The Spread of Buddhism

the spread of buddhism in serindia 109

any case, in 808 a coup d’état gave the throne to a new, Manichaean,

dynasty, the Ädiz. The Manichaean Uighur texts are mostly written

in the same dialect as the Türk runic inscriptions discovered along the

Orkhon river in Mongolia: it is called the Türk or n-dialect.^169 The

Uighur court abode, for prestige reasons, to the Türk dialect of its

imperial predecessors and erected runic inscriptions in Türk dialect in

Tariat, ine-Usu, Tes or Qarabalasun. Buddhist texts in Türk dialect,

or at least with Türk dialectal colouring, are scarce, but some of them

have been identi ed.^170 Even the most archaic texts^171 have been trans-

lated from Chinese originals long after the First Türk Empire. They

must go back to the Uighur Empire, and have to be dated between 750

and 1000. No early Türk Buddhist text has thus been preserved—nor

probably did one ever exist, since Turkic languages were unwritten

until ca. 720.

Under the Manichaean Uighur Khans, Buddhist texts were translated

into Turkic approximately as much from Tocharian as from Chinese or

Sogdian. However, in the most archaic texts the Sogdian in uence is

at its highest. The orthography often follows Sogdian rules rather than

the later Uighur ones, and in any case the Uighur alphabet is but an

adaptation of the Sogdian one. Buddhist Uighur manuscripts in Brhm

are not earlier than the eleventh century and preserve no early linguistic

features: they do not set forth the Tocharian literary. Also, the Buddhist

basic vocabulary (i.e., the non-technical terms) is borrowed mostly from

Sogdian.^172 This vocabulary is stable at all epochs. For the Buddhist tech-

nical vocabulary, the situation is different. The early texts indeed contain

more Sogdian loanwords or Indic loanwords than the later manuscripts,

but they still contain many more Tocharian loanwords.^173

(^169) The n-dialect writes with Runic ñ, Manichaean n, a sound which evolved later
in Uighur y. Cf. von Gabain 1950, p. 5.
(^170) Cf. von Gabain 1976; Maue & Röhrborn 1984–1985, vol. 2, p. 77; Laut 1986,
p. 11; Moriyasu 1990, pp. 150f.
(^171) See, for instance, the Säkiz Yükmäk (TTT VI; Elverskog no. 55), translated from
a Chinese apocryphal stra composed under Empress Wu (690–705), or the Buddha
Biography T II Y 21 + 32 (Elverskog, no. 14), probably translated (maybe via Sogdian)
from the Chinese version T.187 of the Lalitavistara, itself translated in 683 (Müller,
Uigurica 172 II no. 1, 4–7).
For a list, see Asmussen 1965, pp. 145f.; Laut 1986, pp. 143–148. Examples are
pwdyst(w) “Bodhisattva”, or ’zrw’ “Brahma” < Sogd. (’)zrw’ ‘Zurvan’ following the old
Mazdean interpretatio indica. But note a contrario Uigh. knty “forgiveness” < Tocharian
k 173 nti, or Uigh. tnk < Khot. thaga “tax”.
Laut 1986, pp. 93–142 (who rather overemphasises the Sogdian in uence)  nds
no more than 31 Indic loanwords that entered through Sogdian against ca. 350 which
Heirman_f5new_75-129.indd 109 3/13/2007 1:15:57 PM