Title the Identity of the Turkish Rulers to the South of Hindukush from The
Total Page:16
File Type:pdf, Size:1020Kb
The Identity of the Turkish Rulers to the South of Hindukush Title from the 7th to the 9th Centuries A.D. Author(s) INABA, Minoru Citation ZINBUN (2006), 38: 1-19 Issue Date 2006-03 URL https://doi.org/10.14989/48824 © Copyright March 2006, Institute for Research in Humanities Right Kyoto University. Type Departmental Bulletin Paper Textversion publisher Kyoto University ZINBUN 2005 No.38 The Identity of the Turkish Rulers to the South of Hindukush from the 7th to the 9th Centuries A.D.* Minoru IN ABA 1. The Tu rkish Rulers of Eastern Afghanistan At the beginning of the 1990 う 民 m a n y Bactrian documents were discovered in the northern area of Afghanistan which was still in the middle of a civil war. There is no doubt that those documents will ラtogether with the Bactrian inscription from Rabatak near 8urkh Kotal and Buddhist documents from the caves north of Bamiyan , help us to improve our understanding of the history う religion and the society of ancient Afghanistan. A number of those Bactrian documents have been introduced by N. 8ims- Williams with transliterations and translations うwhich will surely stimulate further research (β倒S臼 i m トs - W i ll i a m 2000) Just before that discovery ぅbeing inspired by 8h. K uwayama うs work on the history ofK 瓦pi 釘and the Gandh 瓦ra area prior to the Islamic period (K uwayama 1990) う I published a paper on the history of eastern Afghanistan , specially foュ cusi 時on Ghazni/Z 瓦bulistan from the 7th to the 8th century A. D ・ ぅscrutiniz ing materials available at that tilne (Inaba 1991). A framework of the history of that area during the period in question was deduced as follows: In the latter half of the 6th century , the Hephtalite empire in Tokh 瓦rist 瓦n collapsed due to Tu rkish attack うi.e. Western Tu Jue. Gandh 瓦ra , which had also been under the control of the Hephtalites うand K 瓦pi 釘which was adj acent to Gandh 瓦ra , were greatly affected by that event. The dynasty labeled the Khi 時alides by Kuwayama (or Nezakides by some scholars bωed on the leg- * I would like to express my gratitude to Prof. Y. Yoshida and D r. A. Palumbo , who provided me with invaluable instruction and suggestions ぅand ぅas always , to D r. 8h. Kuwayama who has never failed to kindly and with great generousity discuss with me at great length. And Iam also grateful to Mr. Anthony R. Black for improving my English. 1 MINORD INABA end of the coins) had been established in K 瓦pisl and was exerting its rule over the K 瓦bul valley as far as Gandhara. When in the first half of the 7th century Xuanzang visited there うthe king of K 瓦pisl was of this lineage. In the 650 う h Arab Muslims reached as far as Slstan and intended to make their way further north うwhich resulted in considerable strategic importance being placed on the southern border of the K 瓦pisl kingdom うnamely K 瓦bu l. At that time , K 瓦bul was under the rule of the Turks who had been reported by Xuanzang to have lived in the mountain area between Z 瓦bulist 瓦nand K 瓦pi 釘 うthat is ヲFulishisa tangna 弗 栗 侍 薩 慎 那 . Gradually having extended their power ぅthey came to be independent and うeventually うdepri ved the Khingalide King of almost the whole of his territory. It was they who were referred to as the Turk Sah in the Islamic sources. In the 680 うs , a prince from K 瓦bul fled southward down to Z 瓦bulistan , probュ ably due to the conflict surrounding succession to the throne うand established his independence there (C f. K uwayama 1999). The new r 叫ers of Zabulist 瓦n were denoted in the records of Arab and Persian historians by the name ふ ん rtby l. It has been maintained that this title should be read ム ヂ o znbyl which might have some relation with the cult surrounding the God Zhun (如 n ii of Xua 四ang). Considering the Turkish situation mentioned above , however ぅthis title should be understood as a corrupted form of the Turkish title iltiibiir. 1 Thus うthe two kingdoms of Kabul and Z 瓦bulist 瓦n were of the same Turkish origin and both acted as an impregnable rampart against the Muslims for almost two centuries following. Nevertheless うthere remain several problems unsolved and unmentioned う one of which is the origin of those Tu rks. The Bactrian documents mentioned above and some archaeological findings of the area in question from the last century seem to indicate the Turkish tribe named the Khalaj as a possible candidate for the origin of those Turks. 