Encyclopedia of Chinese Language and Linguistics Volume 3 Men–Ser For use by the Author only | © 2017 Koninklijke Brill NV General Editor Rint Sybesma (Leiden University) Associate Editors Wolfgang Behr (University of Zurich) Yueguo Gu (Chinese Academy of Social Sciences) Zev Handel (University of Washington) C.-T. James Huang (Harvard University) James Myers (National Chung Cheng University) For use by the Author only | © 2017 Koninklijke Brill NV ENCYCLOPEDIA OF CHINESE LANGUAGE AND LINGUISTICS Volume 3 Men–Ser General Editor Rint Sybesma Associate Editors Wolfgang Behr Yueguo Gu Zev Handel C.-T. James Huang James Myers LEIDEN • BOSTON 2017 For use by the Author only | © 2017 Koninklijke Brill NV Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface. ISBN 978-90-04-18643-9 (hardback, set) ISBN 978-90-04-26227-0 (hardback, vol. 1) ISBN 978-90-04-26223-2 (hardback, vol. 2) ISBN 978-90-04-26224-9 (hardback, vol. 3) ISBN 978-90-04-26225-6 (hardback, vol. 4) ISBN 978-90-04-26226-3 (hardback, vol. 5) Copyright 2017 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Brill Nijhoff, Global Oriental and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. This book is printed on acid-free paper and produced in a sustainable manner. For use by the Author only | © 2017 Koninklijke Brill NV 583 Rgyalrong Language Rgyalrong Language 1. Int roduct io n Rgyalrong is a group of four languages spoken in Sìchuān province, People’s Republic of China. These languages, along with Lavrung and Horpa/ Rtau, belong to the Rgyalrongic branch of the Sino-Tibetan family (Sun 2000); they are likely to be closely related to other languages such as Pumi, Muya, Tangut, Queyu and → Qiang within a larger Qiangic branch. The four Rgyalrong languages are known by various names, as some scholars use the Tibetan place names of the area where these languages are spoken, while others use the Chinese transcription of these names. Eastern Rgyalrong or Situ (Chinese Sìtǔ 四土, referring to the four main Tǔsī 土司—a Chinese administrative unit—governing the Rgyalrong area) is the most widely spoken language with about 100,000 speakers, and is the only one to have been put to writing before 1949 (see Btsan. lha 2010). It is spread over a large area from Lǐxiàn 理縣 county (Rnga.ba/Ābà 阿壩 prefecture) in the east to Rong.brag/Dānbā 丹巴 county (Dkar. mdzes/Gānzǐ 甘孜 prefecture) in the west. The main references on this language are Lín’s (1993) grammar, Huáng and Sūn’s (2002) dictionary, Lin (2003) on the verbal system and Lin (2009) on tone and intonation. The other three languages have fewer than 10,000 speakers each. Japhug (Chábǎo 茶堡) is spoken in the north-east of Mbarkhams/Mǎ’ěrkāng 馬爾康 county (Rnga.ba district). The main references are Jacques’s (2008) grammar and Jacques and Chen’s (2010) glossed text collection. Tshobdun (Cǎodēng 草登) is located next to Japhug in Mbarkhams county. No book-length monograph on this language has appeared yet, but numerous articles by Jackson Sun describe various aspects of its phonology and morphosyntax (for instance, Sun 2000, 2003, 2006, 2007). Zbu (Rìbù 日部, Showu) is located in the northeast of Mbarkhams county in pockets in neighboring Ndzamthang county. Few articles have been published on this language (the only reference speciijically devoted to Zbu is Sun 2004). The four Rgyalrong languages have many distinct dialects, each of which deserves detailed documentation, including text collections, dictionaries and reference grammars. In spite of being genetically related to Chinese, the Rgyalrong languages are typologically quite divergent from the isolating type. As we will see, they are the only fully polysynthetic languages in China, or even in all of Asia if we except isolated languages of the sub-polar and polar regions such as Ket, Ainu and Chukchi. These languages are of exceptional interest for both linguistic typology and historical linguistics. 