 Most of the taxons in the list are related to traditional geographic and cultural names. They do not always refer to precise linguistic labels. In the next section, I will propose a classification of the Tibetic family, which is based on the notion of geolinguistic continua and avoids some of the problematic traditional taxons. With a few notable exceptions, spoken Tibetic languages have not been standardized and have not been subject to any significant language planning. Thus, for example, Khams “language,” which refers to a very large group of dialects, does not have a standardized form even if some dialects such as Derge (sde-dge) are often considered as prestigious. A few languages have undergone standardization to some degree. This is the case of Lhasa Tibetan, a variety of Central Tibet (Ü), which is often referred to as spyi-skad (སྤིྱ ་སྐད་) ‘the Common Language.’ It functions as the regional koiné and is also called “Standard (Spoken) Tibetan” (Tournadre and Dorje 1998, 2003). This “common language” is used in Central Tibet (and to a lesser extent in Amdo and Khams, mainly by the educated elite) for spoken communication between speakers of different Tibetic languages or dialects. More importantly, it is used by Tibetan emigrants in the diaspora in India, Nepal or elsewhere in the world.34 Some other Tibetic languages such as Dzongkha, Sikkimese Lhoke or Ladakhi have also achieved a reasonable degree of standardization. If we take in account the number of speakers, prestige and political status, we can identify the following “major” spoken languages: Ü-Tsang, Khams, Amdo

32 The situation is again comparable to the Romance family. While some languages such as Romanian, Spanish and French do not allow a basic conversation. Galician, Asturian and Portuguese allow some mutual intelligibility. The same is true for Piedmontese, Ligurian and Italian. 33 Sherpa for example has only five closely related dialects: Solu, Khumbu, Pharak, Dram, Sikkimese Sherpa. Jirel which is spoken by a small community of less than 5000 people in the area of a single village (Jiri) probably has no significant dialectal diversity. 34 The Common language or Standard Tibetan (spyi skad) should definitely not be called Lhasa Tibetan when it applies to the diaspora. There are some minor lexical and grammatical discrepancies between the diaspora common language and the Lhasa dialect.

120   

   Nicolas Tournadre

(China); Dzongkha (Bhutan); Balti (Pakistan); Ladakhi, Sikkimese Lhoke (India); and Sherpa (Nepal).35 The main written language of the area is Classical Literary Tibetan (CLT), which has been used as the written language of the Tibetic area for more than a thousand years.36 Classical Literary Tibetan is the liturgical language of Bön and Vajrayāna Buddhism, which partly explains why it has played a fundamental role in the area. During the last decade, Classical Literary Tibetan has adapted to the modern information technologies (Unicode, the internet, mobile phones, email etc.) and has gained a real visibility and weight on the internet.37 Nowadays, in some areas, especially in the southern Himalayas, Classical Literary Tibetan is essentially used in the monasteries for religious purpose and is often called for this reason chos-skad (ཆོས་སྐད་) ‘the language of dharma or religion.’38 In Baltistan, literary Tibetan gradually fell into disuse following the conversion of its population to Shia Islam during the 15th century, and was replaced by Persian and later Urdu. In ethnic Tibet (Tibetan Autonomous Region, Qinghai, Sichuan, Gansu, Yunnan), Classical Literary Tibetan in its modernized form is used by both laymen and monks for all written purposes, along with Written Chinese whose influence has been growing rapidly during the last two decades. Thus, the three main spoken languages Ü-Tsang, Amdo and Khams, are normally not used as written languages.39 This situation contrasts sharply with the Tibetic areas outside China, where Ladakhi, Balti, Lhoke, Dzongkha and Sherpa have developed or are developing a written form based on the spoken language. These newly written languages coexist with Classical Literary Tibetan40 (which is still used in the Buddhist and Bönpo monasteries) and with the written national languages such as Hindi, Urdu or Nepali, depending on the area.

35 These eight languages account for the great majority of the speakers. The other Tibetic languages have generally a small number of speakers (from a few hundred to a few thousand thousand.) 36 Of course the literary language has undergone some evolution but it has preserved a very conservative orthography and grammar. 37 For example on Wikipedia, Literary Tibetan has the ranking 135 among the 285 languages in terms of the number of entries as of March 2012. See URL: http://bo.wikipedia.org 38 There are also political reasons not to call it Classical Tibetan. 39 The use of written forms of Amdo, Khams or Ü-Tsang instead of Classical Literary Tibetan is not encouraged by many Tibetans as it would undermine their policital unity. 40 Except in Baltistan (Pakistan) and in the Kargil area (India), mostly inhabited by Muslims.



The Tibetic languages and their classification   

   121

6 The inner classification of the Tibetic family As mentioned earlier, the Tibetic area shows a remarkable dialectal diversity, which can be partly explained by the mountainous terrain and the difficulty of transport across the Plateau and the Himalayas, but this diversity may also be partly explained by language contact with non-Tibetic languages (particularly Bodish, Qiangic and Rgyalrongic languages). Some areas, such as southern Gansu, northern Sichuan, northwestern Yunnan (China) or Ladakh (India) exhibit a very complex dialectal cartography with complex nets of isoglosses. The cladistic approach of “the family tree” can not be applied easily to the Tibetic languages.41 Sun rightly criticizes the traditional methodology used for the classification: In the meantime, the lamentable tendency persists to pigeonhole minor Tibetan dialects into the … major dialects on the shaky basis of shared archaisms (consonant clusters, voiced obstruent initials, non diphtongs, etc.) or global similarities owing to convergent development (syllable cannon reduction, vowel nasalization, tonogenesis, etc.). … All distinct … forms of Tibetan should a priori be placed directly under Old Tibetan as its first order offshoots, unless there are sound reasons for making the flat family tree hierarchical. (Sun 2003a: 796–797)

Among the most influential classifications42 of the entire family, Nishi (1986) should be mentioned. He distinguishes six major groups: Central (or Ü-Tsang), Western Innovative, Western Archaic, Southern, Khams and Amdo. This classification is essentially similar to the one used by Bielmeier in his Comparative Dictionary of Tibetan Dialects (in preparation). The main difference is that the Khams group is divided into Northern Khams and southern Khams. The classification proposed here is essentially based on a genetic approach, but it also includes geographical parameters, migration and language contact factors (Chirkova in press; LaPolla in press). The languages listed in Section 5 can be grouped together at a higher level into 8 major sections. Each section constitutes a geolinguistic continuum. These are: North-Western section (NW); Western section (W); Central section (C); South-Western section (SW); Southern section (S); South-Eastern section (SE); Eastern section (E); North-Eastern section (NE). 41 The phonological isoglosses are deeply intertwingled. They do not match with lexical or morphological isoglosses. It is also impossible to find common innovations that would support a cladistic approach. For example, there is no phonological innovation which is common to the Khams-Hor group or to the Khams sub-group. The traditional Stammbaum is also inappropriate in other languages of the Sino-Tibetan macrofamily (see LaPolla 2001). 42 See also Denwood (1999).

122   

   Nicolas Tournadre

1) North-Western section Ladakhi, Zanskari, Balti, Purki (see Zeisler’s subgroups Shamskat and Kenhat in Zeisler 2011) 2) Western section Spiti, Garzha, Khunu, Jad 3) Central section Ü, Tsang, Phenpo, Lhokha, Tö, Kongpo 4) South-Western section Sherpa and Jirel and other languages/dialects along the Sino-Nepalese border such as Humla, Mugu, Dolpo, Lo-ke, Nubri, Tsum, Langtang, Kyirong,43 Yolmo, Gyalsumdo, Kagate, Lhomi, Walung and Tokpe Gola. 5) Southern section Dzongkha, Drengjong, Tsamang, Dhromo Lakha, Dur Brokkat, Mera Sakteng Brokpa-ke 6) South-Eastern section Hor Nagchu, Hor Bachen, Yushu, Pembar, ‘Northern route,’44 Rongdrak, Minyak, ‘Southern route,’45 Dzayul, Derong-Jol, Chaktreng,46 Muli-Dappa, Semkyi Nyida (for the dialects of Khams, see Suzuki 2009a) 7) Eastern section Drugchu, Khöpokhok, Thewo, Chone, Baima, Sharkhok, Palkyi [Pashi] and Zhongu (see Suzuki 2009a) 8) North Eastern section Amdo, Gserpa, Khalong (see Jackson, 2006, 2007) For some of the above sections, mutual intelligibility is good between adjacent dialects, but it becomes problematic between dialects located at the extreme ends 43 Kyirong is located in China but from both a geographic and a linguistic point of view, Kyirong is related to the South-Western dialects spoken on the Nepalese side and influenced by languages from Nepal (as shown by Huber 2005). 44 The north route includes ‘Chamdo (Chab-mdo), Derge (sde-dge) and Kandze (dkar-mdzes). 45 The southern route includes Markham (smar-khams), Bathang (’ba’-thang), Lithang (li-thang) 46 The literary spelling of Chaktreng is phyag-phreng.



The Tibetic languages and their classification   

   123

of one section. If we take the Central section, Ü and Tsang, Tsang and Tö, Kongpo and Ü generally allow fairly good intelligibility but for Kongpo and Tö dialects, mutual intelligibility is probably limited. Mutual intelligibility is lower in the South-Eastern and Eastern sections, and is very limited between some southern and northern Khams dialects. The same could be said of the North Western section, which includes Ladakhi, Balti, Purik and Zanskari. The eight sections can thus be viewed as complex dialect (quasi-)continua. A lot of authors traditionally make a distinction in their classification between ‘brog-skad ‘pastoralists’ dialects’ and rong-skad ‘cultivators dialects’ (e.g. Gesang Jumian [1964] 2002). This sociolinguistic criterion overlaps with a geographic criterion. The distinction between pastoralists and cultivators is relevant in most of the eight sections, particularly in the North, North-Eastern, Central and NorthWestern. The idea that all the pastoralists (at least in Tibet) can understand each others is frequently heard but it is a myth.47 It is generally true that the pastoralists’ dialects are more conservative than the cultivators’ ones, but they too belong to various sections and do not allow mutual intelligibility. Hor and Khams pastoralists (southern Eastern section) can not communicate easily with pastoralists from Tö (Central section) or Amdo (northern section).

7 Non-coincidence between ethnic groups and linguistic groups of the Tibetic area As mentioned in section (2), the Tibetic languages are spoken not only by Tibetans per se but also by other ethnic groups in India, Nepal and Bhutan who do not consider themselves as Tibetans. As in many other regions of the world, there is no strict coincidence between the ethnic groups or nationalities and the language they speak. The great majority of ethnic Tibetans called bod-rigs བོད་རིགས་ in Tibetan and Zangzu 藏族 in Chinese speak various Tibetic languages (or so-called “Tibetan dialects”). However a minority of ethnic Tibetans do not speak a Tibetic language as their mother tongue, but one or another Tibeto-Burman language

47 See for example the following entry on the Tibetan Wikipedia (http://bo.wikipedia.org/wiki/

བོད་སྐད་): བོད་སྐད་ནི་བོད་ཡུལ་དུ་ བེད་སྤོད ྱ ་བྱེད་པའི་སྐད་དེ། སྤིྱ ར་རོང་འབྲོག་གཉིས་ལས་འབྲོག་པའི་སྐད་ནི་འབྲོག་སྡེ་སྟོད་ ུ སྨད་བར་གསམ་ཀུན་ཏུ ་ཁད ྱ ་པར་མེད་།  “Tibetan is spoken in Tibet. Generally, among the two varieties of pastoralists and cultivators, the dialects of pastoralists spoken in upper, middle and lower Tibet are indistinct. […]”

124   

   Nicolas Tournadre

belonging to the Rgyalrongic, Qiangic or Bodish groups48 (see Poa and LaPolla 2007). The main non-Tibetic languages spoken as mother tongues by ethnic Tibetans are found in the Rgyalrong area in China: Situ, Zbu, Tshobdun, Japhug, Lavrung, Stau, Geshitsa, Stodsde and Nyagrong Minyag. They are all spoken in Sichuan, in Ngaba Prefecture and in a few counties of Kandze Prefecture (Tau, Rongdrak, Dhrango and Nyagrong). Another series of languages spoken by Ethnic Tibetans in Ngaba and Kandze Prefectures (Sichuan) include the following Qiangic languages: Qiang, Prinmi,49 Queyu, Zhaba, Guiqiong, Shixing, Namuyi, Ersu, Minyak and Lüzu. Finally, a Bodish language called Basum or Ba-ke is spoken by Tibetans in the Kongpo area of the Tibetan Autonymous Region. Most of these languages have been heavily influenced by the surrounding Tibetic languages and dialects, as well as by Literary Tibetan, which has long been used as a written language.

8 Language contact Tibetic languages are in contact with many other languages, including members of the following families: Tibeto-Burman (mainly Bodic, Qiangic, Rgyalrongic, Sinitic), Mongolic, Turkic, Indo-European (Indo-Aryan, Iranian) and Burushaski.50 As a result of these contacts, there have been various influences on the vocabulary, phonology, prosody and morphosyntax of various Tibetic languages. Conversely, Tibetic languages have also influenced various neighbouring languages, particularly the Tibetospheric languages.51 Since the end of the 20th century the impact of national languages has been growing in the region through the development of modern education and official media. Within China, Putonghua (i.e. Mandarin Chinese) is now used as a second language by an increasing number of Tibetans, especially in urban areas.52 For

48 However, the speakers of these languages can often speak Amdo or Kham as a second language. 49 Often referred to as Pumi which is the Chinese name of this language. 50 Concerning the influence of Tibetan on some of the neighbouring languages see Hill (2010: 112). About the influences of neighbouring languages on Tibetan see Laufer ([1916] 1987) and Denwood (1999). 51 I.e. the Rgyalrongic languages as well as many Bodish and Qiangic languages. Some of the Tibetic languages such as Khalong or Zhongu have also been influenced by the neighbouring Rgyalrongic or Qiangic languages, see Sun 2002, 2003a. 52 The Chinese dialect used by Tibetans may be a southwestern form of Mandarin, rather than Putonghua strictly speaking.



The Tibetic languages and their classification   

   125

this reason, Chinese is exerting a growing influence on some Tibetic languages and dialects, especially in the eastern regions. Even in Central Tibet, many people in their everyday speech mix Chinese vocabulary with Tibetan vocabulary and grammar. This has given raise to the so-called ramalugskad which translates as ‘half sheep-half goat language.’ Within the Tibetic areas in India and Pakistan, Hindi and Urdu have also become intrusive languages. They are dominant in the school system. Nepali has also gradually become a dominant language not only in Nepal but also in the Indian State of Sikkim. Since the end of the 20th century, another Indo-European language, English, has become widespread amongst the elite of the Tibeticspeaking communities in India, Pakistan, Nepal and Bhutan. It is also used as the education medium in many schools of the southern Himalayas.

9 Conclusions The Tibetic languages form a compact and well-defined language family. These languages allow for the reconstruction of Proto-Tibetic forms which are often very similar to Classical Literary Tibetan but not always identical. I have proposed an inner classification of the family based on the concept of geolinguistic continua and avoiding the traditional cladistic approach of family trees, which is not appropriate to describe the complexity and the diversity of this language family. The recognition of this family will help typologists, comparativists, Sino-Tibetanists, general linguists and anthropologists to have a better representation of the relationships between the various Tibetic languages, as well as between them and other languages of the Tibetosphere.

Abbreviations Am Ba CLT Dz Ho Kh La Sh Ts

Amdo Balti Classical Literary Tibetan Dzongkha Nagchu Hor Khams Ladakhi Sherpa Tsang

126   

   Nicolas Tournadre

References Bauman, James J. 1975. Pronouns and pronominal morphology in Tibeto-Burman. Berkeley: University of California dissertation. Beckwith, Christopher I. 2006. Old Tibetan and the dialects and periodization of Old Chinese. In Christopher I. Beckwith (ed.). Medieval Tibeto-Burman Languages II, 179–200. Leiden: Brill. Beyer, S. 1992. The Classical Tibetan language. New York: State University of New York. Reprint 1993. Bibliotheca Indo-Buddhica series, 116. Delhi: Sri Satguru. Bielmeier, Roland et al. in preparation. Comparative Dictionary of Tibetan Dialects. Vol.I: Verbs, Vol.II: Nouns. (Preprint no.1 1997). Blench, Roger. 2011. How nominal affixes in Austroasiatic and Sino-Tibetan can help us understand Niger-Congo noun classes. Paper given at the 41st CALL conference, Leiden 29–31 August 2011. Chang, Betty Shefts. 1971. The Tibetan causative phonology. Bulletin of the Institute of History and Philology, Academia Sinica 42: 623–763. Chirkova, Katia. 2005. Báimǎ nominal postpositions and their etymology. Linguistics of the Tibeto-Burman Area 28 (2): 1–41. Chirkova, Katia. 2008a. On the Position of Báimǎ within Tibetan. In Alexander Lubotsky, Jos Schaeken, and Jeroen Wiedenhof (eds.). Evidence and Counter-Evidence. Essays in Honour of Frederik Kortlandt, vol. 2, 69–91. Amsterdam: Rodopi. Chirkova, Katia. 2008b. Báimǎyǔ shìzhèng fànchóu jíqí yǔ Zàngyǔ fāngyán de bǐjiào [Evidentials in Baima and Tibetan dialects compared] Mínzú yǔwén 3: 36–43. Chirkova, Katia. in press. On principles and practices of language classification. In Cao Guangshun, Redouane Djamouri, Hilary Chappell, and Thekla Wiebush (eds.). Breaking Down the Barriers: Interdisciplinary Studies in Chinese Linguistics and Beyond. Taipei: Institute of Linguistics, Academia Sinica. DeLancey, Scott. 2010. Towards a history of verb agreement in Tibeto-Burman. Himalayan Linguistics 9 (1): 1–39. DeLancey, Scott. 2012. Still mirative after all these years. Linguistic Typology 16 (3): 529–564. Denwood, Philip. 1999. Tibetan. Amsterdam: John Benjamins Publishing. van Driem, George. 1993. The Proto-Tibeto-Burman verbal agreement system. Bulletin of the School of Oriental and African Studies 56 (2): 292–334. van Driem, George. 2011. Tibeto-Burman subgroups and historical grammar. Himalayan Linguistics 10 (1): 31–39. van Driem, George. 2012. Etyma, shouldered adzes and molecular variants. In Andrea Ender, Adrian Leemann and Bernhard Wälchli (eds.). Methods in Contemporary Linguistics, 335–361. Berlin: Mouton de Gruyter. Gesang Jumian and Gesang Yangjin. 2002. Zàngyǔ fāngyán gàilùn. Beijing: Mínzú chūbǎnshè. Gesang Jumian. 1964. Zàngyǔ fāngyán gàiyào [Concise presentation of The Tibetan dialects]. unpublished manuscript. de Haan, Ferdinand. 2011. Semantic Distinctions of Evidentiality. In Matthew S. Dryer and Martin Haspelmath (eds.). The World Atlas of Language Structures Online. Munich: Max Planck Digital Library, chapter 77. Available online at http://wals.info/chapter/77 Accessed on 2013–03–17. Haller, Felix. 2000. Dialekt und Erzählungen von Shigatse. (Beiträge zur tibetischen Erzählforschung 13.) Bonn: VGH Wissenschaftsverlag.



The Tibetic languages and their classification   

   127

Haller, Felix. 2004. Dialekt und Erzählungen von Themchen: sprachwissenschaftliche Beschreibung eines Nomadendialektes aus Nord-Amdo. (Beiträge zur tibetischen Erzählforschung 14.) Bonn: VGH Wissenschaftsverlag. Haller, Felix. 2009. Switch-reference in Tibetan. Linguistics of the Tibeto-Burman Area 32 (2): 45–106. Hill, Nathan W. 2007. Aspirated and Unaspirated Voiceless Consonants in Old Tibetan. Language and Linguistics 8 (2): 471–493. Hill, Nathan W. 2010. An overview of Old Tibetan synchronic phonology. Transactions of the Philological Society 108 (2): 110–125. Hill, Nathan W. 2011. An Inventory of Tibetan Sound Laws. Journal of the Royal Asiatic Society of Great Britain & Ireland (Third Series) 21 (4): 441–457. Huber, Brigitte. 2005. The Tibetan Dialect of Lende (Kyirong). (Beiträge zur tibetische Erzählforschung 15). Bonn: VGH Wissenschaftsverlag. Jacques, Guillaume. 2004a. The laterals in Tibetan. Paper given at Himalayan Languages Symposium, Thimphu, Bhutan, 1–3 December 2004. Jacques, Guillaume. 2004b. Phonologie et morphologie du Japhug. Paris: Université de Paris dissertation. Jacques, Guillaume. 2010. A possible trace of verbal agreement in Tibetan. Himalayan Linguistics 9 (1): 41–49. Jacques, Guillaume. 2012. Agreement morphology: the case of Rgyalrongic and Kiranti. Language and Linguistics 13: 83–116. LaPolla, Randy J. 1992. On the dating and nature of verb agreement in Tibeto-Burman. Bulletin of the School of Oriental and African Studies 55 (2): 298–315. LaPolla, Randy J. 1994. Parallel Grammaticalizations in Tibeto-Burman: Evidence of Sapir’s ‘Drift.’ Linguistics of the Tibeto-Burman Area 17 (1): 61–80. LaPolla, Randy J. 2001. The Role of Migration and Language Contact in the Development of the Sino-Tibetan Language Family. In R. M. W. Dixon and A. Y. Aikhenvald. Areal Diffusion and Genetic Inheritance: Case Studies in Language Change, 225–254. Oxford: Oxford University Press. LaPolla, Randy J. 2003. Overview of Sino-Tibetan Morphosyntax. In Graham Thurgood and Randy J. LaPolla (eds.). The Sino-Tibetan Languages, 22–42. London: Routledge. LaPolla, Randy J. in press. Subgrouping in Tibeto-Burman: Can an individual-identifying standard be developed? How do we factor in the history of migrations and language contact? In Balthasar Bickel, Lenore A. Grenoble, David A. Peterson, and Alan Timberlake (eds.). What’s Where Why? Language Typology and Historical Contingency. Amsterdam and Philadelphia: John Benjamins Pub Co. Laufer, Berthold [1916] 1987. Loan words in Tibetan. T’oung Pao 17: 404–552. Reprinted in Hartmut Walravens and Lokesh Chandra (eds). Sino-Tibetan Studies II, 483–632. New Delhi: Rakesh Goel. Lewis M. Paul. 2009. Ethnologue: Languages of the World. Sixteenth edition. Dallas: SIL International. http://www.ethnologue.com. Li, Fang-Kuei. 1933. Certain Phonetic Influences of the Tibetan Prefixes upon the Root Initials. Bulletin of the Institute of History and Philology, Academia Sinica 6 (2): 135–157. Matisoff, James A. 1989. A new Sino-Tibetan root *d-yu-k BELONG / TRUST / DEPEND / ACCEPT / TAKE, and a note of caution to megalo-reconstructionists. In David Bradley, Eugénie J.A. Henderson, and Martine Mazaudon (eds.). Prosodic Analysis and Asian Linguistics: to

128   

   Nicolas Tournadre

honour R. K. Sprigg, 265–269. (Pacific Linguistics C-104), Canberra: Australian National University. Matisoff, James A. 1990. On megalocomparison. Language 66 (1): 106–20. Matisoff, James A. 2003. Handbook of Proto-Tibeto-Burman: system and philosophy of Sino-Tibetan reconstruction. Berkeley, Los Angeles, and London: University of California Press. Merriam-Webster dictionary online (http://www.merriam-webster.com/dictionary/dialect) Michailovsky, Boyd. 2011. Les langues tibéto-birmanes. In Emilio Bonvini et al. (eds.). Dictionnaire des langues. Paris: Presses Universitaires de France. Michailovsky, Boyd and Martine Mazaudon. 1994. Preliminary Notes on the Languages of the Bumthang Group (Bhutan). In Per Kværne (ed). Tibetan Studies: proceedings of the 6th Seminar of the International Association for Tibetan Studies, vol. 2, 545–557. Oslo: The Institute of Comparative Research in Human Culture. Nishi, Yoshio. 1986. Gendai Chibettogo hōgen no bunrui [A classification of Tibetan dialects]. Bulletin of the National Musuem of Ethnology 11 (4): 837–900 Noonan, Michael. 2012. Aspects of the historical development of nominalizers in the Tamangic languages. In Foong Ha Yap, Karen Grunow-Hårsta and Janick Wrona (eds.). Nominalization in Asian languages: Diachronic and typological perspectives. Amsterdam and Philadelphia: John Benjamins. Poa, D. and Randy J. LaPolla. 2007. Minority Languages of China. In Osahito Miyaoka and Michael E. Krauss, The Vanishing Languages of the Pacific, 337–354. Oxford: Oxford University Press. Qu Aitang. 1996. Zàngzú de yǔyán hé wénzǐ. Beijing: Mínzú chūbǎnshè. Shafer, Robert. 1950–1951. Studies in the Morphology of the Bodic Verb. Bulletin of the School of Oriental and African Studies 13: 702–724, 1017–1031. Sprigg, Richard Keith. 1972. A Polysystemic Approach, in Proto-Tibetan Reconstruction, to Tone and Syllable-Initial Consonant Clusters. Bulletin of the School of Oriental and African Studies 35 (3): 546–587. Sun, Jackson T. S. 2002. Perfective stem renovation in Khalong Tibetan. Paper given at the 8th Himalayan Languages Symposium, Bern, 19–22 September 2002. Sun, Jackson T. S. 2003a. Phonological profile of Zhongu: A new Tibetan dialect of northern Sichuan. Language and Linguistics 4 (4): 769–836. Sun, Jackson T. S. 2003b. Qiújí zàngyǔde yǔyīn tèzhēng. Mínzú yǔwén (6): 1–6. Sun, Jackson T. S. 2006. Special Linguistic features of gSerpa Tibetan. Linguistics of the TibetoBurman Area 29: 107–125. Sun, Jackson T. S. 2007. Perfective stem renovation in Khalong Tibetan. In Roland Bielmeier and Felix Haller (eds.). Linguistics of the Himalayas and beyond. Berlin and New York: Mouton de Gruyter. Suzuki, Hiroyuki. 2009a. Introduction to the method of the Tibetan linguistic geography – a case study in the Ethnic Corridor of West Sichuan. In Yasuhiko Nagano (ed.). Linguistic Substratum in Tibet – New Perspective towards Historical Methodology (No. 16102001) Report, vol. 3: 15–34. Suzuki Hiroyuki. 2009b. Deux remarques à propos du développement du ra-btags en tibétain parlé. Revue d’étude tibétaine 16: 75–82. Suzuki Hiroyuki. 2009c. Origin of non-Tibetan words in Tibetan dialects of the Ethnic Corridor in West Sichuan. In Yasuhiko Nagano (ed.). Issues in Tibeto-Burman Historical Linguistics, 71–96. Suita: National Museum of Ethnology.



The Tibetic languages and their classification   

   129

Suzuki Hiroyuki. 2009d. Tibetan dialects spoken in Shar khog and Khod po khog. East and West 59 (1–4): 273–283. Suzuki Hiroyuki. 2011a. Deux remarques supplémentaires à propos du développement du ra-btags en tibétain parlé. Revue d’étude tibétaine 20: 123–133. Suzuki Hiroyuki. 2011b. Phonetic analysis of dGudzong Tibetan: The vernacular of Khams Tibetan spoken in the rGyalrong area. Bulletin of National Museum of Ethnology 35 (4): 617–653. Suzuki; Hiroyuki, 2012. A propos du terme ‘riz’ et de l’hypothèse du groupe dialecta Sems-kyinyila en tibétain du Khams. Reveu d’étude tibétaine Vol 23, 107–115. Tournadre, Nicolas. 2005. L‘aire linguistique tibétaine et ses divers dialectes. Lalies 25: 7–56. Tournadre, Nicolas. 2008. Arguments against the Concept of ‘Conjunct’/‘Disjunct’ in Tibetan. In B. Huber, M. Volkart, P. Widmer, P. Schwieger (eds). Chomolangma, Demawend und Kasbek. Festschrift für Roland Bielmeier zu seinem 65. Geburtstag. vol 1, 281–308. International Institute for Tibetan and Buddhist Studies. Tournadre, Nicolas. 2010. The Classical Tibetan cases and their transcategoriality. From sacred grammar to modern linguistics. Himalayan Linguistics 9 (2): 87–125. Tournadre, Nicolas and D. Jamborová. 2009. Taxis: Temps déictique, temps relatif, ordre sequential. In proceedings of Románske štúdie: súčasný stav a perspektívy [Études romanes: situation contemporaine et perspectives], 26.–27. marca 2009 v Banskej Bystrici Studia Romanistica Beliana, 458–466. Banská Bystrica. Tournadre, Nicolas and Konchok Jiatso. 2001. Final auxiliary verbs in literary Tibetan and in the dialects. Person and Evidence in Himalayan Languages [Special issue] Linguistics of the Tibeto-Burman Area 24.1: 177–239. Tournadre, Nicolas and Sangda Dorje. 1998. [Reprints 2003, 2009]. Manuel de tibétain standard, langue et civilisation. Paris: L’Asiathèque. Tournadre, Nicolas and Sangda Dorje. 2003. Manual of Standard Tibetan. Ithaca, New York.: Snowlion. Tournadre, Nicolas and H. Suzuki (with the collaboration of K. Gyatso, X. Becker). in preparation. The Tibetic languages. Zeisler, Bettina. 2004. Relative Tense and aspectual values in Tibetan languages. Berlin and New York: Mouton de Gruyter. Zeisler, Bettina. 2011. Kenhat, the dialects of Upper Ladakh and Zanskar. In Mark Turin and Bettina Zeisler (eds.). Himalayan Languages and Linguistics. Studies in Phonology, Semantics, Morphology and Syntax, 235–310. Leiden: Brill. Zeisler, Bettina. 2012. Practical issues of pragmatic case marking variations in the Kenhat varieties of Ladakh. Linguistics of the Tibeto-Burman Area: 35 (1): 75–106. Zhang Jichuan. 1997. Particularités phonétiques du Baima. Cahiers de linguistique Asie orientale 2 (1): 131–153. Zhang Jichuan. 2009. Zàngyǔ cízú yánjiū [Tibetan Word Family Research]. Beijing: Shèhuì kēxué chūbǎnshè.

Isao Honda

Internal diversity in the Tamangic lexicon 1 Introduction1 Tamangic is a language group comprising Tamang, Gurung, Thakali, Manangba, Seke, and Chantyal, all of which are spoken in Nepal.2 In Shafer (1955), this language group is called the Gurung Branch and is included in his Bodish Section, together with the Tibetan complex and others, such as “Dwags,” “Tsangla” (i.e. Tshangla), and “Gyarong” (i.e. Rgyalrong; cf. Table 1).3 Since then, the proposition that Tamangic languages are close relatives of Tibetan has been generally accepted. Mazaudon for instance, states, “No claim is made here concerning the higher level classification of the family [that is, the Bodish Branch] but many dialects of Tibetan and Tamangish [i.e. Tamangic] are definitely closely related” (2005: 79–80).

1 This study is supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (No. 21520463). I would like to thank Nathan W. Hill, Thomas Owen-Smith and my anonymous reviewers for their comments and suggestions with regard to my data and analysis. All mistakes are, of course, my own. 2 The language group has also been known as “Tamang group” (e.g. Nishi 1982, 1990), “TamangGurung-Thakali” or its abbreviated form, “TGT” (e.g. Pittman and Glover 1970; Watters 2002), “TGTM,” which is an abbreviated form of “Tamang-Gurung-Thakali-Manang” (e.g. Mazaudon 1978, 1996, 2003), or “Tamangish” (Mazaudon 2005). 3 Shafer’s “Dwags,” or “Takpa,” is based on vocabulary listed in Hodgson (1853) and Campbell (1874), and, as many authors noted, his recognition of this language as a Tibetan dialect is an error. Shafer probably confused Hodgson’s Takpa data with the Tibetan district and dialect of Dakpo, spoken in an area south of the Tsangpo and west of the Kongpo. According to Driem (2001) and Michailovsky and Mazaudon (1994), Hodgson’s Takpa is a language spoken in Ta­ wang in Arunachal Pradesh and Tshona in Southern Tibet, and in neighboring areas of Bhutan. Nevertheless, Shafter correctly identified its distinct status of Takpa (or Dakpa) and put it in a distinct language group called East Bodish. The term East Bodish is now widely used to refer to a group covering a number of languages, mainly spoken in Bhutan, and Dakpa is one of them, although van Driem states, “Dakpa appears to be the most aberrant member of East Bodish or, at least, to constitute a group on its own within East Bodish” (2001: 916). It should also be noted that the term Dakpa appears in linguistic literature to refer to various languages/dialects. In Michailovsky and Mazaudon (1994: 555, footnote 2), for instance, the term is used to refer to what Lu (1986) calls the Mama dialect of Cuona Monpa. Meanwhile, the positions of Tshangla and Rgyalrong are still not totally clear, but, as far as Tshangla is concerned, many scholars seem to accept Shafer’s classification. Inclusion of Rgyal­ rong into the Bodish Section, on the other hand, has been questioned by many scholars (e.g. Michailovsky and Mazaudon 1994).

132   

   Isao Honda

The genetic grouping of the languages that are mentioned above as Tamangic languages is less controversial, and the mutual relation among the languages does not appear to be so distant. Yet there exist several different kinds of variance in their phonological, lexical, and grammatical domains. This paper focuses on the lexical variations that are observed in Tamangic languages/dialects and examines what they indicate as to the history of the group and each language/dialect. It is reasonable to assume that these variations are mainly due to language contact. One of the most recent contacts is with Nepali, which is not dealt with in this paper. Contacts with other Tibeto-Burman languages spoken in Nepal date back to much earlier times. These languages include Newar, Thangmi, Kiranti languages (particularly those adjacent to the Tamang language area, such as Sunwar), Magar, Chepang, and Kham. The number of loans from and to these languages appears to be relatively small; nevertheless, they tell us about undocumented contact situations in the past. Tab. 1: Shafer’s (1955, 1966) Bodic Division – Bodic Division – Bodish Section – Bodish Branch – West Bodish [Balti, Purik, Ladakhi ...] – Central Bodish [most Tibetan dialects including Amdo, Kham] – South Bodish [Sikkimese, Tromowa, Dandzongka] – East Bodish [Dwags (Takpa) ] – Tsangla [Tshangla] – Rgyarong Branch [Rgyalrong] – Gurung Branch [Tamangic] – West Himalayish Section – West Central Himalayish Section [Hayu, Chepang, Magar] – East Himalayish Section [Kiranti]

Tibetan dialects have brought more profound impact and influence upon Tamang­ic languages. Tamangic languages contain a large number of lexical items cognate with Written Tibetan and modern Tibetan dialects. These lexical items may be sorted into different groups. One group represents recent loans from modern Tibetan dialects, and another reflects much older contacts, while yet a third may be a set of reflexes derived from a common proto-language. The process of sorting vocabulary into these three groups is not an easy task, and there is a limitation to our knowledge. The main aim of this paper is, thus, not to prove or disprove the validity of the putative Tamangic-Tibetan genetic link, but to discuss the difficulties and obstacles that we have to overcome in order to establish their genetic relation.



