A Study on the Kidarites

A study on the Kidarites: based primarily on documentary sources Xiang Wan Various documentary and numismatic sources mention the Kidarites. This article attempts to reconstruct the chronology of the Kidarites primarily on the basis of documentary sources, especially Chinese Buddhist sources which have not yet been thoroughly studied. The author also discusses the ethno-political affinity of the Kidarites as well as their relationship with other contemporary ethnic groups.

The Kidarites in Chinese Buddhist Sources Frantz Grenet illustrates the history of the Kidarites from an integrated perspective in his article (Grenet 2002: 205-9), as well as in the entry on the lemma “Kidarites” in Encyclopædia Iranica (Grenet 2005b). In the Conclusion of his article he argued that “philology, numismatics and field archeology could and should give one another mutual support” (Grenet 2002: 220), advice he has followed in his writings. Nevertheless, there have been such great divergences between historians and numismatists that “reconstructions of the Kidarite period differ markedly”, as Grenet points out in the commencement of his article concerning the Kidarites (Grenet 2002: 205). On the one hand, he cited the Japanese scholars Kazuo Enoki (1969-70)1 and Shoshin Kuwayama (1989), the former of whom, relying on Chinese and Byzantine chronicles, rejected the early dating of Kidāra’s incursion into Northwest India and the connection of the Kidarites with “Huns”. On the other hand, Grenet discusses the numismatic perspective proposed by Göbl (1967 II: 34-6) based on Göbl’s reading βαγο κιδαρο on the latest series of the gold coins present in the hoard of Tepe Maranjān near Kabul in the reign of Shapur III (r. 383-388), and concluded that “the Kidarites represent the first wave of invaders in Bactria and Northwest India (in the latter half of 4th century). They are identified with the Chionites [i.e. Xiongnu]” (Grenet 2002: 206). This disagreement between documentary and numismatic approach might be reconciled with modifications to either approach. Grenet (ibid., 206) questioned Göbl’s reading of the coin legend mentioned above, and adopted Enoki’s viewpoint (with reserve) that the formation of the Kidarite kingdom in Bactria and the conquest of Gandhāra and ‘Five kingdoms’ to its north be dated to the end of 420’s (ibid., 220). On the contrary, the documentary evidence, first reviewed by Enoki (1969-70), was never beyond dispute. Yu (1986; 2001) criticized Enoki’s arguments, primarily on issues regarding the Hephthalite conquest of the former Kidarite kingdom. In the present article, I will first examine the Chinese Buddhist sources, some of which Enoki and Kuwayama have consulted, in order to reveal the overlooked traces Chinese pilgrim monks and their Indian masters left on events in the Kidarite period.

Faxian’s account of Kidāra 1

It is the argument of Enoki that informs Grenet’s view that the Chinese pilgrim Faxian who visited Northwest India/Gandhāra during the first years of the 5th century “does not mention any” invasion from Bactria (Grenet 2002: 205; 2005b). Nevertheless, after a review of Faxian’s itinerary in the country of Fulousha 弗樓沙 (Puruṣapura) we must cast doubt on this assertion. After a brief sojourn in the country of Jiantuowei 揵陀衛 (*Gandhāvatī)2 Faxian and his companions traveled south to the country of Fulousha: Going southwards from Jiantuowei, the pilgrims in four days arrived at the country of Fulousha. Formerly, when the Buddha was travelling in this country with his disciples, he said to Ānanda, ‘After my parinirvāṇa, there will be a king named Jinijia 罽膩伽, who shall on this spot build a stūpa.’ This Jinijia was afterwards born to the world; and once when he had gone forth to look about him, Śakra, Ruler of Devas, wishing to excite the idea in his mind, assumed the appearance of a little herd-boy, and was making a stūpa right in the way of the king, who asked what sort of a thing he was making. The boy answered, ‘I am making a stūpa for the Buddha.’ The king said, ‘Very good;’ and immediately, right over the boy’s stūpa, he proceeded to rear another, which was more than fourty zhang high, and adorned with layers of all the ratnas. Of all the stūpas and vihāras which the pilgrims saw in their journey, there was no one comparable to this in solemn beauty and majestic grandeur. There is a current saying that this is the finest stūpa in Jambudvīpa. When the king’s stūpa was completed, the little stūpa of the boy came out from its side on the south, rather more than three chi in height. The Buddha’s bowl is in this country. Formerly a king of Yuezhi raised a large force and came to invade this country, wishing to carry the bowl away. When he had subdued the country, the king of Yuezhi3 had sincere faith in the Buddha’s law and wished to take it away. Accordingly they held religious services to triratna on a large scale. When they had done so, they had a large elephant grandiosely decorated, on which they placed the bowl. But the elephant knelt down to the ground, and was unable to go forward. Again they made a four-wheeled cart, in which the bowl was placed, and made eight elephants draw it, but neither were they able to go forward. Then the king knew that the karma-relation between himself and the bowl had not arrived yet, and was deeply ashamed of himself. Forthwith he built a stūpa and a monastery at this place, and left guards to watch the bowl, making all sorts of dedications. The monastery is capable of housing about seven hundred monks. When the sun is about to be in the middle of the sky, many monks bring out the bowl, and hold various kinds of religious services to it together with the laity, then take their midday meal. When they offer incense at the usual rite of sunset, they again do it in the same manner. The bowl can contain about two dou [four liters], and is of various colour, black predominating, with the four rims clearly marked. Its thickness is two fen [about half a centimeter], and it has a bright and glossy luster. When poor people throw a few flowers into it, it becomes immediately full, while some very rich people, wishing to make offerings of many flowers, might not stop till they have thrown in hundreds, thousands, and even myriads of bushels, and still will not be able to fill it. (Legge 1886: 33-5; partly modified and words in brackets added by Kuwayama 1990: 948, 2

slightly emended by the author). Here Faxian related two tales. The one is the legend of King Jinijia 罽膩伽, who undoubtedly referred to the great Kushan king Kaṇīṣka who built the great stūpa next to the city of Puruṣapura; 4

in the other tale Faxian narrated a Buddhist Yuezhi king’s invasion of Puruṣapura. After

capturing the city, he hoped to carry the Buddha’s bowl away but failed to move it however he tried, and so he had to build a stūpa and a monastery there, and left behind a guard and dedications to the bowl. The identification of this Yuezhi king varied among historians. A. H. Dani (1969: 53) and Zhang (1985: 44 n. 18) proposed Kujula Kadphises, the first Kushan king, who occupied Northwest India after annexing four other yabgus according to Chinese chronicles. They agreed with the evidence that the inscriptions on Kujula’s coin (published in Majumdar 1951: 138) read dhramaṭhida and sachadhramaṭhita in Gāndhārī and imply he was a pious Buddhist, in accordance with Faxian’s description of him as having “sincere faith in the Buddha’s law”. Nevertheless, as Lüders has pointed out, the legend “sachadhrama” in Kharoṣṭhī script is actually a sign of Śaivaism, not of Buddhism (cf. Harmatta 1992: 317-8). The prevailing opinion on the Yuezhi king is proposed by Legge (1886: 34), who considered he “was perhaps Kanishka himself”. Enoki (1969-70) gives a sound reasoning for this view, which has so far been accepted by most scholars. Enoki’s argument is based on the repudiation of Martin (1937: 36), who identified the Yuezhi king with “an Ephthalite king”. Enoki raised two questions in his two articles respectively: The passage “in the old days a Yüeh-chih [Yuezhi] king, etc.” can not be taken as meaning an event which took place only a few years earlier, and the Hephthalites’ appearance in Gandhāra took place, as we shall state later, sometime after 477 and before 520. (Enoki 1998: 55, slightly emended) It should be noted that the accounts are written as if these countries were independent of each other and that there had never existed a power to unite them. It contrasts strikingly with Fa-hsien’s [Faxian] clear description of the Middle country (Madhyadeśa), that is, regions along the Gangetic Basin, having been under the rule of a king (of the Gupta Dynasty).5 One should consider, therefore, that the unification of the north and south of the Hindūkush by the Kidarites had not yet been achieved by the time Fa-hsien visited there. (ibid., 80, slightly emended) With the first reason Enoki precludes the identification with the Hephthalites, whom he argued emerged in Northwest India between 470 and 520. With the second argument, he rules out the possibility of the Kidarites, who Enoki believed united both sides of the Hindūkush under one power. Admittedly, Martin’s conclusion is far from correct, but Enoki’s evidence is also in no way impeccable. His viewpoints, though persuasive at the first glance, fail to sustain an in-depth analysis. The first question in Enoki (1969) is, surely there should have been an appropriate period of time between the Yuezhi king’s event and Faxian’s visit in Gandhāra (403).6 The Chinese text translated “in old days” here is “xi” 昔, which Kuwayama, following Legge, translated as 3

“formerly”. It is not an exact number here, unless we turn to another paragraph from the very beginning of Faxian’s itinerary: In former times Faxian was living in Chang’an 法顯昔在長安, deploring the mutilated and imperfect state of vinayapiṭaka, in the second year of Hongshi 弘始 (399-414) era being the Jihai 己亥 year of the cycle, he entered into an engagement with Huijing 慧景, Daozheng 道整, Huiying 慧應, and Huiwei 慧嵬, that they go to India and seek the vinayas. (Legge 1886: 9-10, slightly emended). The Jihai 己亥 year of the cycle is actually not the second but the first year of Hongshi era, namely 399-400 AD (Zhang 1985: 2-3 suggested a mistake of 二 “second” for 元 “first”), whereas Faxian wrote his Foguo Ji 佛國記 in the year 416. A gap of sixteen years thus eludes Enoki’s argument that “in the old days a Yüeh-chih king, etc.” cannot be referring to an event which had taken place only a few years earlier -- sixteen years was indeed not “only a few years” so that Faxian employed “xi” 昔 to express the early date. In fact, “xi” 昔 is a temporal adverb in ancient Chinese language meaning “formerly”, and is an antonym of “jin” 今, meaning “now, today”. Besides, a typical usage of “xi” as a substantive indicates “a decade”. In the Shi Jing 詩經, the most ancient text which we can take into consideration, “xi” is found in the verse No. 167, titled Pick a Fern 小雅·采薇: 昔我往矣，楊柳依依。今我來思，雨雪霏霏。“Long ago, when we started, the willows spread their shade. Now that we turn back, the snowflakes fly.” (Waley 1960: 123) Therefore the identification of Kidāra with the Yuezhi king implies the occupation of Gandhāra in ca. 390 (see below). The time span to the year 416 exceeds 15 years. Accordingly, Enoki’s first argument now seems invalid. Enoki’s second question in his latter article attempts to exhibit the inconsistency between the political situation in Faxian’s time and the Kidarites’ unification of Northwest India. While Faxian was making his peregrination in India, the great Gupta emperor Chandragupta II Vikramāditya (r. 380-415) was on the throne, exemplifying the pax guptica which referred to Madhyadeśa,7 and the feature of unification embodied in Faxian’s witness is proposed by Enoki according to Majumdar (1954) and Kosambi (1965). However, on the same page quoted by Enoki (1998: 80 n. 92), Majumdar (1954: 22) stated: “Unfortunately, he [i.e. Faxian] has not noted anything about the political condition of India; so much so, that he does not even mention the name of the great emperor in whose wide dominions he must have lived for more than five years.” Majumdar (ibid., 346) also implied that the social observations by Faxian were in fact the comments after his itinerary in the country of Mathurā. An image of ideal society vividly presented in the pilgrim’s account was not made in contrast with the milieu in Northwest India as Enoki surmises, but in direct contrast to the poignant reminiscence of Faxian’s patria, the turbulent land of North China. Since the last years of the Western Jin dynasty (281-316), North China had been suffering torments from continuous chaos caused by internal wars to the time of Faxian. Ruthless holocausts, and overwhelming atrocities, as well as the demoralization of saṅgha (which he mourned for the insufficiency of vinayapiṭaka, and thus decided to left for the Buddha’s land), profoundly touched him to the heart. Hereby instead of lamentation he eulogized the tranquil Indian caste society, where neither slaughter nor theft took place. Now Faxian’s record in Northwest India should be reconsidered. Enoki’s argument that the unification of the north and south of the Hindūkush by the Kidarites had not yet been achieved by 4

the time Faxian visited there is not only based on the comparison with the record in Madhyadeśa as we remarked above, but also derives from the assumption that the Kidarite government was a centralized one like its ancestor, the Kushans. However, the reign of the Kidarites in Northwest India was, effectively a loosed one described by Deeg (2005: 111) as “unabhängigen kleinen Reichen unter der mehr oder weniger festen Herrschaft der Kidariten”8 Additionally, the accounts in the Wei Shu 魏書 and the Beishi 北史 discussed below indicate that the countries to the north of Gandhāra were “completely subordinated as a servants [i.e. a vassals]” 盡役屬之9, articulating their status as vassals of the Kidarite kingdom, therefore they had more or less independence than that noted by Enoki as well as by Petech and Deeg. We may observe that Gandhāra proper, with its capital Puruṣapura as a strategic point, was shielded by the prince of the Kidarites, who eventually became king there. Obviously, Faxian did not associate the aforementioned legend of Kaṇīṣka stūpa with the Yuezhi king’s event we will discuss below, and this requires satisfactory explanation. What is more, if Enoki’s identification is accepted, there would remain an inevitable riddle regarding why the Kushan king Kaṇīṣka attacked his own winter-capital Puruṣapura. It is illogical that a king would assault his own capital without a good reason, such as quelling a rebellion. No record suggests either that Kaṇīṣka did capture Puruṣapura from his enemies or that there was a rebellion during his reign. Enoki’s judgment is informed by a legend kept in two Chinese Buddhist documents: Maming Pusa Zhuan 馬鳴菩薩傳 (*Aśvaghosacarita) 10 and Fufazang Yinyuan Zhuan 付法藏因緣傳 (the Account of the Causes and Conditions for the Transmission of the Treasury of the Dharma). 11

He takes the former as a corroboration of the definition of Yuezhi country (Enoki 1998: 84).

The texts extracted from the two documents are cited below: After then, when a king of Little Yuezhi in North India invaded and lay siege to a city in Mid-India, the king of Mid-India sent a letter to him and said, ‘I am ready to pay you if you want anything. Why is it necessary to distress people by staying longer?’ The king of Yuezhi replied to the Indian king, ‘if you agree to surrender, send me three hundred million12 pieces of gold, and you will find pardon’. The Indian king said, ‘There cannot be found even one hundred pieces of gold in every corner of my country. How can I collect such a huge amount of gold?’ The Northern Indian king pointed out that the country had two great treasures: the Buddha’s bowl and a bhikṣu whose talent of preaching the law of Buddha was very skilful. He said, ‘Give me these as substitutes and they will suffice for two hundred million pieces of gold’. Thinking that he could not discard both of these, the Mid-Indian king asked the bhikṣu what to do. But the bhikṣu persuaded the king, saying that bhikṣus were destined to bring salvation to all beings and the king would get a high reputation for allowing him to leave for North India. The king, respecting his words as usual, gave both of the treasures to the Yuezhi king. […] After arriving there the bhikṣu, Aśvaghosa became famous for his skilful propagation of the Buddha’s law. (Taisho 50: 183c-184c, translated by Kuwayama 1990: 961, slightly emended). In Pāṭaliputra there lived nine hundred million people. At that time a king of Yuezhi was very influential and was called Candana Kaṇīṣka.13 His fighting spirit was so ferocious and his bravery was so unrivalled throughout the world that, whenever he made expeditions, he 5

did not fail to subjugate the attacked. So he came to this country at the head of four kinds of forces. After fighting he made the country surrender and promptly asked for nine hundred million pieces of gold. Then the king of Pāṭaliputra offered the king Kaṇīṣka three things, Aśvaghosa and the Buddha’s bowl as well as kind-hearted cock. Each was equivalent to three hundred million pieces of gold. He took them and presented them to the king Kaṇīṣka. Aśvaghosa was especially excellent in his wisdom, and the Buddha’s bowl possessed the merit of Tathāgata, while the cock was kind-hearted, did not drink tainted water and [all of the three] could destroy its enemies. Because of these characteristics, they were worth nine hundred million pieces of gold. The king of Yuezhi was much delighted to receive these, withdrew his troops, and returned to his homeland. (Taisho 50: 315b, translated by Kuwayama 1990: 961). Kuwayama (1990) expounded the context of the Buddha’s bowl in Gandhāra and quoted the two paragraphs above. His translation is an exact conveyance of the original Chinese texts, of which the former was said to have been translated by the great translator Kumārajīva in Chang’an in the beginning of the 5th century,14 and the latter was edited in the early 6th century but in part supplemented later.15 The legends preserved in the two biographies are definitely the same. Although there are dissimilarities between the texts such as the amount of ransom and treasures, the two narratives recount the same frame of one story, as is recognized by Kuwayama. Fufazang Yinyuan Zhuan, the latter text, also provides us with the name of the city under siege as Pāṭaliputra, together with the Yuezhi king’s name Kaṇīṣka. In fact, Kaṇīṣka and Aśvaghosa were contemporaries. Compared with the two biographies, Faxian’s description has several points of resemblance at first sight: a Yuezhi king’s invasion; the king’s faith in Buddha’s law; and the relationship with the Buddha’s bowl. However, the nuances of the narrative are strikingly distinct. Firstly, according to Faxian, the Yuezhi king’s target was Puruṣapura, yet in the two biographies it locates in Mid-India (Madhyadeśa), most likely Pāṭaliputra as the Fufazang Yinyuan Zhuan mentioned. Secondly, Faxian’s account that the Yuezhi king “wishing to carry the bowl away” when the country had been subdued, differs from the ransom demanded immediately after the city had surrendered – three or nine hundred million pieces of gold. Thirdly, none of the legends below appears in the two biographies or in any other Chinese Buddhist sources unless copied from Faxian: the miracle of the immobile Buddha’s bowl, the Yuezhi king’s regret, the construction of a stūpa and a monastery in the city, as well as guards and dedications to the bowl. Lastly, the renowned bhikṣu Aśvaghosa, along with his skilful propagation of the Buddha’s law, was highlighted in the two biographies, but never appeared in Faxian’s record. In addition, Faxian’s journey to India lasted more than a decade (399-412), during the same period when Kumārajīva was in Chang’an translating Buddhist canonical works (401-13). There is, all the same, no clue implying that Faxian had read Kumārajīva’s translation of the Maming Pusa Zhuan before the composition of his itinerary. Kumārajīva’s hero Aśvaghosa was missing in every detail from Faxian’s narrative of the Buddha’s bowl in Northwest India where in his youth Kumārajīva had grown up and studied (see below), indicating complete independence of the tales narrated by the two great figures. Seeing that Faxian related a completely different story we may logically trace an outline concerning the Buddha’s bowl: It was Kaṇīṣka who took the Buddha’s bowl from Mid-India to his 6

capital Puruṣapura, yet after a certain period of time a later Yuezhi king from elsewhere recaptured Puruṣapura, wishing to carry the bowl away, but he failed to move it. The last incident might suggest that his power was less than Kaṇīṣka’s. Consequently, the candidate for the Yuezhi king eventually fell to Kidāra, the king of the “enigmatic Kidarites”16 recounted in the Wei Shu, and the main character in the articles of Enoki: The Ta-yüeh-chih [Da Yuezhi] Country, of which the capital had been situated at Lu-chien-shih [Lujianshi] city 盧監氏城 (according to the Pei-shih 北史, lu is written shêng 賸), lies to the west of Fu-ti-sha[Fudisha] 弗敵沙 (Badakhshan), 14,500 li away from Tai 代 (Dai, the capital of the Wei). In the north it touched the Juan-juan [Ruanruan] 蠕蠕, which invaded (the Ta-yüeh-chih) so many times that the Yüeh-chih had at last to move the capital westwards as far as Po-lo [Boluo] city 薄羅城, 2,100 li away from Fu-ti-sha 弗敵沙 (Badakhshan). The King Chi-to-lo 寄多羅 (Kidāra), who was a brave warrior, at last organized troops and marched to the south to invade Northern India, crossing the Great Mountains [Hindūkush], and completely subjugated five countries to the north of Ch‘ien-t‘o-lo [Qiantuoluo] 乾陁羅 (Gandhāra). … (Enoki 1998: 60, slightly emended). The Hsiao-yüeh-chih [Xiao Yuezhi] have their capital at Fu-lou-sha 富 樓 沙 (Puruṣapura, Peshawar). The King was originally the son of Chi-to-lo 寄多羅, king of the Ta-yüeh-chih. Chi-to-lo was forced to move westwards by the attack of the Hsiung-nu [Xiongnu] 匈奴 and later made his son guard this city. For this reason, the kingdom was named the Hsiao-yüeh-chih. … (ibid., 67). The paragraphs above display a simple fact in spite of the chronically debated problems whether and how Kidāra and his tribe were driven by their foes, whose ethnonym Ruanruan or Xiongnu is still under discussion, especially in the case of Lesser Yuezhi:17 This is that the Great Yuezhi king Kidāra led his troops southward across the Hindūkush,18 subdued Gandhāra and the regions to the north of it. After the conquest of these areas, Kidāra appointed his son to assume the defense of Puruṣapura, named Lesser Yuezhi. The Wei Shu thus provides us with a reliable record of the Kidarites’ conquest of Northwest India, which brought to an end the reign of the last Kūshānshāh there (Grenet 2002: 205, cf. Kuwayama 1999: 38 n. 1). This title, in association with the political arrangement whereby the prince guarded Puruṣapura, is a legacy of the Kushano-Sasanians (Bivar 1979: 330-1). There is marked uniformity with Faxian’s itinerary in Puruṣapura: Kidāra is a Yuezhi king; he invaded and captured Gandhāra (evidently including the city of Puruṣapura); after that he made his son guard the city. It is not difficult to surmise that a Chinese annalist would have filtered the Buddhist legends, while a pilgrim monk like Faxian exaggerated the wonders he had seen in the Buddha’s land despite the political situation as he did the same in the Madhyadeśa. Even though all the texts vary, we may discover the similarity with respect to both contexts.