2. Ph o n o lo g y Unlike modern Chinese languages which have a very simple syllable structure with few or no initial clusters, Rgyalrong languages have the most complex syllabic structure of all Sino-Tibetan languages. Rgyalrong languages typically have large consonant inventories; for instance, Japhug has the following consonantal system with 49 phonemes: p pʰ b mb m w t tʰ d nd ts tsʰ dz ndz n s z l ɬ tʂ tʂʰ dʐ ndʐ tɕ tɕʰ dʑ ndʑ ʂ ɕ ʑ r c cʰ ɟ ɲɟ k kʰ g ŋg ɲ ŋ x ɣ q qʰ ɴɢ χ ʁ j More than 340 consonant clusters are attested in Japhug, including clusters of up to three or even four consonants. Some of the complex clusters of Old Tibetan which are preserved in none of the attested modern → Tibetan languages are still present in Rgyalrong. For instance, Old Tibetan bsgyur(d) ‘to change (past stem)’ is borrowed as βzɟɯr in Japhug. For use by the Author only | © 2017 Koninklijke Brill NV Rgyalrong Language Rgyalrong languages (except Japhug) are tonal languages, and present morphological tone assignment rules (see Sun 2005; Lin 2009). Some of the consonantal contrasts found in Rgyalrong languages are typologically quite rare. We will only mention three of the most unusual ones. First, Tshobdun and Japhug have a series of / fricative+prenasalized voiced stop+medial/ clusters, as in Japhug ʁmbɣi ‘sun’, ʑŋgri ‘star’ etc. Second, Japhug has an intriguing ijive-way contrast between /velar stops+j/ as in kjo ‘to cause to glide’, /palatal stops/ co ‘valley’, /dental affricates+j/ ɯ-mtsjoʁ ‘beak’, /alveolo-palatal affricates/ as in tɕoʁtsi ‘table’ and /uvular stops+j/ as in qjoʁ ‘to vomit’. Palatal stops can be followed by -ɣ-, -r- or -l- as in ɲcʰɣaʁ ‘birch bark’. Third, Zbu has a very complex vowel system, with a series of velarized vs. plain vowels (Sun 2000, 2004:272). 3. M orp h osy nt ax Rgyalrong languages have a complex verbal morphology, unlike anything else found in the Sino-Tibetan family. They present all the characteristics of polysynthetic languages: 1) A high word to morpheme ratio (most verbal forms include at least three morphemes, and can have up to eight); 2) Head-marking typology (e.g., grammatical relationships are marked mainly on the verb, and within the noun phrase possession is marked on the possessed rather than on the possessor); 3) Presence of nominal incorporation. For want of space, we restrict this section to a description of four of the typologically most unusual features of Rgyalrong languages: stem alternation, inverse marking, incorporation, and word-order typology. 3.1 Verb stem alternations Verb stem alternations in Rgyalrong languages were ijirst discovered by Sun (2000, 2004). All four Rgyalrong languages have alternations of some kind, with Zbu presenting by far the most complex system. Three stems are found across the four languages: stem 1 (basic stem), stem 2 (aorist), 584 stem 3 (non-past transitive, singular non-inverse agent). In Zbu, the alternations are largely unpredictable (as with strong verbs in Germanic). Some of these segmental and tonal alternations present polarity: Table 1. Stem alternations with polarity in Zbu (Sun 2004:276) alternation stem 1 stem 2 meaning i-ə ə-i i-e e-i smə̄ vī ʁlē qʰjîv to be cooked to come to speak ill of to be bitter smî və̄ ʁlî qʰjêv Some alternations are partially productive, as demonstrated by the fact that they apply to Tibetan (or even Chinese) loanwords. For instance, the Zbu verb nkrós ‘to discuss’, a denominal verb derived (by the preijix n-) from the borrowed Tibetan noun gros ‘discussion’, has the following three stems: nkrós, nkhrús, nkrə́m (personal ijieldwork). In some cases, it can be proven that some stem alternations are secondary from a diachronic point of view. For instance, the Japhug a/e, o/e, ɯ/i vowel alternations between stem 1 and stem 3 have been shown to originate in the fusion of a sufijix *-jə (still attested in Tshobdun, see Sun 2003) with the verbal stem (see Jacques 2008:234–235). In most cases, the irregular alternations go back to proto-Rgyalrong and possibly to an even earlier stage. 