   133

Internal diversity in the Tamangic lexicon   

2 Recent loans from a modern Tibetan dialect When examining the Tamangic lexicon, one can soon realize that there are several different kinds of sound correspondences among the Tamangic languages/dialects and between the Tamangic languages/dialects and Tibetan. One type of correspondence appears to reflect a relatively recent loan from a neighboring modern Tibetan dialect. One of the most useful clues to differentiate recent loans from a modern Central Tibetan dialect from others is tone, since in most of the Tamangic languages/dialects and most of the modern Central Tibetan dialects there is a highlow tone distinction, including for sonorant-initial words. In every Tamangic language/dialect, except in Chantyal, four distinct tones are identified. The phonetic value of the four tones varies from dialect to dialect, which is shown in Table 2 excerpted from Mazaudon (1978: 165) where the pitch of the four tones in each dialect is described by using Chao Yuen-Ren’s (1930) notation from 1 (lowest) to 5 (highest). The four tones in each dialect are identified as tone 1, tone 2, tone 3, and tone 4 in Mazaudon (1973a, 1973b, 1978) on the basis of “regular” correspondences found among the Tamangic dialects; that is, the word for ‘I’ (first person singular pronoun), for instance, is [⁵⁴ŋa] in Risiangku Tamang (RI) and [³³ŋa] in Manangba (Ngawal: NG), both of which are then described as tone 1. Similarly, the word for ‘drum’ occurs under tone [33] in Risiangku ([³³ŋaː]) and under tone [54] in Ngawal, both of which are then called tone 3, and so on. Tab.2: Tamangic Tone (from Mazaudon 1978: 165) Tone

RI

TL

SA

GH

NG

TU

MA

SY

1 2 3 4

54 44 33/22 211

55/44 43 33/22 51

44 54 11 32

33 54 11 12

33 45 54 31

54 44/33 11 121

43 45 33/22 51

43 45 11 33/22

Except in Mawatan Thakali (MA: Marpha), Manangba, and in Taglung (TL: Eastern Tamang), tone 1 and tone 2 are high resister tones and with clear voice, whereas tone 3 and tone 4 are low register tones and with lax or breathy voice. This highlow distinction is, in many cases, relatively salient and is not so difficult to detect. The high-low tone distinction is also a common feature among modern Central Tibetan dialects (with an exception of Kyirong where there is a three-fold tone distinction, i.e. high, mid, and low, Huber 2002, 2003), although the exact nature of the tone system varies from dialect to dialect. In the case of Southern Mustang (or Baragaonle), a Tibetan dialect spoken in an area adjacent to the

134   

   Isao Honda

Seke-Thakali-Manangba speaking area, for instance, there are only two tones, high and low, but no contour tone. In the following example (1) ‘servant, slave,’ the forms in (1A), with a high tone, which are all from Seke, are most likely recent loans from Southern Mustang, whereas the forms in (1B), with the initial ky- and tone 4 (or an unspecified low tone, i.e. either tone 3 or 4), and the (1C) form are not. Similarly, the Manangba form for ‘roast, fry’ in (3B) with tone 2, which is a high tone [45], could be a recent loan from Tibetan if this tone description is accurate, whereas the forms for ‘male yak’ in (2) and the forms in (3A) cannot be recent loans since their tones are different from those in Southern Mustang. (1) ‘servant, slave’ A Seke (TA) ᴴyokpo; Seke (CH) ᴴyokpa; Seke (TE) ᴴyokpu B E. Tamang (RI) ⁴kyapa; T. Thakali ⁴kyapa; Y. Thakali ᴸkyap; Seke (TA) ᴸ kyappa C Gurung ¹keb cf. WT g.yog-po; SM ᴴyokpo (2) ‘male yak’ W. Tamang ⁴yaː; Manangba (Hi) ³ya; T. Thakali ⁴yā; Y. Thakali ᴸya; Seke (TA, TE, CH) ᴸyā cf. WT g.yag; SM ᴴyak cf. PTB (M) *g-yak ‘sheep/yak’ (3) ‘roast, fry’4 A Gurung, T. Thakali, M. Thakali (G) ³ŋo-; Chantyal ɦõ-; Y. Thakali, Seke (TA, CH, TE) ᴸŋo B MN (Hi) ²ŋwo cf. WT rŋod-; SM ᴴŋö cf. PTB *r-ŋaw- However, this may not be an absolute criterion because we cannot presuppose that borrowed lexical items in a recipient language show a tone most similar to that of the original lexical item in a donor language. In fact, there are many instances where some kind of tone shift or tone mutation is suspected to have 4 In many neighboring Tibeto-Burman languages, we find a form considered to be a modern reflex of the suggested Proto-Tibeto-Burman form, e.g. Bantawa  ŋü-(u); Kulung  ŋə-ma; Yamphu nuˑ-ma; Limbu ŋɔ:-ma, nɔ:-ma (Opgenort 2005); Chepang ŋaw- ‘roast, fry.’ Watters (2003) regards Kham ŋoh- ‘roast’ as a loan from Thakali (cf. footnote 12).



Internal diversity in the Tamangic lexicon   

   135

occurred when or after a lexical item was borrowed. Such instances will be discussed in the next section. Another useful clue to pick out most recent loans is word-initial consonant clusters. Since modern Central Tibetan dialects show a set of particular developments from the Written Tibetan forms, we can guess, for instance, that the forms for ‘hand (Honorific)’ in (4A), with the initial high tone chy- (IPA [tshj-]) cluster, are probably recent loans, whereas the Eastern Tamang form ¹phyaː ‘prostration’ in (4B) is probably not. It is suspected that the latter was borrowed at a much earlier stage, or that the donor dialects of (4A) and (4B) are different.5 A similar analysis can be made of (5) ‘rich’ and (6) ‘intelligent, skillful’ where (5A) and (6) are not considered to be recent loans. (4) ‘hand (Honorific)’ A E. Tamang (KTI) ᴴchyaː; Y. Thakali ¹chyā; M. Thakali, Seke (TA) ᴴchyā; Seke (CH, TE) ᴴchyāk B E. Tamang (RI) ¹phyaː ‘prostration’ cf. WT phyag; SM tšhak (5) ‘rich’ A E. Tamang (KTI) ᴴphyuk-pa; W. Tamang ¹phyukpa cf. WT phyug-po; KY ᴴtɕhukpo; SM ᴴtšhukpo B Gurung, T. Thakali, M. Thakali (G) ⁴plo-; Manangba (Hi) ⁴phlo-; Y. Thakali, Seke (CH) ᴸplo-; Seke (TA, TE) ᴸbro(6) ‘intelligent, skillful’ Manangba (Hi) ²pyaŋ-; T. Thakali ²pyāŋ-; M. Thakali (G) ¹pyaŋ-; Y. Thakali ᴴ pyaŋ‑; Seke (TA, CH, TE) ᴴpyāŋ cf. WT spyaŋ-; SM ᴴtšaŋ(a), ᴴtšaŋpa The next set of examples illustrates cases with an initial bilabial + r cluster. We suspect that (7A) ‘rosary’ and (9A) ‘female yak’ are probably recent loans, but (7B), (8) ‘deliver,’ (9B), and (10) ‘write’ are not.

5 It should be noted that in Kyirong, a modern Tibetan dialect geographically very close to the Tamang speaking area, bilabial + glide clusters are still preserved to a large extent, (and so are Written Tibetan bilabial + r clusters, more consistently; see [7], [9], and [10]), a situation which is different from other modern Central Tibetan dialects (Huber 2002, 2003). Therefore, there may exist a possibility that the (4B) form is a relatively recent loan from this dialect. It is, however, not known how likely this possibility is, as this word is not found in Huber (2002).

136   

   Isao Honda

(7) ‘rosary’ A Seke (CH) ᴴʈhaka B T. Thakali ²phramo; M. Thakali (G) ¹phraŋpa; Y. Thakali ᴴphraŋma; Seke (TA) ᴴphrāŋmo; Seke (CH) ᴴphrāŋpa; Seke (TE) ᴴphraŋpaŋ cf. WT ʼphreŋ(-ba); SM ᴴʈhaŋa; KY phrāŋā (8) ‘deliver, send’ M. Thakali, Y. Thakali, Seke (TA, CH) ᴴpreː- cf. WT sprod(9) ‘female yak’ A Seke (CH) ᴸʈima; Seke (TE) ᴸʈimo B T. Thakali, Y. Thakali, Seke (TA) ᴴpri; M. Thakali (G) ²pri cf. WT ʼbri-mo; SM ɖimo; KY bri̱ (10) ‘write’ E. Tamang (RI) ³pri(ː)-; W. Tamang, Gurung, Manangba (Hi), M. Thakali (G) ³ pri-; T. Thakali ³prup-; Y. Thakali ᴸpri(ː)-; Seke (TA, CH, TE) ᴸpri cf. WT ʼbri-; SM ᴸʈi-; KY pri̱ cf. PTB *riy = *rəy ʻdraw (picture)’ It is more difficult to identify the origins of the other sets of cognates in the above examples and in many others. In the following two sections, I will discuss what kinds of obstacles we have to overcome for a better understanding of the history of the Tamangic group with reference to two kinds of variations found in Tamangic dialects: those concerned with tone, and those with an initial consonant cluster. Since there are so many issues related to initial clusters in Tamangic dialects, in the current article I chose to focus on velar + r clusters.

3 Irregular tone correspondences As mentioned in the previous section, the Tamangic tones 1, 2, 3, and 4 are identified on the basis of the “regular” correspondences found among Tamangic dialects. However, the correspondences are, in fact, not totally regular because there are a number of tone disagreements. One quite common type of irregular correspondence is between the two high tones and between the two low tones; that is, a tone 1 word in a given dialect corresponds to a tone 2 word in another dialect, and a tone 3 word in a given



Internal diversity in the Tamangic lexicon   

   137

dialect corresponds to a tone 4 word in another dialect. Even more problematic are cases where a high tone word in one dialect corresponds to a low tone word in another dialect. This type of irregular correspondence is relatively rare, yet a number of instances have been found. One of the instances was already shown in (1) (between [1B] and [1C]) and (3). Here are some more examples: (11) ‘copper’ A Manangba (Hi) ⁴khyor; T. Thakali, Y. Thakali ⁴kyor; cf. M. Thakali (G) ²kyor B Gurung ²kyuraː cf. PTB *grəy; WT gri ‘knife’ (Matisoff 2003: 189 relates this to the PTB form) C E. Tamang (KTI) ᴸsaŋpo; Seke (TA) ³sāŋ; Seke (CH, TE) ᴸsaŋ D W. Tamang ¹saŋmo cf. WT zaŋs; SM ᴸsaŋa; KY sa̱ŋ (12) ‘give (Honorific)’ A T. Thakali, M. Thakali (G) ³naŋ-; Y. Thakali, Seke (TA, TE, CH) ᴸnaŋ B E. Tamang (RI) ¹naŋ cf. WT gnaŋ-; SM ᴴnaŋ(13) ‘west’ A E. Tamang (KTI) ᴸnup; T. Thakali ⁴nup; Y. Thakali ᴸnu; Seke (TA, CH, TE) ᴸ nuk B W. Tamang ¹nup cf. WT nub; SM ᴸnup; Sherpa (K) ᴸnup (14) ‘corner, edge’ A W. Tamang ⁴suːr; T. Thakali ³sur; Y. Thakali, Seke (CH) ᴸsu (cf. CH ᴸsu-ri ‘at the corner’); Seke (TA) ᴸsur B E. Tamang (RI) ²sur cf. WT zur; SM ᴸsur (15) ‘door-lock, latch’6 A Y. Thakali, Seke (TE) ᴸkiliŋ; Seke (TA) ᴸkilip; Seke (CH) ᴸkilik B E. Tamang (KTI) ᴴkiliŋ cf. SM ᴸkilik, ᴸsilik

6 Confer WT ku-lig ‘key, lock’ (Jäschke 1881 [1995]: 3).

138   

   Isao Honda

(16) ‘middle, between’ A T. Thakali ²par-ri ‘during;’ Y. Thakali ᴴpar; Seke (TA, TE, CH) ᴴpa-ri ‘in the middle’ B Manangba (Hi) ³pʌɾ, ³pʌɾi cf. WT bar; SM ᴸparla; Sherpa ᴸpar-lā (17) ‘silver’ A E. Tamang (RI) ⁴mui; Manangba (Hi) ⁴mwîʃʌ; M. Thakali (G) ⁴muy; Y. Thakali, Seke (TA) ᴸmui B W. Tamang ²mui; Seke (CH, TE) ᴴmyui cf. WT dŋul, mul; SM ᴴŋul cf. PTB *ŋul = *(d)ŋul; (M) *mul (18) ‘sink’ A E. Tamang ³plum B T. Thakali ²plum ⁴ya cf. WT ʼbyiŋ-; cf. Khaling blam cf. PTB *lip; (M) *lip = *lup There are several scenarios which have been proposed to account for tone disagreements among Tamangic dialects. One of them is, as stated in the previous section, that the high tone forms and the low tone forms are both loans from Tibetan but from different dialects or/and at different historical stages. Another possibility is that some forms are loans and the others are old retentions. Some of the tone disagreements could be due to a descriptive error. The Mawatan Thakali form for ‘copper’ with tone 2 in (11A) is, I suspect, such an instance (see also M. Thakali [G] ²tuŋ- ‘close’ in [20]). As indicated in Table  2, tone 4 in Mawatan Thakali is, unlike in most of the Tamangic dialects, not a low tone but the highest in pitch, i.e. [51], and word-initial stops and affricates under tone 4 are realized as voiceless, which makes it difficult to differentiate tone  4 from tones 1 and 2. A totally different type of explanation may also be possible; that is, when or after a lexical item was borrowed, its original tonal features were lost or changed. The change might be sporadic and unpredictable. As has often been pointed out (e.g. Hildebrandt 2004: 31), there appears to be a tendency that nowadays many speakers of Tamangic languages, particularly younger, urban-based ones, fail to realise the tones properly, i.e. as they used to be pronounced in their ancestral (rural) communities. This applies at least to Eastern Tamang, Manangba, Mawatan Thakali, Yhulkasom Thakali, and Seke, on all of which I myself have conducted fieldwork.



Internal diversity in the Tamangic lexicon   

   139

It is also possible, however, that the tone mutation is not totally irregular. This speculation comes from the following observations. First, in many of the above examples, and also in (1B-C), a high tone is found only in one dialect, and thus I suspect that if the cause of the irregular tone correspondence is due to a recent tone change, the change must be, or is more likely to occur from a low tone to a high tone but not vice versa. Second, there are several cases, including (1C) ¹keb ‘slave’ and (11B) ²kyuraː ‘copper,’ where tone 4 words in other Tamangic dialects correspond to a high tone word in Gurung. The following are some of the examples, although the correspondence is not obvious in (20) and (21). (19) ‘snow’ A Gurung ²ɬyĩ B E. Tamang, W. Tamang ⁴kliŋ; Chantyal khiliŋ; Manangba (Hi) ⁴khiŋ; T. Thakali ⁴kin; M. Thakali ⁴lin; Y. Thakali ᴸlim; Seke (TA) ᴸŋin; Seke (CH) ᴸliŋ; Seke (TE) ᴸkiŋ (20) ‘close’ A Gurung ²tu B E. Tamang (KTI) ᴸʈuŋ-; W. Tamang ⁴ʈuŋ-; M. Thakali, Y. Thakali ⁴tuŋ-; cf. M. Thakali (G) ²tuŋ-; Seke (TA, CH, TE) ᴸtuŋ(21) ‘exchange’ A Gurung ²ti B E. Tamang (RI), W. Tamang ⁴tep-; T. Thakali ⁴to-; M. Thakali ⁴?to-; Y. Thakali ᴸ to(ː) cf. WT sdebThere are also a couple of cases where Gurung not only has a tone 4 cognate, but also has a high tone cognate or variant; e.g. (22) ‘one’ Gurung ²kri; ⁴gri (for cognates in other Tamangic dialects, cf. Table 3) (23) ‘bee’ A Gurung ¹kwe; ⁴koe, ⁴kwe B Manangba ⁴kwhe; T. Thakali ⁴koy ‘bee, honey;’ M. Thakali, Y. Thakali ⁴koi ‘honey;’ Seke (TA) ᴸkoi, ᴸkoe ‘honey;’ Seke (CH) ᴸkyui ‘honey;’ Seke (TE) ᴸke ‘honey’

140   

   Isao Honda

(24) ‘vulture’7 Gurung ¹kwre; ⁴kure (for possible cognates in other Tamangic dialects, cf. ‘eagle,’ ‘hawk/falcon?’ in Table 3) (25) ‘daughter’ A Gurung ²camĩ, ²cami; ³ca(ː)mĩ B E. Tamang (RI) ⁴came; Chantyal cɦame; Manangba tsàme; M. Thakali ⁴? came; Y. Thakali, Seke (TA) ᴸcame; Seke (CH, TE) ᴸceme There is no explanation in the previous literature on Gurung of why there are such variations, but they lead us to suspect that in Gurung a tone change is under way sporadically on tone 4 words, or tone 4 is more prone than other tones to mutate into a tone with much higher pitch and without breathiness. As mentioned earlier, tone 4 in Mawatan Thakali, Manangba, and the Taglung dialect of Eastern Tamang is a high or mid tone, a fact which is considered to be the result of dialect-independent developments (Mazaudon 1978: 171). Therefore, it would not be particularly surprising if the above mentioned tone mutation actually occurred or is now under way in Gurung, or that it might occur in other Tamangic dialects too. However, this is no more than just speculation at this moment.

4 Initial clusters As noted before, the mutual relation among the Tamangic languages does not appear very distant, and a number of Proto-Tamangic forms have been proposed (Mazaudon 1978, 1985, 1988). This section will examine the Proto-Tamangic forms with an initial *kr- or *gr- cluster proposed in Mazaudon (1978, 1985, 1988), which are listed in Table 3. The Proto-Tamangic forms proposed by Mazaudon (1978, 1985, 1988) are the first four, i.e. ‘head/hair,’ ‘cubit,’ ‘waist,’ and ‘burn,’ and from the seventh to the eleventh, i.e. ‘cheek,’ ‘body dirt,’ ‘intestine,’ ‘enemy,’ and ‘one.’ As can be seen, there is a regular sound correspondence among Tamangic dialects; that is, while the original cluster is preserved in Risiangku (RI: Eastern Tamang), Sahu (SA: Western Tamang), Ghachok (GH: Gurung), Manangba (MN), Tetang (TE: Seke), 7 Instead of ¹kwre and ⁴kure, a different Gurung form ⁴krwe ‘vulture’ is listed in Hale (1973), in which the contributors of the Gurung data are identical to the authors of the dictionary listing the former forms, i.e. Glover et al. (1977).



Internal diversity in the Tamangic lexicon   

   141

Chuksang (CH: Seke), and only marginally in Tangbe (TA: Seke), it changed into ky- or k- in Marpha (MA: Mawatan Thakali) and Syang (SY: Yhulkasom Thakali), or into ʈ- in Tukche (TU: Tamang Thakali). In Chuksang, and to some extent in Manangba too, an epenthetic vowel is often inserted to break the cluster. In Manangba, all of the old initial unaspirated stops/affricates changed into aspirated stops/affricates in tone 4, and those aspirated stops/affricates then further underwent fricativization (Mazaudon 1988: 6). The Seke forms for ‘cheek’ and ‘intestine’ are probably recent loans from a Tibetan dialect, and so are the forms for ‘enemy’ in Seke, Syang, Marpha, and possibly in Tukche too. It is not clear to me whether the Sahu form for ‘waist,’ i.e. ¹keppa, is a loan or not, but Mazau­ don (1988: 4, footnotes 5 and 9) considers it and also RI ¹keː to be the results of an innovation, *kr- > k- before *i or *e. This also accounts for ⁴kik and ⁴kiː, the Risiang­ku and Sahu forms for ‘one’ (cf. Tetang and Chuksang, where r has been lost only before *i). The aspiration found in the Risiangku and Sahu forms for ‘body dirt,’ i.e. ¹khiti and khriti (tone unknown), are explained by Mazaudon (1988: 9) in a similar way, i.e. *kr- > khr- before *i.8 In two other entries, ‘wheat’ and ‘eagle,’ Mazaudon (1978, 1985) proposes the Proto-Tamangic forms with the initial *grw-, i.e. *grwaA/B ‘wheat’ and *grwatB ‘eagle,’ and explains the lack of the medial r in the Risiangku forms ⁴kwa ‘wheat’ and ⁴kwat ‘eagle’ by the semi-vowel w that follows.9 Some forms for ‘hawk/falcon?’ are listed just as in M ­ azaudon (1978), but no regular sound correspondence is found between those forms and the forms for ‘eagle;’ also cf. Syang, Marpha ᴴtikyā, both of which must be cognate with Ma­nangba (Hi) tikyʌ ‘hawk/raptor/eagle;’ Tangbe, Tetang ᴴʈa and Tangbe ᴴtiŋkyur, all of which refer to one of the birds of prey. There are at least two more entries, i.e. ‘cry, weep,’ ‘climb,’ which show the correspondence indicated in Table 3. In other entries listed, i.e. ‘similar,’ ‘stir,’ and ‘shadow,’ the correspondence is not regular, or it is not clear whether it is regular or not due to a paucity of data. If we assume that the medial r has been lost before *i in Chuksang and Tetang (cf. the Chuksang and Tetang forms for ‘one’), the Chuksang and Tetang forms for ‘shadow,’ ᴸkruksi and ᴸkrip, cannot be reflexes of the Proto-Tamangic cluster.

8 The second syllable -ti must also be explained. Mazaudon (1988: 9) proposes the PTAM *gri(ti) and suggests that *-ti has been lost in other Tamangic dialects. However, I suspect that the Risiangku form ¹khiti might be a loan from Newar (cf. C. Newar [J] khiti ‘secretion, excretion’ < khi ‘feces’ + ti ‘water’) though this does not accout for the Sahu form khriti. 9 In the current paper, the Proto-Tamangic tones *A and *B are based on Mazaudon (1978, 2005), where the proto-tone *A corresponds to the modern tones 1 and 3 whereas the proto-tone *B to the modern tones 2 and 4, rather than on Mazaudon (1985), where the correspondence is the opposite.

142   

   Isao Honda

Tab. 3: Tamangic SET 1 /kr/ ((TA), CH, TE, MN, GH, SA, RI) ~ /ky/ or /k/ (SY, MA) ~ /ʈ/ (TU) *NOTE: Under tone 4, MN [hř], [hr] or [ʃ]; cf. Mazaudon (1988) TA

CH

cubit waist

¹ka ᴴkara [⁴³kra] (M) ¹ku ᴴkuru ²ke(ː) ᴴkere

burn cry, weep climb cheek

¹k(r)āː²k(r)eː(ᴸʈampa)

head/hair

ᴴkroᴴkrāᴴkreː(ᴸʈempa)

TE

SY

MA

TU

MN

ᴴkra

²kyā (M)

²kyā (G)

ᴴʈa

²krɤ (M)

ᴴkre

¹ku (M) ¹kae (M)

¹ku (M) ¹kae (M)

ᴴyā-ᴴʈu ᴴʈe

¹kru (M) ¹kre (M)

ᴴkroᴴkrāᴴkre(ᴸʈampa)

ᴴkyāᴴkaiᴸkyam

¹kyā- (G) ᴴkai³kyāmpā (G) ³gɤi/gᵻ (M) ³kum

¹ʈo²ʈā²ʈe³ʈampa

¹kro- (Hi) ¹kra- (Hi) ¹kre- (Hi) krʌ́mʌ (Hi) ³kri (M)

(⁴ʈa) (M) ⁴kɤi (M)

intestine

(ᴸkyumo)

(ᴸkyumu)

(ᴸkyuma)

³gɤi, gᵻ (M) ᴸkum

enemy one

(ᴸʈa) ᴸki(ː)

(ᴸʈa) ᴸki(ː)

(ᴸʈa) ᴸki(k)

(ᴸʈa) ⁴gɨ (M)

body dirt

stir

ᴸkup-

ᴸkruk-

ᴸʈuk-

ᴸku-

⁴?ku-

shadow

(ᴴuː) (ᴴu(k)ku) ³ko

(ᴸkruksi)

(ᴸkrip)

ᴸku

⁴?kui

ᴸkro

ᴸkro

⁴ko (M)

⁴ko (M)

eagle

cf. hawk/ falcon?

⁴ʈa (M) ⁴ʈi

ᴸkɤi

similar

wheat

kùruŋ (Hi)

²khyā

¹?khyā (M)

⁴hřʌ (M) ⁴hři (M) ⁴ʃi (Hi)

⁴khri (Hi) ³ʈo (M) ⁴ʈo-sim (M) ⁴ʈe ‘vulture’ ᴴʈha ‘hawk’

⁴ʃɔ (M) ⁴hro (Ho) hřʌ (M)



Internal diversity in the Tamangic lexicon   

   143

GH

SA

RI

PTAM

WT

CT (SM)

PTB

¹kra

¹kra

¹kra

*kraA/B

skra

*s-kra

¹kru ¹kre

²kru ²naŋ ¹keppa

¹kru ¹keː

*kruA *kre(t)A

¹kro¹kro¹kre³kra:mu

¹kro¹kraː¹kret³krampa

¹kro¹kraː¹krat³krampa

*kro(ː)A

khru rked-pa sked-pa

ᴴhra ᴴʈā (KY) ᴴʈhu ᴴkyepa

*grampa/uA

’gram-pa

ᴸʈempa

*gam = *gəm ‘jaw

³kri

khriti (M) ³kruŋ

¹khiti

*gri(ti)A

dreg-pa

ᴸʈima

³kruŋ

*gruŋA

rgyu-ma

ᴸkyuma

⁴kik

*graB *grikB

dgra gcig

ᴸʈa ᴴtši/ᴴtšik

*kriy = *krəy; *ri(y), *d-ri(y) (M) *grwat (M) ‘belly/stomach’ *gra (M) *t(y)ik = *(g)-tyik

dkrug-/ dkrog-/ skroggrib-ma

ᴴʈuk-

ᴸʈip

gro

ᴸʈo

⁴gri/²kri

⁴kiː

³kriba/ ⁴?grinaa (²?rudu-)

²krip

⁴kure/ ¹kwre ‘vulture’

³wa

⁴kwa

*grwaA/B

²wet

⁴kwat

*grwatB

’grag(s)-

*krap-

*(g)rip ~ *(s)rip

144   

   Isao Honda

Tab. 4: Tamangic SET 2 /hr/ (TA, CH, TE, SY, MA, TU) ~ /r/ (GH) ~ /ʈ/ (MN, SA, RI) TA

CH

TE

finger

SY

MA

TU

MN

ᴴya-ᴴhri

²hri (G) ᴴya-ᴴhri

ᴴya-ᴴhre

ᴴhrup (M)²tu (M) ²hrāŋ-

¹ʈi (Ho) ¹ya-³ʈi (Hi) ²tu (M)

thread

ᴴhrup

ᴴhrup

ᴴhrup

²hru (M)

²hru (M)

swell

ᴴhrāŋ-

ᴴhraŋ-

ᴴhraŋ-

ᴴhraŋ-

sweat

ᴴhruk-

ᴴhruk-

²hru

²hraŋ(G) ²hru (G)

²ʈaŋ(Ho) ²ʈu (Ho)

There is another sound correspondence found in Tamangic dialects, which is shown in Table 4; that is, /hr/ (Tangbe, Chuksang, Tetang, Syang, Marpha, Tukche) ~ /r/ (GH) ~ /ʈ/ (Manangba, Sahu, Risiangku), all of which are under a high tone. The correspondence is found so far only in four entries, ‘finger,’ ‘thread,’ ‘to swell,’ and ‘sweat.’ For ‘thread,’ the initial *tr- is reconstructed in Mazaudon (1978: 175) and *sr- in Nishi (1991: 91), but in Honda (2008, 2009) I suggested another possibility that the sound correspondence found in the four entries are reflexes of an older *kr- cluster.10 However, the reconstruction cannot be for the Proto-Tamangic because of the existence of the sound correspondences shown in Table 3, which has much wider distribution and must be traced back to the Proto-Tamangic. In any case, what is important to note is that both sets, the ones shown in Table 3 and Table 4, show no clear correspondence to the Written Tibetan forms listed in the tables.

10 As to retroflex stops found in modern Tibeto-Burman languages, Matisoff (2003: 22) states, “most of these retroflex stops derive from Tibeto-Burman clusters of *velar-plus-liquid.” His reconstruction of PTB *krəw, *kriŋ ‘thread, plait’ is based on this observation and on forms like Written Burmese khraññ and krûi ‘thread, string, chain’ (Matisoff 2003: 199). The reconstruction of *sr- from hr- is also in accordance with Matisoff’s (2003: 53) observation that “... voiceless resonants (hl, hr, hw, hy) in Tibeto-Burman languages generally derive from *resonants preceded by the *s- or *ʔ- prefix,” but the development of ʈ from *sr- is not likely. In fact, the developments of ʈ and hr from an old kr- cluster are both attested in Tibetan. In Southern Mustang, for instance, WT dkr- and bkr- both developed into ʈ, e.g. WT dkrug- > ᴴʈuk- ‘stir;’ WT bkra-shis > ᴴʈaši ‘blessing,’ whereas WT skr- resulted in hr, e.g. WT skra > hra ‘the hair of the head.’ In Kyirong, on the other hand, not only WT dkr- and bkr-, but also skr- ended up ʈ, e.g. WT skra > ʈā ‘hair.’



GH

Internal diversity in the Tamangic lexicon   

SA

RI

PTAM

WT

CT (SM)

PTB

¹ʈup

²ʈup

*trupB

skud-pa

ᴴkupa

²ʈaŋ-

²ʈaŋ-

skraŋ-

ᴴhraŋ-

*krəw, *kriŋ (M)

²ʈuː

²ʈuː

   145

¹yori/¹ri

²rõ-

*krwiy = *khrwəy (PLB) *s-krul = *s-ŋrul (M)

rŋul

Tab. 5: East Bodish kr-, gr- clusters (From Michailovsky and Mazaudon 1994)

hair thread village dirt nit count cry out shadow elbow six warm mule

KT

BT (CK)

BT (CM)

M

DK

WT

TAM

ᴴra ᴴrotman ʈoŋ ʈekpa ᴴrikar ɖaŋkha ɖakɖem ɖumaliŋ ɖo:ʔ

kra kronman kroŋ krekpa kriwit graŋ-

ᴴra ᴴrotman ᴴroŋ ᴴrekpa ᴴriwis ᴸraŋkha ᴸrakᴸrep ᴸrumaŋti ᴸrok ᴸrutgriu (u)

ᴴra ᴴrøp

khrʌ⁵³

skra rgyud grong dreg-pa sro-ma grangs ’grags-pa grib(-ma) gru-mo drug drod drel

¹kra

ɖe:ʔ

grep grumaŋti grok grutɖẽ

ɖɛkpɛʔ ᴴriula

tsoŋ¹³ sep⁵³

grak³⁵ (W)

ɖu:

krum¹³tɕuŋ⁵³ kroʔ⁵³ kro³⁵po⁵³ kreʔ³⁵

¹khiti

²krip ¹kru ‘cubit’ ⁴ʈu ⁴ɖot

The situation is similar with Written Tibetan and East Bodish. Table 5 shows East Bodish words with the initial kr- or gr-, most of which are found in the Choekhor (CM) dialect of Bumthap (BT). Some of them correspond to a Written Tibetan form with kr- or gr-, but the others do not. Like Tamangic languages, East Bodish languages are presumed to have had a long history of contact with Tibetan (Hyslop 2008) and therefore to contain a large amount of loans from Tibetan; some may be quite recent and the others may be much older. This makes it difficult for us to establish the nature of the genetic relationship between East Bodish and Tibetan.

146   

   Isao Honda

Interestingly, some of the Bumthap velar + r clusters, i.e. ‘dirt,’ ‘six,’ ‘warm,’ and ‘mule,’ correspond to the WT dr-. Michailovsky and Mazaudon (1994: 552) make the following observation: Some PTB dental + r clusters seem to have shifted to velar + r in Common Bumthang as noted by Shafer in Dwags ‘six’ (PTB *d-ruk). But in ‘dirt’ we believe it is the WT which is innovative (PTB *kriy).11

An innovation from the Proto-Tibeto-Burman *velar + r to a dental + r, which is supposed by Michailovsky and Mazaudon (1994) in WT dreg-pa ‘dirt,’ must also be assumed in Tamangic, but not in ‘dirt,’ but in ‘thread,’ if we reconstruct, as Mazaudon (1978) proposes, the PTAM *trupB ‘thread,’ and if the reconstruction of the Proto-Tibeto-Burman *krəw or/and *kriŋ proposed in Matisoff (2003) is correct. It seems, however, that Mazaudon has abandoned the reconstruction of the *dental + r cluster at the Proto-Tamangic stage because in her later articles (e.g. Mazaudon 1996: 105) no *dental + r is reconstructed. In any case, the variation between the velar + r and the dental + r that is found among Bodish languages is an interesting subject.

5 Possible loans from other Himalayan languages Language contact must also have occurred between Tamangic languages and other Tibeto-Burman languages now spoken in Nepal.12 Although the number of 11 Benedict (1972) proposes another Proto-Tibeto-Burman form *ri(y) ‘dirt,’ to which he relates the Written Tibetan form. Meanwhile, Matisoff (2003) reconstructs the PTB *d-ri(y), which appears to be also based on Written Tibetan. 12 Watters suggests that a dozen of lexical items, particularly words related to agriculture and food preparation, must have been borrowed from Thakali to Kham and notes, “This accords with tribal origin myths which assert that agriculture was introduced by the Ghartis, founders of the present-day settlement sites at a time when the area was inhabited only by hunters and herdsmen (the Puns and Budhas, respectively)” (2003: 17). Those include the following Kham (KH) words; KH cika: ‘barley’ (cf. T. Thakali ¹cika; M. Thakali ᴴcikā; Y. Thakali ᴴcyakā; though their cognate is also found in Seke [TA, CH, TE] and Chantyal; also cf. SM ᴸtšišak); KH rãːdəi ‘millet’ (NEP kodo) (cf. Proto-Kham *raŋrəi; T. Thakali ²rāŋre; M. Thakali ᴴrāŋʈe; Y. Thakali ᴴraŋʈe; though their cognate is also found in Seke [TA, CH, TE] and Chantyal); KH 'chiː ‘grass’ (cf. T. Thakaki ᴴchi; M. Thakali [G] ¹chi; Y. Thakali [M] ¹tʃhi; though their cognate is found elsewhere); KH 'gohr ‘plow’ (cf. T. Thakali, Y. Thakali ³kor; though their cognate is also found in Seke [TA, CH, TE], Manang­ ba, and Gurung); KH kum ‘oxen yoke’ (cf. T.  Thakali ᴴkum); KH rãːthəi ‘quern’ (cf. T.  Thakali ³ramtā; M. Thakali ³rindā; Y. Thakali ᴸranta; though their cognate is found elsewhere; also cf. WT raŋ-’thag); KH 'kyoːh ‘ladle’ (No clear cognate is found in my Thakali data; cf. Seke [TA] ᴸku;



Internal diversity in the Tamangic lexicon   

   147

loans borrowed from non-Tibetan Tibeto-Burman languages does not appear to be large, they must be excluded from our word list if we are searching for old retentions. This is also not an easy task due to the lack of clear historical records of linguistic contact situations in the region in the past, and because these languages also have a genetic relationship with Tamangic, although may be more distant than Tamangic’s relationship to Tibetan. Possible borrowings from non-Tibetan Tibeto-Burman languages into Tamangic languages are mainly found in E. Tamang and W. Tamang, which are, geographically, close neighbors of Newar, Thangmi, Kiranti languages, Magar, and Chepang. They are considered as loans mainly because no cognate is found in other Tamangic languages/dialects. In many cases, the most likely source language is Newar if they are indeed loans. Those possible loans include the following Tamang words: E. Tamang (RI) ¹ka ‘thread, yarn’ (cf. C. Newar [NB] kā; K.  Newar kā, sukā; cf. Table 4 for other forms found in Tamangic); E. Tamang (KTI) ᴴkaʈhi ‘stick’ (cf. C. Newar [J] kathi; [NB] kaʈhi, kathi; K. Newar kathi); E. Tamang (RI) ²chyomo and W. Tamang ²chyomaŋ ‘bamboo shoot’ (cf. C. Newar [NB] choma; K. Newar chwã:); E. Tamang (RI) ²moʈhe and W. Tamang ²more ‘soybean’ (cf. C. Newar [NB] mvāra, mvāca; but also Thangmi moʈe; Jirel ᴸmoʈe; Sherpa [K] ᴸmote); E. Tamang (RI) ³pilim ‘cockroach’ (cf. K. Newar bili; C. Newar [J] bili ‘wasp;’ Sunwar [B] bīri; Jirel bĩrima; Chepang bhi.rimʔ, bhi.riŋʔ); E. Tamang (RI) ¹kal- ‘to fry’ (cf. K. Newar kāl-); W. Tamang ³yako ‘many’ (cf. C. Newar [NB] yako, yekva, yeko; K. Newar yekwa; in other Tamangic languages, we find a lateral initial form with a low tone, e.g. E. Tamang [KTI] ᴸlaːna(-n); Gurung ⁴le, or a dental initial form with a low tone, e.g. Chantyal tɦanna; T. Thakali ³tana; M. Thakali [G] ³tan; Y. Thakali ᴸtana; Seke [TA, CH, TE] ᴸtena); E.  Tamang (RI) ⁴pu ‘rice field’ and W. Tamang ‘field’ (cf. C. Newar [J; NB] bu; K. Newar pyā̃ːbu ‘wet field;’ gaŋu bũ ‘dry field;’ but also Thangmi pebu; in other Tamangic dialects, a totally different form is found, e.g. Seke [TA] ²ke; Seke [CH] ᴴkiri; Seke [TE] ᴴkre; Y. Thakali ¹le; M. Thakali [G] ¹le; T. Thakali ᴴke; Chantyal kyeləs, keləs, kyiləs, kyelə; Manangba ke; Gurung ³ɬyo ³nẽ; cf. WT zhing, kluŋs; Kaike khye; cf. E. Tamang [RI] ²mraŋ ‘field;’ Gurung ²mrõ).

Manangba khu); KH ŋoh- ‘roast’ (cf. [3]); KH toh- ‘exchange’ (cf. [21]); KH 'phwi:- ‘pump bellows’ (cf. Proto Kham *phut-; T. Thakali ²phuy-; M. Thakali ¹phuy-; Y. Thakali ¹phui- ‘blow;’ though their cognate is found elsewhere); KH məhr ‘a male of the smith caste’ (Watters considers that this is from a Thakali word ‘gold;’ cf. T. Thakali, M. Thakali, Y. Thakali ⁴mar ‘gold;’ though their cognate is found elsewhere except in Seke where a loan from a Tibetan dialect is used). Another possible loan from a Tamangic language is Chepang saŋa ‘a species of millet,’ which must have been borrowed from Tamang (cf. E. Tamang [KTI] ᴴsaŋa; W. Tamang ²saŋa). In other Tamangic languages/ dialects, no cognate is found.