2 The primary purpose of this paper is to verify this identification. As a matter of fact , the idea of identifying the Turks in question with the Khalaj tribe is not very recent. L. Petech (1964) already suggested this identification and referred to the expressions “ H e - d a - I u o - z h i 詞 達 羅 支 " and “Ge da-Iuo-zhi 葛達羅支 J うwhich appeared in Chinese sources. A. Rehman (1988) also connected the Turks of K 瓦bul and Zabulist 瓦n with the Khalaj utilizing the materials cited by J. Marquart (1 1 C f. Bombaci 1970: 59; Rehman 1988: 180; Inaba 1991: 53-55; Sims-Williams 2002: 235. 2 C f. Lee & Sims-Williams 2003; Verardi & Paparatti 2004; Inaba 2004. 2 THE IDENTITY OF THE TURKISH RULERS scholars may be , they seem not to have succeeded in examining it su 白ciently う as Petech never produced a comprehensive comparison of information from the Chinese sources and that of the Arabic and Persian sources うnor did Rehman bring material in the Chinese sources to light. However ぅY. Yoshida (2000 [2003]) recently has suggested the assumption that the Chinese Hedaluozhi and Gedaluozhi can be connected with the word “ k h a r a l a じ が う w h i c h appears on the legends of the N ezak S 瓦 h coins うconsidering the latter to be some kind of Indianized form of the word “Khalaj ." I will take this suggestion as my starting point. 2. Hedaluozhi To begin with うlet us consider the Chinese expressions Hedaluozhi and Gedaluozhi. However , since each has variants うit is first necessary to examine the expressions themselves to decide which should be taken as the correct form or whether there are multiple correct forms. To start with うlet us proceed with a passage from X· 初 旬ngshu Vo l. 221b: “ X i e y u 謝 周 目 li e s southwest of Tokh 瓦ristan. It was formerly called Caojuzha 漕 矩 陀 or Caoju 漕 矩but renamed Hedaluozhi in the Xianqing period. Empress Wu again changed the name to the present one. About 400 li to its east lies Jibin 厨 賓and to its northeast lies Bamiyan at almost the same distance. India lies to the south う Persia lies to the west and Guzgan lies northwards. The king resides in Hexina (鶴悉那 * y a k - s i t - n a ) which is 7 ぅ000 li wide. The king also sometimes resides in the Aponi (阿婆体九 - b u a - n i a ) castle. The land produces plenty of saffron and water springs ir 向ate the fields. People from Tu Jue , Jibin and Tokh 瓦rist 加 ( 吐 火 羅 ) live there. Jibin is recruiting young soldiers from this country to against protect the Arabs. They sent an envoy and presents in the 1st year of the Jingyun era (710 A.D.) and later subjugated themselves to Jibin. In the 8th year of Kaiyuan (720) うthe Emperor mandated Gedalωzhi xielifa Shiquer and made him the king They have sent several envoys until the Tianbao era. 門 Here three names for Zabulistan in the Tang period are mentioned , i.e. う Caojuzha , Hedaluozhi (Gedaluozhi) and Xieyu. The Chinese names applied to Z 瓦bulist 瓦n have been analyzed by Yoshida (Kuwayama 1992: 135-139) ぅac cording to whom Caojuzha and Xieyu , as well as Caoli (Xu g αoseng zhu α η う Vo l. 4) ぅ X i e y u e (Huilin yinyi う V o l. 100) and Zhewutl 刷tana (Xuany 仇yinyi う Vo l. 18) are transcriptions for *dz 瓦ul くZ 瓦b 叫-istan) うwhile Hedaluozhi and Gedal uozhi (and Gelωdazhi in the Jiu t α n g s h u ; see below) must have been 3 MINORD INABA transcriptions for a word completely di 百erent from *dz 瓦ul くZ 瓦bu l. As for the latter transcriptions , E. Chavannes (1969: 160 うn. 4) presumes that Hedaluozhi could be an erroneous transcription for Daluohezhi 達 羅 詞 支which might corュ respond to al- Rukhudh くal- Rukhkhaj appearing in Islamic sources , especially in geographical works written during the 9th and the 10th centuries うas a place name of southern Afghanistan around the present Qandahar. As mentioned above うPetech suggested connecting the form to Khalaj. While Kuwayama (1990: 253-54) , distinguishing Hedaluozhi from Gedaluozhi うsupposes the forュ mer to be a place name and the latter to be a kind of Turkish 0 日 c i a l title , the *Qarata 己i. Hedaluozhi also appears in the Zizhi tongjia 凡 V o l. 200: “On the day guiwei of the 6th month [of 661 ]う e i g h t dudufu and seventy-six zhou 州 had been placed in the sixteen countries ぅ i.e. Tl 山 lOluo (Tok 瓦ra) うYanda (Hephtal) ヲ Jibin ぅBosi (Persia) etc. 門 This account concerns the reorganization of Anxi duhufu and sixteen counュ tries in the West in the 1st year of Longshuo (661) ラwhich is discussed below.