3.2 Inverse marking Direct/Inverse alignment is a type of agreement system in which a marker (the inverse marker) indicates that the agent is lower than the patient on the empathy hierarchy: 1. SAP (Speech Act Participant, 1st or 2nd person) >> human >> animal >> inanimate >> generic Verbal forms where the agent is ijirst or second person and the patient is third person (SAP>3) are called direct, as are forms with two third For use by the Author only | © 2017 Koninklijke Brill NV 585 Rgyalrong Language person arguments (3>3) when the agent is higher than the patient on hierarchy (1). Direct forms never have the inverse preijix in Japhug. The existence of inverse marking in Rgyalrong languages was ijirst noticed by DeLancey (1981) and described by Sun and Shi (2002) and Jacques (2010). The following table presents the paradigms of regular transitive and intransitive verbs in the non-past (the only ijinite verbal form without tense/aspect/mood preijixes) in Japhug Rgyalrong, only taking into account verbal forms where both arguments are singular. We only present here a restricted account of Japhug verbal agreement. The -Ø sufijix marks non-ijirst person singular; it is replaced by other sufijixes in the dual and plural. The symbols Σ1 and Σ3 represent verbal stems 1 and 3, respectively. The rows here refer to person of the singular agent of the transitive verbs (A) or the sole argument of the intransitive ones (S), and the columns to the person of the singular patient (P). For example, the form for a transitive verb with second-person agent and third-person patient (2>3) is found in the second row, third column (tɯ-Σ3-Ø) (See Table 2). The “local scenario” forms where both arguments are SAP, 1>2 and 2>1, are marked by special portmanteau preijixes ta- and kɯ- respectively which are synchronically unanalyzable; the sufijixes in these forms are coreferent with the patient (non-ijirst person singular -Ø for 1>2 and ijirst person singular -a for 2>1). In “mixed scenarios” when one argument is SAP and the other is third person, we ijind two situations. First, when the SAP is the agent, the verbal form is similar to that of an intransitive verb save for the fact that stem 3 appears instead of stem 1. When the SAP is patient, sufijixes and preijixes are still the same as that of intransitive verb forms, but the inverse preijix wɣɯ́ - or wɣ- is placed immediately before the verbal stem. The following examples illustrate 1>3 and 3>1 forms: 2. kɯki mbro ki ju-nɯtsɯm-a this horse this ipf-take.away-1sg I will take away this horse. (Japhug; The Fox, 59) 3. a-rɣa nɯ-mbala kɯ 1sg.poss-neighbor 3pl.poss-bull erg wɣɯ́ -nɯtsɯm-a ɲɯ-ŋu inv-take.away-1sg ipf-be The neighbors’ bull is about to take me away! (Japhug; The Demon, 8) When both arguments are third person (“nonlocal scenario”), the transitive verb has two possible forms, either with the inverse preijix wɣɯ́ - / wɣ- and with stem 1, or without this preijix and with stem 3. The choice between the two forms is dictated by both semantic and pragmatic factors. In the form Σ3-Ø the singular sufijix -Ø marks the agent, while in wɣɯ́ -Σ1-Ø, the same sufijix marks the patient. With an animal agent and human patient, direct forms are possible in some cases; inverse is obligatory with an inanimate agent and animate patient, and with a generic agent (even a human agent) in Japhug; human generic arguments are lower than inanimates on the empathy hierarchy in this language. Inverse