322 HISTORICAL NOTICES OF LOP, SHAN-SHAN, AND LOU-LAN [Chap. IX
Here it must suffice to point out that they all belong to the period of Tibetan domination in Chinese Turkestan, which extended from approximately the last third of the eighth century to about the latter half of the ninth century A. D.
In these documents the local name of Nob occurs with particular frequency, and since the context invariably shows that it must apply to localities identical with, or comprised in the same territory as, Miran, we may recognize in it the long-sought phonetic link between the form which Hsüan-tsang's Na fu po was intended to reproduce and the Loft of Marco Polo.28 As I am neither a Sinologist nor a Tibetan scholar, I must hesitate to express any opinion as to whether the initial I of the mediaeval and modern name, or the initial n of its Chinese and Tibetan representative, indicates more correctly the original local pronunciation. But I may at least point out that whereas the change of initial Skr. and Pkr. l into n is common in a number of Modern Inclo-Aryan vernaculars,29 the opposite conversion of initial n > l is familiar to me in at least one dialectic pronunciation of the Chinese ` Mandarin ' language.3°
For a synopsis of the Tibetan records in which Nob is mentioned, I must refer to section v of chapter xII, which deals with the excavations in the ruined fort of Miran. To judge from the abstracts of documents which Dr. Francke has made available in translation, the most frequent references are to two localities, designated as ' Great Nob ' (Nob-ched) and ` Little Nob ' (Nob-chung) respectively. From a number of indications which the mention of the local name in addresses or the character of the context supplies, but which cannot here be discussed in detail, I conclude that the site of Miran, at which the documents were discovered, is meant by ' Little Nob '.31 If this assumption is right, it is probable that ` Great Nob' must be identified with Charkhlik.
This distinction between ` Little ' and ` Great Nob ' corresponds to that which earlier texts indicate between the two main places of Shan-shan, Yii-ni, ` the Old Town ', to the east, and 1-hsün, or ' the New Town ', these two being now represented, as I believe, by the sites of Miran and Charkhlik respectively.32 In keeping with the character of the Tibetan occupation is the repeated reference made to several castles in Nob.33 The mention once made of ` Lower Nob' shows that subdivisions other than ' Greaf and Little Nob ' were also included in the district, just as in modern times the term Lop in general use comprises the whole region adjoining the terminal course of the Tarim and bordered by the Kuruk-tagh and Altin-tagh ranges.
SECTION II.—SHAN-SHAN BETWEEN TANG AND HAN TIMES
We may now resume the task of tracing back the history of the Lop region in the records available for the periods which preceded the re-establishment of Chinese control over the Tarim Basin in Tang times. Our task is considerably facilitated by the fact that the Lop-nor tract lies on the oldest route connecting China with Central Asia, and that this route retained its importance for traffic even during those times when Chinese political power had ceased to assert itself in the ` Western Regions'.
28 For the first record of this view, see Desert Cathay, i P. 449. The identifications, however, there proposed for Great Nob ' and ' Little Nob ' were erroneous, having suggested themselves at a time when only a small portion of the materials was accessible in translation.
29 Cf. Grierson, Phonology of the Modern Indo-Aryan Vernaculars, Z.D.M.G., ii, Bd. 5o, 1916, p. 15. [For Chinese renderings of 1 by n, see now Pelliot, J. Asiat., 1916, janvier-février, p. I x9, note ; also J. Asiat., 1914, sept.—oct.,
x The Hunanese accent, to which I grew accustomed in colloquial intercourse with Chinese officials and literati in Chinese Turkestan (cf. Desert Cathay, i. p. 143), seemed invariably to turn nan into lan, ring into ling, etc. I observed much the same in the rustic pronunciation of the districts which I visited in Kan-su.
S1 Cf. below, p. 467.
s2 See below, pp. 326 sq.
33 Cf. documents M.1. iii. 7 ; iv. 132 ; viii. I o ; x. 9 ; also below, chap. x11, sec. v.
Tibetan Nob in Miran records.
' Great Nob' and ' Little Nob'.
Great and Little Nob' correspond to Charkhlik and Miran.