148   

   Isao Honda

It is possible that E. Tamang (KTI) ᴴmakar ‘monkey’ is also from Newar, cf. C. Newar (NB) mākarha, but both could ultimately be from Sanskrit markaṭaḥ; cf. NEP mākal, markaʈ; Thangmi makar. E. Tamang (KTI) ᴴmoce, W. Tamang ¹moce, and Gurung ⁴maja: ‘banana’ may also have their origins in Indo-Aryan (cf. Sanskrit moca; Magar mocā; Bahing grámochi). E. Tamang (KTI) ᴴkhasu ‘cloud’ and W. Tamang ¹khaso ‘fog’ find a similar form in Thangmi khaśu and Tulung khase, and less transparently, in C. Newar (NB) śu and Sunwar ᴴgossu, all of which mean ‘cloud.’13 In other Tamangic languages/ dialects, we find 1) a form that must be related to WT rmugs-pa ‘a dense fog’ or, more directly, to Southern Mustang ᴴmukpa ‘fog’ and Kyirong mūkpā ‘cloud;’ i.e. W. Tamang ¹mukpa; Manangba mukpa; Seke (TA, CH, TE) ᴴmukpa ‘fog,’ 2) a similar form with a high tone, e.g. Seke (CH, TE) ᴴmutiŋ; Seke (CH) ᴴmu ‘cloud, sky’ (cf. Seke [TA] ᴴmu; Y. Thakali ¹mo; M. Thakali [G], Gurung, W. Tamang, E. Tamang [RI] ¹mu; T. Thakali ᴴmy, all of which mean ‘sky;’ cf. WT rmu-ba, rmus-pa ‘fog’), 3) a similar form with a low tone, e.g. Seke (TA), Y. Thakali ᴸmo; M. Thakali, T. Thakali ³mo (cf. Manangba [Hi] ³mo ‘sky, rain;’ PTB *(r-)muw = (r-)məw ‘clouds, sky’), and 4) Gurung ³nããmjyo ‘mist, fog, cloud.’ E. Tamang (RI) ³prot ‘taste’ (cf. E. Tamang [KTI] ᴸproi ᴴkha- ‘tasty;’ lit. ‘taste comes’) may be related to Jero broto, Wambule broco, Bahing broba ‘fresh, sweet,’ Thulung brəs- ‘taste nice,’ and Sunwar (O) broo-syo ‘savoury’ but is more likely to be directly from Tibetan (WT brod ‘taste’).14 In other Tamangic languages/dialects, a cognate of WT zhim-, Kham lyahm-, Magar (W) lhem- ‘not salty,’ Kurtoep ᴸlembu, and Dakpa lim¹³po⁵³ is found; i.e. Seke (TA) ᴴlim-; Y. Thakali, M. Thakali (G), T. Thakali ²lim-; Chantyal limpa, lĩwa; Gurung ²lĩ-. It is suspected that E. Tamang (RI?; Michailovsky and Mazaudon 1994: 551) ³lapte ‘leaf’ is a loan from Newar (C. Newar [NB] lapate, rapte, K. Newar lapte ‘large leaf’) because its exact meaning is not ‘leaf’ in general but ‘large leaf,’ which is also the case, at least, for the K. Newar word.15 It is, however, also possible that the relation is more profound because of 1) the lax phonation of the E. Tamang word, which is expected to correspond not to /l/ but to /lh/ in Classical and Kathmandu Newar, and 2) the existence of a cognate le̱pte̥ ‘leaf’ in Phu, which I found recently in Hildebrandt and Perry (2011: 172).16 Mazaudon (1985: 13 The lexical similarities found between Thangmi khaśu and C. Newar śu ‘cloud,’ Thangmi pebu and C. Newar bu ‘field,’ and Thangmi moʈe and C. Newar mvāca ‘soybean,’ are suggested in Turin (2004), which explores a possible link between Thangmi and Newar. 14 I would like to thank my reviewer who pointed this out to me. 15 This is based on my own study on E. Tamang (KTI) ᴸlapte. 16 A similar form is also found in Kyirong Tibetan la̱ptī ‘leaf’ and in Gyalsumdo Tibetan [la̱pʈi], which are proposed in Huber (2002) and in Hildebrandt and Perry (2011) to be loans from Ta­ mang­ic languages (the former from Tamang, and the latter possibly from Phu).



Internal diversity in the Tamangic lexicon   

   149

220) does not consider the E. Tamang word to be a loan and reconstructs PTAM *lapteA ‘leaf.’17

6 Final remarks This paper has examined various kinds of lexical variations found in Tamangic languages/dialects and has suggested that those variations arise from several different sources. These include a limited amount of lexical loans from neighboring non-Tibetan Tibeto-Burman languages and a more significant number of loans borrowed from modern Central Tibetan dialects at a relatively recent period. Dialect-independent innovations are also suspected. To get a better idea about that, we need more data from hitherto undescribed Tamangic dialects. Other layers of the Tamangic lexicon must reflect much older contacts with Tibetan or reflexes of old retentions inherited from a common proto-language. Such layers cannot clearly be identified, but further progress can be made when more data from Tibetan and East Bodish is available, and when more comprehensive comparative studies are conducted on the basis of it. As Tibetan specialists repeatedly caution, such studies should not rely too much on so-called Written Tibetan, “a cover term for the written language over a  millennium of use anywhere in the Tibetan cultural region, often including spellings of  no historical value” (Hill 2009: 178–179), but more on forms found in Old Tibetan texts. This is one of the most significant shortcomings of the current investigation, where only Written Tibetan forms compiled in Jäschke’s (1881 [1995]) dictionary are cited and used for reference.

Abbreviations and sources of data In this paper, the following abbreviations for languages/dialects (locations, i.e. village names) and language groups are used (sources of information are also 17 One of my reviewers called my attention to Jacques (2004) which considers that the E. Tamang form is related to Tibetan ’dab-ma; cf. WT ’dab-ma ‘any leaf, a broad leaf’ (Jäschke 1881 [1995]: 274). It should also be noted that two other PTAM forms for ‘leaf’ have been proposed by Mazaudon; i.e. *HlaA (Mazaudon 1985: 220; cf. Gurung ¹la) and *baːA (Mazaudon 1978: 176; cf. E. Tamang [RI], W. Tamang ³paː; Gurung ³po; Manangba [Hi] ³pa; T. Thakali, M. Thakali [G], Y. Thakali, Seke [TA] ³pā; Seke [CH, TE] ᴸpā). Mazaudon (1985: 220, fn. 17) suggests that *H- of the former form derives from a stop initial which is reconstructed at a much older stage, and that this stop might be *b-; cf. e.g. Kurtöp blaʔma.

150   

   Isao Honda

indicated). All of the forms cited from the originals are modified according to phonemic notations adopted here. It should be noted that TA (M), SY (M), MA (M), TU (M), MN (M), which are from Martine Mazaudon’s data, are probably all phonetically described. Tamangic languages/dialects (roughly from east to west and south to north) E. Tamang: Eastern Tamang RI: Risiangku; Mazaudon (1973a) inter alia KTI: Katike; my own data W. Tamang: Western Tamang (SA); Sahu; Hale (1973) inter alia; Mazaudon (1988) (M) Gurung: GH: Ghachok; Glover et al. (1977) Chantyal CT: Noonan et al. (1998) Manangba MN: Hildebrandt (2004) (Hi); Mazaudon (1978) (M); Hoshi (1984) (Ho); my own data T. Thakali: Tamang Thakali; Tukche (TU); Hale (1973); Mazau­ don (1978) (M) M. Thakali: Mawatan Thakali; Marpha (MA); my own data; Georg (1996) (G); Mazaudon (1978, 1988) (M) Y. Thakali: Yhulkasom Thakali; Syang (SY); my own data; Ma­zau­don (1978), (1988) (M) Seke TA: Tangbe; my own data; Mazaudon (1996) (M) CH: Chuksang; my own data TE: Tetang; my own data PTAM: Proto-Tamangic; Mazaudon (1978), (1985), (1988) TAM: Tamangic Tibetan CT: KY: SM: WT: Jirel: Sherpa:

modern Central Tibetan dialects Kyirong; Huber (2002) Southern Mustang; Kretschmar (1995) Written Tibetan; Jäschke (1881 [1995]) Hale (1973) Hale (1973); Kelly (2004) (K)

Other languages/language groups BT: Bumthap CK Choekhor; Bumthap; Michailovsky and Mazaudon (1994) CM: Chume; Bumthap; Michailovsky and Mazaudon (1994)



Internal diversity in the Tamangic lexicon   

   151

C. Newar: Classical Newar; Jørgensen (1936) (J); Malla and Kansakar et al. (2000) (NB) DK: Dakpa; Michailovsky and Mazaudon (1994) KH: Kham: Watters (2004) K. Newar: Kathmandu Newar; Hale (1973) KT: Kurtöp; Michailovsky and Mazaudon (1994) M: Mangdep (or ’Nyenkha); Michailovsky and Mazaudon (1994) NEP: Nepali; Turner (1931) PLB: Proto-Lolo-Burmese PTB: Proto-Tibeto-Burman; Benedict (1972); Matisoff (2003) (M) Bahing: Opgenort (2005) Chepang: Caughley (2000) Jero: Opgenort (2005) Kaike: My own data Khaling: Hale (1973) Magar: Hale (1973); Watters (2004) (W) Sunwar: Hale (1973); Borchers (2008) (B); Opgenort (2005) (O) Thangmi: Turin (2004) Thulung: Opgenort (2005) Wambule: Opgenort (2005)

References Benedict, Paul K. 1972. Sino-Tibetan: A conspectus. Cambridge: Cambridge University Press. Borchers, Dörte. 2008. A grammar of Sunwar. Leiden and Boston: Brill. Campbell, George. 1874. Specimens of languages of India, including those of the aboriginal tribes of Bengal, the central provinces, and the eastern frontier. Calcutta: Bengal Secretariat Press. Caughley, Ross. 2000. Dictionary of Chepang: A Tibeto-Burman language of Nepal (Pacific Linguistics 502). Canberra: The Australian National University. Chao, Yuen-Ren. 1930. A system of tone letter. Le maître phonétique 45: 24–27. van Driem, George. 2001. Languages of the Himalayas: An Ethnolinguistic Handbook of the Greater Himalayan Region (two volumes). Leiden, Boston and Köln: Brill. Georg, Stefan. 1996. Marphatan Thakali. München and Newcastle: Lincom. Glover, Warren. W., J. R. Glover and Deu Bahadur Gurung. 1977. Gurung-Nepali-English Dictionary with English-Gurung and Nepali-Gurung indexes (Pacific Linguistics C–51). Canberra: The Australian National University. Hale, Austin (ed.). 1973. Clause, sentence, and discourse patterns in selected languages of Nepal, part IV: Word lists. Norman, Oklahoma: Summer Institute of Linguistics.

152   

   Isao Honda

Hildebrandt, Kristine A. 2004. A grammar and glossary of the Manange language. In Carol Genetti (ed.). Tibeto-Burman languages of Nepal: Manange and Sherpa (Pacific Linguistics 557), 1–189. Canberra: The Australian National University. Hildebrandt, Kristine A. and Joseph Perry J. 2011. Preliminary notes on Gyalsumdo, an undocumented Tibetan variety in Manang District, Nepal. [Special Issue in Memory of Michael Noonan and David Watters] Himalayan Linguistics 10 (1): 167–185. Hill, Nathan W. 2009. Review of Handbook of Proto-Tibeto-Burman: System and Philosophy of Sino-Tibetan Reconstruction by James A. Matisoff. Berkeley: University of California Press, 2003. Language and Linguistics 10 (1): 173–195. Hodgson, Brian Houghton. 1853. Sifan and Horsok vocabularies, with another special exposition in the wide range of Mongolian affinities and remarks on the lingual and physical characteristics of the family. Journal of the Asiatic Society of Bengal 22: 121–151. Honda, Isao. 2008. Some observations on the relationship between Kaike and Tamangic. Nepalese Linguistics 23: 83–115. Honda, Isao. 2009. Some notes on ‘gold’ and ‘road’ in Zhangzhung and Tamangic. In Yasuhiko Nagano (ed.). Issues in Tibeto-Burman historical linguistics (Senri Ethnological Studies 75), 99–117. Osaka: National Museum of Ethnology. Hoshi, Michiyo. 1984. A Prakaa vocabulary – a dialect of the Manang language. Monumenta Serindica 12: 133–202. Tokyo: Institute for the Study of Languages and Cutures of Asia and Africa, Tokyo University of Foreign Studies. Huber, Brigitte. 2002. The Lende subdialect of Kyirong Tibetan: A grammatical description with historical annotations. Bern: University of Bern dissertation. Huber, Brigitte. 2003. Tone in Kyirong Tibetan: A conservative phonological development. In Tej Ratna Kansakar and Mark Turin (eds.). Themes in Himalayan languages and linguistics, 65–79. South Asia Institute, Heidelberg and Tribhuvan University, Kathmandu. Hyslop, Gwendolyn. 2008. Kurtoep and the classification of the languages of Bhutan. Chicago Linguistic Society (CLS) 44: 141–152. Jacques, Guillaume 2004. The Laterals in Tibetan. Paper presented at the 10th Himalayan Language Symposium, Thimphu, Bhutan, 1–3 December. Jäschke, Heinrich. [1881] 1995. A Tibetan-English dictionary. Delhi: Motilal Banarsidass. Jørgensen, Hans. 1936. A dictionary of the Classical Newārī. København: Einar Munksgaard. Kelly, Barbara. 2004. A grammar and glossary of the Sherpa language. In Carol Genetti (ed.). Tibeto-Burman languages of Nepal: Manange and Sherpa (Pacific Linguistics 557), 193–324. Canberra: The Australian National University. Kretschmar, Monika. 1995. Erzählungen und Dialekt aus Südmustang: Untersuchung zur Grammatik des Südmustang-Dialekts. Bonn: VGH Wissenschaftsverlag. Lu, Shaozun. 1986. Cuònà ménbāyǔ jiǎnzhì. Beijing: Mínzú chūbǎnshè. Malla, Kamal P., Tej R. Kansakar et al. (eds.). 2000. A dictionary of Classical Newari. Kathmandu: Nepal Bhasa Dictionary Committee, Cwasā Pāsā. Matisoff, James A. 2003. Handbook of Proto-Tibeto-Burman: System and philosophy of Sino-Tibetan reconstruction. Berkeley: University of California Press. Mazaudon, Martine. 1973a. Phonologie Tamang. Paris: Société dʼÉtudes Linguistiques et Anthropologiques de France. Mazaudon, Martine. 1973b. Comparison of six Himalayan dialects of Tibeto-Burmese. Pakha Sanjam 6: 78–91. Mazaudon, Martine. 1978. Consonantal mutation and tonal split in the Tamang subfamily of Tibeto-Burman. Kailash 6 (3): 157–79.



Internal diversity in the Tamangic lexicon   

   153

Mazaudon, Martine. 1985. Proto-Tibeto-Burman as a two-tone language?: Some evidence from Proto-Tamang and Proto-Karen. In Graham Thurgood, James A. Matisoff and David Bradley (eds.). Linguistics of the Sino-Tibetan area: The state of the Art (Pacific Linguistics Series C–87), 201–229. Canberra: Department of Linguistics, The Australian National University. Mazaudon, Martine. 1988. The influence of tone and affrication on manner: Some irregular manner correspondences in the Tamang group. Paper presented at the 21st International Conference on Sino-Tibetan Languages and Linguistics, Lund, 5–9 October. Mazaudon, Martine. 1996. An outline of the historical phonology of the dialects of Nar-Phu (Nepal). Linguistics of the Tibeto-Burman Area 19 (1): 103–14. Mazaudon, Martine. 2003. Tamang. In Graham Thurgood and Randy LaPolla (eds.). The Sino-Tibetan languages, 291–314. London and New York: Routledge. Mazaudon, Martine. 2005. On tone in Tamang and neighbouring languages: Synchrony and diachrony. In Kaji Shigeki (ed.). Proceedings of the Symposium, Cross-Linguistic Studies of Tonal Phenomena: Historical Development, Tone-Syntax Interface, and Descriptive Studies, 79–96. Tokyo: Institute for Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies. Michailovsky, Boyd and Martine Mazaudon. 1994. Preliminary notes on the languages of the Bumthang group. In Per Kvaerne (ed.). Tibetan studies: Proceedings of the 6th Seminar of the International Association for Tibetan Studies, Fagernes, vol 2, 545–557. Oslo: The Institute of Comparative Research in Human Culture. Nishi, Yoshio. 1982. Five Swadesh 100-word lists for the Ghale language – A Report on the trek in the Ghale speaking area in Nepal. Monumenta Serindica 10: 158–194. Nishi, Yoshio. 1990. Can fowls fly hundreds of miles over the Himalayas? In Osamu Sakiyama, Akihiro Sato and Masahiro Aoki (eds.). Ajia-no Shogengo-to Ippan Gengo-gaku [Asian languages and general linguistics], 55–77. Tokyo: Sanseido. Nishi, Yoshio. 1991. Himaraya-shogo no bunpu to bunrui [The distribution and classification of Himalayan languages (part II)]. Bulletin of the National Museum of Ethnology 16 (1): 31–158. Noonan, Michael, Jag Man Chhantyal and William Pagliuca. 1998. Chantyal dictionary and texts. Berlin: Mouton de Gruyter. Opgenort, Jean Robert. 2005. A grammar of Jero. Leiden and Boston: Brill. Pittman, Richard and Jessie Glover. 1970. Proto-Tamang-Gurung-Thakali. In Austin Hale and Kenneth L. Pike (eds.). Tone systems of Tibeto-Burman languages of Nepal, Part II: Lexical lists and comparative studies (Occasional Papers of the Wolfenden Society on TibetoBurman Linguistics, Volume 3), 9–22. Urbana: Department of Linguistics, The University of Illinois. Shafer, Robert. 1955. Classification of the Sino-Tibetan languages. Word 11: 94–111. Shafer, Robert. 1966. Introduction to Sino-Tibetan, Part I. Wiesbaden: Otto Harrassowitz. Turin, Mark. 2004. Newar-Thangmi lexical correspondences and the linguistic classification of Thangmi. Journal of Asian and African Studies 68: 97–120. Turner, Ralph L. [1931] 1996. A comparative and etymological dictionary of the Nepali language. Delhi: Allied Publishers. Watters, David. 2002. A grammar of Kham. Cambridge: Cambridge University Press. Watters, David. 2003. Some preliminary observations on the relationship between Kham, Magar, (and Chepang). Paper presented at the 36th International Conferences on Sino-Tibetan Languages and Linguistics, Melbourne, 28–30 November. Watters, David. 2004. A dictionary of Kham: Taka dialect (A Tibeto-Burman language of Nepal). Kathmandu: Central Department of Linguistics, Tribhuvan University.

Gwendolyn Hyslop

A preliminary reconstruction of East Bodish1 1 Introduction It has often been assumed that the languages of Bhutan are simply dialects of Tibetan (e.g. most of the Bhutanese languages listed in the Ethnologue, Lewis 2009, are classified as “Tibetic”). While Dzongkha, the national language of Bhutan, is linguistically a Tibetan language (Tournadre 2008 inter alia), most of the approximately 18 languages are not, as has been shown by Michailovsky and Mazaudon (1994), DeLancey (2008), Hyslop (2008a) and discussed in van Driem (2001) and elsewhere. An overview of the languages of Bhutan is provided in van Driem (1998) asserting that the languages of Bhutan belong to the Tibeto-Burman family, but a rigorous study of the placement of these languages remains outstanding. The aim of this article is to advance our discussion on the placement of one group of these languages, namely the East Bodish languages. The center of gravity for these languages is Bhutan, but they are also spoken in some areas of Tibet and Arunachal Pradesh. The term “East Bodish” was used by Shafer (1954) to identify a proposed family of languages to which Dakpa, a language spoken southeast of Lhasa, belonged. Shafer (1954) noted that Dakpa, and therefore “East Bodish” languages were closely related to, but not directly descended from Old Tibetan. Since then, several other related languages have been identified, though only a few have ever seen data in print (Bumthap: van Driem 1995; Kurtöp: Busch 2007, DeLancey 2008, Hyslop 2008a, 2009, 2011a, b; Dakpa: Hyslop and Tshering 2010; Dzala: Balodis 2009, Genetti 2009; Mangdep: Nishida 2009; comparisons: Michailovsky and Mazaudon 1994, van Driem 2007, Hyslop 2008b). Van Driem (1998) identifies the following East Bodish languages:2 ’Nyenkha, Bumthap, Kurtöp, Dzala, Dakpa, Chali, Khengkha. There is still little work on the subfamily as a whole, but Hyslop (2013) does provide some evidence that links the languages together in one family. 1 The ideas presented here have grown out of fruitful discussion with Karma Tshering, George van Driem, Nathan W. Hill, Scott DeLancey, and Roger Blench. Tom Owen-Smith provided considerable assistance in helping to finalize this chapter from the field. I am also grateful to two anonymous reviewers for their comments on an earlier version of this chapter. However, I retain responsibility for all material in this article. 2 Van Driem (1998) also mentions Black Mountain Mönpa as an “archaic” East Bodish language but later proposes to remove it from the family, considering it a Tibeto-Burman isolate (van Driem 2011).

156   

   Gwendolyn Hyslop

While some scholars have attempted to address the placement of East Bodish within Tibeto-Burman in general terms, there has yet to be any systematic study of the historical placement or development of East Bodish languages. Rather than continue to compare lexical items or grammatical forms – which may easily be borrowed from Dzongkha or Tibetan (languages which enjoy particular prestige in Bhutan) – from a given East Bodish language, we believe a more fruitful approach is to begin with a preliminary reconstruction of Proto East Bodish. Reconstructed forms can then be compared with other languages for an eventual understanding of the placement of the East Bodish languages within the TibetoBurman family. Bhutan, and probably neighboring Buddhist areas in Tibet and Arunachal Pradesh, have been under heavy and continual influence from Classical Tibetan, which serves as the liturgical language for practitioners of Buddhism in the area. In Bhutan, Dzongkha, the national language, has also been an added source of influence for at least the past hundred years. The influence is so heavy, and multilingualism so rampant, that discerning borrowings from native structures is nearly impossible through superficial study. In order to better understand the relationship between Tibetan and the East Bodish languages, we should first compare individual languages within the East Bodish family with each other. For example, if we find that one East Bodish language has a lexical form that is similar or identical to a Tibetan form, while the rest of the East Bodish languages share another form, we can assume the Tibetanlike form has been borrowed into that particular East Bodish language. This is especially true if we see the non-Tibetan reflex exhibits regular sound correspondences among the other languages. Of course, we may well reconstruct forms that were borrowings into Proto East Bodish from Tibetan, in which case it will be more difficult to discern borrowings from true cognates. In any case, the first step is a reconstruction of Proto East Bodish (PEB). Thus, we propose to address the placement of the East Bodish languages through historical reconstruction of Proto East Bodish by way of the daughters. Many data are still missing; the East Bodish languages remain, by and large, undescribed; detailed lexica, phonological and grammatical descriptions are missing; and this article is the first attempt at any reconstruction. This chapter has the following organization. Section 2 provides an overview of the East Bodish languages, including a preliminary phylogeny. Section 3 discusses the relationship between East Bodish and Central Bodish, offering evidence in support of previous observations that the East Bodish languages are not dialects of Tibetan. The next three sections provide the main body of this article, focusing on a comparison of the East Bodish data and what we can reconstruct to Proto East Bodish. In Section 4 we focus on lexical data; in Section 5 we focus on personal pronouns; and in Section 6 we discuss morphology. Section 7 offers a brief conclusion.



A preliminary reconstruction of East Bodish   

   157

2 East Bodish languages There are at least seven East Bodish languages, summarized in Table 1. For the most part, the designation of these languages is straightforward.3 The approximate geographic location of the East Bodish languages in shown in Figure 1. Tab. 1: East Bodish languages Language

Other names

Estimated number of speakers

Dakpa Dzala

Dwags, Tawang/Northern Monpa Kurtöp, ’Yangtsebikha, i ga brok

50,000 20,000

Chali Phobjip Kurtöp Bumthap Khengkha

Chalipakha Hengke, Phobjikha Zhâke, au gemale Bumthang, Bumthabikha, Monpa Kheng

1,000 3,000 15,000 30,000 40,000

There has been some work on the internal phylogeny of the East Bodish languages. The first observation, that Dakpa and Dzala form a separate subgroup within East Bodish, was first noted in van Driem (2007). Previously Aris (1979) noted the close relationship between Kurtöp, Bumthap, and Khengkha. Building from there, Hyslop (2013) offers a preliminary proposal for the internal phylogeny of East Bodish, as shown in Figure 2 there is sufficient evidence to show that Dakpa and Dzala form a subgroup separate from the subgroup formed by Kurtöp, Khengkha and Bumthap, but the precise placement of Chali and Phobjip is less certain.

3 However, the language Mangdep, described as East Bodish in van Driem (2001), is not included in the list. Instead, we consider Phobjip described by van Driem (2001) to be a divergent dialect of Mangdep. The decision to use Phobjip rather than Mangdep is based on the data we have collected. Our preliminary work with a speaker of Mangdep from Trongsa suggested the language was a dialect of Bumthap, while the data from speakers of Hengke, or Phobjip from Phobjikha clearly indicated a separate language. Indeed, this may turn out to be due to misunderstandings in terminology and designations and not counterevidence to the classification of Mangdep as a language separate from Bumthap. Therefore, the most prudent approach, is to address the Phobjip data but omit discussion of Mangdep in this article. It may turn out that the generalization made for Phobjip can be applied to Mangdep as well, but a more detailed study of Phobjip is needed, perhaps in comparison with Nishida (2009).

158   

   Gwendolyn Hyslop

Tibet

Kurtöp Dzala Bumthap

Nepal

Bhutan

Chali

Phobjip

Khengkha

India Bangladesh

Fig. 1: Approximate geographic location of East Bodish languages

East Bodish

? ?

Dakpa Dzala Phobjip Chali Bumthap Khengkha Kurtöp Fig. 2: Internal phylogeny of East Bodish languages (Hyslop 2013)

Dakpa



A preliminary reconstruction of East Bodish   

   159

The diagram in Figure 2 is based, in part, on several sound changes, which are summarized in Table 2. As we will see below, the sound changes are not exceptionlesss; more work on the internal phylogeny of the East Bodish languages is beyond the scope of this article. Tab. 2: East Bodish internal sound changes (from Proto East Bodish) Sound change

Languages

a>e

Dakpa, Dzala

l>j

Chali, Kurtöp, Bumthap, Khengkha

khw > ɸ

Khengkha

o,u,a,e > ø,y,e,i/ _t

Phobjip

kr, khr, gr > ʈ, ʈh, ɖ

Kurtöp (recent sound change; earlier kr > r (high tone); perhaps also happened elsewhere in East Bodish)

A fully reconstructed phonology of Proto East Bodish is still beyond the scope of this article (though obviously crucial for an eventual full reconstruction), pending full phonemic analyses of more East Bodish languages. Based on what has been presented in van Driem (1995), Michailovksy and Mazaudon (1994), Hyslop (2011a), Hyslop and Tshering (2010) and our field notes, we can offer the following statements. All East Bodish langauges make a three-way voicing (voiceless, voiceless aspirated, voiced) contrast in stops and a two-way voicing contrast in fricatives, with alveolar and palatal fricatives. Modern East Bodish languages have minimally five vowels (e.g. Kurtöp: Hyslop 2011a) but maximally at least 9 (Dakpa: field notes). All East Bodish languages appear to have the same tonal system: contrastive high/low tone following sonorants.

3 East Bodish versus Central Bodish Shafer (1955) posits a “Bodish” branch within Tibeto-Burman, identifying “Southern,” “Central,” “West,” and “East” Bodish languages. The West Bodish languages are represented by Balti and Spiti, which Shafer assumed to be directly traced to older forms of Tibetan. Central and South Bodish languages, on the other hand, were considered to be direct descendants of Old Tibetan. Hill (2010a) recently updated this model, arguing that there is currently no evidence for “West” vs. “Central” vs. “South” and all three can be subsumed under one parent, Old Tibetan, which itself

160   

   Gwendolyn Hyslop

was a sister to Proto East Bodish. For the purposes of this article, we retain the term “Central Bodish” to capture the many modern Tibetan languages which can be shown to have descended from Old Tibetan, and are thus distinct from the East Bodish languages, which are not directly descended from Old Tibetan. Old Tibetan has given rise to dozens of different languages and dialects. The many Tibetan languages have received considerable attention in the literature and an overview of the many important and high quality works is beyond the scope of this article (though see Hyslop to appear for a brief summary, and Tournadre this volume). There has also been considerable synchronic and diachronic study of Old Tibetan itself, such as Li (1933), Coblin (1976), Beckwith (1996), Hill (2010a, 2010b, 2011a, 2011b), and Jacques (2010, 2012), amongst many others. Tournadre (2008) identifies 25 Tibetan dialects and several diagnostics which can be used to identify a Tibetan language. A brief comparison of several Tibetan words with Kurtöp, for example, easily suffices to show that Kurtöp (and therefore the other East Bodish languages) are not dialects of Tibetan. This has already been shown by Shafer (1954), Michailovsky and Mazaudon (1994), DeLancey (2008), Hyslop 2008a) but some data are shown in Table 3 to illustrate the point again. Kurtöp data are presented here in The local orthography described in Hyslop (2011a). Tab. 3: Tibetan (Central Bodish) vs. Kurtöp (East Bodish) Tibetan (diagnostic by Tournadre) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20)

gcig gnyis bzhi bdun bcu btsong ‘bras ‘brog-pa chu dmar(-po) gzhu khrag lcags lo mda’ mye mgo nag(-po) nye phyi

Kurtöp

Gloss

thek zon ble ’nis che ’mui mras nakpo khwe zhinti limi kak ’la: ’neng mya gami guyung nyunti chan(do) bi(to)

‘one’ ‘two’ ‘four’ ‘seven’ ‘ten’ ‘sell’ ‘paddy’ ‘pastoralist’ ‘water’ ‘red’ ‘bow’ ‘blood’ ‘iron’ ‘year’ ‘arrow’ ‘fire’ ‘head’ ‘black’ ‘near’ ‘outside’



(21) (22) (23) (24) (25) (26) (27) (28) (29) (30)

A preliminary reconstruction of East Bodish   

‘phur rdo rkang-pa shi shig shing skudpa so sran-ma srab-(po)

ling gor tawa set se seng ’rotman kwa neme ngapmi, pratmi, ’latmi

   161

‘fly’ ‘stone’ ‘leg’ ‘die’ ‘louse’ ‘wood’ ‘thread’ ‘tooth’ ‘pea’ ‘thin’

Many of the Kurtöp forms are clearly cognate with the Tibetan counterparts. This is true of the words for ‘one,’ ‘four,’ ‘die,’ ‘louse,’ and ‘wood.’ Other forms appear to be related, but the relationship is not as straightforward. For example, Kurtöp and Tibetan ‘paddy,’ ‘outside,’ ‘blood,’ ‘iron,’ and possibly ‘bow’ share many phonetic elements, but we are not yet able to discern whether a regular correspondence exists. Other Kurtöp and Tibetan forms appear to be the reflexes of entirely separate roots. This is obviously the case, for example, for ‘two,’ ‘sell,’ ‘red,’ ‘leg,’ ‘tooth,’ and ‘thin.’ In some cases, it appears as though the Tibetan forms may be palatalized versions of the Kurtöp forms; for example, Tibetan gcig ‘one’ and bzhi ‘four’ could be palatalized versions of Kurtöp thek and ble, respectively (the latter possibly an example of Hill 2011b’s ‘Benedict’s law’). Of course, whatever relationship exists is likely to be complicated (e.g. consider the presence of g- in Tibetan ‘one,’ lacking in Kurtöp). While Kurtöp and the East Bodish languages share a relationship with the Tibetan dialects, these data serve to illustrate that Kurtöp (and therefore the East Bodish languages) are not Tibetan dialects themselves, as has already been noted above. Apart from the lexicon, Kurtöp and Tibetan appear to share – in (mostly) form and function – the genitive/ergative, locative, perfective ‑pa and negative prefix (DeLancey 2008). However, a deeper analysis reveals that much of this is due to borrowing (Hyslop 2011a). For example, the allomorphy of the perfective ‑pa in Kurtöp does not reconstruct within Kurtöp, let alone within Kurtöp’s closest neighbors. Instead, it seems that some dialects of Kurtöp have innovated/borrowed this allomorphy. The seemingly shared genitive/ergative -gi does not reconstruct to the parent language shared by Kurtöp and its closest linguistic relatives, Khengkha and Bumthap, and therefore is unlikely to be an innovation shared by the Central and East Bodish languages. Because of this superficial morphological cognacy, it would be more prudent to first ascertain what reconstructs within East Bodish, and then offer a comparison of these forms to Tibetan. With lexical evidence that Kurtöp and Tibetan cannot be sister languages, we can proceed with an examination of forms found within the East Bodish languages, as the first attempt at reconstruction of any aspect of Proto East Bodish (PEB).

162   

   Gwendolyn Hyslop

4 Lexical items With detailed analysis of most East Bodish languages still outstanding, the discussion at the lexical level is necessarily limited to the data we have available. Nonetheless, there are a few words for which we have cognates for all or many of the East Bodish languages. We will first consider a comparison of lexical items in the East Bodish languages. For ease of comparison, let us examine the correspondences based on semantic domain. In Table 4, known forms for various body part terminology are shown. Tab.4: East Bodish body part terms4   Gloss ‘tooth’ ‘hair’ ‘blood’ ‘waist’ ‘hand’ ‘nose’ ‘knee’ ‘bone’ ‘head’ ‘eye’ ‘leg’

PEB5 *kwa *krɑ *kak *khrɑt *lɑk *nɑ́ ?*pOskOm *rOs *gO*mE-

Da wɑ́

khret lɑ́ː

ruspa* gokti meloŋ lɛmin

Dz 6

wɑ́ ʈɑ keːʔ ʈhet lɑː nɑ́7 pukum rý8 gokte meːʔ leme9

  Ku

Bu

Kh

kwa rɑ́ kɑː khrɑt jɑː nɑ́ pusum rospɑ gujuŋ miː tɑwɑ

kwa krɑ kɑk khrɑt jɑk nɑ́phɑŋ

kwa krɑ kɑː ʈhɑt jɑː nɑbli putmoŋ rotpɑ gujuŋ mek tɑwɑ

gujuŋ

Ch

Ph

rɑ́

wɑ́ ~ ó rɑ́

ketpɑ jɑʔ

lɑː

gujuŋ mé;ʔ

pøhø rotho gunu

We can readily reconstruct forms for ‘tooth,’ ‘hair,’ ‘blood,’ ‘waist,’ ‘hand’ and ‘nose.’ It is clear that, from the data currently available, we can approximate

4 An asterisk (*) in the Dakpa or Dzala column indicates the data were drawn from van Driem (2007). In some instances representation of data have been changed in order to be represented in the International Phonetic Alphabet for the purposes of this paper. 5 The capital indicates a back, round vowel but with uncertain height – further data and analyses are required to assign height quality to the reconstruction. Similarly, indicates a front vowel with unknown height, pending future research. 6 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara. 7 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara. 8 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara. 9 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara.



A preliminary reconstruction of East Bodish   

   163

reconstructions of ‘knee,’ ‘bone,’ ‘head,’ and ‘eye.’ It is clear that ‘bone,’ ‘head,’ and ‘eye’ share the same initial syllable. It would appear that ‘knee’ reconstructs to two syllables in Proto East Bodish, but we cannot yet be certain about the form of the second syllable. A final labial nasal is present in the Dzala and Kurtöp reflexes, suggesting it is relatively old, but Khengkha has final -ŋ and no final nasal is found in the Phobjip reflex. Phobjip has lost many final consonants (and vowels), so the absence of -m is perhaps not surprising. What is yet to be explained, however, is the final rounded vowel in the Phobjip form. Either a coronal coda was present in an older stage of the language, or the rounding is due to assimilation to the vowel in the first syllable. Note also the Dakpa and Dzala reflexes for ‘hand’ and ‘nose’ do not show the change a >e as expected, for reasons as of yet unknown. Turning to animals, Table 5 illustrates the comparative East Bodish terms for various animals, again based on the data that is currently available to us. Of the eight forms, we can confidently reconstruct six, including the first syllable for the word for ‘hen.’ Looking only at three languages (Dakpa, Dzala, Kurtöp), we have three separate roots for ‘mouse,’ making reconstruction impossible at this stage. Similarly, the forms for ‘snake’ suggest two different roots. It is worth noting that a Kurtöp form mreka indicates ‘smile shape,’ which is likely related to the Dzala form. However, whether this form or one with a labial stop initial should be reconstructed to Proto East Bodish is still impossible to know. Tab. 5: East Bodish animals  Gloss

PEB

‘otter’ ‘horse’ ‘hen’ ‘bear’ ‘dog’ ‘mouse’ ‘deer’ ‘snake’

*krɑm *tɑ *khɑ*wɑm *khwi *khɑçɑ

  

Da

Dz

Ku

Bu

Kh

Ch

Ph

te khɑwɑ wɑm khi ʝu khɑçɑ

rɛ́m te khɑː10 wəm khi mɑtsɑŋmɑ khɑçɑ mrekɑliŋ

rɑ́m tɑ



krɑm tɑ khɑgɑ

tɑ khɑgɑ

tɑ khom

khwi

ɸi

khwi

chy

kɑçɑ11 po12

khɑçɑ

wɑm khwi ŋíjɑ khɑçɑ po

by

10 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara. 11 Data from Yangdzom and Arkesteijn (1996). Note that it is possible that the lack of aspiration indicated by this form is either the result of 1) error in transcription; or 2) confusion with the Dzongkha form kɑçɑ. 12 Data from Yangdzom and Arkesteijn (1996).

