国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

the Supiye, Supiya of Kharoṣṭhī documents of about the 3rd cent. A. D. (cf. Boyer, Rapson, and
Senart, Kharoṣṭhī Inscriptions, 377-378, s. v. supiya, supiyade, supiyana, supiye, supiye-
hi); three times (Nos. 119, 324, 722), the Supiya are mentioned alongside of Calmadāna, i. e. Cher-
chen (see « Ciarcian »), and the suvarnākara pa[rvati]yana of No. 578, mentioned after supiyana,
may perhaps be a suvarṇākara parvata, « Gold Mine Mountain », the 'Gold Mountain' of the 'Gold
Race'. As also noticed by Thomas, the same form Supiya occurs in Khotanese (the so-called
'Saka'); cf. Leumann, Nordar. Lehrged., pp. 208, 513; Sten Konow, Saka Studies, 183. The very
form So-byi, identical with Ch. *Suo-b'ji, is used in the Tibetan version of the Prophecy of the Li
country (Thomas, 78), probably made from the Khotanese. So we may hold it quite probable that
such was the Khotanese name of the people whom the Tibetans knew as Sum-pa; the only qualifica-
tion to that probably correct inference being that there are many names and terms in the said Pro-
phecy which go back to Chinese originals, so that there is a remote possibility that the hitherto unique
mention of So-byi may be based on Ch. Su-p'i (*Suo-b'ji) itself. The name Sum-pa occurs many
times in various Tibetan texts, mostly without any epithet, but, in one paragraph of the Inquiry
of Vimalaprabhā, as Ba-dag (or Ba-bdag) Sum-pa (Thomas, 241-243), with a probable allusion to
another name Ba-laú Sum-pa.
While we must be very grateful to Thomas for the many texts he has rendered accessible,
I must say quite frankly that I dissent from the identifications he has proposed. According to Tho-
mas, the Gold Country, or Country of the Gold Race, would be the region of Hunza and Nagar in the
Kanjut (Kunjud) Valley, north of Gilgit, and the Gold Mountain might be the great Rakapushi
itself (pp. 153-156, 165-166; but from the names possibly formed with -dkar, -gar, -sgar, suppress that
of Pho-dkar, which is a late transcription of *Boḫar, Bokhara; see «Bucara»). The Supiya, Su-p'i or
Sum-pa would be originally Hsiung-nu (Hūna, Huns), of Sien-pi origin, and the Sum-pa of Tibetan
texts of the 8th-10th cents. would be «really Hūnas or quasi-Hūnas», Su-p'i and Sum-pa being fun-
damentally the same name. As these Hunnic Sum-pa were great marauders, their name
remained attached to other tribes which used to make incursions upon Khotan, such as the Qarluq
Turks. As to the Ba-dag Sum-pa, they would be people of Badaḫšān, who, on account of their
marauding habits, may «have a real identity of name with the Tibetan Sum-pas, who were actually,
like the Tu-yuk-hun [T'u-yü-hun], of Sien-pi origin» (Thomas, pp. 9-10, 156-159).
Although many points are still far from being elucidated, the ethnical and geographical data of
the Tibetan and Chinese texts are often clear enough, and, in my opinion, cannot be reconciled with
the views expressed by Thomas. Both series of sources are in closer agreement than has been
hitherto admitted. The prophecies are concerned with the decline and fall of Buddhism in Central
Asia, and that of Saṅghavardhana says (Thomas, p. 61) that «the monks of 'An-ce, Gus-tig, Par-
mkhan, and Su-lig, after great sufferings, will go to the Bru-ša (= Bru-ža) land (= Baltistan). Also
the monks of the Tho-kar (= Tokharestan, Tūḫaristān) and of Kashmir, having been vexed by un-
believing people, will give up and go to the Bru-ša country». This retrospective prophecy refers in
fact to the advance of the Mussulmans. «Su-lig» is admittedly Kāšyar, and this Tibetan form is a
rendering of the Chinese name Shu-lo (*Ṣi̯*o-lak; see «Cascar»). Instead of 'An-ce (c = ts), the Pro-
phecy of the Li country more correctly gives 'An-se (Thomas, pp. 77, 78, 82); Thomas rejects Rock-
hill's identification with 安西 An-hsi, because he believes this An-hsi to be the modern An-hsi