morphology is relatively widespread in American languages (see Zúñiga’s 2006 for a recent survey) but rare on other continents. Rgyalrong is the only group of languages in Eurasia with a consistent inverse system, where the same morpheme is used in both mixed and nonlocal forms (and even in 2>1 local forms in Situ and Tshobdun). Table 2. A restricted Japhug verbal paradigm (non-past, both arguments singular) 1P 1 A/S 2 A/S kɯ-Σ1-a 3 A/S wɣɯ-Σ1-a 2P 3P Intr ta-Σ1-Ø Σ3-a Σ1-a tɯ-Σ3-Ø tɯ-Σ1-Ø tɯ-wɣ-Σ1-Ø Σ3-Ø Σ1-Ø wɣɯ-Σ1-Ø For use by the Author only | © 2017 Koninklijke Brill NV Rgyalrong Language 3.3 Incorporation In Japhug and other Rgyalrong languages, we ijind a peculiar incorporation-like construction called indirect incorporation (Jacques 2012c). Verbs with indirect incorporation are compounds comprising a nominal stem with vowel changes (o, a, u > ɤ/a, i > ɯ) and a verbal stem. Unlike prototypical incorporation, indirect incorporation is built not by coalescence of noun and verb, but by denominal derivation of a nominal compound, made up of a nominal root and a verb root, by means of a derivation preijix nɯ-, nɤ-, ɣɯ- or sɯ-; most (but not all) verbs with indirect incorporation still have corresponding compound action noun in Japhug, and this derivation is still synchronically transparent (see Table 3). Verbs derived by indirect incorporation are not numerous—only about thirty in Japhug have been found—but the process is still marginally productive, as some verbs have incorporated a noun borrowed from Chinese (‘earn money’ above, from Chinese piàozi 票子, for instance). The incorporated noun can be either the only argument of an intransitive verb (as in ‘gallop’), the patient of a transitive verb (‘to fell trees’) or an adjunct (the other examples); it is never the agent of a transitive verb. The derived verb is generally intransitive, except in some rare cases like ‘to hit with the head’ when the original verb was transitive and the incorporated noun was an adjunct. The following pair of examples taken from the same story illustrates the object-verb vs. indirect incorporation (Jacques 2012a): 586 4. ɬasa ju-kɯ-ɕe tɕe, nɯ tɕu Lhasa ipfv-genr:S/O-go conj dem conj pɕawtsɯ kɤ-fsoʁ ɲɯ-mbat money inf-earn const-easy ‘If one goes to Lhasa, money is easy to earn there.’ (Lobzang, 22) 5. nɤ-mbro nɤ-rŋɯl 2sg.poss-horse 2sg.poss-silver tu-rke-a tɕe ipfv-put_in[3]-1sg conj kɯ-ɣɯ-pɕawtsɯ-fsoʁ jɤ-ɕe nmlz:S/A-derivation-money-earn imp-go tɕe conj ‘I will prepare a horse and some silver for you, go to earn money.’ (Lobzang, 17) Although many languages around the world present more extensive incorporation than Rgyalrong languages, this phenomenon is exceedingly rare in languages of China and Asia in general (excluding the polar and subpolar regions). 3.4 Word order typology Rgyalrong languages have a highly unusual word order typology, as they are strictly verb ijinal but have mainly preijixing morphology, both in the nominal and in the verbal domain. Below a simpliijied verbal template for Japhug is given, where we ijind fourteen preijixal slots and only three sufijixal slots. Table 3. Examples of indirect incorporation in Japhug noun verb compound action noun derived verb meaning mbro ‘horse’ qʰu ‘back’ si ‘timber’ tɯ-ku ‘head’ pɕawtsɯ ‘money’ rɟɯɣ ‘to run’ it. ru ‘to look’ it. pʰɯt ‘to chop’ tr. tɕʰɯ ‘to gore’ tr. fsoʁ ‘earn’ tr. mbrɤ-rɟɯɣ ‘horse race’ qha-ru ‘a look back’ sɯ-pʰɯt ‘tree felling’ — pɕawtsɯ-fsoʁ ‘earning money’ nɯ-mbrɤ-rɟɯɣ nɤ-qʰa-ru ɣɯ-sɯ-pʰɯt nɤ-kɤ-tɕʰɯ ɣɯ-pɕawtsɯ-fsoʁ ‘to run’ ‘to look back’ ‘to fell trees’ ‘to hit with the head’ ‘to earn money’ For use by the Author only | © 2017 Koninklijke Brill NV 587 rgyalrong language 1 2 3 4 a- mɯ- ɕɯ- tɤ/mɤ- ɣɯ- pɯetc. trnsl cisl irr neg dir 5 6 7 8 9 10 tɯ- wɣ-, ʑɣɤ- sɯ- rɤ- nɤkɯ- asɯ prog inv anti 2 refl caus pass appl 11 12 13 14 15 16 17 18 a- / nɯ- ɣɯ-, incor- verb -t -a -nɯ ɤrɤ- porated stem noun other pst 1sg agr pass auto Complex morpho-phonological rules govern the interaction between preijixes: the morphology is not plainly concatenative. entire world that share strict verb-ijinal order and mainly preijixing morphology with Rgyalrong languages are Yenissean and Athabaskan. 