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

the only available. On the other hand, taking into account the fact that, on their way from China
to the west, and before reaching India, the Yüeh-chih or Kuṣaṇa had been in close touch with
Iranian nations, Lévi thinks that *t'ien-tzŭ* first became *baγpuhr*, and that «it is probably through the
intermediary of *baγpuhr* that *t'ien-tzŭ* became *devaputra*» (p. 18-19). But, in such a case, it is
not the meaning of *t'ien* which ought to be compared with that of *deva*, but the meaning of *baγ*.
Whatever the case may be, the equivalence, direct or indirect, of *t'ien-tzŭ* with *devaputra* was so
well felt that *devaputra* was used later in Indian Buddhist circles as the designation of the Chinese
Emperor, in the same way as *baγpār*, *baγbūr* and *faγfūr* in Iranian and Arabic countries (cf.
Chavannes, *Relig. éminents*, 56, 82; Takakusu, *A Record of the Buddhist Religion*, 136; Lévi,
*loc. cit.*, 17). About the year 1000 there reigned a king in Ladakh who had the title of Lhazi-bu
(= *devaputra*); cf. A. H. Francke, *Antiq. of Indian Tibet*, 1, 41.

Recently, Mrs. Bazin-Foucher (*JA*, 1938, 504) has taken exception to the theory which ascribes
a Chinese origin to the royal title *devaputra* of the Kuṣaṇa. Starting from the «divine character»
of the Seleucid kings, and from the ancient «monarchic cult» inherited from the Achaemenids by
the Arsacids and Sassanids, Mrs. Bazin-Foucher concludes : «So it seems to me purposeless to
look in China, as has been sometimes attempted, for the origin of Skr. *devaputra*.» The Yüeh-chih
did not have to bring this title with them from the borders of Kan-su : they found this royal epithet
already in common use in Iran with exactly the same meaning, in Greek as Theopatôr or in Parthian
Pahlavī as *baγpuhr*.» To me this seems to take the question by the wrong end. The Kuṣaṇa, on
their way from Kan-su to Bactria, are not likely to have been influenced by the Greek term Theo-
patôr. On the other hand, *baγpuhr* is so little attested as a common royal epithet in Parthian Pahlavī
that the only example of it known hitherto is in the sense of «Son of God», with reference to Jesus.
And when Sogdian *baγpūr* and the later *baγbūr*, *faγfūr* make their appearance, it is in connection
with the Emperor of China. The explanation of *devaputra* as modelled on *t'ien-tzŭ*, in the same way
as and perhaps through the intermediary of *baγpūr*, makes, in my opinion, a strong case which
Mrs. Bazin-Foucher's argument has not weakened. Lévi was even tempted to go further and to
establish a connection between the title of *devaputra* of the Kuṣaṇa and the epithet of «Son of
God» used for Jesus in the «first Judaeo-Christian communities» (pp. 19-21); but this is a much
more debatable proposition.

When I say that *t'ien-tzŭ*, «Son of Heaven», and its Iranian and Sanskrit counterparts *baγpūr*
and *devaputra* are originally the designation of the Chinese Emperor, this does not mean that
the use of this title always remained as narrowly restricted. The epigraphy of the Kuṣaṇa already
shows that they had appropriated the designation, and so did the kings of Khotan. Among Altaic
nations, the Hsiung-nu and the Turks, T'u-chüeh as well as Uighur, had done or did the same
(*tängri-dä qut bulmïš*, etc.). The notion of various «Sons of Heaven» was even erected into a
regular system, which is not of pure Chinese origin but which has found its clearest expression
in Chinese texts. We can trace it back to the 3rd cent. A. D., but it is probably older. The earliest
text speaks only of three regions of plenty : plenty of men in China, plenty of jewels in the
Mediterranean Orient (Ta-Ch'in), and plenty of horses among the Yüeh-chih (Kuṣaṇa). But the
theory was soon evolved that there were four «Sons of Heaven» : that of the men in China, that
of the elephants in India, that of the jewels in the Mediterranean Orient, and that of horses among