The Greeks in Bactria and India c22: Appendix 10. Ta-yuan
The Greeks in Bactria and India c22
- Language
- English
Appendix 10. Ta-yuan
c22.2
The Greek occupation of Ta-yuan¹ (Ferghana) has not always been accepted,² and one must consider the evidence. Certain things may be ruled out at once. Naturally I agree with those who have said that Ta-yuan does not mean ‘the great land of the Yavanas’; it is enough that the Ch’ien-han-shu knows of a ‘little Yuan’, Siao-yuan, in the Tarim basin, which had nothing to do with Greeks.³ The vine and the alfalfa grew in Ta-yuan, and a considerable number of scholars have believed that the Chinese terms for these plants, p’u-t’ao and muk-tuk, were respectively βότρυς and Μηδικὴ (πόα), but I do not see how this belief can survive Professor Laufer’s analysis.⁴ Certainly, though Laufer said that the vine was native all over Iran and was not brought to Ferghana by Greeks,⁵ he could produce no evidence for it in Ferghana itself earlier than the Annals of the Ts’in dynasty, fourth century A.D.,⁶ which is not in point. But his evidence for the vine in Margiane and Bactria (the nearest countries) is from Strabo, who never mentions Ferghana; so, for all we know, the vine may have been native there or may have been introduced by Greeks or by anybody else; it is no evidence for anything. The habit of keeping wine in jars for years was of course a Greek custom; but unless it be proved that it was not a native custom in Iran also (as to which I know nothing) it cannot be asserted that Greeks introduced it.⁷
c22.3
¹ See generally Chang-k’ien’s description of Ta-yuan in Ssu-ma Ch’ien (Hirth pp. 95-6), reproduced in the Ch’ien-han-shu (Wylie p. 44, de Groot p. 109), and further Chapter VII.
c22.4
² I think the only writer who has really considered it was V. V. Barthold, The Graeco-Bactrian State and its spread to the North-East, Bulletin (Izvestiya) of the Imp. R. Acad. Sc. 1916 pp. 823-8 (in Russian); he rejected it, but apparently solely on the wine question, and that is not the evidence at all. Professor E. H. Minns kindly supplied me with a synopsis of the main points in this article.
c22.5
³ Wylie p. 28, de Groot p. 64 (Sao-wan). The latter (p. 12) says there is no traceable connection between Ta-wan and Sao-wan; all the less likely, therefore, is it that yuan (or wan) should be Yavana.
c22.6
⁴ Vine: B. Laufer, Sino-Iranica 1919 pp. 225-6; alfalfa, ib. pp. 212-13. Barthold op. cit. had rejected the equation p’u-t’ao = βότρυς three years previously. O. Franke (note in de Groot p. 35) says that muk-tuk (mu-su, buk-suk) must be derived from some unknown Iranian word.
c22.7
⁵ Laufer op. cit. pp. 221 sqq.
c22.8
⁶ Ib. p. 221.
c22.9
⁷ Barthold op. cit. said that there are no Greek words for wine-making in the east-Iranian tongues.
c22.10
This narrows the question down to two points. Ssu-ma Ch’ien says that Chang-k’ien travelled through Ta-yuan on ‘postal roads’.⁸ Persians made post roads before Greeks did, but it cannot be shown that the Persians ever ruled Ferghana. The ‘Sacas’ of the earlier lists of Darius’ empire (Behistun and Persepolis) were in the last list, that of Naks-i-Rustam, divided into their two components, the Sacas wearing pointed caps and the Amyrgian Sacas (haumauarga, ‘preparers of Haoma’),¹ the new province-list of Xerxes² agreeing in this respect with that of Naks-i-Rustam; but Herodotus, in his reproduction of Xerxes’ army list, specifically identifies the two,³ and even if not identical they cannot therefore have lived far apart; I know no reason for Professor Herrmann’s view that the Amyrgian Sacas lived on the Pamirs⁴ (which cannot possibly have belonged to Darius), and the Hellanicus fragment, the source of which must be Persian, says that they lived on a plain.⁵ It is now known from Darius’ gold plate⁶ that the Sacas he ruled, who must be these two bodies, lived para Sugdam, ‘beyond Sogd’, *i.e.* in the country between the mountains north of Samarcand and the Jaxartes; and the question is whether the expression ‘Saca-land beyond Sogd’, primarily no doubt the Chodjend country and the land to the westward, would also include Ferghana. (It would be no objection that Ferghana was an agricultural country; some Saca clans were no longer nomads.) I cannot answer that question; but it is certain that if Ferghana was ever Achaemenid it had been lost before Alexander’s day; he could not have left an outlying Persian province without either occupying it himself or sending a force to do so, and also Ptolemy’s Sogdiana does not include Ferghana.⁷ But supposing that Darius *had* ruled Ferghana and *had* made post roads, is it conceivable that the local population would have kept them in order for nearly four centuries till 128 B.C., when Chang-k’ien saw them? These roads are good evidence for the Greek occupation of Ferghana, but they are not actually conclusive.