164   

   Gwendolyn Hyslop

A short list of comparative East Bodish verbs is shown in Table 6. Note that we can readily reconstruct forms for ‘give,’ ‘come’ and ‘pour’ to Proto East Bodish. We have less data for ‘sleep,’ but the fact that cognate forms are found in Dzala, Kurtöp and Phobjip (with vowel fronting as we would expect) suggests we can confidently reconstruct *dot as the verb ‘sleep’ to Proto East Bodish. With regard to the verb ‘eat,’ we suspect the Khengkha form is innovative and thus a form with a voiced fricative can be reconstructed to Proto East Bodish. Several forms have a high back vowel here, but not all, making it difficult to reconstruct. To further complicate matters, the Dakpa reflex undergoes stem alternation, depending on as of yet unknown factors (see Hyslop and Tshering 2010: 15–16), and it is not yet possible to establish which is the more basic form between the two. Finally, one will note the several roots used for the verb ‘go,’ All languages except Chali have a form with a voiced velar initial and either a diphthong ai or a mid front vowel e. We have shown elsewhere that Kurtöp ai > e can be the result of sound change ultimately from -al (Hyslop 2008a) and in Hyslop (2011a) I show that verbs which take the -le form of the imperative originally had an -l coda. In Kurtöp, the verb ge ‘go’ is a word that takes the -le form of the imperative. This, combined with the fact that the root appears with a diphthong elsewhere in the family, is suggestive of a coda -l present at the Proto East Bodish stage. Other roots for ‘go’ in East Bodish are go, i and brok. The form i is found in other languages of central Bhutan, including Tasha-Sili (where it occurs as e; see Hyslop and Tshering 2009). The form brok in Dzala appears to be innovative and unique to Dzala. Tab. 6: East Bodish verbs Gloss

PEB

Da

Dz

Ku

Bu

Kh

Ch

‘eat’ ‘come’ ‘go’ ‘do’ ‘pour’ ‘sleep’ ‘give’

*zV *rɑ *gɑl

zo ~ ze

cɑp

gai bu jok

gai bu jo

zu rɑ i

bi

lok dø bi

zu rɑ ge, go ŋɑk jok ~ joː dot bi

zu

gai

zau rau brok

*lok *dot *bi

bi

Ph rɑ go, ge, i kjɑp dø

East Bodish words for the colors ‘red,’ ‘white’ and ‘black’ are shown in Table 7. As will be noted, we can only confidently reconstruct the word for ‘white.’ There is a chance that the Dakpa/Dzala forms for ‘red’ are related to the forms found in the Bumthap group, if the change l > ʝ can be postulated. Similarly, there is a chance that the Kurtöp form for ‘black’ is related to the Dakpa/Dzala forms,



   165

A preliminary reconstruction of East Bodish   

perhaps assuming the change ml > mj > ɲ. In both cases, more work is clearly needed. We are not sure what caused the loss of aspiration in the Phobjip form, though it could be under influence from Dzongkha, which has kɑːp ‘white.’ Tab. 7: East Bodish colors 



Gloss

PEB

Da

Dz

Ku

Bu

Kh

‘red’ ‘white’ ‘black’

*khɑr

lju khɛ́r mlɛŋbu*

lew13 kheru mlæŋ:pu*

ʝinti khɑrti ɲunti

ʝinti14

ɟinti khɑrti

Ch

Ph kɑt

In Table 8 we illustrate some comparative East Bodish nouns. Forms for ‘rock’ and ‘arrow’ reconstruct straightforwardly. The proto nature of three forms requires additional speculation, and for six of the eleven we are not able to offer a reconstruction. At first glance, it appears that the Dakpa/Dzala root for ‘water’ is innovative, compared to the forms in the other languages. However, the Dakpa/Dzala form tshi is found in Kurtöp as a word meaning ‘stickiness’ or ‘sap,’ and Yangdzom and Arkesteijn (1996) list tshi as the Khengkha word for water. As Khengkha is spoken in a large area by a relatively large population (approx. 40,000 speakers), presumably there is an as-of-yet unexplored dialect diversity, reflected also in these different reflexes for ‘water.’ The point, though, for this digression is simply to highlight the fact that there appear to be two roots in Proto East Bodish that have since come to have the meaning ‘water’ in East Bodish. It is not possible at present to discern what semantic differences were held between *khwe and *tshi in the proto language. All the languages for which we have data show similar forms for ‘wind.’ In the case of monosyllabic reflexes the form consists of either a rhotic or lateral plus a rounded back vowel. Kurtöp, Bumthap, and Phobjip have disyllables for the reflex, but the second syllable in each of these instances contains a lateral plus rounded back vowel. The initial syllables, however, vary across the languages. There is also some similarity in the forms for ‘baby,’ with only Dzala to be completely different. We tentatively propose to reconstruct a monosyllable with a back rounded vowel and coda nasal for this root. The reconstruction of the remaining six words in Table 8 is more problematic. At first glance it appears the words for ‘snow’ would be related. However, 13 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara. 14 Bumthap data here are from Driem (1995).

166   

   Gwendolyn Hyslop

the Dakpa/Dzala forms have aspiration whereas the forms in the Bumthang group do not. Recall that two forms were reconstructed for Proto East Bodish ‘water,’ one of which consisted of an aspirated labial and labiovelar glide onset cluster. Reflexes for this root appear in all the languages except Dakpa/Dzala, leading us to suspect that the forms for ‘snow’ are actually reflexes of PEB ‘water.’ If so, then they could not be reflexes of the same form represented by the Kurtöp, Bumthap, and Khengkha forms. In the reflexes for ‘fire,’ there is a shared syllable me/i found across the languages for which we have data, but it is not yet clear whether the first syllable ga found in the Kurtöp, Bumthap, and Khengkha forms is innovated or conservative. We might expect gami to be innovative, given the preponderance of forms for ‘fire’ lacking ga in neighboring Himalayan languages. However, before looking outside East Bodish it will be important to look fully within the language to ensure that a form with a ga initial syllable is not to be reconstructed to Proto East Bodish (and subsequently lost in some daughter langauges). There are clearly two separate roots represented in the forms for ‘house.’ We suspect the form represented in languages of the Bumthap group is conservative, with the Dakpa and Dzala form being a borrrowing from Tibetan khyim, but more data is needed to confirm this. The final four words in Table 8 exhibit three roots each. There is little doubt that a form for ‘chili’ cannot be reconstructed to Proto East Bodish. The chili, being only recently brought to Bhutan, is surely younger than Proto East Bodish would be, and therefore the words found in the languages today are independent innovations, which may or may not be useful for further internal subgroupings. It might be worth pointing out, at any rate, that the Chali and Phobjip reflexes (chau and chou, respectively) are clearly cognate with Kurtöp chɑwɑ ‘Sichuan pepper,’ reminiscent of the Dzongkha-Tibetan phenomenon, wherein ’ema, the Tibetan word for ‘Sichuan pepper,’ was used in Bhutan to refer to ‘chili.’ In the four languages for which we have forms for ‘wife,’ we see three obviously different forms, not allowing us to reconstruct a form to the proto language. Finally, there are three obviously different words for ‘sun’ in the East Bodish languages for which we have data. Again, it is difficult to know which root is more original. The root found in the Bumthang languages is most obviously Tibeto-Burman (looking similar to Proto-Tibeto-Burman15 *g-nam), but we have not been able to rule out the possibility that this form has been borrowed into the Bumthang group. The Dakpa/ Dzala root is strikingly similar to Lepcha plang ‘to rise’ (Plaiser 2007), which perhaps indicates it is an older Tibeto-Burman form for which Dakpa/Dzala 15 Proto-Tibeto-Burman forms are drawn from the STEDT database (http://stedt.berkeley. edu/~stedt-cgi/rootcanal.pl) in June, 2011.



   167

A preliminary reconstruction of East Bodish   

have an innovated use. The Chali form is perhaps similar to Tamang ‘tihnyi and Thakali tihngi (Hale 1973), and/or Boro san (personal field notes). Whatever the relationship of the Chali form outside of East Bodish, we are still unable to ascertain which root (or roots) to reconstruct to Proto East Bodish without further research. Tab. 8: East Bodish nouns 

  

Gloss

PEB

Da

Dz

Ku

Bu

Kh

Ch

Ph

‘arrow’ ‘stone’ ‘water’

*mlɑ *gor *khwe/ *tshi *rO? *On?

mlə* gor tshi

mləh* gor16 tshi

mja gor khwe

gor khwe

ɸe

khwe

gor khö

rót bikikaʊ17 khɑwɑ mé* khem*

ʝiluŋ óŋɑ kɑ gɑmi me bɑŋgɑlɑ nésɑŋ ne

‘wind’ ‘baby’ ‘snow’ ‘fire’ ‘house’ ‘chili’ ‘wife’ ‘sun’

ɔnu kho me* khɛm* sol muip* plɑŋ

prɑŋ

ʝoloŋ

luŋ lú ɲulu olo úp kɑ kɑ gɑmi gɑmi me mai bɑŋgɑlɑ bɑŋgɑlɑ chau chou némo keme ni ni thanman

Table 9 illustrates the few time words for which we have a majority of comparative East Bodish data. It will be clear that the Proto East Bodish form for ‘year’ is easy to reconstruct, as is the form for ‘yesterday,’ which has second syllable in some but not all daughter languages. It appears as though the form for ‘today’ should reconstruct with an initial da- syllable, but without more data it will be impossible to speculate further. Finally, the data show at least two disparate roots for the word ‘tomorrow,’ not allowing us to reconstruct one form over another to Proto East Bodish. Again, we see that the Dakpa and Dzala forms for ‘yesterday’ do not have an e vowel, as we would expect.

16 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara. 17 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara.

168   

   Gwendolyn Hyslop

Tab.9: East Bodish time words    Gloss

PEB

Da

Dz

Ku

‘yesterday’

*dɑŋ

dɑŋ

dɑŋ

‘year’ ‘today’ ‘tomorrow’

*néŋ *dɑ-

níŋ dɛç nogor

níŋ

dɑŋniŋ/ dɑŋnɑ néŋ dɑsum jɑmpɑ

noŋor18

Bu

Kh

néŋ dusum jɑmpɑt

Ch

Ph

dɑŋlɑ

dõl

ɲéŋ dɑsum jɑmpɑ

né: dɑsu nɛmbɑ

In Table 10 we show comparative forms for question words ‘what,’ ‘where’ and ‘who.’ It appears that there are two roots for each form, making reconstruction difficult. For ‘what,’ the Phobjip and Dzala forms appear similar while Chali and the languages of the Bumthang group have a different form. The fact that affrication or frication exists in the forms for all the languages may turn out to be important – that is – signal evidence of one shared proto form, but there is not enough evidence at present to reconstruct one form to Proto East Bodish. There are clearly two separate roots in the forms for ‘where’ and ‘who.’ Chali shares a form with Dakpa/Dzala for ‘where’ while it appears that Dakpa/Dzala have a unique form for ‘who’ not shared by the languages. At present, we are unable to reconstruct any question words to Proto East Bodish. Tab. 10: East Bodish question words    Gloss ‘what’ ‘where’ ‘who’

PEB

Da

Dz 19

su

di (dzi) gɑ su

Ku

Bu

Kh

Ch

Ph

ʝɑ áu é:

ʝrɑ áu

ɟɑ áu ae

ɟɑ gɑŋ

ɑ́çi údɑ ɛ

Unlike the question words and other forms we have discussed, the numerals, it turns out, are fairly easy to reconstruct for Proto East Bodish. With the exception of ‘two’ and ‘ten,’ the EB numerals reconstruct straightforwardly, as shown in Table 11.

18 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara. 19 This Dzala form was taken from a lexical database produced by the 2008–2009 Field Methods class at the University of California, Santa Barbara.



A preliminary reconstruction of East Bodish   

   169

Tab. 11: East Bodish numerals Gloss

PEB

Da

Dz

Ku

Bu

Kh

Ch

Ph

‘one’ ‘two’ ‘three’ ‘four’ ‘five’ ‘six’ ‘seven’ ‘eight’ ‘nine’ ‘ten’ ‘twenty’

*thek

thi nei sum bli lɛŋe krò nís kɛ̀t dugu ciŋnəi khəli

the nai sum bri lɛŋe ɖoɁ ní gɛt dugu cí khɑlit

the: zon sum ble jɑŋɑ ɖò: ní(s) ɟɑt dogo che khedi

thek zon sum ble jɑŋɑ grok nyít ɟɑt dogo che khaethek

thek zon sum blɑ jɑŋɑ gro nyít ɟɑt dogo che khaide

the ne sum bre jɑŋɑ ɖoɁ ní ɟɑt dugu che khethe

thi zøn sum bre lɑŋ ɖo nís gæ dok khepche khedi

*sum *ble *lɑŋɑ *grok *nís *gjɑt *dOgO *khɑl(thek)

Note that the numerals also exhibit some of the sound changes listed in Table 2. Dakpa and Dzala have /e/ or /ɛ/ where the Proto East Bodish forms have /ɑ/. Proto East Bodish /e/ changes to /i/ in Dakpa and Dzala forms, with the exception of Dzala ‘one,’ which has retained the mid vowel.20 Kurtöp has innovated a retroflex in favor of a velar-rhotic onset, but so have Chali, Phobjip, and Dzala. It is not yet known if this is a regular sound in these latter three languages or not. The Chali, Kurtöp, Khengkha and Bumthap reflexes for ‘five’ illustrate the change l > j, which has been used a diagnostic sound change for the subgroup comprising these four languages (Hyslop 2013). The reflexes for ‘eight’ suggest that the same languages also underwent the change gj > ɟ, though more data are required to confirm this. We have not yet reconstructed forms for ‘two’ or ‘ten,’ pending more data and further analysis of sound change. Given the preponderance of nasal-initial forms for ‘two’ throughout Tibeto-Burman, we would expect the form zon to be innovative. However, it is found in the Bumthang languages and Phobip, but not Chali, as would be expected given the proposed tree in Figure 2 either Phobjip and the Bumthang languages innovated the word separately, our tree is wrong, or the form is conservative and the presence of nasal-initial forms elsewhere in the language is due to borrowing. There also appear to be two separates roots for ‘ten.’ The Phobjip form khepche is analyzed as twenty-half, making the Kurtöp, Bumthap, Khengkha and Chali forms a clipped version of the compound. Importantly, however, the compound (and therefore the 20 As I show in Hyslop (2013), the sound changes are not exceptionless. This may be due to the fact that the sound changes are still in progress, that non-conforming forms are borrowings, or some other reason as of yet unknown.

170   

   Gwendolyn Hyslop

clipped innovations) appear to be Dzongkha borrowings; the Dzongkha form for ‘half is che while the Kurtöp forms is phretse. Further, khepche itself is a way of expressing ‘ten’ in Dzongkha.21 The Dakpa and Dzala forms, though superficially similar with palatal stops and front vowels, presumably represent a separate root.

5 Pronouns Pronouns for the East Bodish languages are shown in Table 12. We will begin with an examination of the forms for first person, which are shared throughout the languages. Tab. 12: East Bodish pronouns Gloss

PEB

Da

Dz

Ku

Bu

Kh

Ch

Ph

1.sg 1.pl

*ŋɑ ?

ŋe ŋɑr

ŋɑt ner (incl) net (excl)

ŋɑt ŋet

ŋɑ/ŋɑt

ŋɑt ne

ŋɑ

2.sg 2.pl

*i/*nVn

i ir

wit nin

wet win

we

í

yi

3.sg 3.pl

*khi/*ba

be ber

ŋe ŋɑtɑ (incl) ŋɑrɑ (excl) i itɑ(ŋ) (incl) irɑ(ŋ) (excl) be betɑ(ŋ) (incl) berɑ(ŋ) (excl)

khit bot

gon/khit bot

gon

khi

khi

For first person singular, all the languages possess a form with a velar nasal initial. In Dakpa and Dzala the following vowel is the mid front vowel e while for the other languages the vowel is the low back a. Recall from Table 2 that Dakpa and Dzala can be separated from the other East Bodish by the shared sound change a > e and thus the form ŋe as opposed to ŋa, found elsewhere, is not surprising.

21 I am grateful to Karma Tshering for calling this to my attention.



A preliminary reconstruction of East Bodish   

   171

The most unusual aspect of the first person pronouns is the innovative coda -t, which is required in Kurtöp but appears “optionally”22 in Khengkha and perhaps Chali. The precise development of coda -t on the Kurtöp absolutive pronouns remains to be elucidated, but some evidence of its development may come from the fact that -t is the ergative suffix on Mangdep pronouns (Nishida, pc).23 Whatever the source of the coda -t, it is clear that it is innovative and the first person singular pronoun can be reconstructed as *ŋa. Without more data it will be difficult to reconstruct the first person plural pronoun. At present, the Kurtöp, Bumthap and Chali data favor a form with a mid front vowel, while Dakpa and Dzala have a form with a low vowel. The Dakpa and Dzala reflexes, however, could be a change from ŋe to ŋa, perhaps in response to ŋa ‘1’ > ŋe through regular sound change. There are two roots present in the East Bodish second person pronouns. The one that occurs in all the daughter languages today has a front vowel i, with a glide onset in some languages. Given the predominance of the high vowel i over the mid vowel e (present only in Bumthap), we reconstruct one Proto East Bodish second person pronoun as *i, believing the glide onsets in Kurtöp, Khengkha, Bumthap and Phobjip to be innovative. We are still missing second person plural forms for several languages, but tentatively reconstruct nVn as another second person form used in Proto East Bodish. Two roots for third person pronouns are found throughout East Bodish. A form with a velar initial and high, front vowel is found as the singular form in Kurtöp, Bumthap, Chali and Phobjip. Although some data from Khengkha, Phobjip and Chali are still lacking, we have found forms with a labial initial for plural forms in Kurtöp and Bumthap, and for both singular and plural forms in Dakpa and Dzala. The form gon, found in Khengkha and Bumthap is presumed to be innovative, as it still exists as a lexical word ‘friend’ in Kurtöp.

22 We do not mean to imply that -t is used without any change of meaning in these languages, as an alternation between ŋa ~ ŋat most certainly is meaningful; rather, we do not yet understand what the alternation signals in Khengkha. With regard to the presence of the coda -t in the Chali first person pronoun, we have recorded instances where it is present, and no instances of ŋa by itself, but we believe not enough data have been collected to state confidently that the form ŋat is used exclusively. 23 One possible scenario is that proto ergative *-s (cognate with Classical Tibetan ergative -s; DeLancey 2003: 258) > -t in East Bodish (*-s > -t in coda position is a sound change that happened elsewhere in the subfamily; e.g. *rospa ‘bone’ > rotpa ‘bone’ in some dialects of Kurtöp). The ergative, being used pragmatically like the ergative today (cf. Hyslop 2010) could have generalized into the default form of the pronouns, giving rise to the current absolutive (unmarked) pronominal forms in Kurtöp and Bumthap.

172   

   Gwendolyn Hyslop

We therefore reconstruct two roots for the third person pronoun to Proto East Bodish: *khi and *ba. The motivation for the vowel in the first form lies in the fact that all modern reflexes have this vowel. The motivation for the low, back vowel in *ba is as follows. First, both Dakpa and Dzala have a mid front vowel e for the forms, with plurality indicated by a separate plural marker (-r in Dakpa or -ta(ŋ)/-ra(ŋ) in Dzala, depending on clusivity). Given that the sound change a > e has occurred in Dakpa/Dzala, it is plausible that *ba could be reconstructed for the proto language shared by these two languages. In Kurtöp and Bumthap, however, the labial is followed by a rounded vowel o. We speculate the vowel a > o when following labial stops, and thus the proto form giving rise to the Kurtöp and Bumthap forms could also have been *ba. Therefore, the form *ba readily leads to be in Dakpa and Dzala, based on what we know about sound change established for these languages, and plausibly leads to bo in Kurtöp and Bumthap, based on what we speculate. Ideally, we would find additional data elsewhere in Kurtöp and Bumthap that would evidence the change a > o when following a labial initial, as further support. Table 13 shows the reconstructed Proto East Bodish pronouns compared against possible Tibeto-Burman cognates. The first person form is clearly a reflex of the Proto Tibeto-Burman form, and the second person *nVN is probably related to the Proto-Tibeto-Burman second person pronoun. The other form, i, has possible cognates in Tamang 2ai and 2eː. There are also possible cognates for the two third person pronouns we reconstruct to Proto East Bodish. Proto East Bodish *ba could well be related to the Proto-Tani form and Proto-Boro-Garo forms. Note that Wolfenden (1929: 94–95) mentions a pre-Classical Tibetan third person pronoun ba but does not provide evidence for this, and, to our knowledge, no Tibetologists can confirm this form in any texts.24 The second third person form we reconstruct is *khi, which also has possible cognates in Singhpo, to the East, and Chantyal, to the west. There are possible Tibeto-Burman cognates for all the personal pronouns. But unless we can prove any of these (possible) shared forms are shared innovations, they do not offer much help to further our understanding of the placement of the East Bodish languages within Tibeto-Burman.

24 Note that Hill (2010b) mentions a first person pronoun kho-bo, but is unclear whether the second syllable would be related to the third person ba mentioned by Wolfenden.



A preliminary reconstruction of East Bodish   

Tab. 13: Reconstructed pronouns 

   173



Gloss

*PEB

Other TB

1 2 3

ŋɑ i/nVN khi/bɑ

PTB *ŋa (Matisoff 2003) Tamang 2ai, 2eː; PTB *na, naŋ (Matisoff 2003: 639) Pre-classical Tibetan ba25; Proto-Tani *bà26; Proto-Boro-Garo27 u1, bi1, Turung Singpho khi28; Chantyal khi29

6 Morphology Due to the dearth of linguistic analysis available on the East Bodish languages, it is particularly difficult to compare and reconstruct morphology. Nonetheless, we have comparative forms for future tense, question copulas, and ergative and genitive enclitics for several languages, as summarized in Table 14. Tab. 14: Comparative East Bodish Morphology  Gloss fut q.cop erg gen

*PEB -mɑ *lo

Da

Dz 30

-m lo -si32 -ku

31

‑mɑ lo -(g)i33 -ku

  Ku

Bum

Kh

Ch

Ph

‑mɑle jo ‑li, ‑gi, ‑i -gi, -i, -li, -ci, -ti

-mɑlɑ jo ‑li34

‑m jo

do

lo

-li

-u

25 Wolfenden (1929: 94–95) 26 Post and Modi (2011) 27 Burling and Joseph (2006) 28 Morey (2010) 29 Noonan (2003: 320) 30 In our Dakpa data (Hyslop and Tshering 2010: 16), we found -m to correlate with third person future, while -k appeared to correlate with first person future. Given that the study was preliminary, it is too early to tell whether or not the analysis of -m being a marker of third person future, or something else, such as disjunct future, for example. 31 Genetti (2009) 32 Hyslop and Tshering (2010: 14) 33 Genetti (2009) 34 van Driem (1995)

174   

   Gwendolyn Hyslop

Future tense forms with the syllable -ma are found in all the languages for which we have data. The source of the second syllable -la/le on the Kurtöp and Bumthap forms is something of a mystery. One possible scenario is one in which -ma has an original function as a nominalizer (see Busch 2007) and that -la had an original function as a copula.35 Reflexes of Proto East Bodish question copula *lo, a form required to predicate information questions, appear in all the languages. On the surface, it appears that the function of the question copula is the same across all languages, but further research should confirm this. The Chali reflex do is unexpected, based on what we currently know about East Bodish sound change. Given the change l > j for Chali and the Bumthap languages, we would expect the Chali reflex to have an initial glide j, rather than stop d. Nonetheless, we tentatively reconstruct the form *lo. Ergative and genitive forms are more diverse in the languages than the forms discussed above. On the surface, it may appear that ergative -gi could be reconstructed to Proto East Bodish, given its presence in Kurtöp and Dzala. However, deeper analysis of the Kurtöp -gi (Hyslop 2011a) is suggestive of -gi being a borrowing over original -li. Dzala, spoken in a region directly adjacent to the Kurtöp speaking area, may exhibit a similar phenomenon. There is not enough evidence yet to ascertain whether Dakpa -si or Bumthap/Kurtöp -li is found elsewhere in East Bodish. Indeed, further research may show that Dakpa -si is reconstructable within East Bodish and related to the ergative -s in Written Tibetan (Beyer 1992: 265). For example, if we have evidence that Mangdep ergative -t is a reflex of ergative -s(i), it would serve as support of this hypothesis. At present, however, we simply do not have enough data. What is certain, though, is that the ergative -gi found in Kurtöp and Dzala does not reconstruct to Proto East Bodish. For genitive, the Chali, Dakpa, and Dzala forms appear to reflect the same root; presumably the Chali form reflects loss of initial k. We are still missing data for other languages, and the Kurtöp data confuses matters. In elicitation, the forms -gi and -i are used on nouns and pronouns have a -ci formative, which we presumed to be related to -gi ~ -i. In texts, however, -li occasionally crops up on nouns and, even more surprising, the form -ti occasionally occurs on the second person plural pronoun as -ti in ninti. As an added piece to this puzzle, Kurtöp has a future tense form -ci36 and Khengkha has a future tense form -ti37. If these

35 Hyslop (2011a: §15.2.4) discusses the Kurtöp nominalizer -male and briefly discusses the origin of the -le formative. 36 Hyslop (2011a) analyzes es -ki ~ -ci ~ -iki as a hortative, though comparative evidence with Khengkha may suggest the form -ci is a reflex of older -ti and that -ki ~ -iki represents a separate form historically, now merged into the same morphophonological paradigm. 37 Yangdzom and Arkesteijn (1996: 35) list -ti following the verb khor ‘take’ as a future tense form, but do not know what differentiates it from the other future tenses.



A preliminary reconstruction of East Bodish   

   175

forms are historically related, then it is possible that the Kurtöp genitive -ci on pronouns is historically derived from an older form -ti, almost completely lost in the language today. The point of this digression is only to point out that it is not yet possible to reconstruct the original form of the genitive in Proto-Kurtöp, let alone Proto East Bodish. Looking outside East Bodish, there is evidence for cognates to PEB *ma elsewhere in Tibeto-Burman. For example, in NE India Dimasa has a future -ma (Longmailai 2012) and in Nepal Chantyal has -m present/immediate future/future imperfective (Noonan 2003). To our knowledge, the question copula is unique to the East Bodish languages.

7 Conclusions The aim of this article was to begin a reconstruction of Proto East Bodish, in the hope of advancing our knowledge of the placement of East Bodish in TibetoBurman. While we have confidently reconstructed a number of lexical items to Proto East Bodish, we are still unable to say much about the placement of the East Bodish languages within Tibeto-Burman. Following detailed phonological analysis of more East Bodish languages and subsequent comparison, the next important step will be to systematically tease out Tibetan loans, perhaps akin to the seven major criteria Jacques (2004: 83–200; 2008: 123–142) has proposed in order to tease out Tibetan loans from Rgyarong. Jacques’ third criterion, which is to identify words that have undergone Tibetan phonological innovations, is already easy enough for us to apply in many cases. For example, the Kurtöp word for ‘Bhutan,’ druk, is clearly a borrowing, as br > ɖ in Tibetan but has remained as br in Kurtöp (cf. Kurtöp bra: ‘cliff’), as also appears to be the case for other East Bodish languages. Other tools, such as Jacques’ seventh criterion (words that violate the native phonological structure of the language) will have to remain untested until in-depth phonological work is done on more languages. An eventual reconstruction of Proto East Bodish will necessarily involve a comparison of the full phonological systems of each language but in the meantime this article has offered a few reconstructed forms (such as numerals) and points to areas in the subfamily that will be difficult to reconstruct (several core vocabulary items as ‘sun,’ ‘tomorrow,’ ‘house,’ etc.). On the one hand, we hope this chapter has definitively shown that what can be reconstructed for the East Bodish languages is indeed different from Tibetan. The East Bodish languages may appear to be more conservative than Tibetan (e.g. lack of palatalization in numerals *thek ‘one’ and *ble ‘four’) but the relationship between them and Tibetan is still unclear.

176   

   Gwendolyn Hyslop

Certainly, it is not as straightforward as stating that the Central Bodish languages descended from Proto East Bodish. If this were the case, we would expect a systematic comparison between words like PEB *kram ‘otter’ and *kra ‘hair,’ which does not obtain (cf. Written Tibetan sram and skra, respectively). On the other hand, we hope this article has also presented some of the difficulties in reconstructing Proto East Bodish. On top of the basic problem of not having much data, there are lexical items for which there are no obvious cognates across the entire subfamily. While some forms are straightforward (numerals up to nine, for example), for others we have several possible roots to choose from and it is not (yet!) possible to know – with evidence internal to East Bodish – which root to reconstruct to the parent language. This is especially true of the case marking morphology presented in this article. In fact, future research may show that conceiving of the East Bodish languages using the family tree model is a problematic endeavor. It may turn out to be more prudent to conceive of the East Bodish languages as different linguistic substrates united by an as of yet unknown linguistic superstrate.

8 Abbreviations 1 2 3 cop ct erg excl fut gen incl q

First person Second person Third person Copula Counter Ergative Exclusive Future Genitive Inclusive Question marker

References Aris, Michael. 1979. Bhutan, the early history of a Himalayan kingdom. Warminster: Aris and Phillips. Balodis, Uldis. 2009. The numeral system of Dzala, a language of eastern Bhutan presented at the 15th Himalayan Languages Symposium, August 1 2009, Eugene, OR. Beckwith, Christopher I. 1996. The morphological argument for the existence of Sino-Tibetan. Pan-Asiatic linguistics: proceedings of the fourth international symposium on language



A preliminary reconstruction of East Bodish   

   177

and linguistics, Vol III, 812–826. Mahidol: Institute for Language and Culture for Rural Development Mahidol University at Salaya. Beyer, Stephan. 1992. The classical Tibetan language. Albany: State University of New York Press. Burling, Robbins, and U. V. Joseph. 2006. Comparative phonology of the Boro-Garo languages. Mysore: Central Institute of Indian Languages. Busch, John M. 2007. Verbal nominalization in Kurtoep. Eugene: University of Oregon MA thesis. Coblin, Weldon South. 1976. Notes on Tibetan Verbal Morphology. T‘oung Pao 52: 45–70 DeLancey, Scott. 2003. Classical Tibetan. In Graham Thurgood and Randy J. LaPolla (eds.). The Sino-Tibetan languages, 255–269. London and New York: Routledge. DeLancey, Scott. 2008. Kurtöp and Tibetan. In Chomolangma, Demawend und Kasbek: Festschrift für Roland Bielmeier zu seinem 65. Geburtstag, 29–38. Halle (Saale): International Institute for Tibetan and Buddhist Studies GmbH. van Driem, George. 1995. Een eerste grammaticale verkenning van het Bumthang, een taal van Midden-Bhutan: met een overzicht van de talen en volkeren van Bhutan. Leiden: Research School CNWS. van Driem, George. 1998. Dzongkha. Leiden: Research School CNWS. van Driem, George. 2001. Languages of the Himalayas: an ethnolinguistic handbook of the greater Himalayan Region, containing an introduction to the symbiotic theory of language. Leiden and New York: Brill. van Driem, George. 2007. Dakpa and Dzala form a related subgroup within East Bodish, and some related thoughts. In Roland Bielmeier and Felix Haller (eds.). Linguistics of the Himalayas and beyond, 71–84. Berlin and New York: Mouton de Gruyter. van Driem, George. 2011. Tibeto-Burman subgroups and historical grammar. Himalayan Linguistics 10 (1): 31–39. Genetti, Carol. 2009. An introduction to Dzala, an East Bodish language of Bhutan presented at the 15th Himalayan Languages Symposium, August 1, Eugene, OR. Hale, Austin. 1973. Clause, sentence, and discourse patterns in selected languages of Nepal IV: word lists. (Summer Institute of Linguistics Publications in Linguistics and Related Fields 40). Kathmandu, Nepal: SIL and Tribhuvan University Press. Accessed via STEDT database: http://stedt.berkeley.edu/search/ on 2011–06–08. Hill, Nathan W. 2010a. An overview of Old Tibetan synchronic phonology. Transactions of the Philological Society 108 (2): 110–125. Hill, Nathan W. 2010b. Personal pronouns in Old Tibetan. Journal Asiatique 28 (2): 549–571. Hill, Nathan W. 2011a. Multiple origins of Tibetan o. Language and Linguistics 12 (3): 797–721. Hill, Nathan W. 2011b. An inventory of Tibetan sound laws. Journal of the Royal Asiatic Society 21: 441–457. Hyslop, Gwendolyn. 2008a. Kurtöp and the classification of the languages of Bhutan. In Max Bane, Juan Bueno, Tommy Grano, April Grotberg, and Yaron McNabb (eds.). Proceedings from the 44th meeting of the Chicago Linguistic Society, 141–152. Chicago: University of Chicago Press. Hyslop, Gwendolyn. 2008b. Kurtöp phonology in the context of Northeast India. In Stephen Morey and Mark Post (eds.). Northeast Indian Linguistics 1: 3–25. Delhi: Foundation/ Cambridge University Press. Hyslop, Gwendolyn. 2009. Kurtöp tone: a tonogenetic case study. Lingua 119: 827–845.

178   

   Gwendolyn Hyslop

Hyslop, Gwendolyn. 2010. Kurtöp case: the pragmatic ergative and beyond. Linguistics of the Tibeto-Burman Area 33 (1): 1–40. Hyslop, Gwendolyn. 2011a. A grammar of Kurtöp. Eugene, OR: University of Oregon dissertation. Hyslop, Gwendolyn. 2011b. Mirativity in Kurtöp. Journal of South Asian Languages 4.43–60. Hyslop, Gwendolyn. 2013. On the internal phylogeny of the East Bodish languages. In Gwendolyn Hyslop, Stephen Morey, Mark Post (eds.), North East Indian Linguistics 5, 91–109. New Delhi: Cambridge University Press/Foundation. Hyslop, Gwendolyn. to appear. Waves across the Himalayas: On the typological characteristics and history of the Bodic family of Tibeto-Burman. Language and Linguistic Compass. Hyslop, Gwendolyn, and Karma Tshering. 2009. The Tasha-Sili language of Bhutan: a case study in language shift and Bhutanese prehistory. In FEL XIII: Endangered languages and history, 101–108. Bath: Foundation for Endangered Languages. Hyslop, Gwendolyn, and Karma Tshering. 2010. Preliminary notes on Dakpa (Tawang Monpa). In Mark Post and Stephen Morey (eds.). North East Indian Linguistics 2, 3–21. New Delhi: Cambridge University Press/Foundation. Jacques, Guillaume. 2004. Phonologie et morphologie du Japhug (rGyalrong). Paris: Universite´ Paris VII – Denis Diderot dissertation. Jacques, Guillaume. 2008. Jiarongyu yanjiu. Beijing: Mínzú chūbǎnshè. Jacques, Guillaume. 2010. A possible trace of verbal agreement in Tibetan. Himalayan Linguistics 9 (1): 41–49. Jacques, Guillaume. 2012. An internal reconstruction of Tibetan stem alternations. Transactions of the Philological Society 110 (2): 212–224. Lewis, Paul. (ed.). 2009. Ethnologue: a guide to the world’s languages. Dallas: Summer Institute of Linguistics. Li Fang-Kuei. 1933. Certain phonetic influences of the Tibetan prefixes upon the root initials. Bulletin of the Institute of History and Philology 6 (2): 135–157. Longmailai, Monali. 2012. Personal pronouns in Dimasa. In Gwendolyn Hyslop, Stephen Morey and Mark W. Post, (eds.). North East Indian Linguistics 4, 104–118. Delhi: Cambridge University Press/Foundation. Matisoff, James. 2003. Handbook of Proto-Tibeto-Burman: system and philosophy of Sino-Tibetan reconstruction. Berkeley: University of California Press. Mazaudon, Martine. 2003. Tamang. In Graham Thurgood and Randy LaPolla (eds.). The Sino-Tibetan Languages, 291–314. London and New York: Routledge. Michailovsky, Boyd, and Martine Mazaudon. 1994. Preliminary notes on languages of the Bumthang group. In Tibetan Studies: Proceedings of the 6th Seminar of the International Association for Tibetan Studies, 2: 545–557. Fagernes: The Institute for Comparative Research in Human Culture. Morey, Stephen. 2010. Turung – a variety of Singpho language spoken in Assam. Canberra: Pacific Linguistics. Nishida, Fuminobu. 2009. Mandebigo shoki Chōsa Hokōku [A Preliminary Report on Mandebi-kha]. Chiba daigaku jinbun kenkyū 38: 63–74. Noonan, Michael. 2003. Chantyal. In Graham Thurgood and Randy J. LaPolla (eds.). The Sino-Tibetan languages, 315–335. Routledge Language Family Series. London: Routledge. Plaisier, Heleen. 2007. A Grammar of Lepcha. Leiden and Boston: Brill. Post, Mark W. and Yankee Modi. 2011. Language contact and the genetic position of Milang (Eastern Himalaya). Anthropological Linguistics 53 (3): 215–258.



