Chinese etymologies of name.
48 HISTORICAL NOTICES OF KASHGAR [Chap. III
concerning Kashgar. From the time of the Former Han Dynasty, when the states of Central Asia were first opened up to the political influence of China, down to the Tang period the
region of the present Kashgar was generally known by the name of a, read Su-/g or
Shu-lê (according to Wade's transcription) 2. The name Sha-lê id) J is given to Kashgar in the accounts of the journeys of the pilgrims Sung Yün, Kumarajiva (circa 400 A.D.), Fa-yung (420 A.D.), Dharmagupta (circa 593-595 A. D.) and Wu-kung; the latter's itinerary distinctly records the identical application of Sha-lê and Su-/g3.
In Hsüan-tsang's Hsi-pi-chi, and in a passage of the Tang Annals evidently reproduced from it, we meet with the name Chia-sha ft 0,, (transcribed Kia-cha by Julien, K`iu-cha by M. Chavannes, and ICa-sha by Dr. Franke), in which it is impossible to mistake the phonetic rendering of some form connected with the present Kashgar 4. The name of the town Chia-shih p of-j, transcribed Kia-che by M. Chavannes and Ka-shih by Dr. Franke, which the same passage of the Tang Annals refers to as the residence of the king of Kashgar, is evidently also closely related in origin 6. Earlier attempts to reproduce that old local name may be recognized in the Ch`i-sha
*•' of the pilgrim Chih-mêng (circa 404 A. D.), and in Fa-hsien's Chieh-ch`a , which, by
M. Chavannes' brilliant and convincing analysis of the itineraries of these two pilgrims, have been shown to be identical with Kashgar 6.
The form Kashgar itself is attested in Muhammadan sources from the earliest period when the Arab conquest reached this part of Central Asia 7. Spelt and pronounced in a variety of fashions (as Kashghar, Kashgar, Qashgar, &c.) in accordance with the phonetic latitude allowed by Turki languages in respect of certain consonants 8, this name has continued in sole use ever since Eastern Turkestan was lost to the Tang dynasty at the close of the eighth century. Chinese records subsequent to that period invariably designate the territory and town by transcriptions of the name Kashgar, though its identity with the earlier Su-16 was never lost sight of by the learned in China 9.
That the several names of so important a territory should have formed the subject of learned etymologies and conjectures both in the East and in the West can scarcely surprise us. Two early attempts to account for the name Su-16 (or Shu-lê) are found in Chinese Buddhist texts, and have recently been discussed at considerable length and with much critical care by
2 The various Chinese designations of Kashgar have been critically discussed in Dr. O. Franke's paper ' Kaschgar and die Kharosthi,' Silzungsber. der kön. preuss. Akad. der Wissenschafien, Berlin, 1903, pp. 184 sqq.
The form Su-k. or Shu-lé is found also in an interesting notice of the three great routes connecting China with the West, which goes back to the information collected by Pei Chu (circ. 605-606 A. D.) and which Baron Richthofen, China, i. p. 53o, note, has extracted. The form Liu-la, given in this extract as the name of Kâshgar, is, according to the information kindly supplied to me by M. Chavannes, based upon a misreading of the usual 4 it.
Compare Chavannes, Voyage de Song Yun, pp. 23, 54, 57, 62, and L'Ilinéraire d'Ou-kong, p. 26.
4 Compare Julien, Mimoires, ii. pp. 219, 509 ; Chavannes, Turcs occid., pp. 121, 339 (where the pronunciation ria-cha (Ch'ia-sha) is quoted from K'ang-hsi's Dictionary); Franke, Sb.P.A. W., 1903, p. 186.
6 See Chavannes, Turcs occid., p. 12i ; Franke, Sb.P.A. W.,
1903, p. 186..
6 Compare Chavannes, Voyage de Song Yun, pp. 54 sq. also below, chap. rrr. sec. iv.
7 The capture of Kashgar by the Arab general Qotayba, in 715-716 A. D., is recorded by Tabari (tenth century), and references to Kâshgar are found in other Muhammadan authors from the tenth century onwards ; compare Bretschneider, Mediaeval Researches, ii. pp. 45 sq.
8 For these variations of spelling compare Shaw, ' Vocabulary of the Turki language,' in J.A.S.B., Extra No. 1878, p. 164 ; also above, note 1.
g See for such transcriptions (ro-shih-ha-li in Yuan shih of 1369 A. D., Ha-shih-ha-rh in Ming Annals of seventeenth century), Bretschneider, Med. Researches, ii. pp. 148, 245; also the modern Chinese records extracted in Ritter, Asien,
v pp. 409 sqq.