The Buddhist faith of Kidāra Nevertheless, another issue should be well examined. That is the devotional Buddhist faith of 7

the Yuezhi king in Faxian’s record. On this occasion it must be presented in another source, i.e. the Dunhuang document Ф. 209 in Opisanie kitajskikh rukopisej Dun’-huanskogo fonda Instituta narodov Azii, named “a segment of Itinerary to the Indian sacred land” by L. N. Menshikov. Lines 20-27 and 37-43 of this document detail the itinerary in Dachu Si 大處寺 (?= Skt. Mahāvihāra) 19

and Xiluo Si 醯羅寺 (Skt. Hiḍḍa vihāra) respectively: 20

… from the Maitreya statue go straight to the Dachu 大處

21

vihāra 寺, a tree is here, where the Buddha tamed the heavenly demon. branches and leaves of the poplar

22

do not sprout until the first day in the seventh month of the year; on the fifteenth day, all the bhikṣus assemble

23

under the tree, preach the Buddha’s law in the presence of all the kings, offered with a hundred ingredients of aliments, and

24

donated with saptaratna. Were the kings present, the branches and leaves of the tree

25

grow spontaneously, the shade covers the kings; were the kings absent, the branches and leaves do not grow. Tianzi 天 子 (devaputra) of that country

26

venerates the tree with saptaratna as well. The assembly of saṅgha was indeed

27

eighty-five thousand in number. …

…… 37 38

…go eastwards three months and some days to Xiwumandi 奚吳曼地 city, to the east of the city there is Xiluo Si [醯]羅寺,20

39

the Tathāgata’s uṣṇiṣa, cakṣu, khakkhara, trīnicīvarāni, kuṇḍikā and pātra21 were all suspended in the air

40

having never fallen. Tianzi and the king donate all in the present time, and

41

appoint fifty thousand musicians playing music, twenty-four hours round the clock with

42

such music donating, as well as command five million soldiers defend the place. The saptaratna was extremely22

43

radiant day and night, needless of lamp and candle. …

It is noted in another essay of mine (Wan, 2009) that Dachu Si is located in the country of Uḍḍyāna, i.e. the Swat region in Northwest Frontier Province, Pakistan; while Xiluo Si, where Buddha’s sacred relics were preserved in Xuanzang’s 玄 奘 record, 23 has been accurately determined by Cunningham (1871: 44-5) to have been in Haḍḍa, eastern Afghanistan, a small town five miles south of the city of Jalālābād. Uḍḍyāna belongs to the five kingdoms to the north of Gandhāra and, as a result of the invasion it formed part of the Kidarites’ dominion (Kuwayama 1989: 110), the same fate befell to Nagarāhara where Xiluo Si or Xiluo city situated, just as Petech (1950: 60) suggested in his comments on the Shuijing Zhu 水經注. As the author has argued elsewhere, the simultaneous presence of the titles Tianzi 天子 ( = Skt. devaputra) and Wang 王 (i.e. King) in Uḍḍyāna and Nagarāhara provides evidence that the itinerary (or geographical work) excerpted in document Ф. 209 was recorded when a political power, of which the lord was no doubt a patron of Buddhism, dominated both of the two places above. With detailed explication of the concept Tianzi and its Kushan origin,24 in Faxian25 and 8

other sources, document Ф. 209 has been dated between 412 and 433, a time span when the Kidarites exercised overlordship firmly in Northwest India.26 At that time the title Tianzi could not be Kidāra himself in the Wei Shu but his descendant, who (together with the ‘king’ narrated to be at Nagarāhara, presumably his son) deployed battalions of troops in Nagarāhara according to document Ф. 209, in a way similar to Kidāra’s disposal in order to prevent the threat from the neighboring Alchonid27 chiefdom, against whom Kidāra had not fought (cf. Alram 1996: 520-6). As shown above, document Ф. 209 portrays the Buddhist beliefs of the Kidarites, and accordingly, demonstrates that Faxian’s Yuezhi king was Kidāra. Faxian’s account is a reflection to Kidāra’s pious faith to Buddha’s law, as Max Deeg (2005: 240) indicates, “die späte Kushan Dynastie Kidariten wohl durchaus Interesse daran haben mußte, wenigstens im religiösen Bereich ihre Legitimation durch Legenden um den berühmtesten ihrer Ahnen, und seiner Verbindung mit dem saṅgha zu festigen, zumal ihnen im Südosten eine starke Gupta Dynastie benachbart war.”28 It should be observed that document Ф. 209 has also proved the assumptions held by Tucci (1958: 288-9) and Zeimal (1996:132-3) about the political situation in Swat and the Kidarite kingdom, respectively.

The Buddha’s bowl and the dating of the Kidarites The tale of the Buddha’s bowl in Faxian’s record was hitherto unfinished. After arriving in Ceylon, Faxian heard another legend related to the bowl: In this country Faxian heard an Indian monk 天竺道人, who was reciting a sūtra from the pulpit, say: Buddha’s alms bowl was at first in Vaiśālī, and now it is in Gandhāvatī. After so many hundred years - he gave, when Faxian heard him, the actual number of years, but he has forgotten it - it will go westward to the state of Yuezhi; after so many hundred years, to Khotan; after so many hundred years, to Kucha; after so many hundred years, to the land of Han; after so many hundred years, to Siṁhala; after so many hundred years, it will return to Mid-India. After that, it will ascend to the Tuṣita heaven. … (Legge 1886: 109-10, slightly modified by Kuwayama 1990: 960. Also cf. Enoki 1998: 85-6, Kuwayama 1989: 107). Kuwayama (1990: 960) scrutinized the details of the Chinese text, asserted that the term “laidao” 來到, which means “come and arrive”, suggests that a Chinese fabrication of the passage. He further deduced that the local place names in the legend “cannot be accepted as face value, since a monk in Sri Lanka would not have known them.” His conclusion, though meticulously examined, skips the fact that Faxian depicted the monk as an Indian, i.e. Tianzhu daoren 天竺道人. Even if Li Daoyuan 酈道元, the annotator of the Shuijing Zhu, could not tell Sri Lanka from India,29 the pilgrim Faxian who traversed both in a ten-year period, expressly pointed out that the sūtra was heard from an Indian monk, not an indigenous Ceylonese. The phrase, “come [and arrive at] the land of Han” 來到漢地, unsurprisingly conveys the fact that Faxian had by then returned to his homeland. Consequently, the entire legend in Faxian’s record should not be interpreted as “spurious”. In conclusion, as narrated by Faxian, the Buddha’s bowl was at that time (410-1) in the region of Gandhāvatī which, albeit different from the prior itinerary, undoubtedly refers to Gandhāra. 9

More to the point, Kuwayama (ibid., 968; 1989: 97-102) also painstakingly analyzed the biography of the monk Narendrayaśas (490-589) from Uḍḍyāna in Northwest India, and concluded that it was in 514 AD when Gandhāra was ruled by the Hephthalite that he visited the Buddhist sites in Northwest India. In his biography the Buddha’s bowl was kept apart from other relics, which is deemed to be true since Narendrayaśas was a native in Northwest India.30 The question is, however, was the Buddha’s bowl once kept in the city of Xiluo, or precisely, together with the uṣṇiṣa, etc. during the time mentioned by Faxian and the visit of Narendrayaśas? Only one source dated to the time between the two monks might provide a positive clue, which is in the Foguo Ji (A Record of the Buddhistic Kingdoms, title identical to Faxian’s work) by Zhu Fawei 竺法維 (probably of Indian origin because of his surname Zhu), quoted and hence merely preserved in the Shuijing Zhu: The Buddha’s bowl is in the country of Great Yuezhi. A Buddhastūpa is standing there. The height is thirty Zhang and it has seven storeys. The bowl is on the first floor. The bowl which is entwined with golden threads is suspended by a chain [The other interpretation is possible that the bowl is suspended by a chain which is entwined with golden threads]. The bowl is made of blue stone. It is said that the bowl had been suspended in the air, but the Subhuti put it on the golden table. The Buddha’s Footprint is at the same place together with the Buddha’s bowl. The kings and the subjects are all holding Brahmapiṇḍāta, saptaratna and fine jade, donated the stūpa and the Footprint. The site of the pagoda, Buddha’s teeth, kāṣāya, and uṣṇiṣa śarīra, are all kept in the country of Puruṣapura. (quoted from the Shuijing Zhu, Bk. 2, translated by Kuwayama 1990: 950, while the last sentence is an original translation. Cf. also Petech 1950: 60). The country Greater Yuezhi can well be identified with the country of Puruṣapura mentioned thereafter according to Petech (1988b: 314), as he takes the Waiguoshi 外 國 事 (Matters concerning Foreign Countries) by another monk, Zhi Sengzai 支僧載 (probably of Yuezhi origin for his surname Zhi), into account. In the opinion of Petech, Zhu Fawei implied that Nagarāhara was part of the Yuezhi country, i.e. the Kidarite dynasty, whose capital must be Puruṣapura.31 It is quite curious that Kuwayama also cites Zhi Sengzai yet without comparing the two ethnonyms in both records, either deliberately or accidentally: The Buddha’s bowl is in the country of Great Yuezhi, otherwise called Fulüsuoyue [*Puruṣavatī] that means the royal city of devaputra. A Buddhastūpa is four Zhang high and has seven storeys. On the wall of all four sides are enshrined golden and silver images of Buddha, the size of which is as large as a man. The bowl is in the center of the second floor. The bowl is entwined with golden threads, being suspended by a chain. The bowl is made of stone and the color of it is blue. (quoted from the Yiwen Leiju 藝文類聚, Bk. 73, translated by Kuwayama 1990: 950-1, slightly emended). The dates of Zhu Fawei and Zhi Sengzai have hitherto remained obscure.32 Fortunately, another Buddhist source traces the history of their time. Edited in the Shinzan Zokuzokyo, the Mingseng Zhuan Chao 名僧傳抄 preserved biographies of eminent monks,33 among which the biography of Sengbiao 僧表, contemporary with Zhu Fawei according to the Gaoseng Zhuan 高僧傳,34 10

provides us with vital information: Sengbiao had as his original surname Gao 高. He was born in Liangzhou 涼州, devoted and courageous. He heard that the Buddha’s bowl was in the country of Fulousha 弗 樓沙 (Puruṣapura), but is now in the Tai Si 臺寺 of Jibin 罽賓. There are constantly five hundred arhats making offerings to the bowl. The bowl had sped across the sky, arrived at Liangzhou 涼洲,35 stayed there along with twelve arhats; after six years they returned to Jibin. Sengbiao regretted not viewing the bowl. Therefore he trekked over the Congling 蔥嶺 (Pamirs), wishing to show his piety to the bowl, and arrived at the country of Jibin. In the meantime, the road to Jibin was blocked. Holding Sengbiao for his full piety, the king of Jibin forged a replica of the bowl for him, and asked him whether he had more wishes. He responded, “It is eulogized that there is the statue of Baosheng 寶勝 (i.e. Tathāgata) in Mojialuo 摩伽羅. Among the foreigners it is said to be closest to the true looks. We hope to keep it for offerings.” Then the king ordained skilful craftsmen to build a statue wrapped with gold foil. The splendid gold statue was one zhang in height, with a genuine śarīra on top of it. On the road back after receiving the donations, Sengbiao sought to turn to Huaihai 淮海, being aware that the land of Liangzhou was to be conquered. He passed by Xinping 欣平 County in Shu 蜀 region, where the śramaṇa Daowang 道汪 sought to keep the bowl and the statue for offerings. They are in Longhua Si 龍華寺 there at present. Sengbiao has been there, worshipping the stone statue. His sojourn there lasted two years and he died in the monastery…. (Shinzan Zokuzokyo 77: 358b). Two remarkable events were mentioned in Sengbiao’s biography: one being that the road of Jibin was obstructed; and the other being that the land of Liangzhou was to be conquered. These two toponyms need to be examined. According to Kuwayama (1987: 708-12), Jibin in the Gaoseng Zhuan is not its Sanskrit equivalent Kaśmīra but Gandhāra. Mingseng Zhuan Chao, which was written before the Gaoseng Zhuan, mentioned Jibin in the biographies of Buddhabhadra 佛駄跋陀 (Shinzan Zokuzokyo 77: 355a) and Zhiyan 智嚴 (Shinzan Zokuzokyo 77: 358b-c), whose stories were edited in the Gaoseng Zhuan as well (in Taisho 50: 334b-335c; 339a-c respectively, where Buddhabhadra was written 佛馱跋陀羅). The accounts in the two biographies here tally with those in the Gaoseng Zhuan. Therefore Jibin in Baochang’s biography of Sengbiao should also stand for Gandhāra (perhaps together with the Swat region) where the Buddha’s bowl was celebrated. However, the road of Jibin was then blocked so that Sengbiao, after descending from the Pamirs, had to rest somewhere in Jibin but certainly not Puruṣapura. Probably he was delayed on the other side of Kabul River, i.e. in Puṣkalāvatī,36 or in the Swat region where a residual Kidarite kingdom is thought to have been located.37 The land of Liangzhou here is the dynasty of Bei Liang 北 涼 (i.e. Northern Liang, 395-439)38. There were three monarchs in Northern Liang, the second of whom, Juqu Mengxun 沮渠蒙遜 (d. 432), was the mightiest. Yet he had no capable heir. In consequence, his death implied the collapse of the dynasty, which took place in 439 when the last Northern Liang ruler, Mengxun’s son Juqu Mujian 沮渠牧犍 held sway over Liangzhou. In view of the dating above, it may have been slightly before 439 that Sengbiao recognized the upcoming subjugation of Northern Liang. Now that the dating of the conquest of Liang is fixed, it 11

can be inferred that the other event, the blocking of the road to Jibin happened shortly before 439. The reason why the road to Puruṣapura was blocked can be assumed to be war, in which Puruṣapura became a frontier city confronting the risk of warfare,39 which took the form of occupation by the Alchonids led by Khingila. With the absolute dating of Sengbiao, the date of Zhu Fawei can be established. According to the sequences presented by the Gaoseng Zhuan, the Shijia Fangzhi 釋迦方志 and the Mingseng Zhuan Mulu 名僧傳目錄, Zhu Fawei came directly after Faxian and before Sengbiao, that is, circa 415-439. This date tallies with the hypothesis of Petech (1950: 60), because it is when the Kidarites controlled Puruṣapura. On the other issue, Zhi Sengzai’s dating can also be fixed to the first half of the fifth century, as suggested by Petech (1988b: 313).40 Both monks recorded the country of Great Yuezhi, about which Kuwayama does not give a clue in his publications, unlike the toponym Jibin related in the biography of Sengbiao. Nevertheless, the name Yuezhi is not only a synonym of the Kidarites in the Wei Shu but also a name closely related to the very existence of the Kidarites in Northwest India as indicated in the following section.

The Kidarites and the country of Yuezhi in the 4th and 5th centuries in Chinese Buddhist sources As stated above, Zhu Fawei and Zhi Sengzai recounted the country of Great Yuezhi, where the Buddha’s bowl was in custody. Dr. Kuwayama contributed three insightful articles (Kuwayama 1987, 1989 and 1990) on the Buddhist relics in Northwest India. Besides the bowl, other relics of the Buddha, e.g. the bottle, the shadow, the cranium, the staff, the rope, and the tooth, were said to be preserved in Yuezhi country in the Buddhist literature. Actually the main site other than Puruṣapura was Nagarāhara as mentioned above, while the bottle as well as the broom of the Buddha was situated in Balkh. The name of Yuezhi often omitted, at times added, a prefix, either Greater (Da 大) or Lesser (Xiao 小). To distinguish the three meanings of Great Yuezhi, Dr. Enoki provided an effective discussion in his inspiring treatise on the provenance of the Hephthalites (Enoki 1959: 11-12). In contrast their counterpart, the Lesser Yuezhi 小月氏, may also be categorized into three tribes chronologically: The first tribe, documented by the Shi Ji (fasc. 123) 史記·大宛列傳 and the Han Shu (Chap. 96) 漢書·西域傳,41 were the remnants after the removal of Greater Yuezhi, who lived in Gansu in the second half of the 2nd century BCE. The third meaning, recorded in the Wei Shu, has already been mentioned in this article, namely the tribe under Kidāra’s son in the region around Puruṣapura. Whereas the second meaning of Lesser Yuezhi, remarked in the Aśvaghosacarita discussed above, was the unique account linking Lesser Yuezhi with Kaṇīṣka though innominately. We should in this respect be well aware that Lesser Yuezhi, as well as Greater Yuezhi mentioned in Kumarajīva’s translation of the Dazhidu Lun 大 智 度 論 (Mahāprajñāpāramitopadeśa, Taisho 25: 243a and 141c, 126b-c), has been meticulously examined by Enoki (1970 = 1998: 81-85). After a conclusive retrospect on previous scholars, Enoki finally assigned Lesser Yuezhi to “the Tokhārestān region, north of the Hindūkush”, while its counterpart Greater Yuezhi, or simply Yuezhi, to “the Gandhāra and Swāt regions” because these areas “had been the most prosperous since the Kushan Dynasty”. Nevertheless, his conclusion that “Kumārajīva distinguishes Gandhāra from Tokhārestān by mentioning the former 12

as the Ta-yüeh-chih and the latter as the Hsiao-yüeh-chih is to show that the north and south of the Hindūkush had not yet been united under a single political power during the period of 402-406” is no more than a matter of conjecture. That the two regions were named Greater vs. Lesser Yuezhi does not necessarily mean that they were under different lordships. Taking the statement of the Wei Shu into consideration, an adverse juxtaposition of the Greater vs. Lesser Yuezhi was attested in Bactria and Gandhāra whereas no scholar yet considers them to be other than “a single political power”.

42

In addition, although Kumārajīva mentioned Yuezhi elsewhere in the Dazhidu Lun as indicating Gandhāra proper as will be discussed below, it seems that his biography in the Chu Sanzang Ji Ji 出三藏記集 (Taisho 55: 100b; cf. Taisho 50: 330b) did not result in another conclusion of Enoki (1998: 85) that Gandhāra was known as Yuezhi to the Chinese in 365 when Kumārajīva’s mother brought the twelve year old child back to Kuči via the mountains to the north of Yuezhi. On the contrary, this fact only proves that Kumārajīva employed Yuezhi for Gandhāra at his time.43 Having adopted the placement of Greater Yuezhi or Yuezhi by Enoki (though differing in reason and dating), we could list the works of Buddhist authors in the fourth and fifth centuries in which the proper name Yuezhi has a particular relation with the well-known Buddhist relic sites (shown in Appendix I). Before 385 when Dao’an 道安 died, the seat of the Buddha’s bowl only appears in the name of Gantuoyue/Jiantuoyue 乾陀越/揵陀越 (*Gandhāvatī), and that of the Buddha’s cranium (uṣṇiṣa) and shadow was [Xi]tuluoyue [醯]吐羅越 (*Hiḍḍavatī)44 in the accounts of Dao’an and Buddhadeva. Moreover, according to Buddhadeva (mid-4th century),45 the Buddha’s bottle was then in the land of Yuezhi. As mentioned above, the Buddha’s bottle was situated in Balkh until the time of Xuanzang. Besides, the Buddha’s broom was also in Balkh at that time according to a passage in Dao’an’s Shishi Xiyu Ji 釋氏西域記.46 Therefore, the country of Yuezhi (neither Greater nor Lesser, implying integrity) was then located in Bactria/Tukharestan, excluding both Northwest India and Kapiśa-Kabul to the south of Hindūkush.47 At the time when Faxian traveled to Ceylon, Huiyuan 慧遠 heard about the Buddha’s shadow in the Yuezhi country from the Indian monk Buddhabhadra (Zürcher 1952: 224). The Kidarites had already occupied Gandhāra, while a decade earlier the site of the Buddha’s cranium Nagarāhara was attested to be West of Greater Yuezhi 大月氏西 in the account of Kumārajīva. At almost the same time, in the first decades of 5th century, Zhi Sengzai and Zhu Fawei recounted that Great Yuezhi was the country where the Buddha’s bowl was preserved, i.e. Puruṣapura. The last record of Yuezhi in Buddhist literature is in the biography of Fayong 法勇 (Dharmavikrama), who departed for India in 420 and returned to southern China in 453 with an itinerary of five volumes.48 One can infer from the above that between the 380s and 430s, when Northwest India was named the Yuezhi country, the Kidarites exercised their administration in and around Puruṣapura. After that, the country of Yuezhi was not seen in the records of Chinese Buddhist monks. Preserved in the Mingseng Zhuan Chao, the biographies of Sengbiao and Fasheng 法盛 did not document the visit of relics in Puruṣapura or Nagarāhara. Slightly before 439 when they arrived at Northwest India, the sacred land was suffering from the flames of combat. Consequently, Sengbiao was hindered in Jibin, while Fasheng’s biography barely stressed his journey in Uḍḍyāna, where he visited the colossal statue of Maitreya.49 13

In conclusion of this section, we can make the following observations about the Yuezhi country from Chinese Buddhist sources of between the 4th and 5th centuries: 1. The name Yuezhi was at first applied to Bactria/Tukharestan in middle 4th century by Buddhadeva and Dao’an. 2. The name Yuezhi applied to Northwest India was not employed in Chinese sources until Kumārajīva adopted it as the name for the country of Nagarāhara in his translation of the Dazhidu Lun at the beginning of the 5th century, while Faxian and other monks employed it for regions about Puruṣapura. 3. It is the war before 439 that blocked Puruṣapura and Nagarāhara from Chinese pilgrim monks. At the same time, the name Yuezhi fell out of use in Chinese Buddhist literature.