A preliminary reconstruction of East Bodish   

   179

Shafer, Robert. 1954. The linguistic position of Dwags. Oriens, Zeitschrift der Internationalen Gesellschraft für Orientforschung 7: 348–356. Shafer, Robert. 1955. Classification of the Sino-Tibetan Languages. Word 11 (1): 94–111. Sino-Tibetan Etymological Dictionary and Thesaurus (STEDT). http://stedt.berkeley.edu/ Taylor, Doreen M. 1973. Clause Patterns in Tamang. In Austin Hale and David Watters (eds.). Clause, Sentence and Discourse Patterns in Selected Languages of Nepal. Part 2: Clause, 81–174. Norman: Summer Institute of Linguistics and University of Oklahoma. Tournadre, Nicolas. 2008. The notion of scale in linguistic classification: is “Tibetan” a language? or a family of languages? Paper presented at the 14th Himalayan Languages Symposium, August 21, University of Gothenburg, Sweden. Wolfenden, Stuart. 1929. Outlines of Tibeto-Burman linguistic morphology, with special reference to the prefixes, infixes and suffixes of Classical Tibetan, and the languages of Kachin, Bodo, Naga, Kuki-chin, and Burma groups. (Prize Publications 12). London: Royal Asiatic Society. Yangdzom, Deki and Marlen Arkesteijn. 1996. Khengkha Lessonbook. Unpublished manuscript. SNV, Thimphu.

Ilija Čašule

Burushaski kinship terminology of ­Indo-European origin 1 Introduction Kinship terms are often considered reliable evidence for a genetic relationship. In this paper we look at some 30 kinship terms in the Burushaski language that can be derived directly from Indo-European following the established systematic phonological correspondences and are not borrowings from Old Indian, the neighbouring Indo-Aryan or the Iranian languages. The coherence of the Burushaski kinship system and its preservation of the original Indo-European kinship terms advance further the position that Burushaski is genetically related to Indo-European.

1.1 Sources and previous studies Burushaski, still considered to be a language isolate, is spoken by around 90,000 people (Berger 1990: 567) in the Karakoram area in North-Western Pakistan at the junction of three linguistic families – Indo-European (Indo-Aryan and Iranian branches), Sino-Tibetan and the Turkic. It has three very closely related dialects: Hunza and Nager which have minimal differences between them, and the Yasin dialect which exhibits differential traits, but is still mutually intelligible with the former two. For a brief profile of Burushaski, see Čašule (2003b: 21–24) or Berger (1990). The earliest, very limited material for Burushaski is from the mid to late 19th century (e.g. Cunningham 1854, Hayward 1871, Biddulph 1880, Leitner 1889). The limited dialectal differentiation and the lack of older attestations make the internal historical reconstruction difficult. The fundamental sources for the description and study of Burushaski are considerable and of a very high quality. We note here only those directly used as sources for the kinship terminology. For a more extensive survey, see Bashir (2000), Tiffou (2000) and Čašule (2010). Most notable and authoritative is Berger’s (1998) three-volume work on the Hunza-Nager dialect (grammar, texts and a Burushaski-German dictionary, henceforth “B”).1 Still very relevant is Lorimer’s earlier groundbreaking 1 See abbreviations for other frequently cited sources at the beginning of the references section.

182   

   Ilija Čašule

three-volume work (1935–1938) on Hunza-Nager and (1962) on Yasin Burushaski. Edel’man and Klimov’s (1970) monograph and Willson’s (1999) compact basic Burushaski vocabulary are also very useful. Backstrom and Radloff (1992) have a very pertinent sociolinguistic survey on Burushaski language maintenance and give basic comparative data with the surrounding languages, like Shina, Ḍomaaki, Wakhi and Balti. Varma (1941) is helpful in the description of the dialectal differences between Hunza and Nager Burushaski. Fundamental for the study of Yasin Burushaski are Berger’s (1974), Tiffou and Pesot’s (1989), Tiffou and Morin’s (1989) and Zarubin’s (1927) grammars and vocabularies. Very important in establishing aspects of the historical phonology and morphology of Burushaski and its internal reconstruction is Berger’s (2008) posthumously published synthesis. In our work (Čašule 1998, 2003a, 2003b, 2004, 2009a, 2009b, 2010, 2012a, 2012b), we have correlated Burushaski with Indo-European, outside of Indic and Iranian, and in our etymological analyses we have uncovered consistent and systematic lexical, phonological and most importantly, extensive and fundamental grammatical correspondences – the latter are outlined in Čašule (2003b: 69–79) and expanded in Čašule (2012b: 134–153). On the basis of the analysis of over 500 etymologies and the highly significant correspondences in over sixty mostly grammatical but also derivational morphemes (nominal case endings, nominal plural endings, verbal prefixes, suffixes and endings, the complete non-finite system, all of the adjectival suffixes, the entire system of demonstratives, personal pronouns, postpositions, adverbs, etc.) we conclude that Burushaski displays characteristics of a language which could have had an early relationship or contact in its history with the North-Western Indo-European group (see esp. Čašule 2004, on the possible correlation with Phrygian). The Burushaski phonological system, internal variation and phonological correspondences with Indo-European are outlined and systematised in Čašule (2003b: 24–42) and Čašule (2012b: 64–98). The correspondences (over 70 of them) in the core vocabulary of names of body parts and functions can be found in Čašule (2003a). Eight new correspondences in this semantic field in Čašule (2009b) bring the total to circa 80. In Čašule (2003b), we provide an in-depth analysis of the Burushaski laryngeals and their consistent and direct correspondence with the Indo-European laryngeals. For a recent appraisal and support of this evidence, see Alonso de la Fuente (2006). We find a relatively close correlation of the Burushaski numeral system with Indo-European in Čašule (2009b). In an extensive analysis and comparison of Burushaski’s shepherd vocabulary with Indo-European (Čašule 2009a), we concluded that almost in its entirety it is autochthonous Indo-European – we identified 32 pastoral terms of Indo-European (non-Indo-Iranian) origin in Burushaski,



Burushaski kinship terminology of I­ ndo-European origin   

   183

ten of which find direct correspondences with the substratal (Thracian? Phrygian?) shepherd vocabulary in Albanian, Romanian and Aromanian. Čašule (2010) is a comprehensive analysis of the systematic phonological (and derivational) correspondences involving mostly core Burushaski vocabulary, which contain reflexes of the Indo-European gutturals (the velars, labiovelars and palatovelars). This monograph also provides a synthesis of the mounting evidence that indicates that Burushaski may be a North-Western Indo-European language, i.e. concludes that Burushaski shows the greatest number of correlations with the Ancient Balkan languages (Phrygian, Thracian, Ancient Macedonian) and Albanian, on the one hand, and with Balto-Slavic and Germanic on the other. In Čašule (2012b), in our discussion article in the issue of The Journal of Indo-European Studies dedicated to the Burushaski-Indo-European debate, we show that the entire Burushaski system of personal pronouns and demonstrative pronouns and adverbs can be correlated closely with Indo-European. In Hamp’s (2012: 155) response he supports our evidence and proposes a deeper antiquity of the genetic relationship between Burushaski and Indo-European. The correlations between Burushaski and substratal and archaic Modern Macedonian and Balkan Slavic vocabulary are discussed in Čašule (2012a). Hamp (2011) in the review of this article based on our full body of evidence, states his support for our position: “Burushaski is at bottom Indo-European [italics Eric Hamp] – more correctly in relation to Indo-European or Indo-Hittite, maybe (needs more proof) IB [Indo-Burushaski]” and further conjectures: “I have wondered if Burushaski is a creolized derivative; now I ask (Čašule 2009a) is it a shepherd creole? (as in ancient Britain).” This can be compared with our tentative conclusion that Burushaski might be “a language that has been transformed typologically at a stage of its development through language contact.” (Čašule 2010: 70).

1.2 The Burushaski phonological system For easy reference, we reproduce Berger’s table of the phonological system of Hz Ng Burushaski, which is essentially valid for the Yasin dialect as well (Yasin Burushaski does not have the phoneme c̣h – for Ys Bur see also Tiffou and Pesot (1989: 7–9):

184   

   Ilija Čašule

Tab. 1: Phonological system of Burushaski (B: I, 13). a e i



o u

h

l

qh q ġ r

kh k g ṅ

ṭh ṭ ḍ

th t d n





s

c̣h c̣ j̣

ćh ć j

ċh ċ z

ph p b m

Notes on the transcription system: 1. All five vowels can be phonetically long, but for phonological and prosodical reasons Berger marks them as double (two component) vowels, in order to mark the position of the stress. This notation system was developed by Buddruss and Berger to indicate the pitch contours, which they consider as a result of first- or second-mora stress (Bashir p.c.). 2. Retroflex consonants are marked with an underdot. 3. w and y are allophones of u and i. 4. ċ = ts in Lorimer and c in Tiffou and Pesot (1989). 5. ġ = γ in Lorimer and Tiffou and Pesot (1989). It is a voiced fricative velar /ɣ/. See Čašule (2010) on the extensive variation between ġ and g. 6. ṅ = [ŋ], [ng] or [nk] 7. The posterior q is similar to the Arabic qāf. “q ist ein stimmloser dorsaler Verschlußlaut, der weiter hinten als k gebildet wird” (B: I, 21–22). 8. The aspirated posterior qh is found only in Hunza and Nager. In Yasin to the latter corresponds a voiceless velar fricative x, similar to the German ch, as in Bach (Tiffou 2004: 10). 9. ỵ is a retroflex, articulated somewhere between a r grasséyé and a γ or rather a fricative r with the tongue in a retroflex position’ (Morgenstierne 1945: 68–9). 10. A hyphen preceding a Burushaski word indicates that it is used only with the pronominal prefixes. 11. “The double hyphens indicate the lengthened strong grade of the pronominal possessive prefix, e.g. móo-skir ‘her father-in-law’” (Tikkanen 2001: 479). 12. A stress on the hyphen -́ indicates that the stress is on the pronominal prefix. For the internal variation and alternations in Burushaski, see Čašule (2003b: 24–29; 2010: 5–11, 14–19).



Burushaski kinship terminology of I­ ndo-European origin   

   185

1.3 S  ummary of phonological correspondences between IndoEuropean and Burushaski We summarise in Table 2 the phonological correspondences we have established between Burushaski and Indo-European (for the exemplification, discussion and analysis, refer to Čašule 2003b: 24–42; with special regard to the laryngeals see Čašule 2012b: 64–98; and for the gutturals see Čašule 2010). Tab. 2: Summary of Burushaski-Indo-European phonological correspondences

IE a > Bur a IE e > Bur e : Hz, Ng i; IE e (unstr.) > Bur a; IE ē > Bur ée, i IE o > Bur ó ; IE o (unstr.) > Bur a, u; IE ō > Bur oó, óo IE i > Bur i IE u > Bur u IE ai, ei, oi, eu > Bur a IE au, ou > Bur u PIE h1- > Bur h-; PIE h1e- > Bur he-; PIE h1u̯er- > Bur har- : -war- : herPIE h2- > Bur h-; PIE h2e- > Bur ha-; PIE h2u̯e- > Bur -we- : -wa-; PIE h1/2i- > Bur iPIE ha- > Bur h-; PIE hae- > haa- > Bur haPIE h4- > Bur h-; PIE h4e- > h4a- > Bur ha- PIE h3- > Bur h-; PIE h3e- > h3o- > Bur hoPIE hx- > Bur hIE l, m, n, r > Bur l, m, n, r IE u̯ > Bur -w/-u; IE u̯- > Bur b-, also m-; IE i̯ > Bur y/i IE m̥ > Bur –um, -am; IE n̥ > Bur -un, -an; IE r̥ > Bur -ur, -ar; IE l̥ > Bur –ul, -al IE p > Bur p, ph, also bIE b > Bur b, also m (rare); IE bh > Bur b, also m (rare) IE t > Bur t : th (rare) : ṭ , and dIE d > Bur d; IE dh- > Bur d-; IE VdhV > Bur -t-, -ṭIE k > Bur k : kh, k : q2 IE kw > Bur k; IE k̂ > Bur k : kh, k : q IE g > Bur ġ; IE gh > Bur g 2 For a detailed analysis of the alternations k:kh, k:q, k:qh, kh:q, q:qh and g:ġ, see Čašule (2010: 14–18).

186   

   Ilija Čašule

IE gw > Bur ġ; IE gwh > Bur ġ IE ĝ > Bur g, ġ; IE ĝh- > Bur g, ġ IE s > Bur s or s : ċ , ċh IE ks > Bur ś The sources for the Indo-European material are Pokorny (1959), Watkins (2000), Mallory-Adams (1997), Mallory-Adams (2006), Buck (1949), and more specifically Benveniste (1973), Friedrich (1966), Hetterich (1985), Blažek (2001) and Kullanda (2002). It is important to note that the Burushaski material has already been sifted carefully for Persian, Urdu and other Indo-Aryan loanwords by Berger, Lorimer, Morgenstierne, Zarubin, Edel’man, Klimov, Varma, Tiffou, Buddruss, Tikkanen and other scholars who have studied the language – their findings are conveniently incorporated in Berger (1998). The main source we have used for further comparison with Indo-Aryan is Turner (1966) and with Persian, Steingass ([1892] 1999).

2 Burushaski kinship terminology 2.1 Loanwords from Indo-Aryan and Tibetan The only synthetic studies of Burushaski kinship terminology, to our knowledge, are Parkin (1987a), who discusses the Tibeto-Burman and Indo-European loans in Burushaski in this semantic field and Parkin (1987b) who looks at the Burushaski kinship system from an anthropological perspective (see the discussion in Section 3). Tikkanen (2001) is a brief analysis of only two Burushaski etyma (see 2.2.9). We refer to Parkin’s work in our analyses, but would like to highlight here the kinship terms he noted that can be considered with some certainty loanwords from Tibetan or Indo-European. We do not refer to the examples where there is confusion between a possible Tibetan honorific prefix and the Burushaski pronominal form a- or to the examples that are considered more or less tentative or problematic. An important such Tibetan loanword is Bur -̇c̣o, Ys -̇c̣u ‘(a man’s) brother; a woman’s sister’ (L: 8, 267; B: III, 104–105), although Berger lists it as an original Burushaski word. Parkin (1987a: 327) compares it with Purik a-co, Ladakhi ə-čo, Kinnauri acho, Bhotia, Sherpa ajo, Tibetan jo all: ‘elder brother’ (“probably originally an honorific”).



Burushaski kinship terminology of I­ ndo-European origin   

   187

Another Burushaski word for ‘brother’ is also a loanword, this time from Indo-Aryan : kaáko pl. : kaákoċaro ‘older brother; polite term for young and old people’ (B: III, 237), also kaká ‘older brother’ (term used only in Rajah families), also gaká (B: III, 240). Berger correlates it with Sh kaáko and T 2998. Parkin states that the common Indic term is usually applied to ‘father’s younger brother,’ “but in this general area appears as ‘elder brother’” (1987a: 328). There is also Bur kaáki, Hz also gaáki pl.: kaákiċaro ‘older sister’ (in the micro-dialects of Ganesh and Altit: ‘only in Rajah families’) (L: 226; B: III, 237). Although this is a typical nursery word, note the exact formal and semantic match (‘older’) with Blg kaka ‘older sister, older woman,’ ‘older daughter-in-law’ (Georgiev 1971: II, 149). This means that the basic words for ‘brother’ and ‘sister’ are borrowed into Burushaski, with the exception of the more specific Bur –yás ‘a man’s sister; a man’s wife’s sister’ (see 2.2.4) and Bur –húles ‘a woman’s brother, husband’s brother’ (see 2.2.5). Among the other loanwords from Indo-Aryan, Parkin (1987a: 328–329) mentions Bur dáado ‘grandfather (on both sides)’ (same in Shina, T 6261; B: III, 108) and Ys Bur sari ‘wife’s brother,’ possibly ‘wife’s sister’ < same in Shina; Ys Bur saroni ‘husband’s brother, wife’s sister, brother’s wife (male speaking), sister’s husband (female speaking)’ < Shina ‘wife’s brother.’ Burushaski táti ‘father’ is also found in Wakhi and Khowar. The plural of táti is tátištaru (given as an original Yasin word in Tiffou and Pesot 1987: 149), found also in Hz, Ng (B: III, 424, with the remark however “actually Khowar”) which suggests to us an older singular form *tátis which could indicate that the Burushaski word is of greater antiquity, i.e. from IE *t-at- : Luw tātis, Skt tatá, Gk tatâ, NWels tad, Lat (inscriptional) tata, all: ‘father’ (M-A2: 211) and Mcd tatko (basic word for ‘father’).

2.2 B  urushaski autochthonous kinship terminology of IndoEuropean origin According to our analysis, the overwhelming majority of Burushaski kinship terms (some 30–32) is of non-Indo-Iranian, autochthonous Indo-European origin. There is a vast body of literature on the Indo-European kinship and family terms. We refer more directly to the synthetic works published in recent years, as it is not our aim to discuss here the Indo-European system in its own right or any outstanding issues.

188   

   Ilija Čašule

2.2.1 ‘mother’ 2.2.1.1 B  ur ‘mother,’ ‘aunt (on mother’s side),’ ‘mother’s sister’ : IE ‘mother; (rare) sister’ Bur máma, pl. mámaċaro ‘mother, aunt (on mother’s side)’ (L: 253: also māma; B: III, 277, and ‘in reference to older women;’ Will: 84, also ‘mother’s sister; a paternal aunt’), and Bur NH mámo ‘Mutter, Mama,’ and mámayo ‘mommy (endearing form of address to girls and young women).’ Note further Bur -́mi (pl. -́miċaro) ‘mother; mother’s sister, aunt, on mother’s side’ (L: 265, 266; B: III, 286–287). There is a direct correspondence with IE *méhatēr- (M-A2: 213; in W-I-S: 457 *máh2ter- / *máh2tr-) ‘mother.’ “Based ultimately on the baby-talk form mā- with the kinship term suffix *-ter-” (Wat: 51; Pok: 700: *mātér-) : Lat *māter, Gk mé̄tēr, Phrg matar, Arm mayr, Av mātar-, Skt mātár-, OIr māthair, OEng mōdor, OPrus mothe, OChSl mati (in oblique cases matere), TochA mācar, TochB mācer, all: ‘mother,’ yet Alb motër ‘sister’ (according to Pokorny, the Albanian word originally referred to “an older sister taking the mother’s place”), Lith mótė ‘woman, wife’ (also ‘mother’ in some dialects, see Kullanda 2002: 92). Burushaski -́mi ‘mother’ may indicate a derivation from a lengthened grade *mē- (after the loss of the laryngeal in a stressed syllable?), perhaps by application of Eichner’s Law which is formulated as “preservation of the timbre of lengthened grade ē in the vicinity of H2” (Eichner 1972: 78).3 For ‘aunt’ see 2.2.8.2 and maybe 2.2.8.3. 3 Consider further in this regard the Burushaski reflex of IE *bhehaĝos ‘beech’ (Gaul bāgos ‘beech,’ Lat fāgus ‘beech,’ OEng bōc ‘beech; book,’ OHG buohha ~ buocha ‘beech,’ Alb bung ‘durmast oak [Quercus petraea],’ Gk phēgós ‘Valonia oak [Quercus aigilops],’ and perhaps Russ buz ‘elder,’ “perhaps a taxonomically broader class of trees” [M-A 58–60]) : Bur Ys behék, Hz Ng biík ‘willow (tree); silver willow tree (Silberweide)’ (with preservation in Yasin of the laryngeal in the inlaut, which prevented the change e > a or ē > i [B: III, 51; see Čašule 2010: 55–56, for the full etymological analysis]).  For ē > i, note also Bur Ys gí-, Hz Ng 1gíỵ- ‘go into, enter into, project o-self, charge’ and giá- pl. of 1 gíỵ- also ‘fall; appear (rainbow); (of vermin) come out’ and -̇-gíỵ- ‘sow, plant’ (B: III, 155), which derives from IE *ĝheh1- ‘release, let go’ (Wat) ‘leave’ (M-A): Gk kikhā́nō ‘meet with, arrive at, light upon,’ Av zizāmi ‘leaves off,’ OEng gān ‘go,’ Skt jáhāti ‘leaves, lets go,’ hīna ‘inferior’ (Wat: 28; M-A: 349) (*ĝheh1- > *ĝhē- > 1gíỵ- (with –y- < h1 and ē > i).  Further, from IE *ĝerh2- ‘grow, age, mature; grain’ : OChSl zǔrěti ‘ripen,’ Gk gēráskō ‘age, grow old,’ Skt jī́ryati ~ jū́ryati ‘grows old, becomes decrepit,’ TochAB kwär- ‘age, grow old’ (M-A2: 190 “widespread and old;” Pok: 390–391) we have Bur ġirġāār man-́ and ġaġár ní- ‘(of corn, walnuts) ripen in great quantities’ (B: III, 176) with a typical reduplicated stem as in Indo-European: Gk gergérimos ‘ripe figs or olives falling of their own account’ (Pok: 391) and erh2 giving both āār and ir. Also (from the zero-grade), Bur gur ‘wheat’ (L: 174; B: III, 161).



Burushaski kinship terminology of I­ ndo-European origin   

   189

From the Indo-European stem *mā-, and reduplicated *māmā-, ‘mother’ we have also Lat mamma ‘mother, midwife, grandmother; mother’s breast, nipple,’ Gk mámma ‘mother, grandmother; mother’s breast, nipple,’ Alb mëmë ‘mother,’ Arm mamm ‘grandmother’ (Pok: 694: 3mā-). As noted by Huld (1984: 96) in regard to the Albanian development above: “In Omaha-kinship systems of terminology as is reconstructed for Proto-IndoEuropean, the term ‘mother’ extends to the mother’s sister and her daughters as well as ego’s own mother.” This is partially the case with Burushaski. It is also very significant that (as in Latin and Greek) we have the correlation ‘mother’ : ‘breast’ from the same stem in Burushaski: –mámut, Ng maámo, Ys -mámu ‘a woman’s breast; nipple (male or female)’ (B: III, 277; L: 254) and by further semantic extension Bur mamúto ‘sucking, immature; suckling’ and Bur mamú ‘milk’ (B: III, 276–277). The Burushaski words –mámut and mamúto can be compared with the Latin derivatives mammātus ‘furnished with protuberances or spouts’ and mammeātus ‘full-breasted,’ pf. participles in –ātus (E-M: 381), although the Burushaski forms do not appear to reflect an –eh2stem. The PIE adjective forming suffix *–to is well represented in Burushaski (also in Shina): e.g. Bur durgasúuṭo ‘weak, exhausted, played out,’ which Berger (B: III, 126) derives from dúrgas ‘ghost,’ which in turn we have traced to IE *dhroughos ‘phantom’ (M-A: 538), also from bambú ‘ball, balloon,’ there is bambúto ‘thick’ (B: I, 210); Berger points that in some cases the suffix is preceeded by a vowel). Usually the forms we discuss are considered nursery words found across cultures (a reduplication from infant babbling). Yet it is curious that according to both Lorimer and Berger, none of Burushaski’s neighbours share this development (although the word is found in colloquial speech in other NIA languages [Bashir, Tikkanen p.c.]). Moreover, in Burushaski it is the basic word for mother. The decisive clue that the correspondence with Indo-European goes beyond sound symbolism or nursery word coincidence, apart from the phonological, semantic and derivational considerations above, is the Burushaski plural h(x) suffix –ċaro which is added mainly to words denoting relations (B: I, 48), and is a variation of the suffix –taro, which we analyse as –tar- + -o (the latter –o is an x pl ending in its own right). It can be seen as the outcome of a morphological reanalysis, whereby an originally singular derivational suffix was retained only in the plural ending. In a preliminary analysis of the numerous Burushaski plural endings we have found cases where a morpheme or phoneme from the singular form has been retained in the plural, but lost in the singular. Consider in this respect Bur ha ‘house’ (< *hak by Berger 2008: 25) (pl. hakíćaṅ, B: III, 184) which is also of indig-

190   

   Ilija Čašule

enous Indo-European origin4 and the Nager pl. joṭarko, of joṭ ‘small, young man,’ which can be analysed as joṭ-ar-ko ‘he (they) who is (are) young’ – compare derivationally with Latin *new-er-ko > noverca ‘stepmother’ (‘she who is new,’ Wat: 58); or in Bur -̇i-, sg. ‘daughter,’ pl. Hz Ng –yúguśanċ, Ys pl. –yúguśiṅa ‘daughters,’ also ‘brother’s daughters’ (B: III, 210; L: 12, 386), which consists of –yu- and gus ‘woman’ + pl. suffix (the full stem is kept only in the plural form); or Bur ge, also gye in L and gye in Cunn ‘snow,’ Ys ge, gye (B: III, 151), which derives from IE *ĝhi̯em- ‘winter, snow’(Wat: 29; W-I-S: 162–169: *ĝhei̯-om-, *ĝh(i)i̯-em-, *ĝh i-m-) where these forms could go back to forms with –m, as the Ng pl. ending here is -́miṅ, i.e. the plural form is g(y)émiṅ from a sg. *gyem (Čašule 2010: ex. [125]). It is highly significant that the Burushaski suffix -taro corresponds directly with the Indo-European suffix *-ter regarded as classifier of the lexical class of kinship terms by Benveniste (1973: 171). See also Section 3.

2.2.1.2 Bur ‘mother’ (in Royal Families) : Phrg ‘Great Mother’ Bur zizí pl. zizíċaro ‘Mother! Form of address used only in the families of the Rajas and (in Nager) Saiyids. Foster mothers, being of lower rank, are called mama. The corresponding term for father is babá’ (B: III, 27). This term is used properly only in Royal Families (L: 63). Berger points to Urdu bābā. These terms are said to have come down from the time of Alexander the Great (Lorimer 1935: II, 30; 238). They are in use in Shina and Khowar and in Balti zizi (B also zi) and bawa (L: 391), yet no one has been able to provide an Indo-Aryan or Tibetan etymological explanation for zizí. We have compared Burushaski zizí with Phrygian Zizimene – “The epithet of Zizimene is frequently used to designate the Great Mother in various centres of Asia Minor” (…) “In an altar from Sizma, a village of Lycaonia situated near Iconium and Laodicea, whose four faces bear reliefs representing deities and inscriptions on side D, we read the dedication Mētri Zizimmēnē. Ramsay believed it is a dialectal form of Dindymene, in at least one case the Mother of the Gods is explicitly called Zizimene” (Gasparro 1985: 35, also Calder 1912: 72–74). The ab­breviated form of Zizimene is Zizi, Ziezi. (For further discussion, see Čašule 2004: 86, ex. [24]). Considering the close correspondence between Phrygian and Burushaski in the fields of ritual, myth, burial and onomastics and the use of Burushaski zizí by the Royal Families, this is a correlation with some merit. If the 4 We derive it from Bur hákin- ‘learn,’ ‘to teach’ (B: III, 186), from IE *h1euk- ‘become accustomed’ (in Bur eu > a, see Čašule 2010): OChSl učiti ‘teach, learn,’ Arm usanim ‘learn, be used to,’ OInd úcyati ‘is accustomed,’ ókas- ‘abode, home, dwelling,’ Lith jùnkstu ‘become accustomed,’ ū́kis ‘farm,’ OIr do-ucci ‘understands,’ Goth bi-ūhts ‘used to,’ TochA okñas ‘learn, teach’ (M-A: 4).



Burushaski kinship terminology of I­ ndo-European origin   

   191

Burushaski word preserves an ancient “cult” term for the Great Mother it could have easily been the ultimate source for those terms in Shina, Khowar and Balti.

2.2.2 ‘father’ 2.2.2.1 ‘father,’ ‘prince,’ ‘grandfather’ : IE ‘father,’ ‘chief,’ ‘grandfather’ Bur Ng 1bapó pl. 1bapóċaro ‘prince’ (B: III, 37). Berger gives separately 2bapó pl. 2 bapóċaro ‘grandfather (paternal or maternal); “father” in addressing,’ according to the Burushaski scholar and poet Nasiruddin Hunzai (Berger’s main informant) also ‘a respectful term of address’ (Sh bapó) (L: 62, who notes that “this word is appropriated to Royal Families”). Parkin (1987: 328) notes also Ys bap ‘grandfather,’ di-bap ‘great-grandfather’ (considered a loan from Khowar by Berger 1974), and Dameli bàp, Lahnda bāpū, Ksh bab, Panj bābbā and Ḍom babo. While Berger (B: III, 37) appears to deem the Burushaski Hz Ng forms to be autochthonous, he notes appropriately as a borrowing from Indo-Aryan: Bur baábo pl. baáboċaro ‘father’s brother’ (Sh baábo, T 9209; B: III, 25). Bur 1bapó and 2bapó might both derive from a blend of bap + IE *ph2tér(W-I-S: 554–562) : OInd pitár, Av pitar-, Arm hayr, OIr athir, OEng fæder, TochA pācar, TochB pācer, all ‘father’ (also in compounds, Pok: 829) or from a reduplicated form *paph2ter- (with p- > b-).5 Note here also IE *papa ‘child’s word for father,’ e.g. Gk pappos ‘grandfather,’ pappās ‘father’ > ‘pope’ (Pok 789: *pap(p)a) which would have merged with the local Indo-Aryan forms. The –o auslaut could have come about by analogy with baábo. The additional and unique Burushaski meaning of ‘prince,’ considered a homonym by Berger, correlates well with developments elsewhere from IE *ph2ter-, e.g. with Lat pater familias or pater ‘senator,’ as Ernout and Meillet (1959: 487) put it: “Pater has a social significance.” or with Gk patḗr which is “a term whose significance is mainly social and religious” (Chantraine 1947: 235,

5 Note Bur badá ‘sole of foot; step, pace’ (B: III, 29) < OInd padá ‘footstep, track, pace,’ yet a newer form in Ys bayá ‘same’ < (?) Pkt paya- ‘footstep, foot’ < IE *ped-, *pod- (nom. root) ‘foot’ and from IE *ped-, *pod- (verbal root) ‘fall, stumble:’ Ys baḍáṅ -wál- ‘fall down on one‘s back, fall over, faint’ (BYs: 131; Wat: 62). Nonetheless, even in such examples we find the alternation p:b as Burushaski also has padáaỵ -̇t- ‘kick a stone with the foot’ (which Berger links with badá above, B: III, 310).  And further examples of a p > b historical change: IE *pī-nu- ‘pine tree’ < IE *pī- in words for ‘fat, sap, pitch’ (Wat: 62) : Bur Ng buyṹũ ‘a tree similar to the pine, not found in Hz’ (in Sh buyṍõ) (B:  III, 66) (possibly here also Bur biỵ̣ ‘butter,’ bis n. ‘fat’ (B: III, 56–57). IE *perk̂- ‘to dig out’ (Wat: 66) > Bur biráq- ‘to dig, to dig anywhere’ (B: III, 54).

192   

   Ilija Čašule

apud Kullanda 2002: 91) and that patḗr “is a designation of the chief of a group” (Ermont, Meillet 1959; 487), or in reference to the gods as in Lat Jup(p)iter. We have proposed that the Burushaski plural suffix -ċaro, variant form of -taro characteristic of nouns denoting relations, came about as a result of a change – staro > -sċaro > -ċaro, thus (under one interpretation) 1bapóċaro < *papos-taro, and note esp. the intermediate stage of such a change in 2.2.4: Bur –yás pl. – yásċaro, Hz pl. also ‑yástaro, Ys -yást, pl. -yástaru ‘a man’s sister; a man’s wife’s sister’ (L: 380; B: III, 474) or -́uỵ (< IE *h2euh2ii̯os) pl. -́uỵċaro and ‘father; father’s brother,’ Ng also ‘polite form of address to superiors’ in pl. ‘forefathers’ (L: 59; B: III, 460), which we derive from *hui̯o/us-ċaro < *hui̯os-tar-o (2.2.2.2).

2.2.2.2 B  ur ‘father,’ ‘father’s brother,’ ‘forefather’ : IE ‘grandfather,’ ‘ancestor on father’s side’ Bur -́uỵ (< IE *h2euh2ii̯os) pl. -́uỵċaro and -́u (< IE *h2éuh2-) pl. -́uċaro ‘father; father’s brother,’ Ng also ‘polite form of address to superiors’ in pl. ‘forefathers’ (L: 59; B: III, 460). There is a direct correlation of Bur -́u with IE *h2éuh2- ‘grandfather:’ Hitt huhha ‘ancestor on father’s side,’ Arm hav ‘grandfather,’ Lat avus ‘grandfather, greatgrandfather,’ avunculus ‘uncle’ (< *h2euh2Vn-tlo-s > Fr oncle, Alb ungj ‘uncle’), avia ‘grandmother,’ Lith avýnas ‘uncle,’ PSl *ujĭna ‘aunt’ (‘who belongs to the uncle,’ formed with IE poss. –ĭ̄no). Bur -́uỵ derives from IE *h2euh2ii̯os ‘father’s father, ancestor on father’s side, uncle:’ PSl *ujĭ ‘uncle (on mother’s side),’ and further derivatives like OIr aue ‘grandson,’ ON āi ‘grandfather’ (Pok: 89: < *au̯o-s; M-A: 609; in Wat: 6, without a laryngeal: *au̯o-) and possibly Alb vëlla (pl. vëllezër) ‘brother’ (see 2.2.5) (Gluhak 1993: 651–652). This form is not found in Indo-Iranian. See also 2.2.5, 2.2.6 and 2.2.8, i.e. Bur Ys –húles, –húlus ‘woman’s brother, husband’s brother’ (B: III, 455) possibly from the same stem, where the laryngeal is retained. In Burushaski, IE ii̯ (or i̯ i̯) > ỵ, as in huyés ‘sheep and goats’ + suffix –yo- = huỵóo ‘wool-bearing animal, sheep’ (see Čašule 2009: 155, 179–180). The laryngeals can be lost in Hunza and Nager after the pronominal prefixes and in the inlaut. There is also Bur Hz Ng áỵa, NH also Ng áỵayo ‘my father!’ “a vocative used in addressing fathers and father equivalents, e.g. paternal uncles, step-father, foster-father, father-in-law, but it is also said that in recent times it has ceased to be used for uncles:” “because there is only one father,” also simply ‘my father’ (with the first p. sg. prefix a-) (L: 6; B: III, 460; Will: 17: ‘father; father-in-law; father’s brother’ [also used as a term of address]) and áỵaku ‘elders, parents.’ Bur



Burushaski kinship terminology of I­ ndo-European origin   

   193

áỵa can be derived possibly from a variation of *h2euh2ii̯os (eu > a) > *h2ai-i̯as (yet with problems in chronology), although it may well be of different origin. It could simply be an assimilated form of a Bur *-́uỵ-o > *-́uỵ-a > *aỵa. Consider also the Burushaski patronymic suffix –áaỵ (B: I, 208–209) which reflects the oldest form and most likely derives from áỵa – the vowel length in the suffix might be a consequence of the lost laryngeal.

2.2.3 ‘son,’ ‘daughter’ While Burushaski does not have a kinship term corresponding to the more widely attested Indo-European stems for ‘son,’ i.e. *suhxnus- and *suhxyús, there is a possibility that it may actually have the underlying Indo-European verb *seuhx‘bear, beget’ (M-A: 533) in Bur súas, súyas, dusúas, dusúyas vt. ‘to bring; to take, fetch; to procure; to buy,’ also d-̇-ċ-, Ng abs. d-̇-ċu- ‘bring; to carry a load’ (B: III, 383; Will: 103). For the semantics, note the meaning of IE *bher- ‘to bear a child’ and ‘to bear a load’ (Wat: 10).

2.2.3.1 Bur ‘son,’ ‘daughter,’ ‘children’ : IE ‘young’ The original Burushaski terms for ‘son’ and ‘daughter’ derive from the same root and are thus discussed together. Hz Ng –í, Ys –yé, pl. –yú also –yúa ‘son,’ pl. also ‘children’ (B: III, 210; L: 41, 386, also with –ū-). For ‘daughter:’ -̇i-, pl. Hz Ng –yúguśanċ, Ys pl. –yúguśiṅa ‘daughter,’ also ‘brother’s daughter’ (B: III, 210; L: 12, 386). The plural form consists of –yu- and gus ‘woman’ + plural suffix (another example of the singular stem retained in the plural). Considering the Burushaski plural forms in –yu-, we propose a derivation from a zero-grade form of IE *hai̯eu̯- ‘young’ (in W-I-S: 277–287: *h2ei̯-u-, *h2oi̯u-, *h2i̯-eu̯-) : OIr ōa ‘younger,’ MWels ieu ‘younger,’ OEng geong ‘young,’ Goth juggs ‘young,’ Lat iuvenis ‘young,’ Lith jáunas ‘young,’ Lett jaûns ‘young,’ OChSl junŭ ‘young,’ Av yvan- ‘youth,’ OInd yúvan ‘young,’ originally from IE *haói̯us ‘strength’ (M-A: 655). For the semantic shift ‘young’ > ‘son,’ note OEng mago ‘son, man,’ OWels map ‘son’ < IE *maghus- ‘young man’ or Gk kóros ‘boy, son’ < *koru̯os ‘youth’ (M-A: 656). The laryngeals in Hunza and Nagar can be lost after the pronominal prefixes.6 6 Compare with the development in Bur –ík (Cunn “goyak,” Hayward 1871 “gúyek” – the latter < *gu-yek ‘your name’), Bur Ys –yék ‘name; good name, good reputation,’ iík dilá Ng ‘one says, it is said,’ -é ... –ík ét-, óos- ‘...to name, to call, to give a name’ (L: 42; B: III, 211) with –k < devoiced

194   

   Ilija Čašule

There is also Bur Ys yóṭes ‘young of animal,’ with initial y- in contrast to Hz Ng jóṭis ‘small child’ (B: III, 228, who derives it from joṭ ‘small, young man’ and correlates it with T 5071, without forms in the surrounding Indo-Aryan languages). The Nager pl. of joṭ is joṭarko and joṭumuċ. The form joṭ-arko can be analysed as joṭ-ar-ko ‘he (they) who is (are) young’ – compare with Latin *neu̯-er-ko > noverca ‘stepmother’ (‘she who is new’) (Wat: 58). Refer further to the alternation j : y in Ys jaġá, Hz Ng -yaġá (B: III, 470), j̣ú- : d-̇y(a)- ‘to come’ (of Indo-European origin), or Bur yaqhú < Turk jakki (B: III, 472), which indicate a possible change y > j. Berger (2008: 44) gives also Ys ten-jó < *ten-yo, dan-jó < *dan-yó and j : ỵ in ġajámiśo : ġaỵámiśo (B: III, 166). The Yasin form with y-, the old Indo-European nom. ending -is and the Nager plural suggest a merger between two forms. There is also Bur juáan, Ng also juwáan ‘young man; soldier’ (B: III, 229) which is considered a loanword from Urdu jawān.

