The Spread of Buddhism

104 xavier tremblay

for instance by a king’s daughter, who was probably Kumraj va’s

niece. Kumraj va himself translated some texts of this tradition.

With the sixth century,^140 all paintings or Chinese reports pointing to

Mahynist beliefs wane progressively. By the time of the Chinese pil-

grim Xuanzang (602–664), Kucha seems to have become a stronghold

of the (Mla)Sarvstivdins.^141 The Kuchean manuscripts (5th–9th

century?) all belong to this school—they contain no clear trace of a

former religion.^142 The reasons for this reorientation remain obscure.

3.4. Turfan before 791

Of all the Tarim Basin cities Turfan was the most cosmopolitan:

the Chinese Former Han dynasty had already disputed the Jushi

kingdom with the Xiongnu since 73 BC, when they divided the

kingdom into an “Anterior Jushi” with capital Yar-khoto near Turfan,

under Chinese in uence, and a “Posterior” kingdom in Guen under

Xiongnu overlordship;^143 already in 313 Turfan was for the Sogdians

the “Chinese town”. However, a permanent Chinese dominance upon

Turfan was not achieved before 327, when the Former Liang

established a command post,^144 and they seem not to have ousted the

local dynasty before 448.^145 In 498, The Ruanruan enfeoffed the house

(^140) Earlier paintings, such as the paintings in the Qizil Cave 83 (maybe ca. 450 AD)
portray Maitreya with other bodhisattvas and miracles performed by Buddha after his
death (cf., e.g., Yaldiz 1987, pp. 90–93 and Plate 56).
(^141) T.2087.51.870a25 (tr. Beal, 1884, book 1, p. 19); Waldschmidt 1926, p. 76; Schmidt
1985, with further reading. The distinction between Sarvstivda and Mlasarvstivda
is largely illusory (Iwamoto 1964; Enomoto 1984). See also Schmithausen 1970,
p. 112 n. 257 [manuscript from Gilgit]; von Hinüber 1985, p. 71; Pinault 1989, p. 13;
Schmidt 1998, p. 76; 1999, p. 100. 142
The designation for “god” (A ñkät, B ñakte < *ni-ghuH-tó-, cf. ved. h- “to invoke a
god during sacri ce”, engl. God, German Gott, and so on, cf. Normier 1980, pp. 267ff.)
betrays nothing more than that the Tocharian Pre-Buddhistic religion had probably
Indo-European roots—not a highly informative  nding. See further Pinault 1998a, pp.
358f. Traces of Mazdeism (maybe imported from Sogdiana or Bactriana) have been
advocated (Liu Mau-Tsai 1969, p. 33; Tremblay 2001, p. 160 n. 261, with reservations;
contra Pinault 1998b, p. 16).
(^143) Shimazaki 1969, pp. 36, 50–54.
(^144) Pelliot 2003, p. 29 n. 100.
(^145) Zhang Guang-da 1996, pp. 303f. The beginning of Tocharian literacy in
Kharo
h and later in Brhm in Turfan might predate the Former Liang dynasty.
Some Kharo
h wood-tablets probably in Tocharian have been found in Tumuq near
Turfan. Tocharian A seems to have been a dead language in Turfan at the time of
copying the preserved manuscripts (Winter 1955; 1961, p. 280; 1963, p. 245; 1984,
pp. 16–17; Tremblay 2001, p. 36 n. 55, pp. 37–44).
Heirman_f5new_75-129.indd 104 3/13/2007 1:15:56 PM