[Pramāṇa](https://indike.org/pramana/) › [Studies](https://indike.org/pramana/ins) › [General history and culture](https://indike.org/pramana/ins.general) › [The Greeks in Bactria and India](https://indike.org/pramana/ins.general.tarn-1938) › c33 # The Greeks in Bactria and India c33: Appendix 21. The Greek Names of the Tochari The Greeks in Bactria and India c33 - **Language:** English ### Appendix 21. The Greek Names of the Tochari ### [c33.1](https://indike.org/pramana/ins.general.tarn-1938.c33.1) It is claimed (p. 289) that the Tochari who invaded Bactria cannot have spoken Dialects A and/or B, because their name was aspirated and these dialects have no aspirate. It seems advisable therefore to collect and examine all the Greek forms of the name, as this has not been done, and the Greek forms are much earlier than most of the Oriental ones. ### [c33.2](https://indike.org/pramana/ins.general.tarn-1938.c33.2) Apollodorus, c. 100 B.C., has Τόχαροι (Tocharoi);¹ this form was popularised by Strabo and has passed into common use as this people’s name. Ptolemy, VI, 11, 6, has this form in connection with Bactria, and also, VI, 12, 4, a form Τάχοροι (Tachoroi), with metathesis of the vowels, in connection with Sogdiana. In this form the aspirate comes in the second syllable, not the first. This placing of the aspirate is also found in the Sanscrit Tukhāra, and again in the name of Bactria (in various languages) from the fourth to the eighth century A.D., Tocharistan (presumably taken from Τόχαροι), and in forms derived from Tocharistan, like toxrï (toxarï or toyari) which is found later in Central Asian documents as the name of the Saca speech of the Kushans of Tocharistan (p. 290), and Hsüan Tsiang’s Tu-ho-lo (Tuoxuâlâ, Bailey) in the seventh century A.D.² ### [c33.3](https://indike.org/pramana/ins.general.tarn-1938.c33.3) ¹ Strabo XI, 511; see p. 284 n. 4. ### [c33.4](https://indike.org/pramana/ins.general.tarn-1938.c33.4) ² H. W. Bailey, Ttaugara, BSOS VIII, 1936, p. 887. ### [c33.5](https://indike.org/pramana/ins.general.tarn-1938.c33.5) What form was used by the other Greek historian, ‘Trogus’ source’, c. 85 B.C., can only be deduced, but certainly it was not Τόχαροι. The MSS of Trogus Prol. XLII give Thocarorum, Thodarorum, Thoclarorum, Toclarorum,³ to which the best MSS add, in Justin XLII, 2, 2, the form Thogariis. It is possible therefore, but by no means certain, that this Greek historian’s form had th and was Thocaroi or Thogaroi; that is, that the aspirate came in the first syllable, not the second. This position of the aspirate is found later in the much-quoted forms Θαγοῦροι (Thagouroi) and Θογάρα (Thogara) in Ptolemy (p. 285 n. 6) for a people and town in Kan-su on the way to Sera Metropolis, forms derived from agents of Maes Titianus and therefore second century A.D.; and again in the Thibetan forms⁴ Tho-kar, Tho-gar, Thod-khar, Phod-kar, whose date I do not know, for a people in Kan-su. The form Phod-kar might guarantee Pliny’s Focari⁵ (in Chinese Turkestan), usually taken to be a corruption of Thocari; this passage in Pliny is most probably Hellenistic, and his form, and not the MSS of Trogus, is the likeliest evidence for the aspirated first syllable being Hellenistic.¹ The forms Thagouroi—Thogara show the same metathesis of the vowels as has already been seen in Tocharoi—Tachoroi. ### [c33.6](https://indike.org/pramana/ins.general.tarn-1938.c33.6) ³ This list from Ruehl’s edition, 1886, p. lxii; O. Seel’s recent edition, 1935, omits Thoclarorum. In the text of Prol. XLII Jeep printed Thogarorum, Ruehl Thocarorum; Seel prints Apollodorus’ form, Tocharorum, which has no place in Trogus. ### [c33.7](https://indike.org/pramana/ins.general.tarn-1938.c33.7) ⁴ F. W. Thomas, JRAS 1931 pp. 834-5. ### [c33.8](https://indike.org/pramana/ins.general.tarn-1938.c33.8) ⁵ Pliny VI, 55, on which see p. 84. ### [c33.9](https://indike.org/pramana/ins.general.tarn-1938.c33.9) ¹ P. Pelliot, \*T’oung Pao\* XXXII, 1936, pp. 261-2, has expressed doubts about the existence of an aspirate in Θαγοῦροι and Θοάνα (p. 519), but did not apparently consider Trogus and Pliny. It would simplify matters if the first syllable of the word had never been aspirated, but there is a good deal of evidence. ### [c33.10](https://indike.org/pramana/ins.general.tarn-1938.c33.10) There is still a third type of spelling which is almost certainly Hellenistic, and might well be as early as the form Tocharoi. Pliny VI, 22 says that a people called Tagorae² crossed the