c22.11
⁸ Hirth p. 94 §9.
c22.12
¹ O. Leuze, *Die Satrapieneinteilung in Syrien und im Zweistromlande von 520-320*, 1935, p. 250 [94], seems to me conclusive that it was such a division and not a new conquest.
c22.13
² E. Herzfeld, *Arch. Mitt. aus Iran* VIII, 1936, p. 58 ll. 26, 27. The list is before 480 B.C.
c22.14
³ Herod. VII, 64, the Sacas wearing pointed κυρβασίαι are called by Persians Amyrgian Sacas.
c22.15
⁴ A. Herrmann, *Sakai* in PW; see Marquart’s criticism, *Untersuchungen zur Geschichte von Eran* II, 1904, p. 140. See however p. 477.
c22.16
⁵ In Stephanus: ‘Ἀμύργιον· πεδίον Σακῶν.
c22.17
⁶ S. Smith, *JRAS* 1926 p. 435; E. Herzfeld, *Memoirs of the Arch. Survey of India* XXXIV, 1928.
c22.18
⁷ Cf. on this Berthelot p. 191.
c22.19
What is conclusive is that Chang-k’ien found the country, as he found Bactria, full of ‘walled towns’; he gives ‘fully 70’ in Ferghana alone, a small country. What they were has already been explained (pp. 121 *sq.*); most were large native villages walled round. But the phenomenon, precisely as in Bactria (which Chang-k’ien compares with Ferghana throughout), means Greek occupation and nothing else; for eastern Iranians did not build towns and did not wall-in villages. This comes out clearly enough in the Alexander-story. There is one exception in that story, and it is given as something exceptional—the seven fortresses which the Achaemenids had built to guard the Jaxartes frontier against nomads.¹ These were military fortresses, not villages. Yet six of them had only low mud ramparts,² and Alexander took them as he pleased; only of the largest one, Cyropolis, is it stated that it had a ‘higher wall’,³ compelling him to bring up his siege train. If there was one place in eastern Iran in Alexander’s day large enough to be called a town it was Bactra itself; and Bactra was not walled. When Spitamenes attacked it behind Alexander’s back,⁴ the commandant in the place had only a few details and the sick; on the other hand, Spitamenes had only light Sogdian and Saca horse, useless against walls. Had there been a wall, the commandant only had to close the gates and Spitamenes was helpless; instead, he led out what men he had and they were naturally annihilated; therefore there was no wall, and Spitamenes could have taken Bactra but for Craterus’ timely arrival. Spitamenes, just before, had taken one of the border forts which guarded the Bactrian frontier;⁵ if his horsemen could take it, it had nothing but a low mud rampart, at best. It was Greeks who turned Bactra into a great fortress and filled Bactria with ‘walled towns’ which nomad horsemen could not take (see generally Chap. III); and no one else could have done the same thing in Ferghana.
c22.20
¹ This story is in Arrian IV, 2 sqq.
c22.21
² *Ib.* IV, 2, 3, τῷ τείχει γηίνῳ τε καὶ οὐκ ὑψηλῷ ὄντι.
c22.22
³ *Ib.* IV, 3, 1, τετειχισμένη...ὑψηλοτέρῳ τείχει ἥπερ αἱ ἄλλαι.
c22.23
⁴ *Ib.* IV, 16, 6 sqq.
c22.24
⁵ *Ib.* IV, 16, 4–5.