The Kidarites in Northwest India and the Problem of Jibin Jibin: A shift from Gandhāra to Kaśmira Concerning the invaders to Northwest India from the north, it is vital to regard Chinese documents as the most reliable sources, especially after Sasanian Iran blocked direct contact of the Roman-Byzantine Orient and the subcontinent. As for the Chinese sources on Northwest India after the Han Dynasty, there are two parallel traditions. The one is the dynastic histories composed by court officials; the other is Buddhist literature written by translators or pilgrim monks. It is important to know that the nomenclatures of both traditions intertwined at times, while a striking difference is suggested by many scholars on one of the pivotal toponyms: Jibin 罽賓, the one that perplexed generations of eminent researchers.50 Twenty years ago Kuwayama (1987) convincingly identified Jibin with Gandhāra in 4-5 century Chinese Buddhist literature. But in a later treatise on the Buddha’s bowl, Kuwayama (1990: 948 n. 4) returns to the conventional view proposed by Shiratori (1917) and Petech (1950: 71ff.) that Jibin indicates Kaśmīra in the narratives of pilgrimage compiled either by the pilgrims themselves or by authors who could use the original narratives, whereas in the Gaoseng Zhuan it is an exception to identify Gandhāra as Jibin on account of the ignorance of the compiler Huijiao 慧皎 regarding the location of the Buddha’s relics, as is revealed in a concluding remark at the end of Zhimeng’s 智猛 biography (Taisho 50: 343c). Several examples are enumerated to lend support for this view. On the one hand, as examples of translations in which Jibin can be identified as Kaśmīra he cites An Faqin’s Aśōkarājāvadāna (trans. ca. 285-306),51 Paramārtha’s Commentary of Abhidharmakoṣaśāstra (trans. 564-567),52 and Xuanzang’s translation of the Abhidharmakoṣaśāstra (trans. 656-660).53 On the other hand, as for the “narrative of pilgrimage” he mentions the Luoyang Qielan Ji 洛陽伽藍記 and Faxian’s Foguo Ji; the latter work states that “Sengshao 僧韶 alone went to Jibin following a barbarian monk”, with no more reference,54 while the former described the pugnacity of the Hephthalite tegin in Gandhāra (supposed to be Mihirakula) when he was competing with Jibin for territory in the year 520. One more quotation from the Gaoseng Zhuan in Kuwayama (1990: 953) notes that Fotucheng 佛圖澄 (Buddhadāna) had been twice in Jibin before 310 (in fact he went to Jibin only once, 14

another mention of Jibin occurred when he was in China). Unfortunately, this piece of information is too brief to reach any conclusion on the identity of Jibin. The documents cited by Dr. Kuwayama illustrate, along with other Chinese Buddhist sources in 4th-5th century, an explicit shift of the toponym Jibin from Gandhāra in early sources, to Kaśmira as attested the translation of Paramārtha and Xuanzang’s translations. It is noteworthy that the Kidarites were coeval with this critical shift, and serve as a momentous factor of this shift. The earliest mention of Jibin, in An Faqin’s translation of the Aśōkarājāvadāna, appears as follows: Formerly the Buddha subdued Naga Apalāla in the country of Wuchang 烏 長 (Uḍḍyāna), converted the Fanzhishi 梵志師 in the country of Jibin, overcame Zhentuoluo 真陀羅 (Caṇḍāla) in the country of Qiantuowei 乾陀衛 (*Gandhāvatī), and vanquished the Naga Gopāla in the country of Qiantuoluo 乾陀羅 (Gandhāra). (Taisho 50: 102b). Fanzhishi 梵 志 師 appears in the different translations of the Aśōkarājāvadāna in the Samyuktagamasūtra 雜阿含經 Vol. 23 (Taisho 2: 165b, translated by Guṇabhadra 求那跋陀羅 in 435-453) and the Aśokasūtra 阿育王經 Vol. 2 (Taisho 50: 135b, translated by Saṅghavarman 僧伽婆羅 in 506-518) as Taoshi 陶師 (kumbhakāla). The two later translations do not mention Jibin. The conversion of Taoshi is recorded in the Mūlasarvāstivādin Bhaiṣajyavastu 根本說一切 有部毘奈耶藥事 as having taken place in Jiliyiduo 及理逸多, which is identical with Libatuo 隸 跋陀 in the Mahāprajñāpāramitopadeśa Vol. 9: Once the Buddha temporarily flew to the immortal’s mount Libatuo 隸跋陀仙人山 in Jibin; Halting in the air he converted the immortal. The immortal said ‘I am glad to dwell here, wishing the Buddha grant me His hair and finger nails. (I shall) build a stūpa and observe them.’ The stūpa is extant nowadays. (commentary by Kumārajīva: descending the mount there is Liyue 離越 vihāra, Liyue should be named Libatuo). (Taisho 25: 126c). The late Buddhist scholar Shi Yinshun55 (2000: p120-1) persuasively concluded that Jiliyiduo 及 理逸多, Libatuo 隸跋陀 and Liyue 離越 are the Chinese renditions of the Sanskrit Revata, and that the legend about the conversion of Fanzhishi 梵志師 is identical with the legend of the immortal Libatuo 隸跋陀仙人. As for the Buddha’s hair and finger nails, it can be found that there were three stūpas in Bactria and Northwest India with the Buddha’s hair and finger nails in them: two of them were situated in the cities northwest and north of Balkh recorded by Xuanzang, Bk. 1 (Ji 1985: 123 n. 1, 2), the other one was established in the mountainous area southwest of Nagarāhara visited by both Faxian and Xuanzang, which the author thinks to be the same stūpa mentioned by Kumārajīva. Shi Yinshun (2000: 124) further indicated that in the city of Jiliyiduo 及理逸多 i.e. Revata was stated closely between the city of Koṣṭhā 稻穀樓閣城 i.e. Maṅgali 瞢 揭厘 in Xuanzang and the city of Nagarāhara. He also presented two piece of evidence from the Ceylon canon: one is that the famous legend of Madhyāntika 末闡提 subduing the Naga Aravāla in the country of Jibin or Jibin-Jiantuo 罽賓揵陀 seen in the Samantapāsādikā 善見律毘婆沙 (Taisho 24: 685a-b, translated by Saṅghabhadra 僧伽跋陀羅 in 488-489),56 is of different origin from the tale of the subjugation of Naga Hulucha 虎嚕茶 or Hunong 忽弄 in the Bhaiṣajyavastu (Taisho 24: 40c) and the Mūlasarvāstivādin Kṣudrakavastu 根 本 說 一 切 有 部 毘 奈 耶 雜 事 15

(Taisho 24: 411a-b); the other is that the yakṣas Pañcika and Harītī in Jibin who converted to Buddhism were in the country of Gandhāra according to the Kṣudrakavastu (Taisho 24: 361a-c). After Shi Yinshun, the conventional view that Jibin in Chinese Buddhist literature was identical with Kaśmira is no longer defensible. As shown above, Jibin, as attested in the translation of An Faqin, Kumārajīva and Saṅghabhadra, is identical to the broad conception of Gandhāra (or Gandhāra-Uḍḍyāna, as proposed by Shi Yinshun). This gives rise to an interesting question: when and why the exact meaning of Jibin transferred to Kaśmira, as is seen in Paramārtha and Xuanzang? We shall leave these questions to later. The Kidarites and problems relating to Jibin Adopting the same method as Kuwayama (1990) in situating the Buddha’s bowl, we have confirmed the emergence of the Kidarites in Gandhāra. One may surmise from this that Jibin was once identical with the Kidarite kingdom in Gandhāra, taking into account the numismatic sequence of Göbl (1967) for the Kidarite coins excavated from this region. Nonetheless, it requires further scrutiny to prove whether this surmise is tenable. Although Kumārajīva and Faxian in 402 (as well as Buddhabhadra in the same period) demonstrated the Kidarites’ (i.e. Yuezhi) presence in Gandhāra, their advent must be attested directly. Scholars have expendeds great efforts on the etymology and geographic identification of Jibin. However, except for the unanimous idea that Jibin played a crucial role in the Buddhist interchanges between China and India in the 4th-5th century, the political-religious mechanism of Jibin has yet been sketched. After Buddhadāna, Kumārajīva was the earliest monk from beyond the Pamirs to have studied in Jibin. At the age of nine (ca. 353) Kumārajīva began to study the Dharma under Bandhudatta, a younger cousin of the king of Jibin (cf. Kuwayama 1987: 706). The biographies of Kumārajīva in the Gaoseng Zhuan (GZ) and the Chu Sanzang Ji Ji (CSJJ) state: Bandhudatta often engaged in discussion with Kumārajīva, and esteemed him. (CSJJ) / Bandhudatta often praised Kumārajīva [for his] smart and outstanding [qualities]. (GZ) So the fame of Kumārajīva was known to the king [of Jibin]. The king invited him to the court and summoned the non-Buddhist gurus to engage him in argument. The gurus disparaged the child and spoke rudely to him. Taking advantage of their weak points, Kumārajīva frustrated them. They turned convinced, humiliated and speechless. The king then admired him even more…. (Taisho 50: 330b; Taisho 55: 100b). We may infer from this paragraph that the king of Jibin was tolerant, and not only an adherent of Buddhism, whereas Bandhudatta might have played an active role in the court of the Jibin king, indicating that t Buddhism was favored at the court. The association of Buddhism and the Jibin court did not fade away but rather strengthened in the late 4th century as attested by the biographies of Guṇavarman 求那跋摩, a remarkable character who died in the 8th year of Yuanjia 元嘉 era (431) at the age of sixty five. He was of Kṣatriya origin from the country of Jibin, and his biographies in both Gaoseng Zhuan and Chu Sanzang Ji Ji record that he was the descendant of the Jibin royal family.57 According to the Gaoseng Zhuan he became a Bhikṣu when he was twenty (386), and the Chu Sanzang Ji Ji tells us that he left home to be a śramana when he was fifteen (381). An attractive description is provided 16

in the two biographies below: When Guṇavarman was aged thirty the king of Jibin died heirless. The council conferred and considered Guṇavarman, intelligent and lofty, a descendant of the royal court, and thus invited him to accede to the throne. Hundreds of officials requested several times but he did not accept. He took leave of his teacher, shunned the mass, and sheltered in a hermitage. (GZ)/ The ministers conferred and decided to elect him to the throne for his royal ancestry,58 but he was concerned that he was being compelled. (CSJJ) Guṇavarman then went far to Ceylon, viewed the custom there and carried forward the Dharma. … Then he went to Java … (Taisho 50: 340a-b; Taisho 55: 104b). After his journey to Java Guṇavarman was invited to South China and was well respected by the Emperor Wendi of the Song Dynasty. It was in the year 396 that he refused to inherit the throne and travelled to Ceylon. Among the verses followed his biography in the Gaoseng Zhuan and said to be his last words in China he stated “Shunning the chaos I floated on the sea, whereupon I passed through Java and Linyi 林邑 (South Vietnam).”(Taisho 50: 342b). It appears that the ‘chaos’ refers to some turbulence in Ceylon, but actually both Ceylon and Gupta India were then undoubtly tranquil. It might be inferred that his allusion to chaos was in fact a reference to the political disturbance caused by the advent of Kidāra in Gandhāra, his very homeland. Moreover, it was around the year 396 when the former king of Jibin lost his life, throne and lands (perhaps in battle?). As a result of the unexpected death, the court hastily proclaimed Guṇavarman heir to the king. His refusal of the throne was probably a shrewd decision, since almost the whole Northwest India was then annexed by Kidāra. The problem continues when Kidāra occupied most of Jibin’s lands. Was there a residual government of the former Jibin court and where was it situated? Due to its inaccessible location and special geographical features, Kaśmira was plausibly the last refuge of the former nobles. Taking Faxian’s record of Jibin as an example we may surmise that Sengshao’s route was directly towards Kaśmira, whereas Faxian’s itinerary stretches, from Upper Swat to Puruṣapura, then extends to Nagarāhara – it is doubtful whether Faxian visited the city of Taxila, where the Jātakas of head-alms and tiger-alms were said to have occurred.59 If we recall the anachronistic but scientific report of Giri fort by Sir John Marshall (Marshall 1951: 342-3, cf. Dani 1986: 77-8), the last Sasanian coins dated from Shapur III to Wahram IV (r. 388-99) indicate the unstable situation in the last years of the fourth century – caused by the Kidarites, not the Hephthalite invaders. Therefore the name Jibin shifted, sometimes after the Kidarite invasion, to Kaśmira, where a residual court of former Jibin country might have existed for a short period. At the same time, the original land of Jibin, the fertile alluvial valleys of Kabul River and Swat River around the mighty metropolis of Puruṣapura, was occupied by the Kidarites. Additionally, Xuanzang’s record in Kaśmira documents another legend, which the author assumes was probably an allusion to Kidāra, that of king Ximodaluo 呬摩呾羅 of the country Duhuoluo 覩貨邏 (Bk. 12), the latter is evidently the equivalent of the Sanskrit Tukhāra, whereas the king’s appellation, as well as the country of Ximodaluo in Xuanzang’s Bk. 3 (cf. also Taisho 50: 248), is proposed by Enoki (1959: 34-5) as a deformation of the Sanskrit word Hematāla. The country Hematāla is suggested the last capital of the Hephthalites, but the king has not been identified with a historical monarch. It should be noted that in the record of Kaśmira 17

Xuanzang narrated a series of three monarchs devoted to Buddhism, i.e. Aśoka, Kaṇiṣka and king Ximodaluo. He used in each passage of the three kings parallel eulogies to demonstrate their exploits. This fervent panegyric alludes to the respect for Buddhism and the piety demonstrated by all three kings. The devotion of Aśoka and Kaṇiṣka to Buddhism has been documented in many sources, while the devotion of a later ruler – “six hundred years after the parinirvāṇa of Tathāgata”60 – as shown at the beginning of the present article, can be identified with Kidāra, the only king devoted to Buddhism mentioned in Chinese Buddhist literature. Regardless of the dating given in the Appendix IV, Xuanzang related that his base was Tukharestan, and the style of his exploits against the former king was quite different from the Alchonid devastation, like Toramaṇa and Mihirakula as attested in Indian inscriptions. We may recall the association between Kidāra and the King Nara or Kinnara in the Rājataraṅgiṇī as proposed by Harmatta (1980: 185-9). Although his argument demonstrates his erudition especially in linguistics, it has been almost consigned to oblivion given his lack of further evidence. This article does not intend to accommodate this hyperthetical chronology, but Harmatta reveals the value of the Rājataraṅgiṇī for the chronology before the year 813,61 even if he tentatively adopted two questionable datings: that of the year 134 for the Kaṇiṣka Era and of the Kidarite invasion by Dr. Enoki.62 The accentuated description of Kinnara in the Rājataraṅgiṇī – lengthier than that of Mihirakula, indicates that Kinnara was a powerful figure in the history narrated by Kalhana. If we compare the historical evidence fixed by contemporary scholars, the only figure – besides Kidāra – who significantly influenced early medieval Northwest India was Khiṅgila, the first invader from the upper Kabul valley, whose prestige was probably reinforced by one of his powerful descendants, Narendrāditya Khiṅgila (cf. Grenet 2002: 217; alias in numismatics Naraṇa cf. Alram 1996: 528, 531). It cannot be excluded that King Nara or Kinnara was Khiṅgila for he was pious to Saivaism, but the association to Kidāra is still favored by the author, because in the Rājataraṅgiṇī the usage of Kinnara – ‘part human’ as indicated by Ranjit Sitaram Pandit (1968: 38 n. 274), describing a race which ‘had a human body with the face of the horse’, tallies well with the record in the Tang Gaoseng Zhuan 唐高僧傳 about Xuanzang’s journey in the city of Hiḍḍa, in which the compiler narrated that ‘recently a northern barbarian king of Greater Yuezhi 近有北狄大月支王, wondering about the karma (from the Buddha’s cranium), sought for the appearance by a gandha, and it showed up an image of a horse.’ (Taisho 50: 448b) This Yuezhi king, known by the compiler Daoxuan in the biography of Xuanzang, both living in the 8th century, refers quite probably to the last renowned King Kidāra. Now it is certain that the Kidarite invasion into Northwest India occurs from around the year 390,63 we turn to discuss subsequent events. Since Faxian’s account of Jibin is ambiguous, a more reliable source is Zhimeng who departed for India in 404. He went to Jibin where he saw the Buddha’s bowl shortly after the start of his journey (Taisho 55: 113b). This specific document has convinced most scholars that Jibin was then Gandhāra with Puruṣapura included (e.g. Kuwayama 1990: 948-9; Deeg 2005: 103). This could also explain that the Kidarites had assumed the great name Jibin, a valiant nation which slew envoys from the mighty Han Empire on several occasions; most of the Buddhist Nikayas – especially the Sarvāstivātin School who greatly influenced Chinese Buddhism, originated and spread in the country of Jibin. Here we encounter a toponym that has both geographic and ethno-politic significance. The Kidarites assumed this name to connect their lineage with the gradually faded glory of the old tradition, as well as relating to its geographic features. 18

On the track of Faxian and Zhimeng, the pilgrim monks Huilan 慧覽 and Dharmavikrama went to Northwest India and they arrived in Jibin worshipping the Buddha’s bowl in Puruṣapura, whereas the Indian monks who visited China, such as Kumārajīva, Zhi Sengzai and Zhu Fawei, mentioned the same city as being in the country of Greater Yuezhi. It can be surmised from this phenomenon that Jibin was solely a Chinese equivalent, which was not adopted by the contemporary Indian monks, while they emphasized the Yuezhi were ruling the country. Around the year 439 when Sengbiao visited Jibin the situation had changed dramatically, reflecting the entire political situation at that time. The author will discuss this in the next section. As a result of the invasion from the west, Puruṣapura was inaccessible from the route via Jibin. It was later that Fasheng made his peregrination to Northwest India, but unfortunately his record preserved in the Mingseng Zhuan Chao (Shinzan Zokuzokyo 77: 358c) does not provide any clues concerning Jibin or the Buddha’s bowl. Jibin was henceforth absent from the biographies of Chinese pilgrim monks, until Song Yun’s 宋雲 account. The name Jibin thus gradually passed into oblivion in Chinese Buddhist sources. Just like Faxian, the great travellers Song Yun and Xuanzang mentioned Jibin only once each. Since most Indian monks avoided using Jibin for a toponym in India,64 this name rarely appeared later. However, in the dynastic histories Jibin never lost its prestige. The first dynastic history after the Han Dynasty, the Weilue 魏略 adopted the same location of Jibin as that of Han dynasty, i.e. Gandhāra. In the Wei Shu, however, Jibin’s depicted geographical features are identical to those of Kaśmira. This reveals a change in usage. Since Jibin is recorded to have sent envoys to the Wei court from 451 to 517, dating them to within the range of the Alchonid reign as suggested by Alram (1996) and Grenet (2002), we may infer that the Alchonids assumed the name Jibin, or there was a contemporary of the Alchonids in Northwest India who adopted Jibin. Song Yun’s account may provide corroboration for the latter view. Now since it is widely accepted that Juchang 居常 and Juduoluo 車多羅 are aliases of the Kidarites who sent envoys to the Wei court in 459, 460 and 477, the identification of Jibin as the Kidarites is not feasible. It can be envisaged that there was a court ruling Kaśmira in the latter half of the 5th century, and this court clashed with Mihirakula as witnessed by Song Yun. Since the Alchonid invasion broke the link between the Kidarites in north Gandhāra and Uḍḍyāna with their principalities in other regions, those in Kaśmira might have been independent after Khiṅgila’s assault. Having defeated the residual Kidarites north of the Indus River sometime after 477, the Alchonids occupied the whole of Northwest India except for Kaśmira. Almost half a century later Mihirakula fled to Kaśmira and died there. After Mihirakula, King Narendrāditya Khiṅgila who unified Gandhāra and Kaśmira ruled Jibin until 580 (according to Alram 1996, or 570 according to Harmatta 1969: 403). It was during his reign that the Alchon and Nēzak clans clashed in Kapiśa-Kabul, as shown by Alram (1996: 530-2). Surveying the clash described by Kuwayama (1999: 25-36) in which the heretics worshipping the deity Sūrya expelled those worshipping Zhuna before the arrival of Xuanzang, the king of the latter was denoted by his bull-headed crown, it is not unaccepted to define the Zhuna group as Nēzak and the Sūrya group as Alchon-Narana (cf. Grenet 2002: 217 with some reservations). It was from this time that Jibin was used to designate Kapiśa. According to Kuwayama (1999: 65-6) this toponym was assigned to Kabul at some time between 666 and 683. The name Jibin has a variant history of usage in Chinese sources. It would take a monograph to explicate the entire process. Here it should be observed that the alteration of its name had a 19