6. ɲɯ-tɯ-ɤ́<wɣ>sɯ-zgroʁ const-2-prog<inv>-attach He is attaching you. (elicitation) 4. H i s t o r i c a l L i n g ui s t i c s Although some combinations are not possible (for instance the inverse preijix cannot appear with the reflexive, which causes the verb to become intransitive), it is quite common to ijind examples with ijive preijixes in a row in traditional stories: 7. a-ɣɯ-lɤ-kɯ-sɯ-mtshám-a irr-cisl-prf:upstream-2>1-caus-hear-1sg You will come here to tell me. (Japhug; The three sisters, 132) With elicited examples it could be possible to obtain eight preijixes or even more. It may be noted that while the sufijixes have no or few irregular forms, the preijixes present many irregularities, which suggests that they underwent grammaticalization signiijicantly earlier. This pattern of preijixation is unusual given the strict verb-ijinal word order, which usually correlates with mainly or exclusively sufijixing morphology. The only language families in the The importance of Rgyalrong languages for the study of comparative Sino-Tibetan linguistics has long been noticed (see for instance DeLancey 1981). In phonology, Rgyalrong languages appear to be extremely conservative: they preserve archaic clusters only found in a few other languages or in reconstructed languages such as Proto-LoloBurmese or Old Chinese (see Table 4). The fact that loanwords from Tibetan into Rgyalrong languages preserve many features lost in all Tibetan languages of China (such as ijinal -s, clusters of the type bsC- or brC-, etc.) is a further conijirmation that these languages have a generally conservative phonology. The phonological conservatism of Rgyalrong languages makes them a useful model for the reconstruction of other subgroups of the SinoTibetan family (see in particular the reconstruction of the Naish group, Jacques and Michaud 2011). In derivational morphology (see for instance Sun 2006; Jacques 2012a), Rgyalrong languages preserve many afijixes and morphological Table 4. Japhug Rgyalrong compared to other ST languages Meaning Japhug Tibetan louse shame body to steal zrɯɣ -zraʁ -skʰrɯ mɯrkɯ shig Proto-Lolo-Burmese (Bradley/Matisoff) *s-rakL sku rku *ko² Old Chinese (Baxter and Sagart) *srik > shī 蝨 *srˁɨk > sè 色 *qʰ(r)ˁo > qū 軀 *kʰˁ(r)o-s > kòu 寇 For use by the Author only | © 2017 Koninklijke Brill NV rgyalrong language processes which may be reconstructible to Proto-Sino-Tibetan, such as the causative sɯ-, sɯɣ-, z- (with three regular and four irregular allomorphs), anticausative prenasalization and the oblique nominalization preijix sɤ-, agent or patient nominalization kɯ-/kɤ-, action noun nominalization in tɯ-, passive a- < *ŋa-, etc. In inflectional morphology, Rgyalrong languages constitute an important piece of evidence in the debate regarding the antiquity of verbal agreement in the Sino-Tibetan family. The Rgyalrong agreement system presents many similarities with that of Kiranti, a group of SinoTibetan languages spoken in Nepal, and most scholars accept the fact that at least some of the verbal morphology found in Rgyalrong and Kiranti is cognate (Jacques 2012b). 