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

319. QUENGIANFU

*chandianfo* V *guengyanfu, guengyenfu* FB *qengunfu* Ft
*chanziafu* VB *guenyasu* LTr *quegianfu* F, TA³
*gemgunphu* VL *gyanfu* P *quengianfu* F, FA, L, TA¹
*gianfu* Pr *margara* VA *quengienfu* FAt
*guanciansu, guengiasu* LT *qençanfu, quençanfu* Z *quenzanfu* R
*guengianfu* FBt *qengiufu* Fr

There is no doubt that Polo applies this name to Hsi-an-fu, the metropolis of Shàn-hsi.
Already in 1897, DEVÉRIA called attention to the fact that a similar form occurred in the Persian
vocabulary of the College of Translators of the Ming dynasty. Polo follows here, as usual, the topo-
nymy then current among Persian-speaking people. DEVÉRIA's paper has already been quoted
by CORDIER (Y, II, 29; III, 77-78), but the facts have not been presented quite accurately. In the
Persian vocabulary of the College of Translators, کنجانفو Kinjanfu, with the phonetic transcription
in Chinese 欽 張 夫 Ch'in-chang-fu (金 *chin* in *BEFEO*, IV, 771, is a misprint, which has passed
since into *JA*, 1912, I, 594), has for Chinese equivalent not 京 兆 府 Ching-chao-fu, as has been
said, but 陝 西 Shàn-hsi, that is, the name of Shàn-hsi province. The same equivalence, Shàn-
hsi, occurs in the Turkish vocabulary of the Ming period, written only in Chinese characters, which
is in the Library of the School of Oriental Studies in London (in the series « *Kō Kwō Yī Yü* »).
In this work, Shàn-hsi is rendered by 勤 昌 府 Ch'in-ch'ang-fu, seemingly *Kinčangfu, which
would probably also be Kinjanfu if we had it in Arabic writing (the transcriptions of the vocabulary
are not very strict). The name appears in Rašidu-'d-Dīn, written once کن جانفو Kin-janfu (*Bl*, II,
495), but elsewhere کنجانفو Kinjanfu (*Bl*, II, 598; also in Rašīd's Ms. History of China; the
form « Kenjangfu » of *Y*¹, 127, 128, is only due to an arbitrary transcription by KLAPROTH, who
himself gives the name in Arabic letters without *-g-*). In BEREZIN's translation (*Ber*, III, 28),
Rašidu-'d-Dīn speaks of Hsi-an-fu as « Čzin-čžao », *i. e.* Chíng-chao, but this is a « learned » correc-
tion due to the editor and translator. Instead of کنک جاو King-Jayu, adopted in the Persian
text (p. 46), the various readings clearly establish that we must adopt کنک جانفو King-Janfu.
Kinjanfu is also mentioned in the *Zafar-nāmāh* (cf. *Not. et Extr.* XIV, 500; *Y*¹, I, 175), and was
still used about 1545 by the author of *Ta'rīḥ-i-Rašīdī* (transl. ELIAS and ROSS, 404); we find it *in
situ* in a Sino-Arabic inscription of 1545 (*TP*, 1905, 279, 284; instead of کنجانفوی Kin,āfafūyī,
read کنجانفوی Kinjanfūyī, an ethnic derivative of Kinjanfu; HUART wrongly thought of Kan-
chou). In the fantastic itinerary of Ibn Baṭṭūṭah across China, قنجنفو Qinjanfu probably also
represents Kinjanfu (cf. *Fe*, 428, where FERRAND is misled by DULAURIER into believing that
Polo's « Quengianfu » is Chên-chiang on the Yang-tzŭ, for which see « Cinghianfu »).
YULE (*Y*¹, II, 246) has taken it for granted that it was Kinjanfu which appeared in Odoric