special significance for Buddhism. To the Chinese people who came to respect Buddhism from the Jin dynasty (265-420) onwards, Jibin conjured the image of a land full of sacrosanct relics of the Tathāgata, miraculous legends in Jātakas and lofty Buddhist masters. This reached its summit when the Kidarites, though expelling the former rulers, maintained the tradition of tolerance and patronage of Buddhism that came down from the age of Kaṇiṣka. Admittedly the prestige of Buddhism had not been vitiated by the time Song Yun visited Gandhāra in the year 520, but the warfare following the invasion of Khiṇgila in 439 simply dented the opulence of Buddhist monasteries and the saṅgha. Later, Buddhism in Gandhāra faded as a consequence of the wholesale damage inflicted by Mihirakula. Kuwayama (1989: 90-111) utterly refuted all the previous theories on the Hephthalite destruction of Buddhism, but his threefold argument does not validly disprove this calamity, which was confined to the reign of Mihirakula, esp. after 520. First of all, Kuwayama is convinced that Buddhism had not suffered from an official ban by the time of Song Yun’s visit. But after the record of the visit there is no mention of Gandhāra in Chinese sources. Although others such as Dani (1969: 69) and Yu (1986: 148) cited the legend in Kura that “all the queens, princes and princesses” of Toramaṇa endowed a Buddhist monastery, the political situation may have changed after 520during the reign of Mihirakula. Shi Yinshun has provided an analysis of the destruction of Buddhism within the Buddhist context and in association with contemporary historical facts (Yinshun 2000: 286-323). Although his dating of Mihirakula is not accurate, Shi Yinshun revealed that the destruction of Buddhism had a definite relationship with King Mihirakula, despite the corruption of the saṅgha themselves. It cannot be excluded that Mihirakula began to persecute Buddhist monks sometime after 520. It is logical that he plundered the treasuries of monasteries after he became embroiled in warfare. This is confirmed by numismatic evidence. The Alchonid drachme coins issued in Gandhāra continue to debase during the reign of Mihirakula until the next ruler, many of the drachmes were even minted from copper (Alram 1996: 528). This sharp debasement of coins is a sign of turbulence.65 Next, his proof that the relics of the Buddha in Nagarāhara and Puruṣapura still existed in early sixth century does not conform their later fate. Even he himself envisaged a possible relocation of the Buddhist relics in Nagarāhara (Kuwayama 1989: 105). Morevoer, the mention by Xuanzang of the loss of the Buddha’s bowl in Puruṣapura serves further proof. The affluent Buddhist saṅghas might well have been plundered since they possessed many jewels and precious metals. Furthermore, the third argument of Kuwayama that the records of Song Yun and Narendrayaśas attest the prosperity of Buddhism in the first half of the sixth century in Uḍḍyāna is not convincing evidence for the situation in Gandhāra. It is widely accepted that the Alchonids marched into Gandhāra from the Kapiśa-Kabul region, not from the north i.e. Uḍḍyāna. His quotation from the Huisheng Xing Zhuan 慧生行傳 (Kuwayama 1989: 110) that, since other countries between Tukharestan and Gandhāra were under the Hephthalites, Uḍḍyāna could not have been an exception overlooks the fact that Gandhāra was occupied by the Alchonid tegin from Kapiśa-Kabul, whereas Chitral and Tashkurgan simply followed the Hephthalites, but were not governed by them. By virtue of its mountainous location Uḍḍyāna might ward off threats from both directions, and this is perhaps the reason why the route linking China and Northwest India shifted after Song Yun’s visit as observed by Kuwayama (1987: 717-22). If we recall that Jinagupta 闍那崛多 chose the route via Bāmiyān instead of the former route when departing for 20

China in 554 (Taisho 50: 433b), it would seem that the decline of Buddhism in Gandhāra had actually taken place. Moreover, Xuanzang depicted the decrease and abandonment of Buddhist monasteries in Uḍḍyāna (Datang Xiyu Ji 大唐西域記, Bk. 3). After Jinagupta the route via Uḍḍyāna was not taken by the pilgrim monks, while the center of the trade route shifted to upper Kabul valley and Bāmiyān. As is stated by Kuwayama, such a detour was surprising and explicit. He attributes this change to the disintegration of the Hephthalite dominion. However, it is unfounded to assign the desertion of the trade route through Chitral – Swat (Uḍḍyāna) to the “Turkish attitude to leave Gandhāra untouched” (Kuwayama 1999: 40-1) for the reason must firstly have been an internal problem in Gandhāra, to say nothing of the fact that the route had already shifted in 554. The legend of the destruction by Mihirakula reflects historical facts, and regardless of how brief this period of damage, he ravaged vast regions of the subcontinent. It would not have been surprising if his Juggernaut-like determination drove him to seize loot from the accumulated treasures of the Buddhist sanctuary. Seeing that the former kings like Toramaṇa showed some respect to the Buddhist presence, and that the land thereafter recovered from chaos, it may be a better explanation that the catastrophe occurred only under his decree – this is the difference between the present article and the traditional view of the Hephthalite destruction of Buddhism. The biographies of Jinagupta and Dharmagupta 達摩笈多 in the Tang Gaoseng Zhuan, as well as the prologue of the Xiyu Tuji 西域圖記 preserved in the Sui Shu 隋書 Bk. 67, significantly speak of the route via Kapiśa and Bāmiyān (ibid., 39), yet none of them mentions the well-known Jibin. It is noteworthy that in the Jiu Tang Shu 舊唐書 Bk. 198 the envoy of Jibin was absent from the court of Emperor Yangdi of the Sui dynasty. All the above show a decline of Buddhism in Gandhāra, while in the biography of Dharmagupta it is recorded that he once resided in a vihāra called Wang Si 王寺 (*rājavihāra) in Kapiśa. This name can be associated with the Jiuwang Qielan 舊王伽藍 (Old King’s Saṅghārāma) in the Datang Xiyu Ji or Guwang Si 古王 寺 (Old king’s vihāra) in the Shijia Fangzhi 釋迦方志 (Taisho 51: 954a). It is also documented in the Tang Gaoseng Zhuan that the king of Kapiśa served Jinagupta well and invited him to be royal tutor shortly after his departure from Gandhāra, where he had studied in the Mahāvanasaṅgharāma in the mountainous areas distant from the urban areas in Mingora and Puruṣapura; the decline of Buddhism in Gandhāra sharply contrasts with the honor and respect it gained in Kapiśa. That is why the latter assumed the name Jibin, a country renowned for its Buddhist piety according to the conventional Chinese point of view. Being more conservative, the Buddhist writers like Xuanzang and Daoxuan rarely used Jibin, and both took it to be a corrupt Sanskrit form of Kaśmira; whereas in dynastic history the envoy of Kapiśa was sent frequently to the Tang court in name of Jibin. This secular use was eventually accepted by the last famous pilgrim monks Huichao 慧超 and Wukong 悟空, at that time when the kingdom of Kapiśa had been usurped by the Turkish military leaders (cf. Kuwayama 1999). There was among all the Chinese Buddhist writers referred to above a universal perspective that Buddhism was greatly respected and miraculously flourishing in Jibin. It was this country, as well as Tianzhu 天竺 i.e. India, which were envisioned by the Chinese Buddhist tradition as the Land of the True Law. With this image or more accurately faith, the Chinese could not have accepted the fact that Jibin was a country of heretics, even if there had been Brahmans and Zoroastrians there long before the Chinese accessed the land. Thus the name shifted for different dynasties about the upper Indus valley, from modern Afghanistan to Kashmir. The features of 21

the states designated Jibin are markedly uniform – obeisance to the True Law during the periods of the Kushan Empire, the Kidarite kingdom, Kaśmir, Kapiśa or Kabul. Among those dynasties of varied ethnic, cultural and political origin modern scholars can only find this single distinctive feature. Among these dynasties the Kidarites played a special role as already pointed out by many scholars: a group of ethnically mixed invaders who attempted to sustain the long-established religious-political order of the Kushan Empire but who eventually failed in the face of the new nomadic hegemony. Their religious tolerance was inherited by their successors, with the exception of the plunderer of Buddhism Mihirakula, whose transitory pillage of Buddhist wealth broke the historical pattern, and was most acutely seared into the memory of the Buddhist adherents. Admittedly, this pillage was largely an individual act by Mihirakula, and so it is too hasty and ill-informed a generalization to speak of the Hephthalite destruction of Buddhism.

The date and ethnic affinity of the Kidarites Methodology and documentary sources Having established the identity of the Yuezhi as the Kidarites in Chinese Buddhist sources, a synthetic investigation of all the related materials, documents, numismatics and archaeology should be undertaken in order to unveil the historical facts concerning the dating and ethnic affinities of the Kidarites. To begin with, after his exhaustive analysis of the documentary sources Enoki (1998: 102) argued that “in order to fix the date of Kidāra and the Kidarites, we should put more reliance on documentary sources than on the types of their coins, and, only after some conclusion has been reached, may we check it with the numismatic evidence.” Whereas Alram (2004) emphasized,66 “The literary evidence is not decisive, since reports by Chinese pilgrims and records by Indian authors are at times ambiguous; and the statements of the Greek and Roman historians, who hardly knew how to deal with the various Hunnic people of the remote eastern lands, are vague. In the absence of authentic evidence, the coins issued by leaders of those people constitute one of the most reliable primary sources for the history of the ‘Iranian Huns’.” A sharp divergence of views exists not only among the conclusions but also the starting points of scholars who rely upon different sources. Proceeding mainly from documentary sources, the present author’s conclusion is perhaps far from satisfactory from a numismatic aspect, although the author has attempted to ease much of the chronic friction. Apart from questions of chronology and dating, the ethnic affinity of the Kidarites needs to be established. If we review the scholarly literature, we find something interesting: Chinese and Japanese scholars, i.e. Enoki (1998: 100-3), Kuwayama (1989: 116-7), and Yu (1986), have all labeled the Kidarites as Kushan/Guishuang 貴霜, whereas western scholars, working from either numismatic or documentary perspectives, have viewed the Kidarites a Hunnish clan. An exception is the late Boris Marshak who, in one of his latest articles (Marshak 2002: 14-5), concluded that Kidāra “probably was of Kushan origin”, and “tried to reconstruct the Kingdom of the Kushanshas, relying on the Huns, who were his servicemen”. The present article agrees with Marshak’s eclectic view. There is another divergence between eastern and western scholars, explained by the following 22

facts: Firstly, the main documentary sources differ as starting points between eastern and western scholars – The Wei Shu styled them Yuezhi, while Priscus called them the Kidarite Huns. Secondly, eastern scholars put more emphasis on the documentary sources, where no clue can be found especially in the Chinese sources, whereas western scholars who attach importance to both documentary and numismatic sources naturally connect the so-called Kidarite Huns with Hunnish coins. And thirdly, there are certain paradigms on both sides, for instance in Chinese sources the Yuezhi/Kushans are irreconcilably hostile towards the Xiongnu/Huns, and with this principle in mind those who rely on Chinese sources cannot imagine a Yuezhi king Kidāra subsequently claiming to a Hun; on the other hand, the paradigm of Hunnish invasion confines the Kidarites to one of the invasive waves from the Steppe, although this is only an assumed association. A solution of this problem must be attained via methodological change. The view held by Dr. Enoki is cited and adopted by the present author. We have in the previous sections established a chronology based on Chinese sources. This chronology is integrated and independent during the years between the mid-4th century and 439, while the chronology from 439 onwards is not clear in Chinese sources. Hereafter the author gives an outline acknowledging the need to strive to: 1, identify the events documented in Chinese sources before 439 with those in western sources; 2, extend the chronology around and after 439 primarily based on Western sources; 3, compare the documentary sources with the numismatic/archaeological evidence in every specific period. Now it is necessary to introduce the documentary sources to be discussed. Contemporary documentary sources on the Kidarites vary among countries of Eurasia, of which the Chinese sources have been reviewed in the previous sections of this paper. Byzantine (primarily Greek, occasionally Latin and Syriac) and Armenian sources have been thoroughly discussed by western scholars. Due to the remote distances and vague concepts used to treat eastern lands, they are secondary to the Chinese sources in reliability.67 It is nonetheless worth mentioning that there are discernible differences between the Byzantine and Armenian sources on the Kidarites. Considering Priscus’s Οὔννοι οἱ Κιδαρίται (καλουµένοι/λεγοµένοι) and Κιδαρίται Οὔννοι, and Ps. Joshua the Stylite’s kšwny’ dhnwn hwny’,68 these chronicles in both languages associated the Kidarites with the Huns with whom they were well acquainted.69 In contrast, the Armenian authors stressed the label “Kushans”, for instance Łazar Pʿarpecʿi, recounting the Sassanian war against the Kushans, who are identified by Grenet (2005b) as Kidarites; whereas Ełišē mentioned Kushans and Huns confusedly. Having adopted the identification of the Kidarite Huns in Priscus with the Kidarites themselves by Grenet (2002: 209), it is not unlikely to accept that the Kushans reported in Ełišē and Ps. Joshua the Stylite are in fact Kidarites. All in all, the confusion of such ethnonyms as Kushan, Hun, and Kidarites represents a lack of awareness among the western writers, of whom the Byzantines emphasized the relation with the Huns indicating their acquaintance with them, while the Armenian added Kushan for their familiarity with the geography. The documents which appeared several centuries after the Kidarites’ presence, such as the sources written in Arabic, are often contradictory, for they usually contain late knowledge and confusing records of different waves of nomadic invaders in northeast of Iran. Nevertheless, these 23

documents have the advantage of honesty; they seldom aggrandize the valor of the Persian kings as pointed out by Frye (1979: 227).70 Contemporary with them, the New Persian works are thus considered as being of less authenticity. The only chronicle of ancient India in Sanskrit, the Rājataraṅgiṇī, is connected with Kidāra by Harmatta, as has been discussed in the previous section. Newly excavated inscriptions and manuscripts in the Indo-Iranian languages have been increasingly vital for studies of the nomadic invasions. The Huns are mentioned in Avestan, Pahlavi and Parthian traditions, and even though almost none of them specifically referred to Kidāra the significant inscriptions bore traces of the eastern neighbors of Iran, especially the Kūšānšahr. The ambiguous Brahmī inscriptions issued by Gupta rulers, Samudragupta (r. circa 340-380) and Skandagupta (r. since 454) also provide us with an impression of Hunnic invasion in India. It must be pointed out that Bactrian scrolls and seals were recently discovered and being studied, and hopefully texts from the heartland of the Kidarites may provide us with more information as time goes by.71

Chronology before 439 After a brief introduction to the methodology and sources, we now turn to the main theme of the Kidarites. Two aspects will be discussed: the chronological outline and the ethno-political affinity of the Kidarites. Since they are at times intertwined, one cannot define either without discussing the other. The author will narrate the historical outline with special points indicated. At first sight we should concentrate on Tukharestan in the pre-Kidarite period, i.e. the reign of of the last assumed Kūshānshāhs. Dani and Litvinsky (1996: 103-8) presented a list of the Kūshānshāhs based on the research of previous scholars like Herzfeld, Harmatta and Bivar. Having accepted that Ardashīr I first conquered Tukharestan in 233, it cannot be refuted that the reign of Kūshānshāh was immediately preceded by the fall of the Kushan Empire of the Kaṇiṣka stock. Besides, it should be emphasized that this stock was almost impossible to restore in Tukharestan, although some ruler in the 8th century claimed ancestry from Kaṇiṣka. A brief interlude is introduced during the reign of the Sassanian kings Wahram II, Wahram III and Narseh (276-302). The history of Wahram II’s reign is described comprehensively by Bivar (1979: 324-7). It can be presumed from his account that the then considerably independent Kushan territories, with the unprecedented title Kūshānshāhānshāh assumed by the Kūshānshāh Ohrmazd (his coins were found at Giri fort in Taxila), were within the Sassanid dominion, and that the independence of the Kushan territories simply resulted from the internecine war between the heirs of the great king Shapur I. Later in the same manner, Narseh’s usurpation of the throne of Wahram III (r. 293) led to an insecure situation and the title Kūshānshāh was again recounted as an independent force,72 but no clue implied an extra-Sassanian ruler until the reign of Shapur III (r. 383-388) when the last Sasanian Kūshānshāhs, Wahram II and Kay Wahram, appeared in the east lands predating Kidāra. For the years 356-358 an account of the Kushans appeared in Ammianus Marcellinus (Am. XVI.9.4) under the name *Cuseni (cf. Marquart 1901: 36 n. 5), together with the first ever mention of Chionitae as rivals of Shapur II. The above events are mentioned by Göbl (1967 I: 19-21) as the starting point of his numismatic chronology. In so far as this identification of the 24

Chionites with that of the Kidarites as is proposed by Göbl, no more documentary source exists. Nevertheless it is unambiguous that Chionitae and *Cuseni are discerned in Ammianus, which means that he could tell the Huns (if we acknowledge the identity of the Huns and the Chionites proposed by de la Vaissière, 2005b) from the Kushans. Moreover, in a later account of the Chionites, Ammianus first recounted that in 358 (Am. XVII.5.1) the Chionites and the Gelani were serving in the army of Shapur II; he then related in the account of Shapur’s siege of Amida in 359 that the Chionites fought as “one of four main contingents” of the Persian army (Am. XIX.2.3). Their king, Grumbates, lost a son in the battle, and the subsequent funeral ceremonies and cremation (a non-Iranian custom) were vividly described by Ammianus (Am. XIX.1.7--2.1). In the latter account the Kushans or *Cuseni never appear, but the Chionites played an important role in the siege of Amida. In the Bactrian documents the supposed name beside Grumbates in Bactrian, γοραµβαδο, is attributed to the prince of the Rōb Kingdom in ca. 470 (Sims-Williams 2002: 232-3). This name has nothing to do with the Kushans or the Kidarites in both linguistic and geographic contexts. That is, to date no name in the Greater Kushan or Kidarite house has been discovered which is the phonological equivalent of the name γοραµβαδο, even a partial adaption, while the Rōb Kingdom has not been proved to be the residence of a Kidarite prince around the year 470. 73 It can thus be concluded that all sources have hitherto described unequivocal differences between the Chionites and the Kushans (*Cuseni), the latter having claimed their independence after their campaign against Shapur II, and having never become an ally of the Sassanids as proposed by Zeimal (1996: 121). Subsequently, Faustus Byzantinus (V. VII; V. XXXVII = Fr. Hist. Gr. V, 285-6, 298-9) in the year 367/8 provided a clue that the Kushans resident in Balkh, who caused severe damage to the army of Shapur II, were of Arsacid origin. Though it was almost certainly a false attribution to the Arsacids, the narrative of Faustus elucidated the residence of the Kushans as well as their hostility and advantage to the Persian army. No evidence suggests whether the Kushans mentioned by the Armenian historian are the same as *Cuseni in Ammianus. Nevertheless, these Kushans could well be supposed to be the predecessors of those led by Kidāra, whose Iranian political identity was directly inherited from them rather than their adversary Sassanids – as stated above, the land of Kūšānšahr in Tukharestan had benefited autonomy or independence since the beginning of the fourth century. It is therefore not reasonable to suppose that Kidāra was once subjugated to the Sassanian king, and even in the reign of Shapur II the Kushans had been an independent and formidable adversary of the Sassanids. The next stage, the 380s, was a decade of tumult rarely mentioned in the documentary sources. The short reigns of Ardashīr II (r. 379-383) and Shapur III were a consequence of the unstable situation. A case in point is that, Shapur III is the last king bore the title Šāhān Šāh Ērān ud Anērān carried on down from Shapur I (Gnoli 1987: 510-1). The title Šāhān Šāh Ērān was immediately preceded by Ardashīr I. Later, after Shapur I’s reconquest of the rebel countries he added ud Anērān to the title, denoting his sovereignty over the eastern countries, as explained in the Res Gestae Divi Saporis (260). In this work the author ventured to surmise that the assumption of this title was a result of the Sasanian conquest of the east, especially of the Kushan territories, and the later abnegation of this title is due to the loss of the eastern provinces. A reasonable testimony is that Ērānšahr in the same inscription is described to include “Armenia, Iberia, Machelonia, Albania and Balāsagān up to Mt. Caucasus and the Gate of the Alans and all the mountain of Parēš(a)hvar” (Alemany 2000: 342-4), embracing almost all of the Sassanian 25