Tanais-Don from east to west with other ‘Scyths’, while across the Tanais had been another people called Inapaei who had utterly perished, viritim deletos; but in VI, 50 he puts this latter people among another set of ‘Scyths’ altogether, across (north of) the Tanais-Jaxartes—ibi Napaei interisse dicuntur. He has therefore confused, as is common enough, the two rivers called Tanais in antiquity, and has also confused two groups of ‘Scyths’ connected with the two rivers; it is uncertain therefore, from Pliny alone, whether the Tagorae lived east of and crossed the Don or lived north of and crossed the Jaxartes. Fortunately Ptolemy settles it. He knew of a people Τακοραῖοι (Takoraioi) north of the Imaos,³ and gives them again as Ταπουραῖοι (Tapouraioi) living near the Τάπουρα ὄρη;⁴ Professor Herrmann has conjectured Ταγουραῖοι (Tagouraioi) for Ταπουραῖοι,⁵ but did not notice Takoraioi, which seems to make his conjecture certain. He identified the mountain chain near which the Tagouraioi lived (on the plains to the north of it) with the Alexandrovski range west of Lake Issyk Kul, and saw in the Tagouraioi that fraction of the Tochari-Yueh-chi who remained in the Sai-wang country when the Wu-sun and Hiung-nu drove out the main body of the Yueh-chi (p. 276) and who were ultimately absorbed by the Wu-sun, a brilliant explanation which may be gratefully accepted. ### [c33.11](https://indike.org/pramana/ins.general.tarn-1938.c33.11) ² Apparently no \*v.l.\* ### [c33.12](https://indike.org/pramana/ins.general.tarn-1938.c33.12) ³ Ptol. VII, 2, 15. ### [c33.13](https://indike.org/pramana/ins.general.tarn-1938.c33.13) ⁴ Ptol. VI, 14, 7-14, with \*v.l.\* Ταποῦρει. ### [c33.14](https://indike.org/pramana/ins.general.tarn-1938.c33.14) ⁵ A. Herrmann, Τάπουρα ὄρη in PW. The name can have nothing to do with the Tapuri, far away in Hyrcania and the Elburz. ### [c33.15](https://indike.org/pramana/ins.general.tarn-1938.c33.15) Certain sources, then, or a certain source, of Pliny and Ptolemy knew the Tochari as Tagorae or Takor(aioi) or Tagour(aioi), all unaspirated forms. But such source or sources also knew of the sojourn of the Tochari in the Lake Issyk Kul district and of their crossing the Jaxartes southward. The information here, then, in Pliny and Ptolemy does not come from merchants’ itineraries or anything similar; it must go back ultimately to some historian who related the movements of the Tochari, and, as Apollodorus used the form Tocharoi, the only possible historian is ‘Trogus’ source’, to whom, as this book has shown, a good deal in Ptolemy ultimately goes back; therefore, though it is only deduction, it is at least as likely that this historian, our best authority, used some unaspirated form like Tagorae as that he used the aspirated form Thocari or Thogari. If Professor Bailey’s suggestion⁶ that the Chinese in the second century B.C. were trying to copy a form τογαρα be right, the unaspirated form is very old, which supports the virtual certainty that the form Tagorae is Hellenistic. The form occurs again, as ttaugara, in a Khotan Saca document of c. A.D. 800.¹ ### [c33.16](https://indike.org/pramana/ins.general.tarn-1938.c33.16) ⁶ \*Op. cit.\* pp. 885-6. ### [c33.17](https://indike.org/pramana/ins.general.tarn-1938.c33.17) ¹ Bailey op. cit. pp. 884-5. ### [c33.18](https://indike.org/pramana/ins.general.tarn-1938.c33.18) Before going on I must notice one thing about Ptolemy, as it is alluded to in my text (p. 286). I have described his method of work elsewhere (p. 231); he knew nothing about Central Asia himself, but he meant to get in all his (or Marinus’) notes, and he has given five notices of the Tochari, belonging to different times and places and with different spellings—Thagouroi in Kan-su, Takoraioi north of Imaos, Tagouraioi near Lake Issyk Kul, Tachoroi in Sogdiana, Tocharoi in Bactria—without suspecting that he is recording the odyssey of one and the same people. It is very characteristic; but the point is that this odyssey is that of the Yueh-chi as given by the Chinese sources (Chap. VII), and it makes the identification of Tochari and Yueh-chi certain. ### [c33.19](https://indike.org/pramana/ins.general.tarn-1938.c33.19) To sum up the Greek side. We have forms of the name with the second syllable aspirated but not the first, one of which is certainly Hellenistic; forms with the first syllable aspirated but not the second, one or more of which may be Hellenistic; and forms with no aspirate at all, one of which almost certainly