2.2.3.2 Bur ‘boy,’ ‘young lad’ : IE ‘boy,’ ‘young man,’ ‘son’ Bur Ys dúlas ‘boy, young lad’ (BYs: 142), Hz Ng hilés ‘boy, lad, youth (unmarried); child, infant’ (L: 201–202; B: III, 198). The last form is considered secondary by Berger (2008: 55), derived from the form with d-. Compare with Lett dêls ‘son,’ Alb djalë ‘boy, young man, son, young animal’ reconstructed by C̣abej 1976 (apud Desnickaja 1984), from OAlb *delās ‘a boy,’ whereas in Demiraj (1997: 134) the reconstruction is from IE *dhih1-a(o)l- and Lat fīlius ‘son’ (for the Latin form see the alternative interpretation, discussed in 2.2.3.4.). These are derived from IE *dhē(i)- ‘to breastfeed’ (Pok: 242), also with the -l- formant, i.e. < IE *dhh1ileha- ‘teat, breast,’ *dheh1lus ‘nourishing, suckling’ (M-A: 82, who consider this form a North-Western and late Indo-European word for ‘teat, breast’). In Burushaski dúlas < *dhh1ileha-s > -díl, with the alternation i : u /_l, r (for the process and examples, see Berger 2008: 10–11). The stem is strongly represented in Burushaski and cannot be traced phonologically or semantically to Indic or Iranian. Other derivatives are: Ys –díl (BYs: 142), Hz Ng –ndíl ‘breast, chest’ (L: 276; B: III, 302) (cp. with MIr deil ‘teat,’ OHG tila ‘woman’s breast,’ OEng delu ‘breast, teat,’ ON dilkr ‘lamb’). Bur díltar ‘buttermilk’ : e.g. Alb dhallë ‘buttermilk’ (yet note the different etymology by Demiraj [1997: 153] from IE *ĝ(l̥)lákt ‘milk’) also Alb dele ‘sheep’ (Demiraj 1997: 127) and further Illyr dalm- ‘sheep’ (Pok: 242). –g in word final position, which we derive from IE *h1eĝ- or *h1eh1ĝ- ‘say’ (Pok: 290: *ēĝ-) : Lat āiō ‘say,’ adāgium ‘proverb,’ axāre ‘+/- call by name, give a name to,’ Gk ē ‘say,’ Arm asem ‘say,’ TochAB āks- ‘announce, proclaim, instruct’ (M-A: 535 – widespread and old in Indo-European).



Burushaski kinship terminology of I­ ndo-European origin   

   195

Bur du ‘kid (up to one year old of either sex),’ NH also dúdo (L: 139; B: III, 123) (see the full analysis of this shepherd term in Čašule 2009a: 161). Possibly here also Bur dúno ‘teat of udder’ (B: III, 125) (cf. from the same stem OIr denaid ‘teat’ [< *dhi-na-ti], dīnu ‘lamb’).

2.2.3.3 B  ur ‘(strong) young man’ : IE ‘free man,’ ‘children,’ ‘people,’ ‘strong man’ Bur láuḍar ‘strong, young (of men, lads); strong young man (15 to 20 years old)’ (also in Sh) (B: III, 264). There is a direct correspondence with IE *h1léudheros ‘belonging to the state of being a free man,’ (W-I-S: 245–246: *h1leu̯dh-ero/(ah2)-) e.g. Lat līber ‘free,’ līberī ‘children’ (as opposed to slaves), Gk eleútheros ‘free’ and with *h1léudhis‘people,’ e.g. OEng lēod ‘person,’ OHG liut ‘person,’ NHG Leute ‘people,’ Lith liáudis ‘common people,’ PSl *ljudĭje ‘people,’ Srb, Croat ljudi ‘people,’ Russ ljudi ‘people, servants,’ found throughout Slavic, and PSl *l’udĭ ‘same’ – Srb, Croat ljuda (arch.) ‘very fat and strong but clumsy man,’ ljudo ‘corpulent, but clumsy man,’ ljudina, ljudeskara ‘a huge man’ < IE *h1leudh- ‘grow,’ from which we have: OEng lēodan ‘spring, grow,’ OInd ródhati ‘grows,’ Av raoδaiti ‘grows’ (M-A: 248; Pok: 684–685; Trubačev 1974, XV: 194–200), also in the name of Lydia: Lydia, Lydos ‘Lydian’ (Gluhak 1993: 392). Demiraj (1997: 417) derives from this stem Alb vëlla ‘brother’ – see the discussion in 2.2.5. As pointed out by Mallory and Adams (1997: 248), “the focus of this word’s meaning was apparently the growth to maturity of humans. Widespread and old in Indo-European.” The precise correspondence with the Burushaski meaning is remarkable. Semantically, the Burushaski word encapsulates the basic original meaning postulated for Indo-European and parallels further specific developments directly (especially with Slavic). There are some phonological difficulties, as we would have expected a change dh > t in the inlaut in Burushaski and monophthonghisation of the diphthong eu > a. In the case of the diphthong, the position of the stress and the initial laryngeal could have prevented the monophthongisation. The form of the suffix could have been influenced by Bur ḍar ‘quite, totally; full (of moon)’ (L: 113; B: III, 131). On the other hand, it may be the case that this is a very ancient loanword into Burushaski. The length of the etymon and the close semantic match, even if with some phonological irregularities, have led us to include it in our comparisons.

196   

   Ilija Čašule

2.2.3.4 Bur ‘bride, newlywed woman’ : IE ‘daughter’ Bur biléeli ‘bride, newlywed woman.’ In Sh biléelo, fem. biléeli ‘darling,’ with the adjectival suffix –eelo, productive in Shina and in Burushaski. Berger gives no other derivatives in Indo-Aryan, which indicates that this is most likely an original Burushaski word. Allowing for a small shift in meaning, we can compare it with Messapian bilia, biliva ‘daughter,’ which has been correlated with Alb bijë, Alb dial. bilë ‘daughter’ (Neroznak 1978: 167). Hamp (1972: 220–225) equated the Messapic and Albanian words with Lat fīlia ‘daughter,’ all from *bhwīliā (Huld 1984: 44 reconstructs */bh(u)eil-ieA2/), i.e. from the root *bheuhx- ‘to be, exist, grow’ (Wat: 11, 18; in W-I-S: 46–58: *bhu̯eh2-, *bhuh2-). As there is an alternation i : u in Burushaski before r and l (Berger 2008: 10–11), biléeli could derive from *bul-eeli. Note further in Burushaski biličan ‘(state of) being’ (“probably based on the 3. y sg form of the verb ba-” L: 81; B: III, 53 biličar < bá-, b- ‘to be,’ which we derive from IE *bheuhx‘to be,’ with the regular change eu > a). Burushaski also has bíiro ‘male (of animals),’ NH also ‘courageous’ (L: 82: biru ‘male;’ B: III, 52), which is most likely a borrowing from Indo-Aryan. Berger notes Sh bíiro ‘male (of animals)’ and points to T 12056, i.e. vīrá ‘man, hero, son,’ RV ‘male (of animals).’ Note, however, the coincidence with Alb bir ‘son,’ which is derived from *bher- ‘carry, bring,’ although Ribezzo (q. in Huld 1984: 44) regarded the form a secondary masculinization of Alb bijë ‘daughter.’

2.2.4 Bur ‘a man’s sister; a man’s wife’s sister’ : IE ‘sister’ Bur –yás pl. –yásċaro, Hz pl. also –yástaro, Ys –yást, pl. –yástaru ‘a man’s sister; a man’s wife’s sister’ (L: 380; B: III, 474), yás-muúlus ‘sister and brother’ (Will: 125). Can be compared with IE *su̯ésōr ‘sister,’ e.g. OIr siur, Lat soror, Eng sister, OPruss swestro, OChSl sestra, Arm k’oyr, Av xvaŋhar-, Skt svásar-, TochB ṣer, all ‘sister’ and Gk éor ‘cousin’s daughter’ (absent in Albanian and Anatolian – Hittite uses the unique forms nēgna and neka resp. for ‘brother’ and ‘sister;’ M-A2: 214). In Burushaski we suggest a dissimilation from the zero-grade form *su̯éstr > *yestr- > *yes-taro > *yas-taro (s-s > y-s). Consider in this respect the loss of –safter the pronominal prefixes in Bur d-̇sil- and d-̇sili- ‘to wet (of water, rain)’ in the Hz Ng form di-íl-, Ys di-híl-, pl. x du-húlja-, Hz also di-síl-, Ng also di-ċhíl- ‘to become wet, get wet, to be soaked’ (B: III, 212).7 In Burushaski ye > ya – consider 7 We derive these verbal forms, together with Bur ċhil, Ys: ċel ‘water; juice, sap’ (B: III, 76) and du-súlġu- ‘become fluid, watery,’ d-̇ċhulġu-, d-̇ċhilġu- ‘make watery’ (B: III, 77, 384), from IE *su̯el-,



Burushaski kinship terminology of I­ ndo-European origin   

   197

the Burushaski alternation –yénum : –yánum (B: III, 477), or Hz Ng ye, Ys ya (B: III, 477). In Berger (1998) there are only 2 words with ye- and in both cases there is an alternation with ya-, and only 3 examples with the pronominal prefixes, making a historical change ye- > ya- feasible. The Ys form –yást might be an indication that the –t- is part of the stem, later reinterpreted as part of the plural form, i.e. *yestr- > *yest- + *-taro.

2.2.5 Bur ‘woman’s or husband’s brother’ : IE (Alb) ‘brother’ Bur –úlus, pl. –úluċaro, Ng pl. –úlusċaro, (very significantly) in Ys with the laryngeal preserved: -húles, -húlus ‘a woman’s brother, husband’s brother’ (L: 54; B: III, 455). The semantic correlation (Burushaski reflects exactly the reconstructed semantics), and surface phonological resemblance with the Albanian vëlla ‘brother’ (derived by Huld 1984: 128–129 [in his notation] from PIE *A1euH-m-tleA2-, with metathesis (sem. < ‘cousin’ < ‘mother’s brother’(s son)’ (see 2.2.2.2.) is highly significant. Demiraj (1997: 417) derives the Albanian word possibly from *u̯ai-láu(d)- and this from *su̯oi-h1loudhi- (an old compound with the reflexive pronoun). In Mallory and Adams (2006: 609), the Albanian word for ‘brother’ is reconstructed from *waládā < *awádlā, ultimately from *h2eu̯óh2-dhlo- (‘mother’s brother?’). For the Burushaski word, however, we propose a much simpler derivation, from *h2au-lo-s, considering the Burushaski nom. and adj. suffix -éelo (found also apocopated in other original Indo-European words like ġaśíl, -díl, see Čašule 2010: 61), as in Bur karéelo ‘ram’ (also in Sh) (B: III, 242; L: 229: karéelu) which is of autochthonous Indo-European origin (see Čašule 2009a: 166–168), or nams ‘greed, greediness’ > nams-eélo, Ng namsiílo ‘greedy,’ also ćhar ‘stone,’ ćhar-eélo ‘climber’ (B: I, 210). The suffix is perhaps relatable to IE *-lo-, secondary suffix, forming diminutives, and a nom. and adj. suffix: Lat -ellus dim. suffix, -ilis adj. suffix, OEng *-ling dim. suffix and nom. suffix (Wat: 50). It would seem that we may have both a borrowed suffix from Shina and a more ancient original suffixal formation. The common semantic development, the retention of the laryngeal and the Indo-European nominative sg. ending –us support strongly this reconstruction.

*sul- ‘to wet, moisten; flow; (noun) liquid, fluid, moisture, sap,’ *seu-, *sū – ‘juice’ (Pok: 912–91; Mann 1984–1987: 1334–1335).

198   

   Ilija Čašule

2.2.6 B  ur ‘grandchild, grandson, granddaughter’ : IE (Sl, Grmc) ‘grandchild, grandson, granddaughter’ Bur -̇mis Hz Ng, -̇mes Ys ‘grandson; granddaughter’ (L: 17; B: III, 289). And further -pháṭ -̇mis NH ‘grandson,’ the first word only registered in this syntagm. Parkin (1987a: 326) compares tentatively Bur -̇mis to Tib mes ‘grandfather,’ but apart from the semantic divergence (Bur ‘grandchild’ : Tib ‘grandfather’), the word is not found in Balti, the Tibetan language with which Burushaski is in contact. We derive Bur -̇mis from *ngis or *nges and thus, considering the Burushaski recent change ng, nk > m,8 it is correlatable internally with Bur -́ṅgo, pl. -́ṅgoċaro, Ys -́ṅgu ‘uncle (paternal or maternal)’ (L: 276; B: III, 306), with the suffix –is, and with a possible semantic shift from ‘nephew’ to ‘grandson’ (as in Macedonian below). We propose a derivation *h4en-ki-s (after aphaeresis of a-) > *nkis > -̇mis. There is a direct correlation with OChSl vnukŭ ‘grandson,’ vnuka ‘granddaughter,’ e.g. Mcd vnuk, dial. mnuk ‘grandson’ and ‘nephew,’ Bulg vnuk, Serb, Croat unuk, Sln vnuk, Russ vnuk, Cz, Slk vnuk, Pl wnuk, USorb wnuk all: ‘grandson.’ The Slavic words have been derived from IE *h4en-ou-ko-s, i.e. a diminutive or possessive to IE *h4eno(u)- ‘grandfather,’ derivationally the same as OHG enenchelī, OHG eninkel, enenkel, Grm Enkel ‘grandson’ are to OHG ano ‘ancestor, grandfather, greatgrandfather,’ and related further to Lith anýta ‘mother-in-law,’ OHG ana ‘grandmother, great-grandmother, female ancestor,’ Lat anus ‘old woman,’ Hitt hanaš ‘grandmother,’ Illyr ána ‘relation’ (Georgiev 1971: I, 167), also OPers nyākā ‘grandmother,’ Pashto anā ‘mother,’ Arm han ‘grandmother,’ Gk annis ‘grandmother.’ While the original stem *h4en- is a nursery word, the precise derivational Burushaski pattern that parallels Slavic and Germanic argues very strongly for an Indo-European, North-Western origin. (See also 2.2.8.2.) The isolated Bur -pháṭ in -pháṭ -̇mis ‘grandson’ (B: III, 327) can be correlated with IE *népōts (fem. *neptī-) and the derivative *neptiyos ‘grandson, nephew,’ e.g. Lat nepōs ‘grandson, descendant’ (in later Latin, also ‘nephew’), OEng nefa ‘sister’s son, grandson,’ OIr nia ‘sister’s son, grandson, descendant,’ Lith nepuotìs ‘grandson,’ OChSl netijĭ ‘nephew,’ Alb nip ‘grandson, nephew,’ Gk 8 For example: Bur Ys -núṅus vs. Hz Ng –dúmus (in Tiffou andMorin 1989: 40, Ys also -númus) ‘knee, hock’ (L: 149; B: III, 125). And further -múṅgus ‘corner’ (B: III, 294, with –m- from the pron. prefixes), which we derive from IE *ĝonu- (gen. *ĝenus) ‘knee, also angle’ (M-A2: 183), i.e from *gnun-g/kus. Also: Bur Ys -yúṅus, Hz Ng -úmus ‘tongue’ (B: III, 455–456), derived from IE *dn̥ĝhuha- ‘tongue’ (M-A2: 175), Hamp (p.c.): *(d)n̥ĝhuha-. The Yasin form is older, as the phonological change is ṅ > m and Hayward gives an older Hunza Nager form “unas,” which Berger explains < *u-úṅus. And further, the Burushaski adjectival suffix –um goes back to –uṅ, e.g. Bur Cunn matung, (B: III, 284): matúm ‘black.’ We derive the Burushaski suffix from the IE adjectival compound suffix *-enko, *-n̥ko-, as in OEng, Eng suffix –ing < Grmc *-inga-, *-unga- (Wat: 36).



Burushaski kinship terminology of I­ ndo-European origin   

   199

népodes ‘descendants,’ OPers napā ‘grandson, descendant,’ Skt nápāt ‘grandson, descendant’ (M-A2: 211; Wat: 58). Pokorny (1959: 764) suggests that it is a compound word consisting of ne- ‘not’ and potis ‘powerful, lord,’ with a meaning of ‘underage, dependent’ (e.g. Lett pats ‘master of the house,’ Russ gospodĭ ‘host,’ Alb zot ‘master of the house,’ Gk pósis ‘husband,’ Hitt pat ‘self,’ Av paiti‘husband,’ Skt páti- ‘husband, master,’ TochA pats ‘husband’ [M-A2: 207]), The a-vocalism in Bur -pháṭ could be an indication of an Indic or Iranian loanword (yet note the Burushaski retroflex), or a merging of two very similar words. As this word is used with the pronominal prefixes, the loss of an initial *ne- cannot be ruled out. Nevertheless, in the construction -pháṭ -̇mis ‘grandson,’ we appear to have a meaning -pháṭ ‘the master’s’ (from *potis) + -̇mis NH ‘grandson,’ which would be an original Burushaski development. Very tentatively, we should mention here Bur -pháṭi ‘forehead’ (L: 154; B: III, 327) – consider for the semantics, e.g. Mcd na čelo (lit. ‘at the forehead’) na grupata (‘of the group’) ‘leading the group.’

2.2.7 Bur ‘husband,’ ‘married man’ : IE ‘man,’ ‘husband’ Bur Hz Ng -úyar pl. -úyariśo (L 57 also –uyer), Ys –yúhar ‘husband, married man’ (B 460) (the Yasin form most probably from *uyhar by metathesis). There is a direct correlation with IE *u̯ihxrós ‘man, husband’ (Wat: 101 *wiəro) OEng wer ‘man, husband,’ ON verr ‘husband,’ Lat vir ‘man, husband,’ OIr fer ‘man, husband,’ Lith výras ‘man, husband,’ Av vīra- ‘man, person (as opposed to animals),’ Skt vīrá ‘hero; (eminent) man, husband’ (M-A2: 202; Pok: 1177), also TochA wir ‘young, fresh’ (W-I-S: 726- 729). This word is one of the few in Yasin Burushaski where the laryngeal is retained in the inlaut in an original Burushaski word. This is apart from the verbal forms or nouns after the pronominal prefixes, i.e. at a morpheme boundary, as in Ys -húlus, Hz Ng -úlus ‘a woman’s brother, husband’s brother’ above, or e.g. in Ys d-̈hémia- : Hz Ng d-̇-mi- ‘gather’ (B: III, 287) < IE *h1em- ‘take, gather’ (M-A: 564). Thus it is very unlikely that it is a loanword from Indo-Aryan or Iranian. For example, in Burushaski we have Ys bihái, Hz Ng biái ‘illness, disease’ (B: III, 50), where the retention of the laryngeal clearly points to an original Burushaski word, from IE *pehx(i)- ‘misfortune,’ *pihx- ‘revile’ (e.g. Av pāman- ‘skin disease,’ OInd pāmán- ‘skin disease,’ OInd pī́yati ‘insults’ (elsewhere in IndoEuropean with a meaning of ‘suffering; hate’) (M-A: 313). In addition, the Burushaski word is not found in the neighbouring IndoAryan languages, and Lorimer (L: 57) indicates that the place of -úyar is very often taken by the Urdu borrowing jamáat ‘husband, wife’ (B: III, 222).

200   

   Ilija Čašule

2.2.8 ‘uncle,’ ‘aunt’ 2.2.8.1 Bur ‘uncle (paternal or maternal)’ : IE ‘uncle’ Bur -́ṅgo, pl. -́ṅgoċaro, Ys -́ṅgu ‘uncle (paternal or maternal)’ (L: 276; B: III, 306). We can relate this term to the same Indo-European stem as in 2.2.2.2. : Bur -́u (< IE *h2éuh2-), pl. -́uċaro ‘father; father’s brother,’ Ng also ‘polite form of address to superiors’ in pl. ‘forefathers’ (L: 59; B: III, 460), with aphaeresis of the initial u-. Note furthermore that no Burushaski word begins with ṅg-. In this instance, the derivation could be perhaps < IE *h2euh2o(n)- (as in OHG ōheim ‘mother’s brother,’ derived from IE *h2euh2o(n)-+ *koimos, i.e. ‘living in grandfather’s home’ (M-A: 609) + suffix *-ko or more simply < IE *h2euh2ino-, as in Lith avýnas ‘uncle,’ PSl *ujĭna ‘aunt’ (‘who belongs to the uncle,’ formed with IE poss. –ĭ̄no) + *-ko.9 The exact semantic and precise phonological match of Bur -́ṅgo, with Alb ungj ‘uncle,’ or for that matter also with the Albanian source, i.e. Lat avunculus ‘mother’s brother; mother’s sister’s husband,’ should be noted, even though the derivational histories cannot be reconciled at this stage. Corresponding to the IE adj. suffix *(i)ko, Burushaski has the adj. suffix –ko (e.g. datú ‘autumn’ – datúko ‘autumn’ (adj), also -kus (< IE *-ko-s) datúkus ‘autumn season’ (B: I, 206–207; B: III, 116–117) and derivatives from the Indo-European composite suffix *-isko-, in Burushaski with assimilation –iski etc. as in Bur hir ‘man,’ hiríski adj. ‘man’s, male’ (B: III, 200), and of course Burúśin ‘Burushaski speaker,’ rel. poss. adj. Burúśaski ‘the Burushaski language’ (B: III, 491). Burushaski also preserves the IE adj. compound suffix *–enko, *-n̥ko > Bur –um < -uṅ as in matúm ‘black’ (Cunn. older matung) or datúmo ‘autumn-‘ (< *datúngo). Burushaski also has the adj. and nom. suffix –(e)n, e.g. Ys ġḗ- ‘to steal’ : ġḗn ‘thief’ (B: III, 175), which corresponds to IE –en- ‘suffix forming nouns and adjectives (with many variants)’ (Wat: 23). Note further Bur -́umo ‘fatherly; kinship, relations on father’s side’ (B: III, 460), where we have the adj. suffix –mo < *-n̥go < *-n̥ko, i.e. from an older *-́ungo, which could also explain Bur -́ṅgo. In Burushaski we would have –nk- > -ng- and aphaeresis of the originally initial u- due to the obligatory pronominal prefixes (as in e.g. d-̇-ṅġur- : do-óṅġur(B: III, 307). The loss of the initial u- could also be a result of disambiguation with the 2p. sg. pron. uṅgó ‘you here’ (B: III, 456). In addition to the etymological solutions proposed above, Bur -́ṅgo could simply be from *u- (< *h2au-) + the IE adj. ending *–enko, *-n̥ko , in the sense of ‘belonging to the uncle,’ as in the Lithuanian and Slavic examples above. 9 For the use of –ko as a kinship suffix in Indo-European note e.g. Mcd tatko ‘father,’ vujko ‘uncle (on mother’s side),’ striko ‘uncle (on father’s side),’ čičko ‘uncle (on father’s side).’



Burushaski kinship terminology of I­ ndo-European origin   

   201

2.2.8.2 Bur ‘uncle,’ ‘aunt,’ ‘older woman’ : IE ‘grandmother,’ ‘aunt,’ ‘old woman,’ ‘nurse’ Bur nána also nané (L: “the form nane is properly used only in Rajas’ families, nana being used by aristocrats and commoners”), pl. nánaċaro ‘uncle (on mother’s and father’s side); aunt’ (in L: “not recorded for maternal aunt”), ‘respectful reference to older men or women’ (L: 274; B: III, 300). It is noted as an original word by Berger, who doesn’t give parallels in the neighbouring Indo-Aryan languages. It can be compared to IE *n-h4en- ‘mother’ (considered by some a reduplication of *h4en-) : NWels nain ‘grandmother,’ Late Lat nonnus ‘nurse,’ Late and Medieval Latin nonna ‘aunt, old woman; nun,’ Alb nëne ‘mother,’ Russ njanja ‘nurse,’ NPers nana ‘mother,’ Skt nanā- ‘mother,’ and especially the remarkable direct correspondence with Gk nánnē ‘female cousin, aunt’ (also nanna ‘aunt,’ nannos and nānos ‘uncle’) (M-A2: 213; Pok: 754; Wat: 57). Under one interpretation, Bur Hz Ng -́nċo , Ys -́nzu pl. Hz Ng -́nċoċaro ‘aunt on father’s side; father’s sister,’ also ‘mother’s brother’s wife’ (L: 281; B: III, 302), could derive by aphaeresis < *h4en- and thus be related to Bur nána. (See further 2.2.1. and 2.2.8.3.) Parkin (1987a: 327) notes in a general way, without Shina or Khowar examples, that Bur nána is similar to Indo-European nani, but more probably related to the “practically universal” Tibeto-Burman root represented in Newari ni-ni, Balti nēne, Kinnauri nāne, aṇe, Bhotia, Sherpa, Tamang ani, Tibetan ne, from PTB *ni ~ *nei all meaning ‘paternal aunt’ ,” which is somewhat different to Burushaski, where it means both ‘uncle (on both sides)’ and ‘aunt (mainly paternal).’ Considering the nursery word origin of such words, it is not surprising that they could appear independently in various languages and later merge in some ways through language contact. They cannot be significant evidence for a relationship on their own, but the possibility that they could be a part of the original vocabulary of Burushaski should not be rejected outright.

2.2.8.3 B  ur ‘aunt on father‘s side,’ ‘father‘s sister,’ ‘mother‘s brother‘s wife’ : IE ‘son‘s wife’ Hz Ng -́nċo , Ys -́nzu pl. Hz Ng -́nċoċaro (Lorimer gives all the plural forms with –u-, as in Yasin: -ntsutsaro, -ntsumuts, -ntsušo) ‘aunt on father’s side; father’s sister,’ and importantly also ‘mother’s brother’s wife’ (L: 281; B: III, 302). This word can be compared very tentatively with IE *snusós ‘daughter-in-law,’ e.g. Lat snurus ‘son’s/grandson’s wife,’ OEng snoru ‘son’s wife,’ Russ snokhá, Gk nuós ‘son’s wife; bride,’ Alb nuse ‘bride,’ Arm nu ‘son’s wife,’ Skt snuṣā́ ‘son’s

202   

   Ilija Čašule

wife’ (M-A2: 215). Note the Middle Dutch semantic extension/generalisation in snare ‘daughter-in-law, sister-in-law, mother-in-law.’ Even though there is a significant semantic shift ‘daughter-in-law’ > ‘aunt’ the formal correspondence appears to be very close and warrants attention : *snusos > *snusos > *nusos > *nso/us = sg. -́nċo (o-stem) and pl. Hz Ng *-́nċostaro > *-́nċosċaro > -́nċoċaro. For a different interpretation, refer to 2.2.8.2. Perhaps there is some (derivational and semantic) contamination with IE *h1i̯enha-ter- ‘husband’s brother’s wife,’ e.g. Lat ianitrīcēs ‘brothers’ wives,’ Gk enátēr, Lith jéntė, OChSl je̜try, Arm ner, Skt yātár, all: ‘husband’s brother’s wife’ (M-A2: 216; W-I-S: 204–207: *(H)i̯enh2ter-). 2.2.9 Bur ‘father-in law’ : IE ‘father-in-law’ Bur -̇- skir, pl. -̇- skindaro, Ng pl. -̇- skiriśo ‘father-in-law, wife’s father or wife’s father’s brother or husband’s father’ (B: III, 381; L: 26). Both Lorimer and Berger suggest a derivation from ? + hir ‘man’ (in L 203, also hīr). Note the alternation u : i common in front of r, l (Berger 2008: 10–11): Ys pl. hurí and huríkia, Hz Ng hiríski : Ys huríski ‘pertaining to men, man’s’ (B: III, 200). Compare with IE *su̯ék̂uros ‘father-in-law:’ NWels chwegrwn, Lat socer, OEng swēor all ‘father-in-law,’ Lith šēšuras ‘husband’s father,’ OChSl svekrǔ ‘husband’s father’ (one of the “centum” words in Sl), Alb vjehërr ‘father-in-law,’ Gk hekurós ‘wife’s father,’ Av xvasur ‘father-in-law,’ Skt śváśura ‘father-in-law’ (M-A2: 215, who cite Szemerényi’s suggestion of a deeper etymology from IE *su̯é- ‘own’ + *k̂oru- ‘head’) (W-I-S: 672–675, also Arm skesraiur ‘father-in-law,’ skesowr’ ‘mother-in-law’). The Ng x pl. -̇- skiriśo and the x pl. ending -iśo in general can be re-analysed as *-is-yo, with -is- being the Indo-European animate nom. pl. ending -es i.e. Bur * -̇- skiris+yo < *skires+yo with *-yo correlatable with the Indo-European relational adjectival suffix –i̯o- ‘of, or belonging to’ (Wat: 103). The pl. -̇- skindaro derives from *skir-tar-o by dissimilation, and by analogy to other nouns denoting relations with the suffix –taro. Other examples of IE *su̯- > Bur s- are: IE *su̯erhxK- ‘watch over, be concerned about’ (M-A: 636) > Bur sarké ‘visible, place from which one can watch’ (B: III, 376), IE *su̯el-, *sul- ‘to wet, moisten, flow; liquid, fluid, moisture, sap’ (Pok: 912– 913) > Bur ċhil ‘water’ (B: III, 77), d-̇sil- ‘make wet, water intensively’ (B: III, 384). For Indo-European k̂ > Bur k, see the detailed exemplification in Čašule (2010: 40–50). The Burushaski kinship term has the same prefix s- found in Bur -̇- skus ‘mother-in-law, wife’s mother, husband’s mother’ < gus ‘woman’ (B: III, 162). We



Burushaski kinship terminology of I­ ndo-European origin   

   203

have derived the prefix from IE *su̯é- (also *se-), reflexive pronoun meaning ‘(one) self’ (Fortson 2004: 130; M-A: 455: *séu̯e (acc.) ‘-self’) or *s(u)u̯-o- ‘one’s own’ (“widespread and old in IE”), e.g. Av hva- ~ hava- ‘one’s own,’ OInd svá ‘one’s own,’ TochA ṣñi ‘one’s own,’ Lat sē ‘him-/her-/itself,’ (poss. adj. suus), OChSl se̜ -self,’ OPruss sien ‘self,’ Lett sevi- ‘-self,’ OHG sih, Goth sik ‘him-/her-/itself’ etc. There is also IE *sū- ‘joint family’ from which we have derived Bur Ys ses, Hz Ng sis sg. and pl. ‘people, folk; person, man’ (B: III, 380) (see 2.2.16) and the first element in Bur sukúin ‘kinsman, blood-relation (descended from a common ancestor); near relation, such as cousin’ (B: III, 384) (-kuin is the denominal suffix for derivation of names of professions, B: I, 137). For further derivatives, refer to 2.2.10. Especially in this example, but also in a very small number of other cases it appears that Bur h- may originate from k- (Edel’man-Klimov 1970:25, 29; Morgenstierne 1945: 74; Čašule 2009). It is curious that the Albanian -h- reflex of IE -k̂ in vjehërr ‘father-in-law’ has also been difficult to explain and etymologists have had to resort to a metathesised stem *u̯esk̂uros (see the discussion and analysis in Huld 1984: 130–131). The h- in hir could be a result of dissimilation k – k > h – k,10 i.e. from an older form *ku/ir-ik-o, contained in the plural forms hiríkanċ and huríkia or in the derivatives hírkuṣ and Hz Ng hiríski : Ys huríski. Another explanation is that Bur h- < k- could have been influenced perhaps by Ys –yúhar ‘husband, married man’ (Hz Ng -úyar) (B: III, 460) which we have derived from IE *u̯ihxrós ‘man’ (2.2.7.) (M-A2: 202; Čašule 2003b: 32–33). The alternation in Bur of i : u in front of l, r (Berger 2008: 10–11), together with the Yasin forms, could point to an older form *skur- < *sék̂uros < *su̯ék̂uros. Consider in this respect Bur kurpá ‘older, experienced man, elder’ (B: III, 248), which in all likelihood retains the original stem *kur-. The element -pá could be the demonstrative adverbial suffix with the meaning of ‘side, direction’ (B: III, 309), as in hirúmpa ‘sharp edge of knife’ < hirúm ‘sharp.’ Perhaps here also Bur kuriáp -mán- ‘hold out, overpower’ (B: III, 247; L: 237). There would have been an apocope of the first element, after the shift of the stress onto the pronominal possessive prefixes. Compare for example with Bur dénkus < *dénekus or daltáśko < *daltásiko (see Berger 2008: 88–89). Tikkanen (2001) has suggested that the female form is the basic one, i.e. he derives Bur -̇- skus < *gús-gus ‘woman-woman’ in order to explain the initial 10 Consider further such examples: Bur huk ‘dog’ (B: III, 203) which is related to Ys kukúres, Hz Ng gukúrus ‘puppy’ (L: 173) Sh Gureshi kukúr (T 3329; B: III, 159); Bur Ys hesk, Hz Ng hisk ‘comb, loom, wrist’ (B: III, 200) < *kes-ko and this from IE *kes- ‘to comb, scratch, itch:’ e.g. Hitt kiss- ~ kisā(i)- ‘to comb,’ kiske/a ‘comb,’ Luw kiš ‘to comb,’ Lith kasá ‘braid,’ OChSl česati ‘to scratch, to comb’ (Pok: 585–586).

204   

   Ilija Čašule

–s- and proposes that the form for ‘father-in-law’ was modelled according to the form for ‘mother-in-law’ (Berger 2008: 141 accepts this interpretation). This is a weak etymological and somewhat difficult formal and semantic explanation. Semantically, -̇- skus refers not only to ‘mother-in-law on the wife’s side’ but also to ‘mother-in-law on the husband’s side’ which would preclude the ‘womanwoman’ interpretation. Furthermore we cannot have an isolated structural and derivational example. The etymology of Bur hir (L also hīr), Ys pl. hurí and huríkia, Hz Ng hiríski : Ys huríski (*hur < *kur- ?) is a difficult one. We are inclined to seek a correlation with IE *k̂ouh1ros ~ *k̂uh1ros ‘powerful’ : OIr cōraid ‘heroes,’ Gk kū́rios ‘having power, like a lord or master with full authority’ (Liddell and Scott 1968: 1013), Av súra ‘hero,’ OInd śávīra- ‘strong,’ śú̄ra- ‘hero’ (M-A: 448), under one isolated interpretation by Berneker, cited in Skok (1974 III: 370), a candidate for the second component in *su̯ék̂uros, accepted e.g. by Gołąb (1992: 85, ex. [55] and [56]), with the semantics of ‘lord of the opposite moiety.’ Parkin (1987b: 163) suggests that the Burushaski terms for ‘father-in-law’ and ‘mother-in-law’ may be “new coinings.” Taking into account the Indo-European antiquity of the terms, we consider them remnants from an original asymmetric non-prescriptive kinship terminology.

2.2.10 B  ur ‘female relation’ , ‘husband’s sisters and daughters’ : IE ‘husbands of two sisters’ From IE *se- or *sē- (< *seu̯e) ‘-self,’ we have Bur Ys salén, also selén (BYs: 175) ‘husband’s sisters and daughters,’ correlated tentatively by Berger (B: III, 378) with silaj̣ín ‘female relation, related women-folk’ (L: 314). Consider also -síldir ‘fathers of a married couple in relation to each other’ (the second component derived by Berger (2008: 34) with an inorganic –d- < hir) and -sílgus ‘mothers of a married couple in relation to each other,’ with gus ‘woman’ as the second component (B: III, 379). There is a direct correspondence of these words with developments from IE reflex. *su̯e-lo-, *su̯e-lii̯on or *su̯e-lihxon-, ‘brothers-in-law whose wives are sisters’ as in ON svilar ‘husbands of two sisters,’ Gk aélioi ‘brothers-in-law whose wives are sisters’ (Pok: 1046; M-A: 85 “word of north-west and centre of IE world”). The Burushaski vocalism in these derivations suggests origin both from *se- (in salén) or *sē- < *seu̯e (M-A2: 417) (in silaj̣ín, -síldir and -sílgus). From IE *su̯e-lii̯on > *saliyin > *salijin (and by metathesis) > Bur silaj̣ín is a full and direct correlation.11 11 A connection with OInd syālá ‘wife’s brother’ (T 13871) (< IE *si̯ō(u)ros ‘wife’s brother’) (M-A: 84) has to be excluded both on semantic and phonological grounds.