provinces in the west.74 Furthermore, Peroz (r. 459-484) reassumed this title as attested in his great seal (Skjærvø 2003), presumably indicating his triumph over the last Kidarites in Tukharestan and the annexation of lands of Anērān in 468. The other legacy from Shapur I, the title Kūshānshāh, was no longer adopted by the Sassanian princes. Numismatic evidence has shown that the last Sasanian Kūshānshāh in the 380s, Kay Wahram, had his residence in Gandhāra, most probably in Puruṣapura (Grenet 2002: 206). With the corroboration of the Chinese Buddhist sources examined above, we can conclude that the advent of the Kidarites in Gandhāra can be dated to around the year 390. Since Kidāra unified both Bactria and Gandhāra he was assured, and of course qualified to adopt the title Kūshānshāh – it is interesting that his territory covered both Greater and Lesser Yuezhi, the two centers of the Kushans in both of the Chinese accounts, Wei Shu and Kumārajīva’s record. The route of Kidāra’s incursion as documented in the Wei Shu indicates that he subjugated the five countries to the north of Gandhāra. Kuwayama (1989: 109-10; cf. also Kuwayama 1999: 38-9) picks up a passage in the Wei Shu considered to be quoted from the Huisheng Xing Zhuan stating that the five kingdoms subjugated by the Hephthalites (more than 100 years later) were Yarkand, Tashkurgan, Wakhan, Chitral and Gandhāra, as well as Uḍḍyāna which lay on the route. Despite his reconstruction of the route of the Hephthalites, his suggestion of Kidāra’s route is to a large extent more convincing than the suggestion that the route via Panjshir valley by Grenet (2002: 207 n. 6). That is because the Wei Shu provided an explicit depiction of how Kidāra “organized troops and marched to the south to invade Northern India, crossing the Great Mountains, and completely subjugated five countries to the north of Gandhāra”, 75 and this cannot be considered as a circuitous route via Panjshir-Kapiśa to the west. As a result of the Kidarite expansion, their adversary the Sassanian Iran lost business opportunities in the Chinese market, attested in the half-century hiatus in the sequence of Sassanian coins unearthed in China (Xia 1974: 94-5), with no coin between Shapur III and Yazdigird II (r. 438-457) has been found in China. The Kidarites issued their own currency and completely blocked the trade route between Iran and China. So far there is no data available for the exact trade pattern of the Kidarites, but it is possible to surmise that they might have had close commercial links with Gupta India (cf. Dani 1969: 66-7). A frequently debated question is the reading of the so-called Kidāra coins buried in Tepe Maranjān. Although the author has not been able to access clear image of specimens in good condition,76 my point of view is that the reading of the coins is not as important as it has been argued. Since the Chinese annals provide a clear picture of the route of Kidāra via the Swat region, whether his influence reached Kabul is no longer a significant topic. It may be surmised that the Kidarite forces advanced towards Kabul in the 380s but after a transient occupation they were evicted by the Alchonids. Moreover, it is more convincing that the coins were issued by Kay Wahram (Grenet 2002: 206). There is no evidence that Kidāra ever advanced to Kapiśa-Kabul where the Alchonids developed. The accounts of two Buddhist translators, Kumārajīva and Buddhabhadra, fill the crucial blank in the chronology of the Kidarites between 380s and 420s. With the itinerary of Faxian, a sketch of the Kidarite dominion in Northwest India can be confirmed. If the numismatic evidence is taken into consideration, the regions around Taxila or Kaśmīra can also be counted in. This demonstrates an imperial power, which is in command of the most bustling and strategic route connecting Yarkand-Tashkurgan, Bactria-Sogdiana and Madhyadeśa under pax guptica. As a result of the prosperity during the last decade of the 4th and the first two decades of the 5th 26

century, the sovereign deserved the royal title Tianzi or devaputra as mentioned in Ф. 209 like his precursors, even though he perhaps did not actually have the title. What is more, it should be emphasized from the aspect of Buddhist history that the pilgrim monks from China and their masters from India flourished during this period. Further investigation may certainly reveal that, prior to Gupta India, it was the Kidarite kingdom in Northwest India that became the base of the Indian monks and the major destination of their Chinese partners. The political relations among Sassanid, Alchonid, Kidarite and Gupta rulers are worth noting. Some like Deeg (2005: 240) argue that there was rivalry between the Kidarites and Gupta India, but this would seem to be inaccurate. The peaceful frontier between the two powers in India made the pilgrimage between Gandhāra-Nagarāhara-Uḍḍyāna and Madhyadeśa safe and speedy, the land in Pitu and Mathurā being thriving and populous, and Buddhist faith flourishing, as described by Faxian (Deeg 2005: 129-32).77 Besides, the Indian pillar inscriptions of the Hūṇa invasion dated to after 450s have no connection with the Kidarites.78 Another case in point is that the Kharoṣṭhī script, formerly prevalent in Northwest India, ceased to be used during the Kidarite reign, when the use of Brahmī, the script of Gupta India, began. All these facts suggest intimate relations with Gupta India.79 Given that there is no consensus on the situation in Northwest India during the first twenty years of the 5th century, we find the situation in Bactria even more obscure with the three major political powers locked in frequent belligerency. Kumārajīva referred to Tukharestan by its alias “Lesser Yuezhi”, which might imply that the Kidarites in Bactria were in smaller numbers, or of lesser prominence than those in Northwest India. That Faxian stopped in Nagarāhara demonstrates the limits of the Kidarite dominion. The same dating of the Alchonid chiefdom in Kapiśa-Kabul region is ascertained by Alram (1996: 521-6) based on numismatic evidence. Yet quite curiously, the Wei Shu documented the westward removal of Kidāra from Gandhāra, which Enoki (1998: 103) supposed to be in the mid-5th century. A false orientation to the old record in the Han Shu should first be considered otherwise it may refer to conflict with the Alchonids in Kapiśa-Kabul region (and with the Sassanids). If this were the case, it might be inferred that Kidāra’s destination was Balkh, Bactria’s former capital. In Bactria, the Kidarite reign is recorded in the Wei Shu, generally thought to be a fact until 437, with Sengbiao obstructed somewhere north of Puruṣapura before 439. The Chinese documents tally well with other sources in the same period, as stated below, but remain arbitrary in helping to interpret a record by Ṭabarī during Wahram V’s reign (420-438). During the reign of the legendary Persian king Wahram V Ghūr there was fierce warfare, which is mentioned by Arabo-Persian writers such as Ṭabarī, Mas‘ūdī, Dīnawarī and Firdawsī. Ṭabarī recounted a raid of a Turkish Khāqān into Persia, which was eventually repelled by Wahram. After the conflict, Wahram dispatched a general to the region beyond the Amu Darya to subjugate its inhabitants. In the conquered land he appointed his brother Narseh the Marzbān-i-Kushan at Balkh. The same event is recorded by Mas‘ūdī (1962: 229) who mentioned the route of the Turks as being via aṣ-Ṣoghd. Altheim (1969: 49) also mentions Dīnawarī’s report that Wahram’s pursuit did not end at the Amu Darya, but crossed the river somewhere near Balkh. A later detailed depiction was presented by Firdawsī (1967: 312-3). He recounted that it was the Khāqān of Čīn who invaded Persia and his campaign failed at Kashmihan near Merv. He then established the boundary at Fārāb on the northern side of the river Jīhūn80 between Iran and Turkestan. He related that in the next period the defeats of Peroz by Khušnavāz (Akhšonvāz in 27

Arabic sources) were due to his defiance of the border at the Amu Darya which had been established by his grandfather.81 Both Ṭabarī and Firdawsī composed their works in the light of the epic Xvaδāy-nāmaγ written in Pahlavi and dated to the Sassanian age as is noticed by Yu (1986: 78). Nevertheless, the interpretation of this “Turkish invasion” is far from being agreed upon by all scholars. The Turks here must have been a contemporary “non-Persian tribe who lived to the northeast of Persian territory”.82 Suggestions have been made by Marquart (1901: 52-5), Enoki (1959: 19-22), Harmatta (1969: 392), Marshak (1971: 60-62) and Yu (1986: 76-8; 1987: 59-60). Marshak pointed out that the whole story was fabricated. There are many anachronisms, so it is certainly not reliable, but a majority of scholars regard this work to be historical. Marquart first suggested these people were the Chionites, and then specifically identified them as the Avar or Juan-Juan [Ruanruan] (cf. Enoki 1959: 22 n. 6). Enoki ruled out the possibility that they were the Hephthalites; he considered them to be either the Kidarites or the Chionites, as long as the latter were identical with the Wennasha 溫那沙 in Chinese sources). Harmatta quoting O. Hansen suggested the name Turk was a corruption of *tuγrak i.e. Kushans = Kidarites.83 On the contrary, Yu Taishan denied the previous views and identified the invaders as Hephthalites. It appears that the base of this “Turkish invasion” was initiated in Sogdiana as Marquart indicated.84 On the other hand, Narseh’s acquisition of the title at Balkh as recounted by Ṭabarī and Dīnawarī remains a problem, which evidently contradicts the account of Central Asia before 437 in the Wei Shu. Yu (1987: 60 n. 12) provided the possibility that Wahram occupied Balkh temporarily after his campaign against the “Turks”, but the city was before long recaptured by the Kidarites as a result of Kidāra’s westward movement. Evidence pointed out by Yu also signifies the natural border of the Amu Darya, which suggests a northern provenance for the invasion in the direction of the river. Moreover, the title “Khāqān” or “Khāqān of Čīn” cannot be attested in the Kidarite titles prior to the 420s but would seem to be a title of the Donghu 東胡 (proto-Mongolian) tribes such as the Juan-Juan/Ruanruan and Xianbei 鮮卑. This suggests that the intruders from Sogdiana had some relationship with tribes from the east, for instance Juan-Juan/Ruanruan, whose rapid expansion is a prominent feature of the Inner Asian history in the first half of the 4th century.85 The present opinion is that the invaders were certainly not the Kidarites. Taking into account the Chinese sources, we may draw a political map until 437 in which Sogdiana was controlled by Xiongnu, whereas Tukharestan was controlled by the Kidarites or Yuezhi. But archaeological evidence shows that Northern Tukharestan i.e. Termez and as far as Qunduz, was once occupied by the Chionites (Alchonids?) from the beginning of the 5th century until the 440s (Ilyasov 2003: 135-6 quoting B. I. Vainberg; cf. Alram 1996: 522, 533). Therefore Wahram V might have encountered a raid from in Sogdiana/Northern Tukharestan and he succeeded in defeating the ‘Xiongnu’/Chionites who were or who had aided the invaders and he pursued them via Balkh; then, following the assumption by Yu, the Kidarites recaptured Balkh (recall Kidāra’s western movement according to the Wei Shu). The statement of Priscus that the Sassanian kings paid tribute to the Kidarites (see below) indicates that Wahram V did not subdue them. On the contrary, the Arabo-Persian documents after Xvaδāy-nāmaγ covered up these ignominious events, leaving us a confusing record.

Chronology circa and after 439 (primarily in Tukharestan) 28

In the year438, Wahram V died. The new king, his son Yazdigird II, made peace with the Byzantine Empire as soon as he succeeded to the throne in order to turn his armies eastwards, against the Kitarite Huns. Remarkably, Sengbiao’s visit to Gandhāra coincided with the warfare in the first years of Yazdigird II’s reign. As the road could exactly be identified with the one connecting Shahbaz-garhi with Puruṣapura via Charsadda, the adversary of Jibin in Gandhāra was not the Sassanids but the Alchonids, whose eminent leader Khiṅgila firmly held sway over Gandhāra around the year 440, according to the numismatic analysis of Alram (1996). Thinking of the long-standing confederation between the Hunnish chiefs and Sassanian monarchs from the time of Shapur II onwards, it is not surprising to associate Khiṅgila’s campaign with Yazdigird’s war for they were facing the common enemy, the Kidarites, until then a mighty power in the prosperous regions of both Tukharestan and Gandhāra.86 The campaign of Yazdigird II is documented primarily by two Armenian chroniclers: Ełišē who mentioned the Kushans and the Huns and Łazar Pʿarpecʿi (Thomson 1991: 133) who recorded the defeat of Yazdigird by the Kushans.87 According to Ełišē, Yazdigird set out for “the kingdom of the Huns, whom they call Kushans (Honac aškharhin, zor Kʿušans anowanen)” (Thomson 1982: 66; Ter-Mkrtičjan 1979: 78). Concerning the Huns/Kushans mentioned here, scholars’ views vary. Most scholars tend to believe that there is a common confusion of the Kushans and the Huns in Armenian literature.88 If we read Ełišē’s whole work, “the kingdom of the Huns” (Honac aškharhin) appears only once. In the accounts before and after, the place where Yazdgird II marched and campaigned is called “the land of the Kushans” (erkirn Kʿušanac), “the country of the East” (ašiarhin arewelic) or “the Kushan territory” (terutʿiwnn Kʿušanac). Other Armenian writers, such as Faustus Byzantinus, Mosēs Khorenacʿi, Łazar Pʿarpecʿi and Sebēos, all adopt the name “land of the Kushans” for Tokharestan. Therefore, there does not seem to have been a group of people identifiable as the “Huns also named Kushans”. It might well be possible from the account of Ełišē that the ordinarily termed “land of the Kushans” was then occupied by the Huns. Thus we may conclude that the Huns have occupied lands of the Kushans around the year 440. Here the Kushans are doubtlessly those led by Kidāra, recalling his western migration prior to 437. He might have marched back against the Huns, not the Sassanids – as it is reported by Priscus that Yazdigird II initiated the war in 438. Ełišē subsequently narrated that during the years 442-449 Yazdigird resided in the land of the Kushans. Afterwards, at the beginning of the 12th year of his reign (449), he assembled vast numbers of troops and dispatched them to the land of Talakan.89 Seeing this, the Kushan king did not dare engage him, and fled to an impregnable desert location along with all his troops. The troops of Yazdigird captured a lot of booty, but never eliminated their adversary. According to a Syriac chronicle, Yazdgird defeated the local ruler of Chol at the southeastern bank of the Caspian Sea, and built a fort there (Marshak 1971: 59). The next event is documented by both Łazar Pʿarpecʿi and Ełišē, the former’s record is considered more accurate, and in fact more reliable since he distinguished the Huns, Kushans and Hephthalites. In the 16th year of his reign (453) Yazdigird again invaded in the land of the Kushans from Nishapur. Because of a revolt of the prince Bel of Xailandur he was defeated. All the above events portray an image that before and during the reign of Yazdigird II the formerly secure lands of the Kushans were gradually occupied by the Huns. Undoubtedly the Kushan king 29

remained powerful and organized a large group of troops. It can also be observed that the Kushan king succeeded in uniting and defensing Tukharestan with possible assistance of the Huns (cf. Marshak 1971: 65). Although there is still ambiguity during this period, it may be affirmed that the Kidarites (conceivably the Kushans in Armenian sources) in Tukharestan were waging wars with their contemporary rival, Sassanian Iran. The Kidarites might have been assimilating, and in the meantime inevitably were being assimilated by a particular Hunnish group, while being threatened by another, possibly the Hephthalites. Evidence of this trend can be found in the accounts of Priscus which are of premier significance among the sources on the Kidarites in the middle of the 5th century. The fragments of Priscus’ writings, particularly Fr. 33, have been adopted as important corroboration of Enoki’s view (see Enoki 1998: 71-5). However, a new translation by Blockley (1983: 348-9) has overturned the old suggestions, and this has been surveyed by Grenet (2002: 209 and n. 11): The Persian monarch [i.e. Peroz] … asked Constantius [the Roman envoy] to come to him while he was engaged not in the cities but on the borders between his people and the Kidarite Huns (Gk. Οὔννοι οἱ Κιδαρίται). With these a war had begun, the cause of which was that the Huns were not receiving the tribute monies which the former rulers of the Persians and the Parthians had paid. The father of the monarch [i.e. Yazdigird II] had refused the payment of the tribute and had undertaken the war, which his son had inherited together with the kingdom. (translated by R. C. Blockley, texts in the brackets is suggested by Grenet 2002: 209 and n. 11). Blockley’s notes (Blockley 1983: 396 n. 163) indicate that the tributary was not the Kidarites but the Sassanids before Yazdigird II (i.e. Wahram V and/or the former rulers). Besides, Priscus’ remark about the Parthians may allude to the fact that Parthia was tributary to Yuezhi, or to the subsequent conquest of Indo-Parthia kingdom of Gondophares by the Kushans. Consider further on the account of Priscus. On the one hand, one may agree with the fact that the Kidarites were at first aided by other Hunnish tribes as suggested by Grenet (2002: 209). On the other hand, the name Kidarite Huns may also imply the usurpation of some Hunnish leader of the Kidarite throne in Bactria, and the Hunnish people consequently acquired the ethnonym. After all, it is no more than a hypothesis to provide a generalization of the ethnic identity of the so called Kidarite Huns, for this appellation might refer to a political, rather than ethnic, character of these people (cf. Bivar 2006). Concerning later events, it is almost impossible to deny “combined operations” by the Sassanids and the Hephthalites (Alchonids) against the Kidarites in the early years of Peroz (before 468) that resulted in the expulsion of the Kidarites from their capital Balaam as suggested by Grenet (2002: 209; 2005b). As to the site of Balaam, the more recent proposal by K. Czeglédy, adopted by Marshak (1971), Zeimal (1996: 125-6) and Grenet (2002: 211 n. 13) is favored as these scholars have delivered adequate reasons for their argument. The siege of Balkh signifies the successful cooperation of the anti-Kidarite partners, but in what was a familiar theme in Turko-Persian history, Peroz was addicted to the victory and became in turn the opponent of the Hephthalite rulers. Some scholars e.g. Bivar (2006) and Yu (2001: 195-7) considered that the Kidarite Huns 30

described in Priscus were not the Kidarites but the Hephthalites of Procopius (I, 3). Yu also cites Ps. Joshua the Stylite to support this view. Yet in the opinion of the author, it is the account of Ps. Joshua the Stylite that distinguishes the Kidarites from another Hunnish group, presumably the Hephthalites. In Ps. Joshua the Stylite, (Ch. 9), the Sasanian king Peroz, for the sake of his war against the Kushan Huns (Ḳūšānāyē d-hinnōn Hunnāyē), often received money from the Greeks; whereas in Ch. 10, although he had overwhelmed the Huns and added to his dominion from the Huns’ hands thanks to the financial aid of the Greeks, Peroz eventually became the captive of the Huns (Hunnāyē). Being aware of this, Byzantine Emperor Zēnōn (r. 474-491) paid a ransom with his own money and made the king make peace with the Huns. Emperor Zēnōn succeeded the throne after the death of Leo I (r. 457-474) in 474. Hence Peroz must be redeemed after 474. According to Priscus (Fr. 41), the Sassanian army led by Peroz seized Balaam, the capital city of the Kidarite Huns, in 468. If the Kidarite Huns were the Hephthalites, this would mean that the Hephthalites had been defeated, and so the warfare between them and the Sassanians during Zēnōn’s reign becomes hard to explain. Bearing in mind that the Syrian historian was at his time not able to distinguish the different groups of nomadic Huns, it is not inappropriate to conclude that the Kushan Huns in Ch. 9 are identical with the Kidarite Huns in Priscus, while the Huns in Ch. 10 (contemporary with emperor Zenon i.e. the year 484) are the Hephthalites vividly depicted in Procopius. In short, no source prior to Priscus in 456 AD suggests a valid association of the Kidarites with the Huns/Chionites/Xiongnu. Quite the opposite, they were on many occasions named Kushans, although no evidence shows an association with the house of Kaṇiṣka. The records of Priscus and Ps. Joshua the Stylite testify to confusion of the Kidarites and the Huns in Byzantine literature, and it is obvious that Byzantine chroniclers were not familiar with the historical facts in the East. As for the ethnicity of the Kidarites, an acceptable explanation is cited from Marshak (2002: 14-15): …Kidara (who probably was of Kushan origin) tried to reconstruct the Kingdom of the Kushanshahs, relying on the Huns, who were his servicemen and probably entered the territory of modern Afghanistan during his reign. An blending of the Kushans and the Huns hence came about. The perplexing ethnonym “Kidarite Huns” attests to this situation.