5. Conclusion Given the considerable importance of these languages for both typology and comparative linguistics, it is surprising that so few scholars engage in the description of these highly endangered languages, whose importance for the study of the Sino-Tibetan family as a whole can be aptly compared to the role of Sanskrit in Indo-European. In addition to their importance for understanding the history of Sino-Tibetan as a whole, they open radically new perspectives for Old Chinese historical phonology and morphology. A c k nowle dgment I wish to thank Lin Youjing, Zev Handel, Waltraud Paul, and Jackson T.S. Sun for comments on this article; I am responsible for any remaining mistakes. Bibliography Btsan.lha Ngag.dbang Tshul.khrims, Rgyal.rong dmangs.khrod gtam.tshogs [Popular Rgyalrong texts], Běijīng 北京: Mi.rigs dpe.skrun khang, 2010. DeLancey, Scott, “The Category of Direction in TibetoBurman”, Linguistics of the Tibeto-Burman Area 6/1, 1981, 83–101. Huáng Liángróng 黄良荣 and Sūn Hóngkāi 孙红开, Hàn-Jiāróng cídiǎn 汉嘉戎语词典 [A Chinese- 588 Rgyalrong dictionary], Běijīng 北京: Mínzú 民族 出版社, 2002. Jacques, Guillaume, “Phonologie et morphologie du Japhug (rGyalrong)” [Phonology and morphology of Rgyalrong], dissertation, Université Paris VIIDenis Diderot, 2004. Jacques, Guillaume, Jiāróngyǔ yánjiū 嘉絨語研究 [Study on the Rgyalrong language], Běijīng 北京: Mínzú 民族出版社, 2008. Jacques, Guillaume, “The Inverse in Japhug Rgyalrong”, Language and Linguistics 11/1, 2010, 127–157. Jacques, Guillaume, “Argument Demotion in Japhug Rgyalrong”, in: Gilles Authier and Katharina Haude, eds., Ergativity, Valency and Voice, Berlin: Mouton de Gruyter, 2012a, 199–225. Jacques, Guillaume, “Agreement Morphology: The Case of Rgyalrongic and Kiranti”, Language and Linguistics 13/1, 2012b, 83–116. Jacques, Guillaume, “From Denominal Derivation to Incorporation”, Lingua 122/11, 2012c, 1207–1231. Jacques, Guillaume and Chen Zhen, Une version rgyalrong de l’épopée de Gesar [A Rgyalrong version of the epic of Gesar], Gyalrong Studies 1, Senri Ethnological Reports 93, Osaka: National Museum of Ethnology, 2010. Jacques, Guillaume and Alexis Michaud, “Approaching the Historical Phonology of Three Highly Eroded Sino-Tibetan Languages: Naxi, Na and Laze”, Diachronica 24/4, 2011, 468–498. Lín Xiàngróng 林向荣, Jiāróngyǔ yánjiū 嘉戎語研究 [Study on the Rgyalrong language], Chéngdū 成都: Mínzú 民族出版社, 1993. Lin Youjing, “Tense and Aspect Morphology in the Zhuokeji rGyalrong Verb”, Cahiers de LinguistiqueAsie Orientale 32/2, 2003, 245–286. Lin Youjing, “Units in Zhuokeji rGyalrong Discourse: Prosody and Grammar”, dissertation, University of California at Santa Barbara, 2009. Sun, Jackson T.-S., “Parallelisms in the Verb Morphology of Sidaba rGyalrong and Lavrung in rGyalrongic”, Language and Linguistics 1/1, 2000, 161–190. Sun, Jackson T.-S., “Caodeng rGyalrong”, in: Graham Thurgood and Randy J. LaPolla, eds., Sino-Tibetan Languages, London: Routledge, 2003, 490–502. Sun, Jackson T.-S., “Verb-Stem Variations in Showu rGyalrong”, in: Ying-chin Lin et al., eds., Studies on Sino-Tibetan Languages: Papers in Honor of Professor Hwang-Cherng Gong on His Seventieth Birthday: Language and Linguistics, Taipei: Institute of Linguistics, Academia Sinica, 2004, 269–296. Sun, Jackson T.-S., “Jiāróng yǔzǔ yǔyán de yīngāo: liǎnggè gè’àn yánjiū 嘉戎语组语言的音高：两个 个案研究” [On pitch in the rGyalrongic languages: Two case studies], Yǔyán Yánjiū 语言研究 25/1, 2005, 50–59. Sun, Jackson T.-S., “Jiāróngyǔ dòngcí de pàishēng xíngtài 嘉戎语动词的派生形态” [Derivational morphology in the Rgyalrong verb], Mínzú Yǔwén 民族语文 4, 2006, 3–14. For use by the Author only | © 2017 Koninklijke Brill NV 589 Rhetoric Sun, Jackson T.-S., “The Irrealis Category in rGyalrong”, Language and Linguistics 8/3, 2007, 797–819. Sun, Jackson T.