Burushaski kinship terminology of I­ ndo-European origin   

   205

For the change (alternation) y > : j in Burushaski note: Ys yóṭes : Hz Ng jóṭis (B: III, 228) (for other examples, see 2.2.3.1). It is very likely that Bur –sáġun pl. –sáġundaro and –sáġuyo ‘nephew, niece, child of brother or sister’ (L 306: “originally applied only to sister’s children;” B: III, 371–372), Yasin also ‘grandson’ (BYs: 174) contains the morpheme sa-. The second component would then be correlatable with an o-grade of IE *ĝenh1-, (also *ĝen-, *ĝnē, *ĝnō-, *ĝonh1-, * ĝn̥̄h1-) in words for ‘beget; bear; produce; be born’ (Pok: 373; Wat: 26), also represented in Burushaski. Compare with IE *ĝenh1-tōr ‘father, procreator’ : Lat genitor ‘procreator,’ Gk genétōr, OInd janitár ‘same’ (M-A: 195), or with IE *ĝenh1-ti- : Celt-Iber kentis m. and f. ‘child; son’ (W-I-S: 140) or IE *ĝnh1-ó- : e.g. Lat –gnus in prīvignus ‘stepson’ (W-I-S: 139). Most directly, note the precise correlation between Gk gónos ‘sperm, semen; child, procreation’ and Bur ġunó, Ys ġonó ‘seed (not of cereals); sperm, semen’ (in Sh gunóo) (B: III, 180), Ng gono (L: 186), to which Berger relates also ġunóṅ ‘newly obtained land, in which only grass will be sown’ (B: III, 180). There is also Bur du-ġún- ‘ripen, to mature’ and d-̇squn-, (Ys d-̇sqon-) ‘cause to mature (of sun, of people); have an idea, give a stimulus, make a suggestion’ (B: III, 179), which Berger also links tentatively with Bur ġunó above. (See Čašule 2010, ex. [102].)

2.2.11 Bur ‘relations by marriage’ : IE ‘husband’s sister,’ ‘brother’s wife’ Bur gul, gulgár, gulc̣hé and esp. the form given by NH gulbúl, all h plurals ‘relations by marriage; relationship by marriage,’ Ys gul, gúlkuṣ (B: III, 160; L: 173). Berger tentatively, perhaps because the word is not found in the surrounding Indo-Aryan languages, proposes a correlation with T 3330, i.e. Skt kula ‘herd, troop; race, family; noble family; house,’ H kul ‘herd, clan, caste, family,’ which is inconclusive, as the Sanskrit and Indo-Aryan semantic development, even if very close, seems to indicate more ‘relations within the family, race, caste’ rather than the more specific meaning of ‘relationship by marriage’ as in Burushaski. Such an analysis would leave the other suffixed and extended forms unexplained and would assume k- > g-. Bur gulgár can possibly be a compound, where the second component is Bur gar ‘marriage’ (B: III, 146). Bur gulc̣hé could also be a compound with the second component < Bur –ćhi- ‘to give’ (B: III, 99). With a slight semantic generalisation, we propose a correlation with IE *ĝl̥h3-u̯os ‘husband’s sister’ : Lat glōs ‘husband’s sister, brother’s wife,’ OChSl zǔlŭva ‘husband’s sister,’ Russ zolovka ‘husband’s sister, brother’s wife,’ Gk gálōs ‘same,’ in Hesychius, without ethnic identification: gális ‘husband’s sister, brother’s wife,’ OInd giri- ‘brother’s wife,’ esp. Phrg gellaros ‘brother’s wife,’ Arm tal ‘husband’s sister,’ of PIE status (M-A: 521).

206   

   Ilija Čašule

For the comparison with IE, esp. relevant is NH gulbúl, where assuming w > b ( a regular change in the anlaut) we would have: IE *ĝl̥h3-u̯os > *gul-bos and then by assimilation and partial reduplication: gulbúl.

2.2.12 Bur ‘a man’s wife’s brother,’ (~’brother-in-law’), ‘a man’s sister’s husband;’ ‘a woman’s husband’s sister’ (~ ‘sister-in-law’), ‘a woman’s brother’s wife’ ~ IE ‘young, new’ : ‘man, husband’ Bur –rík, pl. –ríkindaro, Ys –rék ‘a man’s wife’s brother’ (~ ‘brother-in-law’), ‘a man’s sister’s husband;’ ‘a woman’s husband’s sister’ (~ ‘sister-in-law’), ‘a woman’s brother’s wife’ (L: 303; B: III, 365). We have already noted in (2.2.7) the clear and direct correspondence between Bur -yúhar (Ys) and -úyar (Hz Ng) ‘husband, married man’ with IE *u̯ihxros ‘man, husband.’ A more general meaning has been suggested for the root of IE *u̯ihxros, possibly from ‘young, new,’ as in TochA wir ‘young, fresh,’ and Alb ri and re ( ‘neighbour’ or a reflection of the communal family, where the neighbour is the ‘son-in-law.’

2.2.14 Bur ‘spouse,’ ‘ancestor,’ ‘fourth generation marker’ : IE ‘fourth generation marker’ Bur hápkuin ‘spouse, husband or wife’ (B: III, 191; L: 194), also Ys hápa ‘household’ (BYs: 150). There is also -̇pkuċ Hz Ng ‘maternal relatives (related on the mother’s side),’ Ys ‘ancestors on father’s side,’ which Berger derives tentatively from *-hápkuċ, or considers it a plural form of -̇pi ‘grandfather, forefather, paternal ancestor; grandmother’ (B: III, 314). Lorimer explains that his informant in saying Ja ape Zimi is actually talking about his paternal great-great-grandfather (Lorimer 1962: 17), which indicates a direct semantic match with the Indo-European stem below. Parkin (1987: 326) notes only Bur api ‘grandmother’ (the a- is the 1. p. sg. prefix) and correlates it with the “standard and widely occurring Tibeto-Burman root for ‘grandmother’” from PTB *p’i > Balti api, Tib phyi. Berger, certainly aware of Lorimer’s (L: 533) comparison of Balti and Burushaski, considers the Burushaski words autochthonous. Lorimer gives for Balti apo ‘grandfather’ and api ‘grandmother.’ We propose that the new meaning of ‘grandfather’ in Balti is due to the Burushaski meaning, where in addition to ‘grandmother’ (which is the only meaning in Tibeto-Burman) we also have the (primary?) meaning of ‘grandfather, forefather, paternal ancestor’ directly in line with the Indo-European developments below. For Burushaski we consider the words with hap- to be original, and there could have been a blend of two stems.

208   

   Ilija Čašule

There is a direct correlation with PIE *h4ep- (in Pok 323: *epi-), ‘4th generation marker:’ Lat ab-avus ‘great-great-grandfather,’ ab-nepōs ‘great-great-grandson,’ OEng of-spring ‘offspring,’ Gk apó-pappos ‘descendant,’ OPers ap-anyāka‘great-great-grandfather,’ OInd áp-atyam ‘offspring,’ which Mallory and Adams (M-A: 156) consider of PIE status. Semantically, the Burushaski development is closest to Old Persian and Latin. For the semantic correlation consider also IE *h1(?)ab- ‘father, forefather, man’ : Goth aba ‘man, husband’ (Bomhard-Kerns 1994: 572).

2.2.15 Bur ‘free and independent tribe’ : IE ‘freeman’ Bur haráaỵ pl. haráakindaro ‘a free and independent tribe, owing allegiance to no one (there are no such nowadays)’ (B: III, 192) e.g. Diram Haráaỵ. Lorimer (L: 194–195) considers haráaỵ pl. haráaỵiśo, Ng pl. haráaỵićaṅ ‘enclosure, fold (for animals); camping place in high pasture grounds, for flocks and their herdsmen’ a separate word, whereas Berger (B: III, 192) subsumes both under one heading. Compare with PIE *h4erós ‘member of one’s own (ethnic) group, peer, freeman’ : Skrt arí-ḥ ‘devoted, trustworthy, loyal,’ Hitt araš ‘member of one’s own social group, peer, comrade, partner, fellow, friend,’ arawanni- ‘free; freeman,’ Lyc arawa- ‘free from;’ and from *h4eri̯os : OIr aire ‘freeman (whether commoner or noble),’ Av airya ‘Aryan,’ OPers ariya ‘Aryan,’ Irn Alani, OInd arí ‘attached to, faithful; a faithful devoted person; +/- kinsman,’ aryá ‘kind, favourable; attached to, true, devoted,’ ā́rya- ‘Aryan, one who is faithful to the Vedic religion’ (M-A: 213, who derive it from *h4er- ‘put together’). The Burushaski meaning of ‘free and independent tribe’ matches directly the semantic development in Anatolian and Celtic, and differs significantly from the Indo-Iranian meanings of ‘faithful, attached, kind, favourable,’ which would exclude that branch as the source of this word.

2.2.16 Bur ‘people, folk,’ ‘kinsman, blood-relation’ : IE ‘one’s own’ Ys ses, Hz Ng sis sg. and pl. ‘people, folk; person, man’ (B: III, 380) and the first element in Bur sukúin ‘kinsman, blood-relation (descended from a common ancestor); near relation, such as cousin’ (B: III, 384). -kuin is the denominal suffix for derivation of names of professions (B: I, 207). We can derive these forms from IE *su̯é- (also *se-), reflexive pronoun meaning ‘(one)self’ (Fortson 2004: 130; M-A: 455: *séu̯e (acc.) ‘-self’) or *s(u)u̯-o-



Burushaski kinship terminology of I­ ndo-European origin   

   209

‘one’s own’ (“widespread and old in IE”), thus in Burushaski from *su̯é-s. See the broader discussion in 2.2.9 and 2.2.10. The same line of semantic derivation as ours has been applied in the analysis of Alb gjysh ‘grandfather,’ by Liukkonen (1993: 58) who derives it from IE *sau̯isi̯a and relates it to Lith sāvas ‘own.’ Other scholars have derived the Albanian word together with Skt sūṣā́ ‘progenitor’ or ‘paternal grandmother’ from IE *seuhx‘bear, beget’ (Orel 1998: 140; M-A: 238: ?? *suhxsos- ‘grandfather’).

3 Discussion and conclusions Perhaps the single most important trait that pervades the Burushaski kinship terms and strengthens the etymological correspondences with Indo-European is the human category plural suffix –ċaro, a variant of –taro (after –n- : -daro) mainly added to words denoting relations (B: I, 48). It corresponds directly with the Indo-European suffix *–ter, which, as we noted, Benveniste (1973: 171) considered the classifier of the lexical class of kinship terms. Such a close semantic and derivational correlation is of great importance in ascertaining the origin of the Burushaski kinship terms. Although the semantic correlation is indisputable and specific, it may be objected that in Burushaski the morpheme is part of the human plural endings of kinship terms, whereas in Indo-European it is a derivational suffix. Nevertheless, in a preliminary analysis of the numerous Burushaski plural endings we have found other cases where a morpheme from the singular form has been retained only in the plural. (Refer to the examples in 2.2.1.1.) The Burushaski kinship terms ending –taro consists of two morphemes: -tar- + -o, where in fact the basic plural form is –o, an x pl ending in its own right, which reinforces the position that –tar was originally a singular derivational suffix. We have explained the variation –ċaro : –taro as a result of a change st > sċ > ċ with the –s- shifting onto the plural suffix from the IE nom. sg. –is, -us. In Burushaski the suffix is much more productive than in Indo-European, where there are only six words that have it: *méhatēr ‘mother,’ *ph̥tḗr ‘father,’ *dhuĝ(ha)tēr ‘daughter,’ *bhréhater ‘+/− brother,’ *ĝomhx-ter- ‘son-in-law’ and *h1i̯enha-ter- ‘husband’s brother’s wife’ (M-A2: 209–210). This raises the question whether originally in Pre-Indo-European (?) all kinship terms had the suffix, thus signalling that Burushaski preserves an older situation, or that the suffix spread by analogy in the latter (an innovation). Another methodological issue that needs to be addressed is the value of kinship terms deriving from nursery words. The crucial points that will differentiate a com-

210   

   Ilija Čašule

monly occurring coincidental sound symbolic form from a legitimate historic correspondence are the phonological, derivational and inflectional history and the specific semantic developments. Thus, ma or ma-ma can be found in a myriad of languages as the word for ‘mother.’ The reasons why the Burushaski terms (2.2.1.1.): -́mi (pl. -́miċaro) ‘mother; mother’s sister, aunt, on mother’s side’ and máma, pl. mámaċaro ‘mother, aunt (on mother’s side)’ (also māma) (B: III, 277, also ‘in reference to older women;’ Will: 84, also ‘mother’s sister; a paternal aunt’) as well as Bur NH mámo ‘mother, mum’ (B: III, 277), and mámayo ‘mommy (endearing form of address to girls and young women)’ should be correlated with Indo-European are, apart from the ending -ċaro, the specific semantics reflecting an Omaha kinship system (‘mother, mother’s sister’) and the derivative with the IE suffix *–t(o) : – mámut, Ng maámo, Ys -mámu ‘a woman’s breast; nipple (male or female)’ and by further semantic extension Bur mamúto ‘sucking, immature; suckling.’ It is hard to see “nursery word” coincidence in such a complete semantic and derivational correspondence. This argumentation is valid for the other Burushaski and Indo-European kinship terms originating from Lallwörter (e.g. the words for ‘father’ [2.2.2.]). The overwhelming majority of the kinship words discussed are characterised by little to no semantic latitude and with a clear morphological, i.e. derivational correlation and are widespread in Indo-European. These are terms for ‘mother,’ ‘father,’ ‘father’s brother,’ ‘children,’ ‘son,’ ‘daughter,’ ‘boy; young man,’ ‘a man’s sister,’ ‘bride,’ ‘woman’s or husband’s brother,’ ‘grandchild,’ ‘husband,’ ‘uncle,’ ‘aunt,’ ‘father-in-law,’ ‘relations by marriage,’ ‘female relation,’ ‘ancestor,’ ‘free tribe,’ ‘kinsman.’ One of the fundamental kinship and family terms that seems to defy a clear Indo-European explanation is Bur gus, pl. guśíṅanċ, Ng guśíanċ ‘woman (married); female (of animals)’ – gus huk ‘a bitch’ (L: 174–175; B: III, 162) and the derivatives -̇-skus ‘mother-in-law (on both sides)’ (L: 27; B: III, 381), géeskus Hz, géesgus Hz Ng ‘widow, widower’ (B: III, 152). It is very tempting to seek a (banal?) correlation with Bur –ġuṣ ‘woman’s privy parts, vulva’ (L: 188; B: III, 182) (which Tikkanen p.c., suggests could be related to Ys –khús ‘anus’ (BYs: 159), and thus ultimately from IE *kutsós ‘anus, vulva,’ e.g. Gk (Hesychius) kūsós ‘anus, vulva’ and from *kutsnós > Lat cunnus ‘vulva,’ NPers kun ‘vulva’ (see Čašule 2003a). If we were to accept this etymology then almost all fundamental kinship terms in Burushaski are of Indo-European origin. Furthermore, there is a possibility that Bur ġéniṣ, pl. ġéniṣanċ ‘queen, Mir’s wife, rani’ also ‘gold’ (pl. ġénaṅ) (B: III, 175) continues the closest generic Indo-European word for ‘woman’13 *gwénha- (gen. *gwnéhas) (M-A: 648; W-I-S: 178, give (with ?) also 13 Hamp (2012: 154) considers this example one of the most convincing correspondences between Burushaski and Indo-European.



Burushaski kinship terminology of I­ ndo-European origin   

   211

the stem *gwen-iH-) : OIr ben ‘woman,’ OEng cwene ‘woman, prostitute, wife,’ OPruss genna ‘wife,’ OChSl žena ‘wife, woman,’ Gk guné̄ ‘woman, wife,’ Arm kin ‘wife,’ Av gənā ‘woman, wife,’ Skt gná-̄ ‘goddess, divine female,’ TochB śana ‘woman,’ and esp. OEng cwēn ‘woman, wife, queen,’ Eng queen, from the suffixed, IE *gwén-i- (Wat: 34; M-A2: 204–205; W-I-S: 178, cite with ? also OPhrg knais, knays). This would mean that in Burushaski the generic word was “elevated” to the meaning of ‘queen,’ while an initially perhaps vulgar term was used with a generic meaning (‘cunt’ > ‘woman’). One major innovation in Burushaski is the use of the Indo-European root *hai̯eu- ‘young’ to derive the words for ‘son’ and ‘daughter’ (2.2.3.1). At the same time, Burushaski appears to preserve the Indo-European verb *seuhx- ‘bear, beget’ in Bur súas, súyas, dusúas, dusúyas vt. ‘bring; take, fetch; procure; buy,’ also d-̇-ċ-, Ng abs. d-̇-ċu- ‘bring; carry a load’ (B: III, 383). There is a more significant difference with Indo-European in the Burushaski terms for ‘brother’ and ‘sister.’ It does not continue IE *bhréhater ‘+/- brother,’ which is very strongly attested in Indo-European, and only tentatively IE *su̯ésōr ‘sister’ (neither stem is found in Albanian and Anatolian) (2.2.4). Burushaski has borrowed terms for ‘brother’ and ‘sister’ both from Indo-Aryan and esp. Tibetan (Balti). In the indigenous term for ‘brother,’ we find a direct correspondence between Bur –úlus, pl. –úluċaro, Ng pl. –úlusċaro, in Ys with the laryngeal preserved: -húles, -húlus ‘a woman’s brother, husband’s brother’ and Albanian vëlla ‘brother’ (derived by Huld: 128–129, from PIE *A1euH-m-tl-eA2-, with metathesis (sem. < ‘cousin’ < ‘mother’s brother(‘s son)’ (2.2.5) and (2.2.2.2). To this possible Albanian-Burushaski correspondence we can add the AlbanianBaltic-Burushaski isogloss in (2.2.3.2): Ys dúlas ‘boy, young lad,’ which we compare with Lett dêls ‘son,’ Alb djalë ‘boy, young man, son’ (from OAlb *delās ‘a boy’). We note further the Messapic-Albanian-Burushaski correspondence in (2.2.3.4): Bur biléeli ‘bride, newlywed woman’ : Messapian bilia, biliva ‘daughter,’ Alb bijë, Alb dial. bilë ‘daughter.’ See also the discussion in (2.2.1.1) where the Burushaski word for ‘mother’ means also ‘mother’s sister’ providing perhaps a clue as to why the Indo-European word for ‘mother’ has the meaning of ‘sister’ in Albanian. Another correspondence that strengthens a Paleobalkanic correlation is that of (2.2.1.2): Bur zizí (pl. zizíċaro) ‘Mother’ (in Royal Families) with Phrygian Zizimene ‘Great Mother,’ especially as Burushaski legend states that this term originates from the time of Alexander the Great. In Čašule (2004) we found very close lexical correspondences in ritual, myth, burial and onomastics between Burushaski and Phrygian (originally the Ancient Balkan language of Central and Northern Macedonia), with which the findings here are consistent. The correlations with Albanian itself have also increased. Perhaps the most striking grammatical trait Burushaski shares with it within Indo-European is

212   

   Ilija Čašule

its use of kho-, khi- which is prefixed to the distal forms to derive the proximate demonstratives, which parallels the unique Albanian development where the Albanian prefix kë is preposed in the same way before the distal pronouns. In our detailed analysis of the Burushaski demonstrative and personal pronouns we have found correlations between Burushaski and North-Western IndoEuropean (esp. Balto-Slavic, Tocharian) and Armenian (Čašule 2012b). An important Burushaski-Albanian and Slavic commonality can be found in the numeral system. Morphologically, the endings of the Burushaski numbers 2, 4, 5, (most probably also 6) and 9 can be traced to the Indo-European ordinal numbers with -to, in Burushaski also -ti (in the forms used for counting) (the latter perhaps from the IE abstracts in –ti), generally the same as in Albanian and Slavic (Čašule 2009b). These specific grammatical correspondences correlate with a number of Burushaski lexical isoglosses with Slavic, Germanic and North-Western IndoEuropean in general and especially with the Albanian and Romanian ancient Balkan substratal lexis (for which, see e.g. Brâncuṣ (1983), Poghirc (1967), Russu (1967, 1970) and Burushaski. In Časule (2009a), out of the 31 Burushaski autochthonous shepherd terms of (non-Indo-Iranian) Indo-European origin (and with no semantic latitude), 10 correspond closely with the Albanian and Romanian substratal pastoral vocabulary. We have further evidence for a correlation with North-Western Indo-European in (2.2.10) : Bur silaj̣ín ‘female relation, related women-folk’ : IE reflex. *su̯elo-, *su̯e-lii̯on ‘brothers-in-law whose wives are sisters,’ a word of the north-west and centre of the Indo-European world. If the etymological analysis in (2.2.6) is correct, and it is not a Tibetan loanword, in the Burushaski word for ‘grandson, granddaughter’ we might have a Germanic-Slavic-Burushaski isogloss, with identical derivation. Note also the discussion in (2.2.8.1). Our study of the Burushaski kinship terms and their close correlation with Indo-European is clearly a historical comparative linguistic analysis rather than a strictly anthropological approach. For a detailed anthropological study, one can refer to Parkin (1987b) who gives convincing evidence for the prevalence in Burushaski and Shina of a symmetrical, two-line prescriptive kinship system, typical of Tibetan. Parkin makes a strong case on how a non-prescriptive kinship system can be transformed and become a prescriptive one. In our view, the asymmetrical terms in Burushaski for ‘mother-in-law’ and ‘father-in-law’ are actually not innovations as Parkin suggests, but remnants of the original non-prescriptive pattern. We note here the strong support for our work by Paul Friedrich, whose expertise has been Proto-Indo-European kinship (Friedrich 1966), who stated : “Čašule’s analysis is conclusive. At a more concrete level I was struck by the 30 pastoral terms



Burushaski kinship terminology of I­ ndo-European origin   

   213

that are of Indo-European but not Indo-Iranian origin. The full correspondence of 32 kinship forms with Indo-European seems conclusive” (Friedrich 2012). The coherence of the Burushaski kinship system and the comprehensive preservation of 30 (32?) original Indo-European kinship terms, together with the other large body of evidence in our previous work: the laryngeals and general phonological and lexical correspondences (Čašule 2003b, 2012b), the gutturals (Čašule 2010a), the numerals (Čašule 2009b), the personal and demonstrative pronouns, grammatical correspondences in general and in nominal and verbal morphology (Čašule 2012b), and the correspondences in e.g. the anatomical parts (Čašule 2003a) and the shepherd vocabulary (Čašule 2009a) all argue for a (North-Western?) Indo-European origin of the Burushaski language.

4 S  ummary of Burushaski autochthonous kinship terms of Indo-European origin Kinship or related term

Burushaski

Indo-European

‘mother’

máma, pl. mámaċaro -́mi pl. -́miċaro mámo, mámayo zizí pl. zizíċaro in Royal Families Bur ġéniṣ, pl. ġéniṣanċ 1/2 bapó pl. 1/2bapóċaro also ‘prince’ -́uỵ , pl. -́uỵċaro

*méhatēr-

‘[great/royal] mother’ ‘queen’ ‘father,’ ‘grandfather’ ‘father,’ ‘father’s brother,’ ‘forefather’

‘boy,’ ‘young lad’

-́u pl. -́uċaro Hz Ng –í, Ys –yé, pl. –yú and –yúa, pl.‘children’ Ys dúlas ‘boy, young lad’

‘strong young man’

láuḍar

‘daughter’

-̇i-, pl. –yúguśanċ, Ys pl. –yúguśiṅa biléeli

‘son’

‘newlywed woman’ ‘sister’

–yás pl. –yásċaro, Hz also –yástaro, Ys –yást

Zizimene (Great Mother) (Phrygian) *gwen-iH- ‘woman, queen’ *papos-ter- or IA bap + *ph2tér*h2euh2ii̯os *h2éuh2*hai̯eu̯- ‘young’ Lett dêls ‘son,’ Alb djalë ‘boy, young man, son, young animal’ IE *h1léudheros ‘free man,’ Srb ljuda ‘very strong man’ *hai̯eu̯- ‘young’ Alb bijë, dial. bilë ‘daughter’ *su̯ésōr

214   

   Ilija Čašule

‘woman’s or husband’s brother’ ‘nephew,’ ‘niece’ ‘grandson,’ ‘granddaughter’

–úlus, pl. –úluċaro, –úlusċaro, Ys -húles, -húlus –sáġun pl. –sáġundaro and –sáġuyo ‘nephew, niece’ -̇mis Hz Ng, -̇mes Ys -pháṭ in -pháṭ -̇mis ‘grandson’

‘uncle’

-úyar pl. -úyariśo (also – uyer), Ys –yúhar hir, Ys pl. hurí and huríkia, Hz Ng hiríski -́ṅgo, pl. -́ṅgoċaro, Ys -́ṅgu

‘uncle’ ‘aunt’

nána and nané pl. nánaċaro

‘aunt,’ ‘father’s sister’

-́nċo , Ys -́nzu pl. Hz Ng -́nċoċaro, Ys –ntsutsaro -̇- skir, pl. -̇- skindaro, Ng pl. -̇- skiriśo 1. Ys salén, also selén and

‘husband, married man’

‘man’

‘father-in-law’ ‘husband’s sisters and daughters’

‘neighbour’

2. silaj̣ín ‘female relation, related women-folk’ gul, gulgár, gulc̣hé, gulbúl –rík, pl. –ríkindaro ‘a man‘s wife‘s brother’ Bur Ys gamáiċ, grambéśu

‘fourth generation marker’

hápkuin ‘spouse,’

‘relations by marriage’ ‘brother-in-law’

‘freeman’ ‘people, folk’

*h2au-lo-s (or Alb vëlla ‘brother’) *se- +*ĝenh1-ti- or *se- + IE *ĝnh1-ó ‘own procreation’ (?) *h4en-ki-s, ~ OChSl vnukŭ ‘grandson’ *népōts ‘grandson, nephew’ *u̯ihxrós ‘man, husband’

*k̂uh1ros ‘powerful,’ ‘master’ *h2euh2ino- (Lith avýnas ‘uncle,’ (with poss. –ĭ̄no) + *-ko. *n-h4en- > Gk nánnē ‘fem. cousin, aunt’ (nanna ‘aunt,’ nannos and nānos ‘uncle’) *snusós ‘daughter-in-law’ *su̯ék̂uros *su̯e-lo-, *su̯e-lii̯on ‘brothers-in-law whose wives are sisters’ *ĝl̥h3-u̯os ‘husband‘s sister’ (?)*u̯rihx- ‘new’ + suffix –ko. Gk gambrós ‘son-in-law’ < IE *ĝ(e)m(hx)ros ‘sister’s husband’ *h4ep- ‘4th generation marker’

hápa ‘household’ -̇pi ‘grandfather, forefather, paternal ancestor; grandmother, 4th generation marker’ Bur haráaỵ pl. haráakindaro *h4erós, *h4eri̯os ‘freeman’ ‘a free and independent tribe’ Ys ses, Hz Ng sis ‘people, *su̯é-s ‘one’s own’ folk; person, man’



Burushaski kinship terminology of I­ ndo-European origin   

   215

Abbreviations of languages and dialects Alb – Albanian, Arm – Armenian, Av – Avestan, Blg – Bulgarian, Bur – Burushaski, Celt-Iber – Celt-Iberian Croat – Croatian, Cz – Czech, Eng – English, Fr – French, Gk – Greek, Goth – Gothic, Grm – German, Grmc – Germanic, H – Hindi, Hitt – Hittite, Hz – Hunza dialect of Burushaski, IA – Indo-Aryan, IE  – Indo-European, Illyr – Illyrian, Ir – Irish, Irn – Iranian, Ksh – Kashmiri, Lat  – Latin, Lett – Lettish, Lith – Lithuanian, Mcd – Macedonian, MWels – Middle Welsh, Ng – Nager dialect of Burushaski, NH – Nasiruddin Hunzai – Burushaski scholar and poet, Berger’s main informant, NIA – New Indo-Aryan, NPers – New Persian, NWels – New Welsh, OAlb – Old Albanian, OChSl – Old Church Slavonic, OEng – Old English, OHG – Old High German, OInd – Old Indian, ON – Old Norse, OPers – Old Persian, OPrus – Old Prussian, OWels – Old Welsh, Phrg – Phrygian, PIE – Proto-Indo-European, Pl – Polish, PSl – Proto-Slavic, PTB – Proto-TibetoBurman, Russ – Russian, RV – Ṛgveda, Sh – Shina, Skt – Sanskrit, Slk – Slovak, Sln – Slovenian, Srb – Serbian, TB – Tibeto-Burman, Tib – Tibetan, TochA  – Tocharian A, TochB – Tocharian B, U – Urdu, Ys – Yasin dialect of Burushaski

Abbreviations of frequently cited sources B = Berger, 1998. BYs = Berger, H. 1974. Cunn = Cunningham, A. 1854.. L = Lorimer, D. L. R. 1938. LYs = Lorimer, D. L. R. 1962. M-A = Mallory, J. P. and D. Q. Adams (eds.). 1997. M-A2 = Mallory, J. P. and D. Q. Adams. 2006. Pok = Pokorny, J. 1959. T = Turner, R. L. 1966. Wat = Watkins, C. 2000. Will = Willson, S. R. 1999. W-I-S = Wodko, D. S., B. Islinger and C. Schneider. 2008.

216   

   Ilija Čašule

References Alonso de la Fuente, José Andrés. 2006. El burušaski, ¿una lengua aislada? Revista Española de Lingüística 35 (2): 551–579. Backstrom, Peter C. and Carla F. Radloff. 1992. Languages of Northern Areas. Sociolinguistic survey of Northern Pakistan. Vol. 2. Islamabad: National Institute of Pakistan Studies, Quaid-i-Azam University and Summer Institute of Linguistics. Bashir, Elena. 2000. A thematic survey of Burushaski research. History of Language 6 (1): 1–15. Benveniste, Emil. 1973. Indo-European language and society. Coral Gables: University of Miami. Berger, Hermann. 1974. Das Yasin Burushaski (Werchikwar). Wiesbaden: Otto Harrassowitz. Berger, Hermann. 1990. Burushaski. In Ehsan Yarshater (ed.). Encyclopaedia Iranica, vol. 4, 567–568. London and New York: Routledge and Kegan Paul. Berger, Hermann. 1998. Die Burushaski-Sprache von Hunza und Nager. Teil I. Grammatik. Teil II. Texte. Teil III. Wörterbuch. Burushaski-Deutsch; Deutsch-Burushaski. (Neuindische Studien 13.) Wiesbaden: Harrassowitz Verlag. Berger, Hermann. 2008. Beiträge zur historischen Laut- und Formenlehre des Burushaski. (Neuindische Studien 15.) Wiesbaden: Harrassowitz Verlag. Biddulph, Joseph. [1880] 1986. Tribes of the Hindoo Koosh. Calcutta: Government Printing Office. (Reprinted 1986. Lahore: Ali Kamran Publishers.) Blažek, Vaclav. 2001. Indo-European kinship terms in *-eter. In Šefčik, O. and B. Vykypĕl (eds.). Grammaticus: Studia Linguistica Adolfo Erharto Quinque et Septuagenario Oblatu, 24–33. Brno: Masaryk University. Bomhard, Allan R. and John C. Kerns. 1994. The nostratic macrofamily: A study in distant linguistic relationship. New York: Mouton de Gruyter. Brâncuṣ, Grigore. 1983. Vocabularul autohton al limbii Române. [The autochthonous vocabulary of the Romanian language]. Bucharest: Editura Ṣtiinṭifică ṣi Enciclopedică. Buck, Carl D. 1949. A dictionary of selected synonyms in the principal Indo-European languages. Chicago-London: University of Chicago Press. C̹abej, Eqrem. 1976. Studime etimologjike në fushë të shqipes. [Albanian etymological studies] Studime gjuhësore. [Linguistic studies]. I. Prishtinë. Calder, William M. 1912. Inscriptions d’Iconium. Revue de philologie, de littérature et d’histoire anciennnes 36: 72–74. Čašule, Ilija. 1998. Basic Burushaski etymologies. (The Indo-European and Paleobalkanic affinities of Burushaski). Munich-Newcastle: Lincom Europa. Čašule, Ilija. 2003a. Burushaski names of body parts of Indo-European origin. Central Asiatic Journal. 47 (1): 15–74. Čašule, Ilija. 2003b. Evidence for the Indo-European laryngeals in Burushaski and its genetic affiliation with Indo-European. The Journal of Indo-European Studies 31 (1–2): 21–86. Čašule, Ilija. 2004. Burushaski-Phrygian lexical correspondences in ritual, myth, burial and onomastics. Central Asiatic Journal 48 (1): 50–104. Čašule, Ilija. 2009a. Burushaski shepherd vocabulary of Indo-European origin. Acta Orientalia 70: 147–195. Čašule, Ilija. 2009b. Burushaski numerals of Indo-European origin. Central Asiatic Journal 53 (2): 163–183. Čašule, Ilija. 2010. Burushaski as an Indo-European “kentum” language. (Reflexes of the Indo-European gutturals in Burushaski). Munich: Lincom.



Burushaski kinship terminology of I­ ndo-European origin   

   217

Čašule, Ilija. 2012a. Macedonian and South Slavic lexical correspondences with Burushaski. Balkanistica 25: 221–256. Čašule, Ilija. 2012b. Correlation of the Burushaski pronominal system with Indo-European and phonological and grammatical evidence for a genetic relationship. Journal of Indo-European Studies 40 (1–2): 59–154. Chantraine, Pierre. 1947. Les noms du mari et de la femme, du père et de la mère en grec. Revue des études grecques 59/60: 219–250. Cunningham, Alexander. 1854. Ladák, physical, statistical and historical: with notices on the surrounding countries. London: W. H. Allen and co. [pp 398–418 a short Burushaski word list]. Demiraj, Bardhyl. 1997. Albanische Etymologien: Untersuchungen zum albanischen Erbwortschatz. Leiden Studies in Indo-European. Rodopi. Desnickaja, Agnija Vasilʹevna. 1984. Sravnitel’noe jazykoznanie i istorija jazykov.[Comparative linguistics and history of languages]. Leningrad: Nauka. Edel’man, Džoj. I., and Georgij A. Klimov. 1970. Jazyk burušaski. [The Burushaski language]. Moscow: Nauka. Eichner, Heiner 1972. Die Etymologie von heth. mēhur. Münchener Studien zur Sprachwissenschaft 31: 53–107. Ernout, Alfred and Antoine Meillet. 1959. Dictionnaire étymologique de la langue latine. Paris: Librairie Klincksieck. Fortson, Benjamin W., IV. 2004. Indo-European language and culture. Oxford: Blackwell Publishing. Friedrich, Paul. 1966. Proto Indo-European kinship. Ethnology 5: 1–36. Friedrich, Paul. 2012. Letter to Elena Bashir of January 4, 2012. Gasparro, Gulia Sfameni. 1985. Soteriology and mystic aspects in the cult of Cybele and Attis. Leiden: E. J. Brill. Georgiev, Vladimir. 1971-. Bŭlgarski etimologičen rečnik. [Bulgarian etymological dictionary]. Sofia: Bulgarian Academy of Sciences. Gluhak, Alemko. 1993. Hrvatski etimološki rječnik. [Croatian etymological dictionary]. Zagreb: A. Cesarec. Gołab, Zbigniew. 1992. The origins of the Slavs (A linguist‘s view). Columbus: Slavica. Hamp, Eric P. 1972. Albanian. Current Trends in Linguistics, vol. 9: 1626–1692. Hamp, Eric P. 2011. Review of Čašule (2012a) ms. Hamp, Eric P. 2012. Comments on Čašule’s “Correlation of the Burushaski pronominal system with Indo-European and phonological and grammatical evidence for a genetic relationship.” Journal of Indo-European Studies 40 (1–2): 154–156. Hayward, George, W. 1871. Appendices to letters on his explorations in Gilgit and Yassin. Vocabularies of the dialects of Dardistan, Wakhán, Shignán and Roshnan. Journal of the Royal Geographical Society of London 4: 18–28. Hetterich, H. 1985. Indo-European kinship terminology. Anthropological Linguistics 27: 453–480. Huld, Martin E. 1984. Basic Albanian etymologies. Columbus: Slavica. Kullanda, Sergey. 2002. Indo-European “kinship terms” revisited. Current Anthropology 43 (1): 89–111. Leitner, Gottlieb W. 1889. The Hunza and Nagyr handbook. Pt. 1. Calcutta. Liddell, Henry Georg, Robert Scott, Henry Stuart Jones and Roderick McKenzie. [1897] 1968. Greek-English lexicon. Oxford: Clarendon Press.