The Kidarites in Sogdiana and related topics The residual Kidarites north of Hindūkush sheltered from their adversaries in their last haven – Sogdiana, as first suggested by Grenet (2002: 207-8). There they were eventually subdued by the rival Hephthalites. Recent research by Ilyasov (2003) and Yoshida (2003) suggests the Hephthalite origin of the kingship in Sogdian oasis states – the so-called Zhaowu 昭武 courts. Their arguments are copius and persuasive, and the author of the present article will note some points related to their views. Concerning the emergence of the Kidarites in Sogdiana, the research of Dr. Étienne de la Vaissière (2005a) should be cited. According to this exhaustive work on the Sogdian city states, 31

the advent of the Kidarites to Sogdiana occurred between the 440s and 456. Instead of his early view (de la Vaissière 2002) following Enoki (1955) that the King Huni 忽倪 predated the Kidarite entry, he concluded quoting the Tongdian 通典 that Huni was known to the Chinese court from the embassy in 457. Even via a review of the related materials we cannot determine whether the information in the Tongdian is a genuine copy of the original Wei Shu or a sentential realignment of the accounts in different chapters of the Wei Shu.90 Thus it is difficult to choose between the two possible years of the embassy reporting Huni, as has been admitted by Enoki (1955: 60-1). Grenet (2002: 207-8) on one occasion followed the date of the embassy given as 437, but he then turned to the later date (Grenet 2005b). His two chronologies provide almost the same duration for the Kidarite rule in Sogdiana, i.e., between 441 and 509. Another question is whether King Huni, i.e. the Xiongnu king, was of Kidarite origin. The evidence presented by de la Vaissière (2005a: 108; 2005b: 20-1) that the term for Hun, “xwn”, entered the Sogdian onomastic system during the last generation of the Sogdian caravaneers in the 5th century bolsters the evidence for the presence of Huns in Sogdiana, but it is quite doubtful whether these Huns/Xiongnu (if we accept the identity of the two ethnē) were the Kidarites. Surveying the documentary evidence, de la Vaissière (2005b: 20) endorses the presence of Huns/Xiongnu : “une dynastie kidarite qualifiée de Xiongnu pas les sources chinoises et de hunnique par les sources byzantines”. However, Chinese (i.e. the Wei Shu) and Byzantine sources (i.e. Priscus) narrating events respectively in Sogdiana and Bactria failed to form “un contexte idéal pour que le prénom Xwn apparaisse dans l’onomastique sogdienne”. It is admitted that the Chinese sources can be a proper corroboration, but the Byzantine account of Priscus, narrating a totally irrelevant event occurring in the battlefield between the Kidarites and the Sassanids in Tukharestan (notably he also mentioned the neighbor of the Kidarites) with an ambiguous meaning of the Huns – since we bearing in mind that the term “Huns” was a generic name of many different waves of intruders from north/east of Iran, does not provide a positive link between the Kidarites with Sogdiana. An apparent problem is the time span. If we accept the idea that Huni held sway until 457 and the Kidarite arrival occurred in 440s,this only leaves an average of less than six years (including Huni’s reign prior to 457) for the supposed three kings91 reigning at Sogdiana. The most critical doubt is raised by the contradiction between the two Chinese chronicles Wei Shu and Sui Shu if we accept the view that the former employed the term Xiongnu to denote the Kidarites. The Sui Shu states that “The people of Kang 康 (i.e. Samarkand) are descendants of those of Kangju 康居. Their kings originally had the surname Wen 溫. They were Yuezhi men.” (Pulleyblank 1952: 320, transliteration adjusted) Moreover, if we accept the conclusion of the work of Ilyasov (2003: 135-40) that the tamgha S 2 of the Alchonids and the Hephthalites tallies well with the tamgha on the upper Indus graffiti dated in mid-5th century,92 it is not unimaginable to identify the “Huns” if the Karakorum stone inscriptions and the “Xiongnu” of the Chinese histories who reigned in Sogdiana prior to 457 with the Hephthalites, and with those Hephthalites who held sway in Sogdiana after 509 with the king of Cao 漕 (Kapiśa) , i.e. Nēzak shahs.93 Acknowledging the substantial relation of the Hephthalites with Sogdiana, what then is the role of the Kidarites? They act as the adversary of the Hephthalites and occupied Sogdiana for years, and it was Sogdiana which became the final stronghold of the Kidarites. The cultural elites, as well as the technologies and art sthey brought with them from the heartland of Bactria, 32

exemplified the pinnacle of the resplendent culture of Transoxiana admired as far away as the Abbasid Caliphate and Mogul Empire. It is time to reconsider the passages in the Wei Shu and the Sui Shu concerning sovereignty in Sogdiana. The account of the Sui Shu that the people of Kang were descendants of Kangju and the kings were originally Yuezhi is not implausible as long as we bear in mind that the label Yuezhi refers to the Hephthalites (see above), but it is also reasonable that the name Yuezhi was used also for the Kidarites (as in the Wei Shu) and that the Kidarites had been the rulers of the Sogdiana for about half a century (later than 457-509), not less than the duration of the nominal lordship of the Hephthalites (509-560). It was Wei Jie 韋節 who presumably acquired the information on the origin of the Sogdian people during his visit in the early 7th century. What he learned from the local aristocrats was that they were of Kushan (Yuezhi) origin, whereas the local people might be traced to the pastoral Kangju of Han era. Admittedly, Wei Jie misconstrued the meaning of Zhaowu along with the anachronistic westward migration across Pamir, as Yoshida argued, because Wei Jie attempted to reconcile his new information with the old records. In addition to the possible Hephthalite origin of the name Zhaowu proposed by Yoshida (2003: 38 n. 6), a Kidarite origin also cannot be precluded. Although almost indistinguishable ethnically and linguistically (especially after the cultural fusion of the Kidarites and Huns), the two groups are antagonists within the political milieu. To demonstrate this, it is of great importance to consult the work of de la Vaissière (2005a: 108-12) that the Kidarite reign and the Hephthalite lordship differ greatly. In the Kidarite period, Sogdiana achieved the unprecedented integration both politically and militarily. The major city, Kushaniyya, was built as the capital of the Kidarites.94 That Kushaniyya was in a narrow sense Sogd 粟特 in the Wei Shu has been ascertained by Shiratori (1928: 139-42). The advent of the Kidarites brought about both demographic expansion and the development of fortification, which demonstrated that they were assimilated into Sogdian society as the aristocrats. What can be observed in this period is an “enduring stability”, implying that the Kidarites were firmly established before the Hephthalite hegemony. Needless to say, the artistic flowering, the emergence of urban elites, and the disposal of caravaneers mentioned by de la Vaissière displayed a strategic government and an energetic society in Kidarite Sogdiana, which could therefore resist for years and surrender with dignity to the Hephthalites. On the other hand, we have the impression that the Hephthalites were “fabulously rich”. Tremendous ransoms from Sassanian Iran filled the coffers of the Hephthalite kings, whose luxury lifestyle in their heyday was witnessed by Song Yun. What the Hephthalites offered to Sogdiana was in sharp contrast to the human resources and intellectual foundations provided by the Kidarites. The result is that the Sogdians served, after their capitulation to the hegemony,95 as their merchants, collecting treasures from the remote lands, and in turn dispersing Sassanian silver coins along their trade routes. It can be inferred that the Hephthalites did not replace the local Kidarite lords as long as the latter paid them for their submission. This means that the Kidarites succeeded in retaining their superior status in Sogdiana even after 509, unlike the case in Gandhāra where the Alchonids under Hephthalite sovereign eventually established their own direct rule. Bearing in mind the reluctant diversion of the Sogdians’ role from military nobles to submissive traders, one may not be surprised at the egocentric self-image of the Sogdians as well as the absence of trade in their representations.96 Even later, when the Turks penetrated their lands, the Sogdian nobles of Kidarite origin welcomed rather than resisted their annexation as the Turks 33

shattered the yoke of the voracious Hephthalites. That is why scholars discovered an amiable association between the Turks and the Sogdians, the latter accepting Turkish lordship while the new rulers adopted customs from Iran and India. Yu (1986: 60-2) maintains that the kings in Samarkand were not of Hephthalite origin. Quoting the Sui Shu, he argues that the dynasty in Samarkand adopted a surname Wen for a time, and thereafter they restored the original surname Zhaowu (contra Enoki 1955, who suggested that Wen and Zhaowu were different dynasties). It is worth noting that Wen had never been reported a surname of the Zhaowu people i.e. the Sogdians in China. This may reveal one of the following possibilities: Firstly, that the Zhaowu dynasty was of Hephthalite origin, so they adopted the title Zhaowu, and the original title Wen, i.e. Xwn, was archaic. Secondly, the Zhaowu dynasty was once forced to take the title ‘Wen’, and then abandoned the name. However, we cannot trace any vestige of ‘Wen’ in Sogdian onomastics except for the Karakorum petroglyphs. This may suggest a sudden, deliberate name change, rather than a gradual one as in the former situation. Thus we may suppose the following possibilities: Firstly, the Hunnish hegemony gave rise to the adoption of the name ‘Wen’ in Sogdian onomastics, but this adoption was soon dropped. Secondly, there were discernable social structures in Sogdiana linking the Huns and the Sogdians – the merchants travelling across the Karakorum were Huns rather than Sogdians. If the latter is the case, the appearance of Xwn in Karakorum petroglyphs would indicate a retreat of the Hunnish clans; by the same token, their adversaries would have gained control of the oases.97 A recent article by Rahman, Grenet and Sims-Williams (2006) reports that a seal from Swat in Pakistan bearing the legend “bago Ulargo uonano šao o(a)zarko (k)ošanošao samark(and)o (af)r(ig)ano”. This suggests a ruler named Ularg from Samarkand who was both “king of the Huns” and “great Kushanshah”. Quite probably, this is the king of Samarkand whose influence extended as far as Swat. It is reasonable to surmise that it is this Hunnish king who succeeded to the title Kushanshah from the Kidarites, and actually “took the leadership” in the Kidarite realm. This marks an end of the Kidarite reign in Sogdiana and possibly in other regions. The sketch in this paper is so rough that we cannot provide a total bird’s-eye view of Sogdiana in mid-5th century. Inspired by the breakthrough of Dr. de la Vaissière, we may find that the actual situation in Sogdiana is far more complex than one or two dynastic replacements. The city-states and strongholds like Panjkent, Kushaniyya and Paykent are coeval, but quite unlikely to have been under a united government; instead a precarious situation would seem to be indicated. The Kushans or Kidarites were allied with some, but in conflict with others elements, of the Huns/Hephthalites, and they applied their tactics in the oases. It is quite probable that they were active simultaneously in the 440-50s, and then one absorbed the other, by force or other means. Later, their nominal allegiance to the Hephthalite commonwealth was a far from decisive event, for they soon turned to the Turks when the new conquerer arrived.

Concluding Remarks In the Datang Xiyu Ji, Bk. 1, Xuanzang narrated: Traveling out of the Iron Gate I arrive at the country of Tukhara (Duhuoluo 覩貨邏). It spreads more than one thousand li from the south to the north, and three thousand li from the 34

east to the west. Its eastern limit is the Congling 蔥嶺 mountain, and to the west it borders on Persia. The Great Snow Mountain (Daxueshan 大雪山) is situated to the south, while the Iron Gate serves its natural border in the north. The river of Fuzou 縛芻 (Amu Darya) flows through the middle of its territory to the west. Several hundred years ago the royal kinsmen left no descendant. So the local chiefs and despots raised their powers, holding sway in their respective lands…. Xuanzang illustrated the geographical limits of Tukharestan. 98 He also remarked that the disruption of the states had taken place “several hundred years” previously. What we are interested in is the actual date of the disruption. Naturally, the Hephthalite dominion is considered to be an integrated power on Tukharestan. Thus we may propose a date of ca. 560 when the Turks caused it to disintegrate. This hypothesis contradicts the record of Xuanzang who stated that the kingship continued interrupted for several hundred years. If Kidāra is king Ximodaluo, we may be assured that Tukharestan was integrated under his sway in ca. 390. After that a wave of Hunnish invasion was observed in Tukharestan, and some time after 437 the kingship of the Kidarite court was probable interrupted. It is not certain whether the Kidarite Huns in Priscus’ account were of Kidarite or Hunnish origin, but in 468 it is certain that the Kidarites in Tukharestan were defeated. Therefore we may suppose that the disruption narrated by Xuanzang actually occurred between 437 and 468. The elites of Tukharestan migrated to the north of Amu Darya during this period. This is exactly the same time when Sogdiana boomed as the intermediary power along the Silk Road – replacing the former role of Bactria-Tukharestan. It is significant that the Kidarites made a fair contribution to the rise of Sogdiana. Moreover, the Hunnish-Hephthalite immigration as well as the Sassanid invasion resulted in the disturbance of Tukharestan.

35

Appendices I.

Buddhist relic sites and Yuezhi country in contemporary Chinese Buddhist sources

Narrator 99

Buddhadeva Dao’an

Source

Relics

Sites

Date

Y, T

Bowl

*Gandhāvatī

Mid-4th c.

Bottle

Yuezhi

Bowl

*Gandhāvatī

S

Buddhabhadra

100

Rope

*Hiḍḍavatī

T

Broom

Yuezhi

G

Shadow

Nagarahāra

Additional Remarks

Mid-4th c. – 385

in China since ca. 402

Kumārajīva

D

Cranium

West of Great

NW of Gandhāvatī capital

He

told

411-412

402-406

Huiyuan

in

101

102

Yuezhi Faxian

F

Bowl

Puruṣapura

Arrived at the

The relics in Nagarahāra

Cranium/Shadow

Nagarahāra

sites

was of variant places

Y

Bowl

402,

itinerary in 416

etc. Zhi Sengzai

in

Great Yuezhi,

First half of 5th

*Puruṣavatī

was

i.e.

c.

capital of devaputra

the

*Puruṣavatī Zhu Fawei

S

Bowl

Great Yuezhi

Tooth/Rope

Puruṣapura

ca. 415-439

/Cranium etc. 103

Zhimeng

C, G

Bowl Hair/Tooth/Craniu

Jibin104 105

*Nagaravatī

In India during

*Nagaravatī was 1300 li

404-437,

southwest of Jibin

itinerary in 439

m /Shadow/Footprint Dharmavikrama

C, G

Bowl

Jibin

Cranium

Yuezhi

After 420

Yuezhi is to the west of Jibin, and is on west to the Indus river

106

G

Bowl

Jibin

Ca. 420

Sengbiao

M

Bowl

Puruṣapura,

Shortly before

Jibin

439

Huilan

107

Sources C: Chu Sanzang Ji Ji; D: Dazhidu Lun; F: Faxian Zhuan (Foguo Ji); G: Gaoseng Zhuan; M: Mingsengzhuan Chao; S: Shuijing Zhu; T: Taiping Yulan; Y: Yiwen Leiju.

36

II.

The Kidarites in western sources

(From the reign of Shapur II to Peroz) Records in Western languages – Greek, Latin, Armenian and Syriac as well as in Arabic and New Persian, serve as primary sources on the Kidarites and related tribes in addition to the Chinese sources. All of the following sources have been mentioned in the article. To demonstrate an integrated view, hereby they are listed with tentative identification with the Chinese sources and the historical tribes: Chionites, Kidarites and Hephthalites. Date 356-9

Author Ammianus Marcellinus

Lang. Latin

Clan

Leader

Chin.

R. with Iran

identity

*Cuseni

X

Yuezhi

Hostile

Unidentified

Chionites

Grumbates

Hostile

Possible

/Federati

Hephthalites

Faustus/ 370s

Pʿawstos

Armenian

Kushan

Buzand

Of Arsacid origin

Yuezhi

Hostile

Possible Kidarites

Ṭabarī Mas‘ūdī 420s

Arabic Turks

Dīnawarī Firdawsī

Khaqan

New

Khaqan of

Persian

Chin Kushan

438-53

Ełišē

Armenian

X

Hons, also called

Łazar Pʿarpecʿi

456-67

Priscus

457- ca.

Ps. Joshua

484

the Stylites

467-

Procopius 108

Ṭabarī

Firdawsī

Armenian Greek

Kushan Kidarite Huns

Yuezhi

Unidentified

Hostile

Kidarites

Little X

Kushan 453

Little

Hostile

Yuezhi

Hostile

? X Kounkhas

Little Yuezhi Tuhuol uo?

Hephthalites /Chionites

Hostile

Kidarites

Hostile

Kidarites

Kushan Huns

X

Hostile

Kidarites

Huns

X

Hostile

Hephthalites

Greek

White Huns

X

Hostile

Hephthalites

Arabic

Hephthalites

Akhshonwar

Hostile

Hephthalites

Hephthalites

Khushnawaz

Hostile

Hephthalites

Syriac

New Persian

37

III.

A tentative chronology on the Kidarites

Date

Bactria/Sogdiana

Northwest India

356-58

*Cuseni conflicted with Shapur II

367-68

Kushans of Balkh defeated Shapur II

380s

Kidāra held sway over Tukharestan Kidāra

ca. 390

crossed

the

Hindūkush,

occupied Gandhāra ca. 396 390s-410s

The Kidarites subdued Kashmir The Kidarite homeland Tukharestan faced pressure of Huns from the north

420-38

Khorasan faced pressure from the north

402-420s

Faxian and other pilgrim monks visited Northwest India under the Kidarites

ca. 430

The campaign of Wahram V and his entrance to Sogdiana

430-37

Kidāra’s

movement

for

Tukharestan,

perhaps against the Sassanids

Kidāra’s son was assumed defense of Puruṣapura Khiṅgila’s invasion to Gandhāra,

438-40

seized Puruṣapura 438-42

Yazdigird

II’s

first

campaign

to

Tukharestan (Kushan) 442-49

The Kidarite king of Kushan built castles to resist Persians

449-53

The final battle and decisive defeat of Yazdigird II

ca. 440s

The Kidarites entered Sogdiana

ca. 450s

Sogdian citadels and new cities founded

456

Firsit mention of Kidarite Huns in Western sources

ca. 464

The

Pretender

Peroz

resumed

battle

against the Kidarites 468

The Kidarites was expelled from their capital Balkh

477

Last record of the Kidarites (in India) in Chinese sources

509

The Hephthalite conquest of Sogdiana

38

IV.

A calculation of date of the King Ximodaluo, i.e. Kidāra

Here the date of Kidāra’s invasion to Kaśmira (Jibin/Gandhāra) is calculated according to the dates given by Xuanzang (Bk. 3). Since it is unattainable for the time being to fix a convincing date of the Buddha’s mahāparinirvāṇa, the author gives a formula using only the data of the two kings: Aśoka and Kaṇiṣka. Data: King

Years after the mahāparinirvāṇa

Initial year of reign in Kaśmira

Aśoka

100

268 BCE

Kaṇiṣka

400

127 CE

Ximodaluo

600

?

We can simply observe that the difference between Aśoka and Kaṇiṣka was 300 years; whereas that between Kaṇiṣka and Ximodaluo was 200 years. This is certainly not the actual years, but the difference may stand for a proportion between the times of the three kings, i.e. 3:2. Then we may simply calculate the date of King Ximodaluo given the equation below as C denotes the date of King Ximodaluo: C = 127 + (268 + 127 – 1) (200/300) = 390 CE The date of King Ximodaluo tallies with the proposed date of Kidāra. It is certain that no simple calculation of chronology is valid without reference to historical evidence. Therefore the date here is no more than supplementary.

1

In 1998, Dr. Enoki’s opera omnia, Studia Asiatica was published. The editor collected the articles published in 1969 and 1970, then slightly emended and merged them into one article titled “On the Date of the Kidarites”. In the beginning of this last article it is noticed that “The readers are asked to take the author’s opinions expressed here as his latest ones.”(Enoki, 1998: 53 n. 1) Therefore quotations in the present article of Dr. Enoki’s views are from this article rather than the original ones published in MRDTB, 1969 and 1970 unless specially pointed out. 2 See Petech (1950: 58) and Kuwayama (1990: 950) for the adoption of -vatī with the corresponding Chinese translation -wei 衛 or -yue 越. 3 Kuwayama (1990: 948) inserts “and others” here according to the Chinese character ‘等’, whereas Zhang (1985: 39-40) deletes the character as a modification. 4 The great stūpa at Shāh-jī-kī-Ḍhērī near modern Peshawar is the celebrated Queli Futu 雀犁浮 圖 in Chinese sources (cf. Foucher 1901: 331). See Pelliot (1934: 86-87), Fan (1978: 327-8), Ji (1985: 61) and more recently Deeg (2005: 233-6) for the discussion of the name, cf. also Wan (2009) on a new source of the stūpa. 5 Here Dr. Enoki inserts a note: stating that this was under the reign of Chandragupta II. Concerning Faxian’s statements to this effect, see Majumdar (1954: 22, 346); Kosambi (1965: 193). 6 Cf. Kuwayama (1997: 710) for this dating. 39

7

This proper name, the middle land, Mid-India or Madhyadeśa, is defined slightly different by Enoki and Kuwayama. The author of the present article has adopted the definition of Kuwayama (1990: 946 n. 2): modern Bihar and its environs. 8 Deeg’s conclusion tallies with Petech (1991: 462) and Dani (1996: 167) with Tucci (1958: 288-9) at least with regard to Swat. It is curious nonetheless that the argument of Deeg here is falsely attributed to Enoki (1969-70); cf. also Deeg (2005: 240 n. 1196). Whereas Enoki (1970: 37-38) did assign a date for the Kidarites’ entry to NW India in between 412 and 437, long after Faxian’s pilgrimage. 9 See below for the original texts. Enoki’s translation “completely subjugated”, is better than that of Kuwayama, because it presents an unequivocal impression that the conquest of the states in NW India was total, implying the existence of a centralized political system. 10 Reconstructed by Enoki (1998: 84), while Deeg (2005:238) supposed it to be *Aśvaghosa-bodhisattva-avadāna. 11 Name translated by Kuwayama (1989: 91). 12 Here Kuwayama translates as “nine hundred million”, but this is obviously illogical and not attested in the Taisho Tripiṭaka. 13 Emended from Kuwayama’s translation “Candra Kaṇīṣka”. 14 Kuwayama (1990: 960) especially argues that Kumārajīva was in Gandhāra between 358 and 361. 15 See ibid, also cf. Kuwayama (1987: 709). 16 As Sims-Williams (2002: 232) attempts to demonstrate with his examination of loan words to the Bactrian language. It is to some extent surprising that the Kidarites left not a single loan word in Bactrian, the language of ancient Afghanistan where their capital is supposed to be situated. We can compare this with Grenet (2005b). 17 Grenet (2002: 205 n. 1) suggested a reminiscence of the Lesser Yuezhi in the Han Shu. He disputed the association of Xiongnu mentioned here with the Hephthalites, as proposed by Enoki (1998: 69-70). Additionally, Grenet (2005b) presumes another possible interpretation of the accounts on the Lesser Yuezhi, supposed to demonstrate “hostilities in Gandhāra between the Kidarites and some Hunnish predecessors there”. Since the only attested Hunnish invasion led by Khingila occurs after the entrance of the Kidarites (Alram 1996: 526), an imitation of the Han Shu is the most acceptable of the three interpretations. 18 Here Enoki’s rendition, which implies Kidāra himself organized his troops, is closer to the facts discussed below, whereas Kuwayama misinterpreted as he dispatched an expedition. 19 Special thanks for Prof. Wang Bangwei (personal communication, 2007) for this Sanskrit name. 20 The character for “Xi”醯 is absent, but a blank is left in the manuscript. Zheng and Xu (2002: 94, 107) correctly added the character according to Faxian and Xuanzang’s texts. 21 Here it is an erroneous record since the Buddha’s bottle (kuṇḍikā) and bowl (pātra) were at that time kept in Balkh and Puruṣapura, respectively. See Wan (2009). 22 Originally Shu 殊. Zheng and Xu (2002: 93, 107) emended this character unconvincibly as Zhu 珠 (bead). 23 Faxian attributed the name “Xiluo” to the city where the uṣṇiṣa was preserved. 24 As Lévi (1934: 190) determined, the royal title devaputra was only adopted by Kushan kings in the Indian tradition, while in the Chinese tradition the counterpart Tianzi has abundant cultural content (Cf. also Pelliot 1923, Yu 1986: 230-233, especially see Chen 2002 on the origin of this Chinese concept). However, it should be emphasized that from Tang dynasty onwards the imperial title Tianzi was no longer applied to foreign monarchs in Chinese literature, and was an imperial title reserved for the Chinese emperor. 25 In Wan (2009) the author compared the population of saṅgha documented in Ф. 209 with that recorded by Faxian, Song Yun, and Xuanzang, thus concludes that Ф. 209’s description is probably a little later than Faxian’s information. 26 This dating is not possible without evidence in the history of Ceylon, i.e. the Tamir invasion of Ceylon in 433, which saw the destruction of Singhalese Buddhism. 27 Regarding the Alchonids or Alchons, see Ilyasov (2003) for a thorough analysis of the related materials esp. on the tamghas. In the addendum of his article Ilyasov summarized “if to consider that the dynasty of Khingila (i.e. Alchon) is Hephthalite, then we should admit that 40