-S. and Shí Dānluó 石丹羅, “Cǎodēng Jiāróngyǔ yǔ ‘rèntóng děngdì’ xiāngguān de yǔfǎ xiànxiàng 草登嘉絨語與 ‘認同等第’ 相關的語法 現象” [Empathy hierarchy in Cǎodēng rGyalrong grammar], Language and Linguistics 3/1, 2002, 79–99. Zúñiga, Fernando, Deixis and Alignment, Inverse Systems in Indigenous Languages of the Americas, Amsterdam: Benjamins, 2006. Guillaume Jacques Rhetoric There are many deijinitions of rhetoric. In her book, Rhetoric in Ancient China, Xing Lu deijines it as “the art of persuasion, the artistic use of oral and written expressions, for the purpose of changing thought and action at social, political and individual levels” (1998:2). This deijinition will act as our guide in this entry. In Chinese, persuasion was known as shuì 說, explanation as míng 明, and argumentation as biàn 辯. Although there is overlap between these terms (and others), Lu argues that each word has a particular function in conceptualizing and contextualizing persuasive discourse. For example, shuì is associated with face-toface persuasion and míng deals with the use of symbols in social and epistemological contexts. Lu suggests that the term míngbiànxué 明辨學 is comparable to the Western study of rhetoric, with míng aiming to seek truth and justice and biàn concerning the art of persuasion. This term also captures the contradiction inherent in the two key concepts of Western rhetoric, viz. truth and/or persuasion. Distinctions between “Western” and Chinese concepts of rhetoric have been frequently drawn (Harbsmeier 1999; Shankmann and Durrant 2002; Jullien 2004). The distinctions include the origins of rhetoric, with the law courts of Classical Greece providing the impetus for the use of forensic rhetoric while, in China, rhetoric was used to persuade rulers (Graham 1989). This leads to a further distinction of Classical Greek rhetoric, namely its use in Greece among equals, while Chinese rhetoric was more “bottom-up”. This, in turn, meant that the Chinese persuader had to be careful. The Chinese philosopher Guǐgǔzi 鬼谷子 (more or less a contemporary of Aristotle) was clearly aware of this and the influence of the audience upon rhetorical style. He considered persuasion from below (yīn 陰) to above (yáng 陽) to be a disturbance of the natural order of things. Persuasion from below to above or from an inferior to a superior was yīn and required special effort. Persuading from above to below was yáng and required less effort. “Yáng . . . encourages straightforward speaking. Yīn . . . encourages speaking in forked tongue” (Tsao 1985:103). A third distinction which developed as a consequence of the different Chinese and Western rhetorical tradition is that Western rhetoric is considered more “agonistic”, while Chinese is considered more “irenic” (Durrant 2002:283). But, as Durrant also points out, Chinese could be agonistic, giving as examples Wáng Chōng 王充 (27–110 CE) and others who criticized the great historian Sīmǎ Qiān 司馬遷 (c. 145–90 BCE). This is a salutary reminder that there were (and remain) many different ways of “doing” rhetoric in both these great traditions. In this context, while it is true that Chinese rhetoric is more associated with writing than with speech, while Western rhetoric is more associated with speech than with writing, it is certainly not true to say that this was ever at the expense of the other. One reason why it is difijicult to be accurate about the origins of Chinese rhetoric is that rhetoric did not constitute a discrete discipline in Classical China. As illustrated above, different terms described different aspects of rhetoric. There were others, including: yán 言, ‘speech and the use of language’; cí 辭, ‘modes of speech, types of discourse, eloquence, style’; and jiàn 諫, ‘giving advice, persuasion’ (Lu 1998:5). The current Chinese term for rhetoric is xiūcí 修辭, which literally means something like ‘reijining words’. Its ijirst recorded use is the Confucian classic, the Book of Changes (Yìjīng 易經). The term—to mean rhetoric as an independent discipline—is, however, recent. It is ijirst commonly used when Chinese scholars started For use by the Author only | © 2017 Koninklijke Brill NV