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

(*Sjän-pjie; though we have long been accustomed to say « Hsien-pi », Hsien-pei is the normal
pronunciation of the characters) were an Altaic race, which is often said to be Tungus, but which
I consider as Mongol, particularly inasmuch as the T'u-yü-hun are said to be of Hsien-pei descent
(cf. TP, 1921, 325-326). There is no reason, however, to connect that ancient name of tribes living
beyond the borders of north-eastern China with the Sum-pa who invaded Khotan. Whether Sum-pa
be only a Tibetan form of Su-p'i, Supiya, etc., or whether it be a different name, of purely Tibetan
origin, applied by the Tibetans (because of phonetic analogy?) to the Su-p'i, one fact is quite certain:
apart from the prophecies, Sum-pa actually exists in Tibetan geographical nomenclature, with a
well defined value; it is the name of a district in Amdo (cf. Sarat CHANDRA DAS, Dictionary, 1272).
Although there is an Amdo in Central Tibet, south-west of the celebrated Daṅ-la Pass (cf. S. HEDIN,
Southern Tibet, Maps, Sheet XII, D 7/8), the Tibetan spelling of which escapes me, there can be
no doubt that Sarat CHANDRA DAS' Amdo is the one commonly known, Tib. mDo (cf. his Dictionary,
675), sometimes marked on our maps as « Amdoa » (Tib. mDo-wa, « inhabitants of mDo »), i. e
the region south and south-east of the Kökö-nör. In modern times, the polygraph Ye-šes dpal-
'byor, born in Amdo, is often designated as Sum-pa mkhan-po, « the Sum-pa master » or « the
Sum-pa abbot ». Unfortunately, I do not find Sum-pa in the maps at my disposal. For earlier
times, THOMAS has already noticed (p. 156) that a Sum-pa man was mentioned in a Tibetan document
from Chinese Turkestan, and that the name of the Sum-pa also occurred in the Padma-than-yig.
In the Tanjur, two works have been translated with the collaboration of monks from Sum-pa, and
Mss. from Sum-pa have been used to establish the text of a third one (cf. P. CORDIER, Index, III,
78 [No. 397], 455 [No. 26], 500). But the location of Sum-pa south-east of the Kökö-nör confronts
us with a fresh difficulty: it agrees neither with a probable origin for men who invaded Khotan
in what is supposed to be middle T'ang times, nor with the location of the kingdom of Su-p'i, alias
Sun-po (Sum-pa), in Chinese texts of the same period. I see no certain solution to proffer. It
may be that the Su-p'i (Supiya) once held sway over almost the whole of northern Tibet from south
of the Kökö-nör to the passes which led north-west to Khotan, and that the Tibetan conquest reduced
them to a much more restricted area, which is represented by the kingdom of Su-p'i of the texts
of the T'ang period. This earlier great Su-p'i kingdom would be the same not only as the 'Kingdom
of Women', « south of the Onion Range », which has a special notice in the Sui shu, but also as
the 'Kingdom of Women' which the same text occasionally mentions west of Fu-kuo, and,
at a still earlier date, as the 'Kingdom of the Woman King' or 'Kingdom of Women' of the
Wei shu. The Sum-pa district, south of the Kökö-nör, would owe its name either to Su-p'i tribes
left behind when the restricted kingdom of Su-p'i was established under Tibetan suzerainty, or
to Su-p'i people who were transferred there at the time of the Tibetan occupation in the 8th and 9th
centuries. The first hypothesis would be more probable, if we should identify with these Su-p'i
(Sum-pa) south-east of the Kökö-nör the 'Eastern Kingdom of Women' of the T'ang histories.
Hsüan-tsang, writing in the first half of the 7th cent., would still have had an echo of the former
and more extensive 'Kingdom of Women', and the error of the Hsin T'ang shu would be to
ascribe to the restricted 'Eastern Kingdom of Women' the western limits assigned by Hsüan-
tsang to the great Suvarṇagotra. That Suvarṇagotra, Hsüan-tsang's 'Eastern Kingdom of Women',
cannot be looked for, as THOMAS would wish, in the Hunza-Nagar district is sufficiently established,