Alchon-Chionites conquered under the aegis of the Hephthalites lands beyond Hindūkush but kept, their symbolic and tribal name.”(ibid., 154) This statement probably depicts the actual relationship between the two tribes, and the overlordship of Hephthalites in Gandhāra was definitely witnessed by Song Yun. 28 The direct threat was apparently not Gupta India but the chiefdom of Alchonids and its provisional allies, Sassanian Iran, as will be discussed below. 29 As discussed by Petech (1950: 62) in his compilation of the Shuijing Zhu, Bk. 2. 30 Although explicated by Deeg (2005: 241 n. 1197), a possibility that the sites of the relics were unknown to Narendrayaśas cannot be ruled out. This is because it is recorded that “some said that…” 或言 about the sites of the relics in his biography in the Tang Gaoseng Zhuan 唐高僧傳 (Taisho 50: 432b), and the text did not report his personal experience. 31 Except for the similarity with Ф.209’s mistake regarding the location of the Buddha’s bowl, Zhu Fawei’s depiction is also in accordance with Ф.209’s statement in which there is the description of how Buddha’s relics “all in the air remaining never fallen”, implying a similar dating. 32 For instance, Kuwayama (1990: 951) believed that Zhu Fawei lived in the latter half of the 4th century or early 5th century, which is in any case inaccurate. 33 Written by Baochang 寶唱, the text has lost except a catalogue: the Mingsengzhuan-mulu 名僧 傳目錄 (A Catalogue of Mingsengzhuan). The Japanese monk Sōshō 宗性 copied excerpts from the original texts lost nowadays. For an introductory account, see Su (1978). 34 Taisho 50: 337b: “In addition, Zhu Fawei 竺法維 and Shi Sengbiao 釋僧表 had both been to the land of the Buddha.” The encyclopædic Buddhist work Shijia Fangzhi 釋迦方志 reads: “such as Fawei 法維 and Fabiao 法表 left their names without biographies, thus their accounts are difficult to edit.” (Taisho 51: 969c). Cf. Chavannes (1903: 437). 35 The passage has been emended to 涼州 (Liangzhou) here. 36 Puṣkalāvatī was of course then the dominion of Kidarites, i.e. Yuezhi country, see for instance Enoki (1998: 81) where he identifies “Fu-chia-lo [Fujialuo] 弗 迦 羅 or Fu-chia-lo-p’o-to [Fujialuo-poduo] 弗迦羅婆多” city of Yuezhi country in the Dazhidu Lun 大智度論, as Puṣkalāvatī. 37 Göbl (1967: II, 224). 38 There were five dynasties which adopted the name Liang 涼 during the Eastern Jin and Song period (317-479), of which the states which established their capital in Liangzhou were chronologically Qianliang 前涼 (Former Liang, 317-99), Houliang 後涼 (Latter Liang, 386-403) and Northern Liang. See Corradini (2006: 211-20) for detailed information on the Liang states. The dating of Sengbiao, with reference to Daowang mentioned in the Gaoseng Zhuan (Taisho 50: 371c-372a), demonstrates that Liangzhou no doubt referred to Beiliang. 39 Faxian and Ф.209 mentioned the defense of Puruṣapura and Nagarahāra respectively, of which the latter might have been the immediate frontier against the Alchonids in Kapiśa-Kabul region. 40 This also echoes Ф.209’s reference to the royal title ‘devaputra’ adopted by the Kidarites, as implied in Zhi Sengzai’s account. 41 As for the records in the Han Shu see Hulsewé and Loewe (1979), cf. also Hulsewé (1975) for the accounts in the Shi Ji. 42 This adverse juxtaposition is so curious that no sound interpretation has hitherto been proposed. Since the Wei Shu explained the provenance of Lesser Yuezhi, the division by Kumārajīva may be associated with the importance or actual size of the territory in both regions. 43 For the date of compilation of Kumārajīva’s biography, see Enoki (1998: 85 n. 103) who acknowledged it to be either just after his death in 413 or at some time under the Qi 齊 Dynasty (479-501). The latter interpretation seems to replicate an initial assumption, because in 479 the Kidarites are not mentioned in any sources. 44 An emendation with reference to Xuanzang of Xiluoyue 醯羅越 for Xiluo was first proposed by Kuwayama (1990: 952), but as a result the character “tu” 吐 no longer makes sense. To solve this inconsistency, the author proposes that Xi should be added with tu preserved because the Chinese transcription of the Sanskrit aksara “-ḍḍa” can be separated into two syllables, i.e. tuluo 吐羅 (EMC thɔ’-la, in Pulleyblank, 1991), cf. also Deeg (2005: 124-5). 41

45

Although Petech (1988a: 252-3) questioned the identity of Fotudiao 佛圖調 (Buddhadeva) in the Shuijing Zhu as being Zhu Fodiao 竺佛調 whose biography is in the Gaoseng Zhuan (Taisho 50: 387c-388a), for “il n’y a pas mention d’un récit de voyage écrit par lui”, he identifies Fotudiao with Buddhadeva with a dating of “la première moitié du IVe siècle”, as well as making the comment “même si l’auteur n’est pas identique avec Fo-t’ou-teng ou Tchou-fo-t’iao, on peut admettre qu’il appartenait à leur cercle”. The author prefers a later dating to the middle of the 4th century, since after the death of his tutor Fotucheng 佛圖澄 (Buddhadāna) in 348 Fotudiao himself began to preach the dharma. 46 Attributed to Dao’an. Petech (1988a: 243) provides a French translation of the text in the Taiping Yulan 太平御覽, fasc. 765. 47 It is noteworthy that the Yuezhi country mentioned here has nothing to do with that of Kaṇīṣka’s royal house. The last king attested in the Chinese annals, Vasudeva I, sent his envoy to Wei court in the year 230, after which the name Yuezhi disappeared in the Chinese annals for more than a century. Although the Nanzhou Yiwu Zhi 南州異物志 accounted Yuezhi country around the year 250, the name Yuezhi never occurred until that of Kidāra in the Wei Shu (cf. Enoki 1998: 89, 95-6). It is probable that the Chinese were not aware of the actual power transition in the Kushan dominion in the 3rd-5th centuries, so that the compiler of the Wei Shu did not distinguish between Kidāra and the former Kushans, which resulted in a confused narrative in the texts as discussed above. 48 This date is listed in the Lidai Sanbao Ji 歷代三寶記 (Taisho 49: 92c) as “the last year of Yuanjia 元嘉 (424-453) era” whereas his biography in the Chu Sanzang Ji Ji and the Gaoseng Zhuan asserted that his journey in India lasted some twenty years (ending in the 440s). Fayong’s mention of Yuezhi country is again without the title Greater or Lesser, suggesting that the other part of Kidarite kingdom, i.e. Bactria, might have been eliminated before Fayong’s arrival soon after 420. However, his association of Jibin (obviously not Kashmir as proposed by Enoki 1998: 87) with the Buddha’s bowl, together with the association of the Yuezhi country with the uṣṇiṣa, makes it difficult to clarify the situation. 49 The colossal statue of Maitreya, described by Faxian in Tuoli 陀歷 (the Dardic area, cf. Cunningham 1871: 82, Dani 1983: 3-4), is also mentioned in Ф. 209, see Wan (2009). 50 The present article solely discusses Jibin in Chinese sources from the 4th century on. On Jibin in earlier Chinese records, i.e. in the Han Dynasty chronicles cf. Hulsewé and Loewe (1979: 104 n. 203), the two editors unconvincingly identify Jibin as Kaśmira, see Daffinà (1982: 316-8) for an acute criticism of this view, especially against the opinion of Jibin-Kaśmira identity based on the Chinese consonant system by Pulleyblank (1962-63). 51 The date is the period when An Faqin was in China, as documented in the Kaiyuan Shijiao Lu 開元釋教錄 (Taisho 55: 497c). 52 Attested in the prolegomenon of the translation (Taisho 29: 161b). 53 Xianqing 顯慶 era, from the Datang Neidian Lu 大唐內典錄 (Taisho 55: 301a). 54 Deeg (2005: 110 n. 499) suggested Sengshao a Hīnayāna monk who departed from a northern Mahāyāna state (Khotan, Khargalik etc.) southwards for Kaśmira. His opinion is plausible, yet remains unattested by other sources. 55 For a comprehensive biography of Shi Yinshun, see Bingenheimer (2004). 56 Date attested in the Lidai Sanbao Ji (Taisho 49: 95c). 57 The former recounts that his grandfather Harībhadra had been exiled for his uprightness, and that his father Saṅghānanda was a recluse in a hermitage, while the latter states that Guṇavarman was an offspring 支胤 of the Jibin dynasty. At any rate, he was of noble origin from the royal court. 58 The election to the royal throne is seen in the Rājataraṅgiṇī 324, 469-77 in Kaśmira, first mentioned after the death of king Mihirakula. 59 Although Deeg (2005: 124 n. 564) criticized the opinion held by Beal (1887: 202) that Faxian had never been to Taxila, he does not provide valid evidence of the monk’s sojourn there. His suggestion of the route from Gandhāvatī to Taxila and the way back is only a hypothesis. However if we review the text of Faxian’s record of Taxila, it is no more than a clichéd account of a typical city. At any rate Faxian did never travelled further east to Kaśmira, at the time when there might have been a local Jibin state. 42

60

According to Xuanzang (as well as other sources in the Sarvāstivādin tradition as collected and analyzed by Lamotte 1976), The reign of Aśoka (268-237 BCE) fell one hundred years after the mahāparinirvāṇa of the Buddha. Falk (2004) has convincingly demonstrated that the reign of Kaṇiṣka commenced in 127 CE, four hundred years after the mahāparinirvāṇa, but apparently no record of such invasion from north of Hindūkush is preserved for the six hundred year period after the mahāparinirvāṇa. Since the date of the mahāparinirvāṇa has yet to be defined, the calculation of the absolute chronology of king Ximodaluo can be formulated using the certain initial date of Aśoka and Kaṇiṣka (see Appendix IV), and, not surprisingly, that date happens to be around the year 390 CE, coeval with Kidāra’s reign. 61 There are other writers, for example A. Biswas in The Political History of the Huṇas in India, who have expended great efforts on establishing the chronology of the Kashmirian epic. 62 If we predate several years his conditions by adopting the new idea that the Kaṇiṣka era began in 127, the date of Kinnara is identical to that of Kidāra, namely the 390s. 63 In the numismatic field, this view has been proposed by Göbl (1967) and Harmatta (1969: 388-90). 64 One notable exception is the Shishi Xiyu Ji attributed to Dao’an and preserved in the Shuijing Zhu, Bk. 1. Petech (1950: 22) identified it with Kaśmira, but then pointed out that this must be incorrect because Dao’an’s work “is accustomed to take liberties with hydrography”. However, if we accept Dao’an, we may be sure that the Indus River flows from north of Taxila to Gandhāra, and the region adjacent to (and even including!) Taxila might be the correct counterpart of Jibin here in the Shuijing Zhu. 65 It is remarkable that Khiṅgila’s first issue of skyphate coins underwent a similar debasement from gold alloy to inferior silver (Alram 1996: 527), which suggests an unstable situation. 66 Cf. Alram (1996: 525-6). 67 Cf. Marshak (1971) quoted by Zeimal (1996: 125). 68 Harmatta (1969: 391) read this, after F. Altheim, as kywny’ dhnwn hwny’, and he used this quite dubitable reading to establish the Chionite-Kidarite identity. Similar reading is seen in Marquart (1901: 58); cf. Tremblay (2001: 187). 69 De la Vaissière (2005a: 98, 2005b) proposes the identity of the Huns and the Chionites. However, the Kidarites never called themselves Huns, *Chions or Xiongnu, which is still ignored by J. Ilyasov who ascribes the “kušanisch-hunnisches Mischtamga” in Göbl (1967 II: 53, 208; III: pl. 3; IV: S 82) to the Kidarites (Ilyasov 2003: 154). This tamgha is associated with the gold coins in Tepe Maranjān near Kapiśa as well as seal G 16 of the Kushano-Sassanian period. As stated below, the coins in Tepe Maranjān do not belong to Kidāra. Thus no valid archaeological or numismatic evidence so far supports or implies the identity of the Kidarites with the Huns-Chionites. 70 Frye also stressed, from a comparative aspect, that the linear view of the time among Arabic historians may endow their work with extra reliability. 71 See Grenet (2005b) for these achievements. 72 Harmatta (1969: 384-5) regards this title as the residual ruler of the Great Kushans in Gandhāra. This can also be interpreted as an independent ruler of the Sassanian house, cf. Enoki (1998: 98). 73 The author even believes that the Kidarites never extended their power as far as Rōb. A linguistic parallel occurs on the coin legend ΓOBOZIKO in Göbl (II: 56, Em. 32) which is closely related to the Alchonids but has nothing to do with the Kidarites (cf. Ilyasov 2003: 135-8). 74 However in the reign of Ardashir I, the concepts of Ērān and Anērān were slightly different, e.g. in the Kerdīr inscriptions. It is not an exaggeration to observe that the title Kūshānshahr was of the utmost significance in all of the lands of Anērān when Armenia was regarded as part of Ērānshahr (cf. for instance Bivar 1979: 319). 75 Quoted from Enoki’s translation. 76 Göbl (1967 I: 17-8) has read the inscription as “KIOOPO”, but Grenet (2002: 206 and Pl. Ia) gives a reading of “κιοοοοοο» with the explanation that “on the best specimens only the initial sequence κι- followed by a row of undifferentiated ο’s is visible”. Perhaps the number of the οs was a significant factor, since it can be inferred that the Bactrian form of the name Wahram may takes six or more characters, whereas Kidāra’s name, attested in Bactrian Κιδιρο or Κηδδιρο, takes four or five characters after the initial κι-. 77 The fact that Faxian who crossed Pamirs in 402 did not mention any troubles noted by de la 43

Vaissière (2005a: 99) does not necessarily mean that a tranquil situation prevailed in Bactria, because his route through Wakhan, Chitral, Uḍḍyāna and Gandhāra was within the heartland of the Kidarite kingdom; on the contrary this is a reliable confirmation of the harmony prevailing in the subcontinent. 78 Enoki (1998: 77-8) identified the “Mlecchas” in the Junāgaḍh Inscription (ca. 457) and the “Hūṇas” in the Bhitarī Inscription as the Kidarites. His view was accepted by Deeg (2005), but criticized by Yu (1986: 85-7; 2001: 199-200) and Grenet (2002: 211). 79 The Allahabad inscription mentioned the submission of Daivaputra-shāhi-shāhānushāhi to the Gupta king Samudragupta (340-380). Although Dani (1996: 165) suggested that the title was residual from the Great Kushans, it may have also alluded to a connection with the Kidarites, whose advent in Gandhāra could well have led to a collision with Gupta India even if they eventually yielded. Even Dani himself once proposed this point of view (Dani 1969: 66-7). 80 This name is identified by Enoki (1959: 20) as being the Syr Darya, but only the Amu Darya which could have been logical in the context of the trespass of Peroz as mentioned below (cf. Yu 1986: 83 and n. 5). 81 The defiance of Peroz suggested that Wahram V waged wars with the Hephthalites or those who can be described as the same ethno-politic group (the Chionites), but obviously not with the Kidarites. 82 Enoki (1959: 21) pointed out that “northwest” was a mistake for “northeast”. 83 This view is highly doubtful because the people in Sogdiana were never styled “Tocharians” or Kushans. 84 It should be noted that Marquart (1901: 52-3) described the title “Marzbān-i-Kushān” as “Hüter der Mark gegen die Kūšān”, and according to him it by no means served as a substitute for “Kūshānshāh”. 85 It is proposed, for instance, by Yu that it was the Ruanruan/Juan-Juan who droved the Hephthalites then in Sogdiana down to Khorasan. Thinking of the Mongolian affinity proposed by Pulleyblank (1963: 258) and the Xianbei origin suggested by Yu (1986) of the Hephthalites, the identity of the Turkish invasion with the Hephthalites cannot be excluded, although they were presumably named differently. 86 The role played by Khiṅgila, first pointed out by Grenet (2002: 218-20) because of a connection with the fresco and seals of the legendary hero Rustam, is revealed because Khiṅgila had the same features as Rustam: they were both local leaders of Sistan, they both campaigned in the eastern lands and they both warded off the threat from east of the Sassanids (most probably the Kidarites). It is also noteworthy here the enthronement of Peroz who originally assumed the title of “Marzban” took place in Sistan, contiguous to the Alchonids. The collusion of these two in ca. 457 is thus not merely conjecture. 87 Special thanks to my former colleague Yang Xi 楊曦 for the Armenian language support. 88 Robert Thomson believed that the terms Kushans and Huns are inexactly used in Armenian sources (Thomson 1982: 63). E. V. Zeimal, when writing the section on the Kidarites in the History of Civilizations of Central Asia, avoided using Armenian sources for which he gave a detailed reasoning (Zeimal 1996: 124-25). Recent scholars such as Grenet have also avoided using these sources. 89 Based on the account of this writer (Ełišē Bk. 2), Bregel (2003: 12-3) in his atlas of Central Asia illustrated an assault of Yazdigird II in 442 on Talakan in Tukharestan east of Balkh. However, Ter-Mkrtičjan (1979) used this toponym to denote the region of Khorasan west of Balkh. Zeimal (1996: 125) mentioned this conflict in interpretation. It is noteworthy that the tactics of strong fortification were a characteristic strategy of the Kidarites for their achievements in Sogdiana, as summarized by de la Vaissière (2005a: 109). 90 The original edition of the Wei Shu might have been lost when Du You composed the Tongdian. This is still open to question. See for instance Li (2005) on the editing of the Tongdian; in this article Dr. Li Jinxiu remarked that the Tongdian has its own rules governing the selection of sources despite its original statements that it would follow the principles of a dynastic history. On its specific relation with the Wei Shu, Yu (1986: 228-33) keenly criticizes the view that the overall arrangement of the Tongdian reflected the actual features of the Western Regions, whereas the extant Wei Shu provides an exact geopolitical map. However, recent Japanese scholars like Kuwayama (1999: 38 n. 1) tend to appreciate its authenticity. 44

