90 xavier tremblayto set foot in Sogdiana. This attempt has left two traces: (1) the name
of the city of Bukhara was slightly modi ed to resemble rx’r [rxr]
“vihra”.^68 (2) The sole mention of Buddhists in Sogdiana is to be found
in the biography of Xuanzang: two novices were attacked by re when
they attempted to worship Buddha in an empty monastery!^69
It is therefore all the more amazing that art iconographers claim that
Buddhism pros pe red in Sogdiana at least up to the seventh century.
The most positive statement is Marshak’s (apud Azarpay 1981, p. 28):
“Buddhism, which obtained wide recognition, penetrated Sog diana in
the rst centuries AD. Buddhism took such rm root in Sogdiana that
a number of Buddhist terms entered the Sogdian language and were
later used by the Manichaeans in their own religious texts”. Five argu-
ments have been adduced in favour of such a statement:- Some iconographic motives in Sogdian painting have been con-
nected to Buddhist representations: for instance, the riding kings on
temple II at PenÌikent, or the domed structures depicted in some
paintings.^70 But even if the connexion stands scrutiny, we should not
forget that the Sogdian iconography was deeply renewed during the
Hephthalite period and was bound to borrow from Bactria. Roman,
Byzantine or Mesopotamian in uences can be traced as well.^71 An
iconographic motif proves nothing of religious af liation. - Mazdean customs bear traces of Indian in uence: umbrella-shaped
altars with hanging bells,^72 four-armed deities, systematic interpretatio indica
of the gods through Indian counterparts. But the attribution of these
unquestionably Indian in uences to Sogdian Buddhism comes short
by two facts. Firstly, the interpretationes are already present on Kuaa
coins and betray contacts between Mazdeism and Buddhism in Bactria
rather than in Sogdiana. Secondly, an Indian in uence does by no
means imply a Buddhist in uence: a Hindu one is more appropriate to
Mazdeism. The fact that the Sogdian interpreter of the Buddhist text
Vessantara Jtaka could mechanically replace Indra, iva and Brahma
of his Vo rl a g e with the Mazdean gods Ahura Mazd, Vaiiu and Zurvan
(^68) Tremblay 2003, pp. 122f.
(^69) T.2053.50.227c10–22, tr. Li 1995, pp. 44–45.
(^70) Naymark 2001, pp. 313–322, esp. 320. Temple II was even misinterpreted as a
Buddhist temple (Verardi 1982, pp. 275–280). 71
Naymark 2001, esp. pp. 79–80 (Lupa romana); pp. 345–388 (Joseph re-used as a
Mazdean god). 72
Grenet 1994, p. 48 and g. 15.
Heirman_f5new_75-129.indd 90 3/13/2007 1:15:54 PM