154 HISTORICAL NOTICES OF KHOTAN
Early use of name
name. Apart from the inherent improbability of a territory so far removed from the ethnic frontiers of India having first been named in Sanskrit, there is the significant fact that the word ku in the meaning of ` earth ', which the above etymology assumes, is an artificial creation of Sanskrit lexicography, unknown to the genuine old language. That this learned Sanskritization of the local name is, however, far older than Hsüan-tsang's time, can now be shown conclusively from the Kharosthi documents of the third century A. D. discovered by me at the Niya Site. These, in their Prakrit queerly mixed with Sanskrit phrases, actually employ the form Kustana or Kustanaka, by the side of the far more frequent form Khotamna 13.
The latter form, together with its variants Khotana and Khodana, is of special interest 14 ; for it plainly attests the fact that the name of Khotan, whatever its origin, had already in the first centuries of our era assumed a phonetic form closely approaching to, or practically identical with, the present one. This enables us to account without any difficulty for all the varying reproductions of the local name recorded by Hsüan-tsang. Taking first the form Ch`ü-tan, the use of which is ascribed to the Hindus, it is evident that it transcribes a name which must have sounded very much like the present Khotan ; for the syllable Kho we must, in accordance with Hsüan-tsang's transcriptional system, expect a Chinese character sounding ch`ü 15. In Huan-na, which Hsüan-tsang mentions as the form used in current speech, we may, perhaps, recognize a local pronunciation of the name, in which the initial aspirate was reduced to a spirant and the t between the two vowels o and a reduced to an hiatus, as it would have been by regular phonetic change in a Prâkritic language. The form Khodana, attested in the Kharosthi documents, would seem to indicate the first step in this process. Huo-tan is described as the form used by the Hu, i. e. the people of Central Asia whom the Chinese in the time of Hsüan-tsang seem to have comprehended under this general term 16. As Chinese characters sounding huo are used for transcribing a variety of syllables, such as kha, kha, ha, hu, ô, &c.17, it is difficult to indicate the pronunciation really meant. But it is clear that it could only have been a phonetic variation of the form Khotan.
Hsüan-tsang in his list mentions last the name Yü-tien which, he says, was used
formerly. But in reality this is the name by which Khotan is invariably designated in all the Chinese dynastic histories from the period of the Former Hans down to that of the Mings. It appears never to have quite gone out of official use even after the reconquest of Eastern Turkestan by the Manchu dynasty in the eighteenth century 18. That Yü-t`ien was intended as a phonetic
18 I owe the information as to the occurrence of the form Kustanaka, to Prof. E. J. Rapson.
14 For specimens of passages containing the forms Khotamna and Khodana, compare tablets N. i. x 04 + x 06 ; N. xv. 3X8 ; N. iv. 108, tentatively transcribed and translated by Prof. Rapson in his Specimens of Kharosthi inscriptions presented to the Fourteenth Oriental Congress, 1905.
'b Owing to the great rarity of Sanskrit words containing the Aksara kho, Julien's list of Chinese transcriptions gives no instance of ch'i1(k'iu) or chü (kiu) as the phonetic rendering of kho; but examples of ch'ü (k'iu) to represent Skr. go, gho, ku, &c. are frequent ; see Méthode pour transcrire les mots sanscrits, pp. 130 sq. The regular transcription of Skr. kha by kia presents an exact parallel; see ibid., p. 122.
16 As to the varying application of the term Hu in Chinese literature, comp. Lévi, Notes chinoises sur l'Inde, iv.
PP• 43 sqq•
17 Compare Julien, Me'thode pour transcrire, &c., pp. 108 sqq. (Julien's ho represents the sound transcribed by M. Lévi Houo, in Wade's system Huo).
18 In all extracts from the Annals collected in the Pien i tien, as translated by Rémusat, the name Yü-t'ien is used, down to a notice relating to 1573-1619 A. D.; see Ville de Khotan, p. 107 ; also Bretschneider, Mediaeval Researches, pp. 47 sq. From the latter authority it is seen that during the Mongol period (x280-1368 A. n.) the name Yü-t`ien remained quite familiar, though the Chinese historians of the period ` generally try to render the name Khotan, which was in use with the Mongols'. Such renderings are the forms 0-duan, Wu-duan, Wa-duan, Hu-tan quoted by Bretschneider. The first three indicate a pronunciation resembling the Tibetan U-then, for which see below.
The Hsi yü wen chien lu, of 1777 A. D. gives the name as Khotian, but mentions Yü-t'ien also; see Klaproth,