is, and all may be, Hellenistic. Moreover each aspirated form has an unaspirated parallel; with Tocharoi—Tachoroi compare Takor(aioi); with Thagouroi compare Tagour(aioi); with Thogari, Thogara, compare Tagorae. Is this conclusive (as ex hypothesi it ought to be) that the name of the historical Tochari was certainly aspirated? I do not so see it, and one must consider what it means. ### [c33.20](https://indike.org/pramana/ins.general.tarn-1938.c33.20) Will the known phenomena of metathesis and shifting of the aspirate account for all these forms? These phenomena are largely dialectical variations, as for example the several forms of the Thessalian name, and belong to the formative period of the Greek language; but in our case dialectical variations do not come in question, for we are dealing throughout with one form of Greek only, the Hellenistic Koine. As regards the Koine, the interchange of τ and θ is said to be common in Egypt and Asia Minor,² and would account for the variation T—Th in the initial letter of the Tochari name; the interchange of κ and χ is also said to occur in the same countries,³ though I have not seen the evidence given.⁴ Again, the variant Τόχαροι—Thocari would represent a known form of aspirate-shifting, like κιθών for χιτών, βάθρακος for βάτραχος, Φύτιος for Πύθιος,⁵ though, once more, much of this is dialectical; and there are cases of interchange between κ and χ which are due to aspirate-shifting, like Χάλκας—Κάλχας, Καλχηδόνιοι—Χαλκηδόνιοι,¹ though the two forms may sometimes belong to different localities or dialects.² So far, the points to be noted are two. First, all these phenomena, so far as they appear in Hellenistic Greek—and we have to include yet another, the metathesis of the vowels—appear as single cases in some particular word and largely in inscriptions (or papyri); there is no authority for supposing that we can apply them all together to the same word, especially one known only from literary texts, and claim that they will explain the tremendous mix-up we find over the Tochari name, which appears to have no parallel in Hellenistic Greek.³ And secondly, interchanges like that of τ with θ or κ with χ do not enable us to say in any particular case whether the aspirated form is the original one or not.⁴ ### [c33.21](https://indike.org/pramana/ins.general.tarn-1938.c33.21) ² Many instances in K. Dieterich, \*Untersuchungen zur Gesch. d. griech. Sprache\* 1898 pp. 84-5, and add \*SEG\* VI, 281, θέκνα for τέκνα (Phrygia); see also E. Schwyzer, \*Griech. Grammatik\* I (in Müller’s \*Handbuch\* II i), 1934, p. 204. ### [c33.22](https://indike.org/pramana/ins.general.tarn-1938.c33.22) ³ Schwyzer ib. p. 204. Dietrich p. 84 says it does not occur. ### [c33.23](https://indike.org/pramana/ins.general.tarn-1938.c33.23) ⁴ A good case is \*SEG\* VI, 718 (Pamphylia), Μαλάκου and Μαλάχου (the same man). It is said to occur on Attic vase-inscriptions from the fifth century B.C. onwards (G. Meyer, \*Griech. Grammatik\*² 1886 p. 209), but that is hardly in point. ### [c33.24](https://indike.org/pramana/ins.general.tarn-1938.c33.24) ⁵ Lists in R. Kühner, \*Ausführliche Grammatik d. griech. Sprache\* (3rd ed. by Fr. Blass) I, I, 1890, pp. 278-9; Meyer op. cit. p. 209; Schwyzer op. cit. pp. 268-9. Many are old, and in Kühner-Blass are given as variants between Attic and New Ionic. Φύτιος—Πύθιος is Hellenistic, Ditt.³ 710 D. ### [c33.25](https://indike.org/pramana/ins.general.tarn-1938.c33.25) ¹ See last note. ### [c33.26](https://indike.org/pramana/ins.general.tarn-1938.c33.26) ² Cf. G. Daux, \*Delphes\* 1936 pp. 240 \*sqq.\*, on Πλυγονεῖς and Φλυγονεῖς, who in the second century B.C. were different peoples. ### [c33.27](https://indike.org/pramana/ins.general.tarn-1938.c33.27) ³ One word in Greek has as many as eight variants, βάτραχος (list in Kühner-Blass I I p. 289), but they cannot be Hellenistic; of the first four, two are Ionic, and the last four, from Hesychius, must belong to different dialects, though no information is given. ### [c33.28](https://indike.org/pramana/ins.general.tarn-1938.c33.28) ⁴ See the conflicting theories as regards old Greek in Meyer p. 209 and Kühner-Blass I I p. 277. ### [c33.29](https://indike.org/pramana/ins.general.tarn-1938.c33.29) But there are further considerations. One is the persistent γ in the Tochari name; I have found no suggestion anywhere that γ ever interchanges with χ. Another is the fact that an aspirate may just be introduced where no aspirate is; Attic has a tendency to turn unaspirated tenues into the aspirated form,⁵ and the Koine in turn aspirates a good many words which are not aspirated in Attic.