218   

   Ilija Čašule

Liukkonen, Kari. 1993. Albanisch-baltisch-slavische lexikalische Beziehungen und ihre Bedeutung für die historische Phonetik des Albanischen. Studia Slavica Finlandensia 10: 55–61. Lorimer, David Lockhart Robertson. 1935. The Burushaski language. Vol. 1. Introduction and grammar with a preface by G. Morgenstierne. Vol. 2. Texts and translations. Oslo: Institutet for Sammenlignende Kulturforskning. H. Aschehoug&Co. Lorimer, David Lockhart Robertson. 1938. The Burushaski language. Vol. 3. Vocabularies and index. Oslo: Institutet for Sammenlignende Kulturforskning. H. Aschehoug & Co. Lorimer, David Lockhart Robertson. 1962. Werchikwar-English Vocabulary. Oslo: Norwegian Universities Press. Mallory, James P. and Douglas Q. Adams (eds.) 1997. Encyclopedia of Indo-European culture. London – Chicago: Fitzroy Dearborn Publishers. Mallory, James P. and Douglas Q. Adams. 2006. The Oxford introduction to Proto-Indo-European and the Proto-Indo-European world. Oxford: University Press. Mann, Stuart E. 1984–1987. An Indo-European comparative dictionary. Hamburg: Helmut Buske. Morgenstierne, Georg. 1945. Notes on Burushaski phonology. Norsk Tidsskrift for Sprogvidenskap 13: 61–95. Neroznak, Vladimir. 1978. Paleobalkanskie jazyki. [Paleobalkanic languages]. Moscow: Nauka. Orel, Vladimir. 1998. Albanian Etymological Dictionary. Leiden: Brill. Parkin, Robert J. 1987a. Tibeto-Burman and Indo-European loans in Burushaski kinship terminology. Bulletin of the School of Oriental and African Studies 50 (2): 325–329. Parkin, Robert J. 1987b. Kin classification in the Karakoram. Man (New Series) 22 (1): 157–170. Poghirc, Cicerone. 1967. Considérations sur les éléments autochtones de la langue roumaine. Revue Roumaine de linguistique 12: 19–36. Pokorny, Julius. 1959. Indogermanisches etymologisches Wörterbuch. Bern-Munich. Russu, Ion I. 1967. Limba traco-dacilor. [Thraco-Dacian language]. 2nd edition. Bucharest. Russu, Ion I. 1970. Elemente autohtone în limba românǎ. Substratul comun româno-albanez. [Autochthonous elements in the Romanian language. The common Romanian-Albanian substratum]. Bucharest. Skok, Petar. 1974 (I-IV). Etimologijski rječnik hrvatskoga ili srpskoga jezika. [Etymological dictionary of the Croatian or Serbian language], Zagreb: Jugoslavenska akademija znanosti i umjetnosti. Steingass, Francis Joseph. [1892] 1999. A comprehensive Persian-English dictionary, [orig. edition: London: Routledge & K. Paul]. Reprint: Lahore: Sang-e-Meel Publications. Tiffou, Étienne. 2000. Current research in Burushaski: A survey. History of Language [Special issue on Burushaski]: 6 (1): 15–20. Tiffou, Étienne (ed.) 2004. Bourouchaskiana : actes du colloque sur le bourouchaski organisé à l’occasion du XXXVIème Congrès international sur les études asiatiques et nord-africaines (Montréal, 27 août-2 septembre 2002). Louvain-la-Neuve: Peeters. Tiffou, Étienne and Yves C. Morin. 1989. Dictionnaire complèmentaire du Bourouchaski du Yasin. AMI 17. SELAF 304. Paris: Peeters/Selaf. Tiffou, Étienne and Jurgen Pesot. 1989. Contes du Yasin. (Introduction au bourouchaski du Yasin avec grammaire et dictionnaire analytique). Paris: Peeters/Selaf. Tikkanen, Bertil. 2001. Burushaski -̇-skir ‘father-in-law’ and -̇-skus ‘mother-in-law.’ Studia Orientalia 94: 479–482. Trubačev, Oleg. 1974-. Etimologičeskij slovar’ slavjanskix jazykov. [Etymological dictionary of the Slavonic languages]. Moskva: Nauka.



Burushaski kinship terminology of I­ ndo-European origin   

   219

Turner, Ralph L. 1966. A comparative dictionary of the Indo-Aryan languages. London: Oxford University Press. Varma, Siddeshwar. 1941. Studies in Burushaski dialectology. Journal of the Royal Asiatic Society of Bengal (Letters) 7: 133–73. Watkins, Calvert. 2000. The American heritage dictionary of Indo-European roots. (2nd edition). Boston: Houghton Mifflin. Willson, Stephen, R. 1999. Basic Burushaski vocabulary. (Studies in languages of Northern Pakistan, Vol. 6.) Islamabad: National Institute of Pakistan Studies/Summer Institute of Linguistics. Wodko, Dagmar S., Britta Islinger and Carolin Schneider. 2008. Nomina in Indogermanischen Lexikon. Heidelberg: Winter Universitätsverlag. Zarubin, Ivan Ivanovič. 1927. Veršikskoe narečie kandžutskogo jazyka. [The Wershik dialect of the Kanjut language]. Kollegija vostokovedov. Zapiski 2: 275–364.

Christian Huber

Subject and object agreement in Shumcho 1 Introduction Shumcho is a Tibeto-Burman (West Himalayish) language spoken in a handful of villages in the Kinnaur District of Himachal Pradesh in the Indian Himalayas. It is sandwiched roughly between the Tibetan-speaking part of Kinnaur in its north and northeast and the Kinnauri-speaking area in its south. The term Shumcho [ɕʊmtsʰo] (ɕʊm ʻthree,ʼ tsʰo ca. ʻgroup, bunchʼ) refers to the three villages Kanam, Labrang and Spillo. Shumcho is the Kinnauri version of the name for that area, locally the area is called Humcho [hʊmtsʰo]. However, the language is also spoken outside the Shumcho area proper. There does not seem to be an established name for the language among its speakers, although they distinguish it from “Hamkat” (i.e. Kinnauri, or some variety thereof) and other local languages (speakers from the Shumcho area occasionally refer to it as hʊmtsʰo kɐt ʻShumcho languageʼ). For brevityʼs sake I will use the term Shumcho also to refer to the language.1 According to current knowledge Shumcho is spoken in the villages of Kanam, Labrang and Spillo (incl. Karla) by all castes, in the Jangram area it is spoken in Jangi, Lippa and Asrang by the lower castes, the high caste there speaks Jangrami (in many villages in Kinnaur there are two or more indigenous languages, according to castes). In the Ropa valley it is spoken in Shyaso, Rushkalang and Taling (apparently by all castes) and by the lower caste(s) in Sunnam (the high caste there speaks Sunnami; in Ropa and Gyabong Kinnauri “Hamkat” is spoken). Within Kinnaur, Shumcho forms a group with the closely related languages Chhitkuli (spoken in Rakchham and Chhitkul), Jangrami and Kinnauri (field data suggest that Sunnami perhaps does not belong in this group).2 1 Part of this research was funded by the Austrian Science Fund project Documentation of Oral Traditions in Spiti and Upper Kinnaur (P15046). Three field trips (2007, 2009 and 2011) were funded by the Phonogrammarchiv, Austrian Academy of Sciences.  My very special thanks go to Kinnaur: to my Kinnauri brothers Kesar Negi and Kailash Negi, to Jaswant Singh Negi, Ruk Mani Negi, Dev Raj Negi, Mahender Singh Negi, Chhering Medup Negi, Krishna Devi Negi, Lachhem Zangmo Negi (R.I.P.) and all other members of the Bethu and Zela families (Kanam), who have supported my research from the outset, as well as to Sherab Neema Negi (a/k/a Boris Lama) and Dharam Sain Negi (Kanam), Gopi Chand Negi (Jangi), Rattan Mala Negi (Ribba), and too many more to mention who I had the pleasure to work with over the years. I am also indebted to two anonymous referees for their comments. 2 Published Shumcho data are confined to a short word list in Gerard (1842: 548–551), a handful of entries in lists such as those in J.D. Cunningham (1844: 225–228) or A. Cunningham (1854:

222   

   Christian Huber

1.1 Organisation of the article In this article I will deal with some aspects of verbal agreement in Shumcho and provide a descriptive account of Shumcho subject and object agreement morphology as it presents itself in the currently available data.3 Like other West Himalayish languages Shumcho has an articulate agreement system, which however displays features that have not yet been described for other languages of that group. All Shumcho data are from fieldwork, drawn from elicitations, narratives and interviews/conversations, and represent the Kanam variety of Shumcho. In Section 2, which is concerned with subject agreement, I introduce the agreement markers that are found with tensed verbs and copulas. I examine the two occurring marking patterns and also deal with two instances where agreement markers are absent. In addition, I will describe the general pattern of subject agreement marking in imperatives and point out some irregular cases. Section 3 is devoted to the marking of agreement with direct (or indirect) objects and pays special attention to the accompanying morphological and phonological effects in different verb types. Section 4 deals with subject and object agreement marking in complement clauses in logophoric and non-logophoric contexts. It 397–419), or a few scattered forms in works such as Neethivanan (1971), B.R. Sharma (1976) or D.D. Sharma (1988). Further languages spoken in Kinnaur include also the Tibetan variety (or varieties) of upper and lower Kinnaur, varieties of Kinnauri Harijan/Oras boli (Indo-Aryan) spoken by the lower castes in many villages of lower Kinnaur, and Hindi, which is the administrative language and whose influence can be felt at all levels. The non-Tibetan Tibeto-Burman speech varieties of Kinnaur are usually classified as West(ern) Himalayish (e.g. Grimes 2000b: 703, van Driem 2001: 934f., 939). The Jangrami variety of Jangi appears to differ slightly from the LippaAsrang variety. For Chhitkuli see Bailey (1920: 78–86) and D.D. Sharma (1992: 197–304), for Kinnauri see, among others, Konow (1905), Bailey (1909, 1911), Grierson (1909), Joshi (1909, 1911), B.R. Sharma (1976), D.D. Sharma (1988), Saxena (e.g. 1992, 1995, 1997, 2002a, 2002b, 2004, 2007a, 2007b) and Takahashi (2001, 2007). Kinnauri is often the language of the high caste only, there is not yet much known on actual dialectal variation. In this article I use the term “Standard” Kinnauri for the Kinnauri variety described in Bailey (1909, 1911) and Joshi (1911) but I reserve judgement as to whether or not it represents a standard (D.D. Sharma 1988 apparently utilises data from several Kinnauri varieties). In cited data I retain the orthography used in the original publications. 3 In this article I use the term subject for the sole (core) argument of an intransitive verb or the agent argument of a transitive verb, and the term object for the patient/theme argument of a transitive verb. As the citation form of verbs I use the infinitive (VERB-mɐ), as suggested by my consultants. In the transcription of Shumcho data, tones cannot yet be indicated with sufficient certainty (there is minimally a high tone, e.g. tʰá-mɐ ʻbreak (trans.)ʼ vs. tʰa-mɐ ʻask for sth.,ʼ sú-mɐ ʻwipe awayʼ vs. su-mɐ ʻbathe (trans.),ʼ etc.); /a/, /e/, /i/, /ɨ/, /o/, /u/ are long vowels. Certain aspects of Shumcho phonology are still under research. Also, there is some variation among speakers, and even in the usage of individual speakers.



Subject and object agreement in Shumcho   

   223

will be shown that Shumcho has logophoric first person subject and object agreement in embedded clauses that is sensitive to different types of logocentric licensers. A summary in section 5 concludes the article.

1.2 Some preliminaries Before turning to discussing verbal agreement let me provide some background on the Shumcho pronouns and the major types of verbal constructions found in finite clauses. I will also introduce the various tense and aspect markers. However, I cannot give an in-depth account of the temporal-aspectual system here. The pronouns and demonstratives are shown in Tables 1 to 3. The 1st person singular form gɨ is restricted to the absolutive and ergative cases (gɨ-Ø [I-abs], gᵻ-s [I-erg]), all other cases employ the possessive form ɐŋ ʻmyʼ (e.g. ɐŋ-rɐ [I.poss-dat] ʻto me,ʼ ɐŋ-nɐŋ [I.poss-com] ʻwith meʼ). Also the non-honorific 2nd person singular form ka appears only in the absolutive and ergative (ka-Ø [you.sg.nhon-abs], kɐ-s [you.sg.nhon-erg]), otherwise the possessive form kɐn ʻyourʼ is used (e.g. kɐn-rɐ/ kɐnnɐ [you.sg.nhon.poss-dat] ʻto you,ʼ kɐn-nɐŋ [you.sg.nhon.poss-com] ʻwith youʼ). The dual and plural forms contain the numeral nɪɕ ʻtwoʼ and the nominal plural marker -pɐŋ, respectively. In the 2nd person there is a honorificity distinction in all numbers. The plural gɪrɐŋpɐŋ appears to be a by-form that is used only occasionally (to what extent it comes with different semantics is not yet known). In the third person, which shows a tripartite deixis, only the singular has separate honorific forms for all speakers (e.g. honorific dɔte vs. non-honorific dɔ). For most speakers I have consulted, the dual and plural forms are neutral with respect to honorificity, as in Table 2. Forms such as dɔtenɪɕ or dɔtepɐŋ in Table 3 are occasionally met in the speech of older people but are mostly rejected by younger consultants and never occured spontaneously in data elicitation. Likewise, for most encountered speakers only the non-honorific singular forms may also function as preposed demonstrative determiners (e.g. dʑɪ sɐt ʻthis deityʼ) but honorific demonstrative determiners (e.g. dʑɪte sɐt ʻthis.hon deityʼ) are found at times in older peopleʼs speech. Tables 2 and 3 also include the variants involving the discourse-deictic prefix ɔ-. In addition, Shumcho has a separate set of 3rd person pronouns, termed here “4th person,” which serve as reflexive pronouns or indicate coreference with a higher noun phrase or 3rd person matrix clause subject (among other functions; for some examples of their use see section 4).4 Again there is a honorificity distinction 4 The choice of the label fourth person is inspired by the Eskimologist practice of referring to a reflexive/coreferential 3rd person as “fourth” person, see e.g. Thalbitzer 1911: 1021, SchultzLorentzen 1945: 28, Fortescue 1984: 142, 155.

224   

   Christian Huber

in the singular. For most speakers, ɛnpɐŋ serves as a common plural form; however, ɛzɪnpɐŋ has also been encountered as a honorific plural here. Whether there exists also a honorific dual form ɛzɪnɪɕ is currently not known.5 Tab. 1: Pronouns: 1st and 2nd person singular

dual

plural

1st

inclusive exclusive

gɨ, ɐŋ

ʊnɪɕ nɪŋnɪɕ

ʊppɐŋ, ʊlpɐŋ nɪŋpɐŋ

2nd

honorific non-honorific

gɪrɐŋ ka

kɪnɪɕ kɐnɪɕ

kɪnpɐŋ, gɪrɐŋpɐŋ kɐnpɐŋ

Tab. 2: Pronouns, demonstratives: 3rd and 4th person singular honorific 3rd

near far not present

dʑɪte nɔte dɔte

ɔɪte ɔnɔte ɔdɔte

ɛzɪn

4th

dual non-hon.

dʑɪ nɔ dɔ ɛn

ɔɪ ɔnɔ ɔdɔ

plural

neutral/non-hon. neutral/non-hon. dʑɪnɪɕ nɔnɪɕ dɔnɪɕ

ɔɪnɪɕ ɔnɔnɪɕ ɔdɔnɪɕ

ɛnɪɕ

dʑɪpɐŋ nɔpɐŋ dɔpɐŋ

ɔɪpɐŋ ɔnɔpɐŋ ɔdɔpɐŋ

ɛnpɐŋ

Tab. 3: 3rd and 4th person – honorific dual and plural forms

3rd

4th

near far not present

dual

plural

honorific

honorific

dʑɪtenɪɕ nɔtenɪɕ dɔtenɪɕ ?

ɔɪtenɪɕ ɔnɔtenɪɕ ɔdɔtenɪɕ

dʑɪtepɐŋ nɔtepɐŋ dɔtepɐŋ

ɔɪtepɐŋ ɔnɔtepɐŋ ɔdɔtepɐŋ

ɛzɪnpɐŋ

5 dɔ has a by-form dɔt whose status vis-à-vis dɔ is still obscure. In addition, there is a superhonorific item ɛmi, which consultants render as “this person” (cf. mi ʻpersonʼ). However, ɛ does not function as a determiner otherwise and ɛmi also has no plural. Where observable so far, ɛmi is used like a super-honorific 4th person pronoun. The 4th person pronoun ɛn seemingly cannot be used as a demonstrative.



Subject and object agreement in Shumcho   

   225

In a finite clause the main verb may occur in four guises: (i) the main verb may be inflected for tense or mood and subject agreement as in (1) and (2), (ii) it may be marked for aspect but not for subject agreement as in (3) and (4), (iii) it may appear in an analytic construction where the verb is marked for aspect and an auxiliary carries tense and subject agreement, as in (5) and (6) (or, instead, additional aspect marking, as in [7]), and (iv) it may be marked for aspect and have an auxiliary that is marked for tense but not for subject agreement, as in (8) (alternatively, the auxiliary may again be marked for aspect, see [9]). In a converb construction it is the main verb that is marked for tense and subject agreement, as in (1), or minimally for the matrix aspect, as in (3) and (4). (1) gɨ sɪŋ hɐŋ-mɐ dʑɛ-ʊ tʊ-rɔ-kh I wood bring-inf go-conv come-fut-1sg ʻI will go to bring wood and come backʼ (2) tsɪran gɨ pura dʑɪ mɐ-rɪŋ-rɪ-kh tsiran I full this neg-say-ass-1sg ʻThe tsiran (i.e. trance mediumʼs statement), I will not tell you this (in its) full (version)!ʼ (3) gɨ ɔ-dɔ sɪŋ kjʊph-ʊ tʊ-ɪ I dd-that wood carry.on.back-conv come-impf ʻI will come by carrying that woodʼ (4) ɐtɕo-s pɪtɐŋ phɔt-ʊ pɪn-mɪn elder.brother-erg door open.trans-conv send-perf nɐ̃ʊ̃ lɔʈɐŋ dɔ brɪn dʑɛ-ʊ inside:loc corpse that horizontal go-perf ʻthe elder brother had opened the door and the corpse inside toppled overʼ (5) dɐth mɔt narɛnɐs-ʊ=nɪŋ nɔ-sa kɐmtɕɪt=nɪŋ then few N.-gen=etc. that-ʻhavingʼ information=etc. mɐ-gɔ-ɪ tɔ-rɛ-kh neg-hear-impf be-past-1sg ʻ(I was still a child,) so I did not understand that little bit of information about Narenas and all thatʼ

226   

   Christian Huber

(6) ɐpi ɔ-ɕɪ lɔ-ʊ=ɐ=le, grandmother dd-like say-perf=emph=emph gɪrɐŋ=lɪ=tɐ tɕhɐŋ kjʊm-ʊth=ɔ mi, you.sg.hon=also=emph now house-abl=emph person gɪrɐŋ mɔt=tɕe sɛs-ʊ wɐŋ-rɔ-na you.sg.hon few=qty/nr know-prog happen-fut-2sg.hon ʻ(Then earlier I said to granny,) I said like this, “Granny, you are also a person from the house (i.e. a family member), you might know a bit!”ʼ (7) dɐth=te tɕhɐŋkh-pɐŋ nɔ lɔkʈɐs-pɔʊ tɕhɐŋkh=nɪŋ then=emph son-plur that L.-plur:gen son=etc. ɔpɔŋkh-pɔʊ tɕhɐŋkh=nɪŋ, ɐlɐ hʊm pɨ=tɕe wɐŋ-ro-Ø, O.-plur:gen son=etc. more 3 4=qty/nr happen-fut-3nhon dɔpɐŋ gjɐkh-ɕ-ʊ tɔ-ɪ tɕhɐŋ they play-pluract-prog be-impf now ʻThen children, that (is), the son(s) of the Loktas-family (and) the son(s) of the Opongkh-family, ‑ maybe there were ca. 3, 4 more (children) – they were playing, thenʼ (8) gɨ ɔ-ɕɪ go-mɪn to (/ *tɔ-kh) I dd-like hear-perf be(pres) (/ *be-1sg) ʻI have heard soʼ (9) mɛnthɔkɔ-ʊ=nɪŋ nɪŋpɐŋ pura βɪdɪɔ lɐ-ʊ, dɔ-rɐ M.-gen=etc. we.excl full video do-perf that-dat mɐma=nɪŋ-rɐ kɐn-mɪn tɔ-ɪ gɨ maternal.uncle=etc.-dat show-perf be-impf I ʻWe made a complete video of (the) Menthoko(-festival). I have shown that to the uncle.ʼ Subject agreement markers may attach to the tense or mood markers6 -ro (future),7 -re (past), -gjo (past/non-future irrealis)8 and -ri (assertive). They cannot attach 6 Some labels and characterisations here may be subject to change as research proceeds and new insights are gained. 7 -ro, -re, -ri, -rɪʊ (and the imperative ending -rɪɕ) are realised as -to, -te, -ti, -tɪʊ (and -tɪɕ) when following a sibilant or /t/ and will be given as such in segmented examples. However, with the verbs dza-mɐ ʻeatʼ and dʑe-mɐ ʻgoʼ the future tense marker -ro regularly appears as ‑to (likewise, the cohortative marker -re appears as -te with these verbs). 8 -gjo has a cognate in the neighbouring and related language Kinnauri, which is described as expressing distant or reported past (D.D. Sharma 1988: 142, Saxena 1995: 265, 1999: 36). Its use



Subject and object agreement in Shumcho   

   227

to the aspect markers -mɪn,9 -rɪʊ,10 -ʊ (perfective), -ʊ (progressive)11 and -ɪ (imperfective).12 Among these, perfective -mɪn and imperfective -ɪ may appear in constructions involving an auxiliary, progressive -ʊ must appear with an auxiliary, and perfective -ʊ and -rɪʊ do not appear in auxiliary constructions. There is no separate morpheme for the present tense. Present tense forms do not occur with lexical verbs but are restricted to the copulas to and tɐɕ, where present tense is expressed by the use of the bare roots, to which subject agreement markers attach directly (see the discussions around Tables 7 and 9). In elicited data, only subjects of transitive verbs marked for perfective aspect (-ʊ, -mɪn, for some consultants also -rɪʊ) may receive ergative case marking, subjects of tensed verbs or verbs in the imperfective (-ɪ) or progressive (-ʊ) aspect cannot be marked ergative. In actual discourse, however, ergative case marking is also found outside the indicated cases, see e.g. (27) below. in Shumcho seems somewhat different. In my data, forms with -gjo mostly occur in interrogative contexts, if the speaker is speculating or wondering about some state of affairs or potential options, with -gjo apparently indicating insufficient external evidence or information, or insufficient personal knowledge on the speakerʼs part, to decide among options. In declarative contexts -gjo seems to indicate a possibility, option or conclusion for which there is some evidence, or an informed guess. When attached directly to verbs -gjo is restricted to a past (but not necessarily distant past) interpretation. Constructions involving the copulas to or tɐɕ in combination with -gjo with may also include the present. I cannot give a detailed discussion here. 9 -mɪn indicates that an action, transition or process is already completed (or over) at the time of speech. It places events in the (minimally non-recent) past and is also used if some event occurred prior to some other event(s). 10 -rɪʊ is ostensibly composed of the elements /rɪ/ (or /ri/) and /ʊ/. At present no straightforward analysis can be given. -rɪʊ is restricted to (apparently only non-honorific) 3rd person subjects and indicates that some action or event that took place in the recent past is completed or over at the time of speech. The speaker has firsthand knowledge resulting from being present when the action or event took place. 11 Both markers -ʊ are formally indistinguishable, yet occur in mutually exclusive distribution: perfective ‑ʊ cannot occur in an auxiliary construction, progressive -ʊ must occur with an auxiliary; neither can function as a relativiser/nominaliser (in certain discourse contexts, however, the auxiliary can be dropped with progressive -ʊ, e.g. in questions such as ɐmɐ dʑɛ-ʊ [where go‑prog] ʻWhere are you going?ʼ or tɕhɛ lɐ-ʊ [what do-prog] ʻWhat are you doing?ʼ). In addition, there is a converb marker -ʊ which again is formally indistinguishable from perfective and progressive -ʊ, and may in fact be only another instance of one or the other. Since it often is not clear to what extent serialised verbs express separate, successive events or events that overlap, or are inseparable from each other, no safe decision can be made at present. 12 With agentive or eventive verbs, -ɪ mostly results in a general present/habitual or a future interpretation, e.g. tʊ̃ɪ̃ < tʊŋ-ɪ [drink‑impf] (i) ʻ(usually) drink(s),ʼ (ii) ʻwill drink;ʼ with stative verbs, -ɪ expresses the holding of a state, e.g. sɛs-ɪ [know-impf] ʻknow(s).ʼ ‑mɪn and -i also function as perferctive and imperfective relativisers/nominalisers, respectively, e.g. tɕe-mɪn kɪtab [writeperf book] ʻwritten book,ʼ gjɐkʰ-ɪ mi [play-impf person] ʻplaying person, sportsperson.ʼ

228   

   Christian Huber

A verbal form marked for subject agreement consists minimally of a verb stem and a (potentially zero) agreement marker following the stem (see below Table 8). More complex forms may also contain one of the tense or mood markers and be marked for agreement with a 1st or 2nd person object. A verb stem may also contain a valency-changing (anti-causative, reflexive, reciprocal) morpheme following the root. In the perfective (-ʊ, ‑mɪn, -rɪʊ), imperfective (-ɪ) and progressive (-ʊ) aspects and in the infinitive, a stem may also carry the pluractional marker -ɕ, which indicates event pluralisation. In Shumcho, pluractional verbal forms require a (grammatically or semantically) plural subject (see e.g. [7] above). However, the pluractional marker cannot occur in tensed verbs or in imperatives, i.e. it does not co-occur with subject agreement on the main verb.13

2 Subject agreement Subject agreement is found in the configurations given in Table 4, i.e. following a tense (or mood) marker, or directly following a copula or verb, and in the imperative. Tab. 4: AgrS configurations: (VERB/COP)-TENSE-AgrS COP-AgrS (NEG-)VERB-AgrS Imperative:

VERB-AgrS, VERB-ri-AgrS

13 The pluractional marker-ɕ is formally identical with the intransitive (anticausative/reflexive/reciprocal) marker -ɕ but their distribution differs in a number of ways. Among other things, the latter may occur in verbal forms of all aspects and tenses whereas pluractional -ɕ is restricted in the indicated way (some thoughts on a potential common origin were presented in Huber 2012). Both markers do not co-occur if the intransitive marker is realised as /ɕ/ (e.g. *dzʊra-ɕ-ɪɕ-mɐ [prepare-intr-pluract-inf] ʻprepare oneself [pluract.]ʼ vs. dzʊraɕ-mɐ i) [prepare-intr-inf] ʻprepare oneselfʼ or ii) [prepare‑pluract-inf] ʻprepare [pluract.]ʼ). However, pluractional -ɕ may attach to intransitivised stems otherwise, e.g. ɲɐ-kʰ-ɕ-mɐ [hide-intr-pluract-inf] (vs. ɲá-ɕ-mɐ [hide(trans.)-pluract-inf]). Both markers are not to be confused with the 1st/2nd plural and 3rd honorific subject agreement marker(s) -ɕ, for which see Table 5 with discussion. For some preliminary remarks on pluractionality in Shumcho see Huber 2011.



Subject and object agreement in Shumcho   

   229

The subject agreement markers are presented in Table 5 (the imperative will be discussed later). Tab. 5: Subject agreement markers singular 1st 2nd 3rd

inclusive exclusive honorific non-hon. honorific non-hon.

-kh -na -n -ɕ -Ø

(% -ɪ̃) (% -ɪ̃)

dual/plural -ɕ -ɕ, -ɕ -ɕ -ɕ -Ø

-kh

(-ɪ̃) (-ɪ̃) (-ɪ̃) (-ɪ̃) (% -ɪ̃)

As can be seen in Table 5, only the 1st and 2nd singular persons display a true person distinction. In the 3rd person there is no number distinction but a honorificity distinction only. This is also found in other West Himalayish languages such as Kanashi (D.D. Sharma 1992: 378), Chhitkuli (D.D. Sharma 1992: 261) and Kinnauri (D.D. Sharma 1988: 132–133; Saxena 1995: 261–262; Takahashi 2001: 109). However, the marker -ɕ functions also as the subject agreement marker for the 1st and 2nd person plural, rendering these forms non-distinct from honorific 3rd person forms. -ɪ̃ (given in parentheses in Table 5) appears to be a secondary marker whose status vis-à-vis the regular markers is not yet clear. -ɪ̃ seems generally accepted in the 1st and 2nd plural persons, but consultants vary as to its acceptance in honorific 3rd persons and in the honorific 2nd person singular.14 The plural markers are also used for the dual, which has no separate set of markers. The 1st person singular marker -kh can also serve as the 1st person exclusive plural marker (younger speakers however seem to use -ɕ here exclusively), so that there are three ways of marking a 1st person exclusive plural (or dual) subject, as shown in (10): (10) a. nɪŋpɐŋ age prɐtɕar lɐ-mɐ koɕɪɕ lɐ-rɔ-ɪ̃ we.excl further publicity do-inf effort do-fut-1pl ʻwe will try to make further publicityʼ

14 Some consultants consider the marker -ɪ̃ an intrusion from Kinnauri/Hamkat or some other language rather than a genuine Shumcho agreement marker. There is only a handful of instances of -ɪ̃ in my non-elicited data.

230   

   Christian Huber

b. nɪŋpɐŋ nɐmɕa sʊ-rɔ-kh we.excl bride bring-fut-1pl.excl ʻwe will bring the brideʼ c. dʑa nɪŋpɐŋ dedʊm-ʊ tɕɔkti bɛɕa-rɛ-ɕ here we.excl D.-gen water.drop establish-past-1pl (ca.) ʻhere we established Dedumʼs (future) presenceʼ Honorificity in 3rd persons does not appear to be a purely grammatical feature in Shumcho, as can be concluded from the fact that a honorific pronoun such as dɔte ʻs/he.HONʼ can also be the subject of a non-honorific verbal form such as dʑɛre or dʑɛrɪʊ, as illustrated in (11). Honorific 3rd person marking on verbs rather seems semantically conditioned since it is mostly found if the subject is a religious authority or a deity, as in (12). (11) dɔte dʑɛ-rɛ-ɕ / dʑɛ-re-Ø / dʑɛ-rɪʊ s/he.hon go-past-3hon / go-past-3nhon / go-perf(3nhon) ʻs/he wentʼ (12) sɐŋgjɐs ɐlɪ wɐŋ-mɪn nɔte, nɔte=lɪ ɪmɛɐ S. how become-perf s/he.hon s/he.hon=also once mi wɐŋ-rɔ-ɕ person happen/become-fut-3hon ʻHow had he (i.e. god Davla) become Sangyas? Also he was perhaps (a) human once.ʼ Subject agreement comes in two patterns that differ mainly in the expression of the honorific 2nd person singular. In pattern A, which has the 2nd person singular honorific marker -na, subject agreement attaches directly to the tense or mood markers -ro (future), ‑re (past), -ri (assertive) and -gjo (past/non-future irrealis) but not directly to a verb or copula. Pattern A occurs in all verb forms with these markers, and in the past tense forms (-re-AgrS) of the copula to and the past/ non-future irrealis forms (-gjo-AgrS) of the copulas to and tɐɕ. In all cases involving these tense and mood markers subject agreement marking follows the same pattern (the assertive -ri has only 1st person forms).15 Table 6 provides the future paradigm (‑ro-AgrS) of sʊ-mɐ ʻbring.ʼ

15 -ri occurs only in negated forms and seems restricted to 1st person subjects, see e.g. ex. (2). It expresses emphatic certainty that the speaker will not do, or, with a stative verb such as sɛs-mɐ ʻknow,ʼ does not know something. (One consultant accepts also 2nd person forms, however.)



Subject and object agreement in Shumcho   

   231

Tab. 6: AgrS pattern A – future: sʊ-mɐ ʻbringʼ

1st 2nd 3rd

inclusive exclusive honorific non-hon. honorific non-hon.

singular

dual/plural

sʊ-rɔ-kh

sʊ-rɔ-ɕ sʊ-rɔ-ɕ, sʊ-rɔ-ɕ sʊ-rɔ-ɕ sʊ-rɔ-ɕ sʊ-ro-Ø

sʊ-rɔ-na sʊ-rɔ-n sʊ-rɔ-ɕ sʊ-ro-Ø

(% sʊ-rɔ-ɪ̃) (% sʊ-rɔ-ɪ̃)

sʊ-rɔ-kh

(sʊ-rɔ-ɪ̃) (sʊ-rɔ-ɪ̃) (sʊ-rɔ-ɪ̃) (sʊ-rɔ-ɪ̃) (% sʊ-rɔ-ɪ̃)

bring-fut-AgrS ʻI will bringʼ etc.

Pattern B is found only in the present tense forms of the copula to ʻbeʼ and in a particular negated future form of verbs. Here, the subject agreement markers attach directly to the stem without an intervening tense or mood morpheme. As illustrated in (13a), pattern B lacks the 2nd person singular honorific marker -na that occurs e.g. in the corresponding past tense form in (13b) but shows a marker -ɪ instead, whose status will be discussed below. (13) a. gɪrɐŋ tɕhɛ lɐ-ʊ tɔ-ɪ you.sg.hon what do-prog be-ɪ ʻWhat are you doing?ʼ

(/ *tɔ-na) (/ *be-2sg.hon)

b. gɪrɐŋ tɕhɛ lɐ-ʊ tɔ-rɛ-na you.sg.hon what do-prog be-past-2sg.nhon ʻWhat were you doing?ʼ Table 7 shows the present tense paradigm of the copula to ʻbe.ʼ16 16 Shumcho has two overt existential/equational copulas, to and tɐɕ, which occur in the present and past tense and also in the past/non-future irrealis. There are no future forms, the gap is filled by the future forms of the verb wɐŋ-mɐ ʻhappen, becomeʼ (which however often allow also for an epistemic modal reading, see e.g. [6] and [12]). In present and past, only to occurs in all persons, tɐɕ is restricted to 3rd person forms. In present and past 3rd person forms the copulas display a (non-mirative) firsthand evidential vs. mirative (DeLancey 1997) distinction that can roughly be characterised as follows. to indicates that the utterance is based on the speakerʼs internally established knowledge about some state of affairs obtaining at the time of speech (or at the respective time in the past) resulting from the speakerʼs own experience, her/his presence as the event takes (or took) place, or her/his personal familiarity with the situation. tɐɕ indicates that the knowledge about some state of affairs is (or was) freshly obtained by the speaker; there is (or was) no personal familiarity with the respective situation, the speaker has no internally established knowledge based on his/her own experience, the respective state was not previously

232   

   Christian Huber

Tab. 7: AgrS pattern B: COP to+AGR (ʻbe,ʼ present)

1st

inclusive

singular

dual/plural

tɔ-kh

tɔ-ɕ tɔ-ɕ,

exclusive 2nd 3rd

honorific

--- (tɔ-ɪ)

non-hon.

tɔ-n

honorific

tɔ-ɕ

non-hon.

to-Ø

(% tɔ-ɪ̃) (% tɔ-ɪ̃)

(tɔ-ɪ̃) tɔ-kh

(tɔ-ɪ̃)

tɔ-ɕ

(tɔ-ɪ̃)

tɔ-ɕ

(tɔ-ɪ̃)

tɔ-ɕ

(% tɔ-ɪ̃)

to-Ø

cop-AgrS

The attachment of subject agreement markers directly to a verb stem is met almost exclusively with negated verbs, as in (14), the resulting forms are interpreted as a somewhat emphatic negated future (or, with stative verbs such as sɛs-mɐ ʻknow,ʼ non-past or general present). The marker -ɪ̃ seems inapplicable here (*mɐtʊɪ̃, *mɐdʑɛɪ̃ etc.). (14) ɕɪŋkh dɔte gɨ lɐmanɪɐ-pɔʊ nɐmɕa mɐ-dʑe-kh lɔ-ʊ sister s/he.hon I L.-plur:gen bride neg-go-1sg say-perf ‘The sister said, “I will not go as the Lamaniasʼ bride!”’ Some resulting forms are given in Table 8 (“(.)t” in mɐ-tʊ.t-Ø is not an agreement marker but a root augmentation, /ɪ/ in mɐ-pɔs-ɪɕ is epenthetic).17

known to (or experienced by) the speaker. The evidential and mirative properties of to and tɐɕ may also account for the fact that to and tɐɕ are not available in the future tense. With 1st and 2nd person subjects the distinction is neutralised as tɐɕ has no 1st and 2nd person forms but it is fully operative with possessors and dative experiencers of all persons. Constructions without an overt copula (or, involving a zero copula) are unmarked with respect to evidentiality and mirativity. The past/non-future irrealis forms in ‑gjo allow for subjects of all persons. They seem to indicate a guess based on some clue or, in questions, make reference to insufficient knowledge on the speakerʼs part for deciding among available options (see above footnote 8) but details are still under research. 17 A number of verbs exhibit a root augmentation -/t/ (notated “.t”) in certain environments that does not surface in the infinitive or before vocalic aspect markers. Verbs of this class include khɛ-mɐ ʻgive,ʼ sʊ-mɐ ʻbring,ʼ lɔ-mɐ ʻsay,ʼ kɪ-mɐ ʻtake,ʼ lɐ-mɐ ʻdo,ʼ tʊ-mɐ ʻcome,ʼ and tɔ-mɐ ʻbe,ʼ all of which have a short vowel in the infinitive, and the imperative-only root dɔ.t ʻcome along!.ʼ This t‑extension surfaces e.g. before the irrealis mood marker -gjo (e.g. tʊ.t-gjo ʻperhaps s/he/ they cameʼ), when preceding other verbs in a compound formation (e.g. tʊ.t-kja-mɐ [come-finishinf]), in certain emphatic constructions (e.g. sʊ.t-ɔ mɐ-sʊ-ʊ [bring-emph neg-bring-perf] ʻdidnʼt



Subject and object agreement in Shumcho   

   233

Tab. 8: AgrS pattern B: (negated) future short forms tʊ-mɐ ʻcomeʼ

dʑe-mɐ ʻgoʼ

rɪŋ-mɐ ʻsayʼ

pɔs-mɐ ʻsitʼ

hon. nhon.

mɐ-tʊ-kh --- (mɐ-tʊ-ɪ) mɐ-tʊ-n

mɐ-dʑe-kh --- (mɐ-dʑɛ-ɪ) mɐ-dʑe-n

hon.

mɐ-tʊ-ɕ

mɐ-dʑe-ɕ

mɐ-rɪŋ-kh --- (mɐ-rɪ̃ɪ̃