c22.25
It will be seen that I follow the opinion which, prior to Professor Herzfeld’s Sakastan, was universally held, that Chang-k’ien’s Ta-yuan was Ferghana (though, as will appear, it was larger than Ferghana proper), as Ta-yuan admittedly is in Chinese literature from the fourth century A.D. onwards.⁶ Herzfeld however has argued⁷ that Chang-k’ien meant by *his* Ta-yuan not Ferghana but the Pamirs; but though he has rightly pointed out elsewhere that what Chang-k’ien saw and what he heard are different things, he has here omitted what Chang-k’ien saw in Ta-yuan and has based his argument solely on the relative positions by points of the compass which Chang-k’ien gives for the several peoples he mentions, a thing over which a stranger in a strange and unknown world might easily make mistakes; indeed if one considers the mistakes made by Ptolemy, with all the sources of information which he had at his disposal,⁸ it would be extraordinary if Chang-k’ien had *not* made any. And when Chang-k’ien put the Yueh-chi west of Ta-yuan⁹ (which is what Herzfeld is going on) he did just make a mistake (or Ssu-ma Ch'ien made one in transcribing his Report); this is shown clearly enough by Pan-ku, who in transcribing the same passage quietly corrected 'west' to 'south-west' or 'south',¹ and this is one of those cases in which the later historian, who had at his disposal much new information collected by Chinese envoys and caravans, is bound to be right.
c22.26
⁶ For this later evidence see de Groot p. 109.
c22.27
⁷ Sakastan pp. 22 sqq. and plan on p. 23.
c22.28
⁸ E.g. he puts the Massagetae east of Sogdiana instead of west (VI, 13), and Sogdiana both east (VI, 11) and west (VI, 12) of Bactria, both being wrong.
c22.29
⁹ Hirth p. 95 §22.
c22.30
¹ Wylie p. 44 (south); de Groot p. 109 (south-west). South-west will serve. The Yueh-chi had not moved in the meanwhile except to extend southward across the Oxus to occupy Bactria; their capital was still north of the Oxus.
c22.31
But the real point is that, apart from its Saca rulers, Chang-k'ien describes Ta-yuan, which he had seen, as a settled country exactly like Ta-hia (Bactria),² with the same soil, climate, productions, and customs,³ which must mean highly organised agriculture like the Bactrian. He says that the people grew not only wheat but rice,⁴ which means irrigation and a hot summer climate (he does not mention rice in Bactria, though he had heard that it grew in Chaldaea),⁵ and that the vine and alfalfa grew well in Ta-yuan,⁶ and that like Bactria the country was full of 'walled towns', whose number he gives as 'fully 70';⁷ the Chinese subsequently planted an agricultural colony there.⁸ Was there really on the Pamirs a highly organised agricultural state with wheat-fields and rice-fields,⁹ clover-fields and vineyards, and seventy walled towns? Any description of the desolate and windswept Roof of the World will suffice to negative the supposition.
c22.32
² Hirth p. 97 §47.
c22.33
³ Wylie p. 44, de Groot p. 109.
c22.34
⁴ Hirth p. 95 §19.
c22.35
⁵ Ib. p. 97 §41.
c22.36
⁶ Ib. pp. 95 §19, 108 §99.
c22.37
⁷ Ib. p. 95 §20.
c22.38
⁸ Ib. p. 116.
c22.39
⁹ On the Indian side, some inferior rice is said to be grown to-day in the Gilgit valley, about 5000 feet up. But Gilgit is very far from being the Pamirs, which have an average height of 13,000 feet and are under snow for at least half the year.
c22.40
It may however be taken as certain that Chang-k'ien's Ta-yuan, as a state, was larger than Ferghana proper and extended up the river valleys to the Pamir watershed. He says that Ta-yuan borders to the east upon two states, Hau-mi (U-bi), a very small place, and Yu-tien (Hu-tin, Khotan), and that in Yu-tien all the rivers run eastward but that westward of it (i.e. in Ta-yuan) they all run westward.¹⁰ The Saca rulers of Ta-yuan, then, ruled up to the watershed, and may in part have lapped round the Yueh-chi country on its eastern side; and the people in the high valleys may, for all we know, have been part of the Amyrgian Sacas. But this is not what Herzfeld meant (see his plan p. 23); and it is not what Herrmann meant either.
c22.41
¹⁰ Ssu-ma Ch'ien (Hirth p. 95, de Groot p. 12). The parallel passage in the Ch'ien-han-shu, Wylie p. 30, de Groot p. 69, shows that the statement is Chang-k'ien's.
Linked texts (1)
- Ἀλεξάνδρου ἀνάβασις — Λούκιος Φλάουιος Ἀρριανός, Ancient Greek