91

Enoki (1955) even proposes a possibility of four kings, which is criticized by Yu (1986). De la Vaissière (2005b: 20 n. 38, 39) pointed out that the name m’ymrγc appears eight times as forename and five times as a patronym in the upper Indus glyphs. This name was associated with the city of Maymurgh, attested first in Chinese sources in 457 (de la Vaissière 2005a: 107). This would mean that the last date of the graffiti was as late as the 450s (“avant 460”). De la Vaissière thus argued convincingly that the actual end of the Sogdian trade via the upper Indus valley can be dated to around the year 460. This date contradicts the hypothesis that the Kidarites swept away Hunnish rule around the year 440. 93 Kuwayama (1999: 29-32) identified Cao with Kapiśa, but he remained unaware of the right reading of the coins. Surveying Alram (1996), Grenet (2002: 217-8) established the correct linkage. 94 Cf. Grenet (2002: 209). 95 For example, de la Vaissière (2005a: 111) noticed a limited conflict leading to the subjection of Panjikent to the Hephthalites. 96 This point was emphasized by Grenet (2005a) and the author was first made aware of this thanks to the lecture at Peking University given by Dr. Valerie Hansen on May 30th, 2006. 97 Whether those passing through Karakorum bearing the name “Xwn” were the upper class of the Sogdians is so far not clear. Therefore the association of them with the Hunnish kings of Sogdiana is not valid. 98 Congling clearly stands for the Pamirs and the Great Snow Mountain is the Hindūkush, contrary to the identification of Kuwayama (1999). 99 The pupil of Buddhadāna (232-348), Cf. Petech (1988: 255-6). 100 On *Hiḍḍavatī see footnote 44. 101 Cf. Zürcher (1952: 224) and Kuwayama (1989: 102-3). 102 Cf. Enoki (1998: 81). 103 Cf. Kuwayama (1990: 947). 104 Deeg (2005: 103 n. 460) suggested that it was Gandhāra including Swat. 105 Chu Sanzang Ji Ji, Vol. 15, Taisho 55: 113b. The country of Najialuowei [那]迦羅衛 can be reconstructed as Sanskrit *Nagaravatī. The character Na 那 was lost in the Taisho Tripiṭaka, in other editions mistaken by adding Wei between Jia and Luo, thus Jiaweiluowei 迦惟羅衛 (= Kapilavastu). This was an obvious error followed by the Gaoseng Zhuan. Cf. Kuwayama 1987: 710 n. 27; Deeg 2005: 250 n. 1241. 106 Taisho 50: 399a. 107 He died at the age of sixty around the year 460 and his journey to Jibin took place around the year 420. 108 A list of other Arabic sources is provided by Bivar (2006). 92

Reference Primary documentary sources Chinese Non-Buddhist: Shi Jing 詩經 = Waley 1960 Shi Ji 史記 Han Shu 漢書 = Hulsewé 1979 Wei Lüe 魏略 Wei Shu 魏書 Beishi 北史 45

Sui Shu 隋書 Jiu Tang Shu 舊唐書 Tongdian 通典 Nanzhou Yiwu Zhi 南州異物志 Shuijing Zhu 水經注* = Petech 1950 Yiwen Leiju 藝文類聚* Taiping Yulan 太平御覽* Dunhuang Manuscripts in the St. Petersburg Branch of the Russian Institute of Oriental Studies, vol. 4, Document Ф. 209. = Men'šikov and Vorob'ëva-Desyatovskaya 1999 Hulsewé, A. F. P and M. A. N. Loewe, China in Central Asia, the Early Stage: 125 B.C. A.D. 23. An Annotated Translation of Chapters 61 and 96 of the History of the Former Han Dynasty, Sinica Leidensia: 16, Leiden, 1979. Men'šikov, L. N. and M. I. Vorob'ëva-Desyatovskaya, Opisanie kitajskikh rukopisej Dun'huanskogo fonda Instituta narodov Azii, translated into Chinese by Yuan Xizhen and Chen Huaping, Shanghai, 1999. Petech, Luciano, Northern India According to the Shui-Ching-Chu, Serie Orientale Roma: II, Rome, 1950. Waley, Arthur translated, Shi Jing: the book of songs, New York, 1960. *These sources contain literature ascribed to Buddhist monks. Chinese Buddhist: Shi Faxian 釋法顯, Foguo Ji 佛國記 = Zhang 1985, Legge 1886, Deeg 2005 Yang Xuanzhi 楊衒之, Luoyang Jialan Ji 洛陽伽藍記 = Fan 1978, Chavannes 1903 Shi Xuanzang 釋玄奘, Datang Xiyu Ji 大唐西域記 = Ji 1985 Taisho Tripiṭaka 大正新脩大藏經, Electronic Edition 2006, Taipei: Chinese Buddhist Electronic Text Association, 2006. Various Buddhist sources not listed is collected in Taisho Tripiṭaka. Shinzan Zokuzokyo 卍 新 纂 大 日 本 續 藏 經 , Electronic Edition 2006, Taipei: Chinese Buddhist Electronic Text Association, 2006. Mingseng Zhuan Chao 名僧傳抄 is collected in Shinzan Zokuzokyo. Chavannes, Édouard, "Voyage de Song Yun dans l'Udyâna et le Gandhâra (518-522)", Bulletin de l’École française d'Extrême-Orient, 3: 1903, 379-441. Deeg, Max, Das Gaoseng-Faxian-Zhuan als religionsgeschichtliche Quelle: der älteste Bericht eines chinesischen buddhistischen Pilgermönchs über seine Reise nach Indien mit Übersetzung des Textes, Wiesbaden, 2005. Fan, Xiangyong 范祥雍, Luoyang Jialan Ji Jiaozhu 洛陽伽藍記校注, Shanghai, 1978. Ji, Xianlin 季羨林, Datang Xiyuji Jiaozhu 大唐西域記校注, Beijing, 1985. Legge, James, A record of Buddhistic kingdoms: being an account by the Chinese monk Fâ-hien of his travels in India and Ceylon (A.D. 399-414) in search of the Buddhist books of 46

discipline, Oxford, 1886. Zhang, Xun 章巽, Faxian Zhuan Jiaozhu 法顯傳校注, Shanghai, 1985.

Sources in other languages: Altheim, Franz, Geschichte der Hunnen, 2nd edition, vol. 2, Berlin, 1969. Ammianus Marcellinus, with a translation in English by J. C. Rolfe, in 3 vols., Cambridge, Mass., 1935. Blockley, R. C., The Fragmentary Classicizing Historians of the Later Roman Empire. Eunapius, Olympiodorus, Priscus and Malchus, Liverpool, 1983. Firdawsī, The Epic of the Kings: Shāh-nāma, The national epic of Persia by Ferdowsi, translated by Reuben Levy, Chicago, 1967. Fragmenta Historicorum Graecorum: Apollodori bibliotheca cum fragmentis, translated by Theodor Müller and Victor Langlois, eds. Karl Müller, in 5 vols., Paris, 1878-85. Longlois, Victor, Collection des historiens ancients et modernes de l’Arménie, vol. 1, Paris, 1867. Mas‘ūdī, Les Prairies d'Or, translated by Barbier de Meynard and Pavet de Courteille, Paris, 1962. Procopius, with a translation in English by H. B. Dewing, in 7 vols., London, 1914. Ranjit Sitaram Pandit, Rājataraṅgiṇī: The Saga of the Kings of Kaśmīr, New Delhi, 1968. Ṭabarī, "Taiboli Shi suozai Yeda shiliao jianzheng" 泰伯里史所載嚈噠史料箋證 “An annotated translation of sources on the Hephthalites in Ṭabarī”, annotated by Yu Taishan 余太山, translated into Chinese by Song Xian 宋峴, Zhongya Xuekan 中亞學刊 (Journal of Central Asian Studies) 2: 1987, 51-64. Ter-Mkrtičjan, L. H., Armjanskie istochniki o Srednej Azii V - VII vv., Moscow, 1979. The Chronicle of Joshua the Stylite, with a translation into English and notes by W. Wright, Cambridge, 1882. Thomson, Robert W., trans., Eḷishē: History of Vardan and the Armenian War, Cambridge, Mass., 1982. ______, trans., The History of Łazar P'arpec'i, Atlanta, Georgia, 1991. ______, trans., The Armenian History attributed to Sebeos, Part I: Translation and Notes, Liverpool, 1999.

Secondary works Alemany, Agustí V., Sources on the Alans: a critical compilation, Handbuch der Orientalistik: 8, Leiden; Boston, 2000. Alram, Michael, "Alchon und Nēzak, zur Geschichte der Iranischen Hunnen in Mittelasien", in G. Gnoli ed., La Persia e L'Asia Centrale da Alessandro al X Secolo, Atti dei Convegni Lincei: 127, Rome, 1996, 517-54. Beal, Samuel, “Some Remarks on the Narrative of Fâ-hien”, Journal of the Royal Asiatic Society, 19, 1887, 191-206. 47

Bingenheimer, Marcus, Der Mönchsgelehrte Yinshun (*1906) und seine Bedeutung für den Chinesisch-Taiwanischen Buddhismus im 20. Jahrhundert, Heidelberg, 2004. Bivar, A. D. H., "The absolute Chronology of the Kushano-Sasanian Governor in Central Asia", in J. Harmatta ed., Prolegomena to the Sources on the History of Pre-Islamic Central Asia, Budapest, 1979, 317-34. Bregel, Yuri, An Historical Atlas of Central Asia, Handbuch der Orientalistik: 9, Leiden, 2003. Chen, Sanping, "Son of Heaven and Son of God: Interactions among. Ancient Asiatic Cultures Regarding Sacral Kingship and Theophoric Names", Journal of the Royal Asiatic Society, 12: 3, 2002, 289-325. Corradini, Piero, "The Barbarian States in North China", Central Asiatic Journal, 50: 2, 2006, 163-232. Cunningham, Alexander, The Ancient Geography of India, London, 1871. Daffinà, Paolo, "The Han Shu Hsi Yü Chuan Re-translated: A Review Article", T'oung Pao, 68: 4-5, 1982, 309-39. Dani, A. H, Peshawar. Historic City of the Frontier, Peshawar, 1969. ______, Chilas, the city of Nanga Parvat (Dyamar), Islamabad, 1983. ______, The historic city of Taxila, Tokyo, 1986. ______, "Eastern Kushans and Kidarites in Gandhāra and Kāshmir, and Later Hephthalites", in B. A. Litvinsky ed., History of Civilizations of Central Asia, III, Paris, 1996, 163-83. Dani, A. H. and B. A. Litvinsky, "The Kushano-Sasanian Kingdom", in B. A. Litvinsky ed., History of Civilizations of Central Asia, III, Paris, 1996, 103-18. Enoki, Kazuo, "Sogdiana and the Hsiung-nu", Central Asiatic Journal, 1: 1955, 43-62. ______, "On the Nationality of the Ephthalites", Memoirs of the Research Department of the Toyo Bunko, 18: 1959, 1-58. ______, "On the Date of the Kidarites (1)", Memoirs of the Research Department of the Toyo Bunko, 27: 1969, 1-26. ______, "On the Date of the Kidarites (2)", Memoirs of the Research Department of the Toyo Bunko, 28: 1970, 13-38. ______, "On the Date of the Kidarites", in ed., Studia Asiatica: The collected papers in Western languages of the late Dr. Kazuo Enoki, Tokyo, 1998, 54-106. Falk, Harry, "The Kanishka Era in Gupta Records", Silk Road Art and Archaeology, 10: 2004, 167-76. Foucher, A., "Géographie ancienne du Gandhâra", Bulletin de l'École française d'Extrême-Orient, 1: 1901, 332-69. Frye, N., "Islamic Sources for the Pre-Islamic History of Central Asia", in J. Harmatta ed., Prolegomena to the Sources on the History of Pre-Islamic Central Asia, Budapest, 1979, 221-30. Gnoli, G. "ΒΑΣΙΛΕΥΣ ΒΑΣΙΛΕΩΝ ’ΑΡΙΑΝΩΝ" in G. Gnoli ed., Orientalia Iosephi Tucci Memoriae Dicata, Serie Orientale Roma: 56, Rome, 1987, 509-32. Göbl, Robert, Dokumente zur Geschichte der iranischen Hunnen in Baktrien und Indien, 4 vols., Wiesbaden, 1967. Grenet, Frantz, "Regional interaction in Central Asia and Northwest India in the Kidarite and Hephthalite periods", in Nicholas Sims-Williams ed., Indo-Iranian Languages and Peoples, Proceedings of the British Academy: 116, New York, 2002, 203-24. 48

______, "The self-image of the Sogdians", in E. Trombert and É. de la Vaissière ed., Actes du Colloque «Les Sogdiens en Chine» , Etudes thématiques: 17, Paris, 2005a, 123-40. Harmatta, J., "Late Bactrian Inscriptions", Acta Antiqua Academiae Scientiarum Hungaricae, 17: 1969, 297-432. ______, "Kidara and the Kidarite Huns in Kashmir", Acta Antiqua Academiae Scientiarum Hungaricae, 28: 1980, 185-9. ______, "Religions in the Kushan Empire", in J. Harmatta ed., History of Civilizations of Central Asia, II, Paris, 1992, 313-29. Hulsewé, A. F. P, "The problem of the authenticity of Shih chi 123, the memoir on Ta Yüan", T'oung Pao, 61: 1975, 83-147. Ilyasov, Jangar Ya., "On a number of Central-Asian Tamgas 1", Silk Road Art and Archaeology, 9: 2003, 131-57. Kosambi, D. D, Ancient India, New York, 1965. Kuwayama, Shoshin, "Literary Evidence for Dating the Colossi in Bāmiyān", in G. Gnoli ed., Orientalia Iosephi Tucci Memoriae Dicata, Rome, 1987, 703-27. ______, "The Hephthalites in Tokharistan and Northwest India", Zinbun, 1989, 89-134. ______, "The Buddha's bowl in Gandhāra and Relevant Problems", in Maurizio Taddei ed., South Asian archaeology, 1987: proceedings of the ninth International Conference of the Association of South Asian Archaeologists in Western Europe, Rome, 1990, 945-78. ______, "Historical Notes on Kāpiśī and Kābul in the Sixth-Eighth Centuries", Zinbun, 34: 1, 1999, 24-77. la Vaissière, Étienne de, Sogdian Traders, Handbuch der Orientalistik: 12, Leiden; Boston, 2005a. ______, "Huns et Xiongnu", Central Asiatic Journal, 49: 1, 2005b, 3-26. Lamotte, Étienne, Histoire du bouddhisme indien: des origines à l'ère Śaka, Louvain -la-Neuve, 1976. Lévi, Sylvain, "Devaputra", Journal Asiatique, 224: 1934, 1-22. Li, Jinxiu 李錦繡, "Tongdian Bianfangdian ‘Tuhuoluo’ tiao Shiliao Laiyuan yu Xiyu Tuji" 《通典·邊防典》“吐火羅”條史料來源與《西域圖記》 "The Historical Sources of the Entry Tochari in Tong Dian Bian Fang Dian (Frontier Defense Volume of the "Comprehensive Manual") and Xi Yu Tu Ji (An Illustrated Book of the Western Regions) ", Xiyu Yanjiu 西域研究 (The Western Regions Studies), 2005: 4, 25-34. Majumdar, R. C., The History and Culture of the Indian People: 2, The Age of Imperial Unity, Bombay, 1951. ______, The History and Culture of the Indian People: 3, The Classical Age, Bombay, 1954. Marquart, J., Ērānšahr nach der Geographie des Ps. Moses Xorenac'i mit. historisch-kritischem Kommentar und historischen und topographischen Excursen, Berlin, 1901. Marshak, B. I., “K voprosu o vostochnykh protivnikakh Irana v V v”, Strany i narody Vostoka, 10, 1971, 58-66. ______, "Central Asia from the third to the seventh century", in Annette L. Juliano and Judith A. Lerner eds., Nomads, Traders and Holy Men along China’s Silk Road, Silk Road Studies: 7, Turnhout, 2002, 11-22. Marshall, Sir John Hubert, Taxila: an illustrated account of archaeological excavations carried out at Taxila under the orders of the Government of India between the years 1913 and 49

1934, 3 vols., Cambridge, 1951. Martin, M. F. C., "Coins of Kidāra and the Little Kushāns", Journal of the Royal Asiatic Society of Bengal, Letters, 3: 2, Numismatic Supplement, 1937, 23-50. Pelliot, Paul, "La theorie des quatre Fils du Ciel", T'oung Pao, 22: 1923, 97-125. ______, "Tokharien et Koutchéen", Journal Asiatique, 224: 1934, 23-106. Petech, Luciano, "La «Description des Pays d'Occident» de Che Tao-ngan", in his Selected Papers on Asian History, Series Orientale Roma: 60, Rome,1988a, 231-56. ______, "Alcuni Dati di Chih Sêng-Tsai Sull'India", in his Selected Papers on Asian History, Series Orientale Roma: 60, Rome, 1988b, 313-21. ______, "Indien bis zur Mitte des 6. Jahrhunderts", in G. Mann et al. ed., Propyläen Weltgeschichte, Bd. 2: Hochkulturen des Mittleren und östlichen Asiens, Berlin, 1991, 462ff. Pulleyblank, Edwin G., "A Sogdian Colony in Inner Mongolia", T'oung Pao, 41: 1952, 317-56. ______, "The Consonantal System of Old Chinese 1", Asia Minor, NS, IX: 1, 1962, 58-144. ______, "The Consonantal System of Old Chinese 2", Asia Minor, NS, IX: 2, 1963, 206-65. ______, Lexicon of reconstructed pronunciation in early middle Chinese, late middle Chinese, and early Mandarin, Vancouver, 1991. Rahman, Aman ur, Frantz Grenet and Nicolas Sims-Williams, “A Hunnish Kushan-shah”, Journal of Inner Asia Art and Archaeology, 1, 2006, 125-31. Shiratori, Kurakichi 白鳥庫吉, "Kipingoku kō" 罽賓國考 “A Study on Ki-pin”, 東洋學報 The Toyo Gakuho: Reports of the Investigations of the Oriental Society, 7: 1, 1917, 33-102. ______, "A Study on Su-t'e, or Sogdiana", Memoirs of the Research Department of the Toyo Bunko, 2: 1928, 81-145. Sims-Williams, Nicholas, Bactrian documents from Northern Afghanistan, 1: Legal and economic documents, Corpus Inscriptionum Iranicarum: 6, Oxford, 2000. ______, "Ancient Afghanistan and its invaders: Linguistic evidence from the Bactrian documents and inscriptions", in Nicholas Sims-Williams ed., Indo-Iranian Languages and Peoples, Proceedings of the British Academy: 116, New York, 2002, 225-42. Skjærvø, Prods Oktor, "The Great Seal of Peroz", Studia Iranica, 2003, 281-86. Su, Jinren 蘇晉仁, "Mingseng Zhuan ji Mingseng Zhuan Chao" 《名僧傳》及《名僧傳抄》 “Biographies of Famous Monks and A Copy of Biographies of Famous Monks”, in Zhang Mantao ed., Zhongguo Fojiao Shixueshi Lunji 中國佛教史學史論集 (A Collection of Papers on Chinese Buddhist Historiography), Xiandai Fojiao Xueshu Congkan 現代佛教學術叢刊 (Series of Modern Buddhist Academics): 50, Taibei, 1978, 23-58. Tremblay, Xavier, Pour une histoire de la Sérinde: Le manichéisme parmi les peuples et religions d'Asie Centrale d'après les sources primaires, Veröffentlichungen der Kommission für Iranistik: 28, Vienna, 2001. Tucci, G., "Preliminary Report on an Archaeological Survey in Swat", East and West, 9: 1958, 279-328. Wan, Xiang 萬翔, “E cang Dunhuang Yishu Ф. 209 hao kaoding” 俄藏敦煌遺書 Ф. 209 號 考訂 “Textual analysis of the Dunhuang document Ф. 209 preserved in Russia”, Journal of the Dunhuang and Turfan Studies 敦煌吐魯番研究 11: 2009, 345-68. Xia, Nai 夏鼐, "Zongshu Zhongguo Chutu de Bosi Sashan Chao Yinbi" 綜述中國出土的波 斯薩珊朝銀幣 “A survey of Sassanian silver coins found in China”, Kaogu Xuebao 考古學報 50

(The Chinese Journal of Archaeology), 40: 1974, 91-110, 2pts. Yinshun 釋印順, Fojiao shidi kaolun 佛教史地考論 (Studies on the history and geography of Buddhism), Xinzhu, 2000. Yoshida, Yutaka, "On the Origin of the Sogdian Surname Zhaowu 昭武 and Related Problems", Journal Asiatique, 291: 1-2, 2003, 35-67. Yu, Taishan 余太山, Yeda Shi Yanjiu 嚈噠史研究 (A study on the Hephthalites), Jinan, 1986. ______, "Yeda Shi Ruogan Wenti de zai Yanjiu" 嚈噠史若干問題的再研究

“A re

-examination on some points of the history of the Hephthalites”, Zhongguo Shehui Kexueyuan Lishi Yanjiusuo Jikan 中 國 社 會 科 學 院 歷 史 研 究 所 集 刊 (Journal of Institute of History, Chinese Academy of Social Sciences), 1: 2001, 180-210. Zeimal, E. V., "The Kidarite Kingdom in Central Asia", in B. A. Litvinsky ed., History of Civilizations of Central Asia, III, Paris, 1996, 119-33. Zheng, Binglin 鄭炳林 and Xu Xiaoli 徐曉麗, "E cang Yijing Xifangji yanjiu" 俄藏義淨 《西方記》研究 “A study on Western Record by Yijing”, Universal Gate Buddhist Journal 普門 學報, 10: 2002, 89-120. Zürcher, Erik, The Buddhist Conquest of China: the spread and adaptation of Buddhism in early medieval China, 2nd edition, 2 vols., Leiden, 1972. Web pages: (All from Encyclopædia Iranica online edition 2006, http://www.iranica.com/, ed. by Ehsan Yarshater, last access: Mar. 10th, 2007.) Alram, Michael, Hunnic Coinage, date: Aug 9, 2004. http://www.iranica.com/newsite/articles/v12f6/v12f6007.html Bivar, A. D. H., Hephthalites, undated (2006). http://www.iranica.com/articles/v12f2/v12f2036.html Grenet, Frantz, Kidarites, date: July 29, 2005b. http://www.iranica.com/newsite/articles/ot_grp8/ot_kidarites_20050729.html

51