⁶ I have come across two cases,⁷ not in Egypt or Asia Minor and too recent to be cited in any grammar, where a certain κ has just been turned into χ; and there is the well-known instance of the Indian Pātaliputra becoming Παλίβοθρα in all Greek writers.⁸ This is not the only case where the Greeks of the Farther East just inserted an aspirate; the \*Sp\* in the Iranian name Spalirises (Chap. VIII) is certain from all the Kharoshthi and most of the Greek legends on his coins, but there are coins (see App. 18) on which the Greek legend is spelt Shpalirises (\*p\* for Sigma). Indeed it must be remembered that in common speech in many languages (and some of the Tochari-forms must come from unlearned traders) the aspirate is an unstable thing; Catullus’ Arrius is own brother to John Leech’s ’Arry. As regards the Greek evidence, then, this cursory examination may suggest that it is difficult to say offhand whether in the second century B.C. there \*was\* an aspirate in the Tochari name or not; and even if it be supposed that θ or χ or both really belong to the original name, we are faced with a further question: in the period under consideration (second century B.C. to second century A.D.) is it certain that θ φ χ in Greek were aspirated forms at all? The latest authority has pronounced the evidence on this point to be quite indeterminate.¹ ### [c33.30](https://indike.org/pramana/ins.general.tarn-1938.c33.30) ⁵ Meyer pp. 209, 242; Kühner-Blass I I p. 111. ### [c33.31](https://indike.org/pramana/ins.general.tarn-1938.c33.31) ⁶ Kühner-Blass I I p. 112. ### [c33.32](https://indike.org/pramana/ins.general.tarn-1938.c33.32) ⁷ SEG VII, 142 (Palmyra, A.D. 81) Σπασίνου Χάραχος, for Χάρακος (spelt correctly in no. 135); J. Coupry and M. Feyel, \*BCH\* LX, 1936, p. 42 (Philippi, A.D. 41), χ(ολωνίαι) as a transliteration of coloniae. ### [c33.33](https://indike.org/pramana/ins.general.tarn-1938.c33.33) ⁸ Schwyzer \*op. cit.\* p. 204 calls attention to this. ### [c33.34](https://indike.org/pramana/ins.general.tarn-1938.c33.34) ¹ Schwyzer p. 205 sums up the evidence as ‘teils positiv teils negativ’. He himself, not so long ago, would have written ‘theils’. ### [c33.35](https://indike.org/pramana/ins.general.tarn-1938.c33.35) In fact, the Greeks never had a settled name for the Tochari (even the two contemporary historians use different forms) but only a medley of forms which fall into three classes; and the reason, whatever it was, must go deeper than such Greek linguistic phenomena as metathesis or shifting of the aspirate. It might mean that members of the horde themselves pronounced their name in different ways, and, if so, one possibility (there are probably others) would be that in the language of the Tochari we are dealing with a dying aspirate, as the sh sound died in classical Greek; compare the mute h in modern French, and the dying aspirate in wh in Britain to-day, where in the Gaelic-speaking Highlands one still hears ‘which’ and ‘when’, while in England it is common enough to hear ‘wich’ and ‘wen’. A curious point may be noticed here. It has recently been claimed² that, beside A and B, a third ‘Tocharian’ dialect, C, existed in the Shan-Shan kingdom (near Khotan) in the third century A.D.; and some words contain the spirant kh, which would be equivalent to the Greek χ in Τόχαροι. Since A and B contain no aspirates, these words have been put down as foreign;³ but I wonder whether in some cases they may not be the remains of a dying aspirate? ### [c33.36](https://indike.org/pramana/ins.general.tarn-1938.c33.36) ² T. Burrow, JRAS 1935 pp. 667, 675. ### [c33.37](https://indike.org/pramana/ins.general.tarn-1938.c33.37) ³ Ib. p. 669 ### [c33.38](https://indike.org/pramana/ins.general.tarn-1938.c33.38) I must just mention that another name, Θοάνα, has been brought into discussion and claimed as the one relic, apart from their own name, of the speech of the historical Tochari.⁴ But what Ptolemy does give is a people Θόανοι somewhere on the Silk Route with a town Θοάνα;⁵ and the Θόανοι have been omitted from discussion. Θοάνα is said to be Sogdian drw’n;⁶ but unless some Sogdian document actually makes drw’n a town of the Tochari, Ptolemy’s sixth book is prima facie conclusive against any connection.⁷ ### [c33.39](https://indike.org/pramana/ins.general.tarn-1938.c33.39) ⁴ Bailey op. cit. pp. 888, 916; cf. Pelliot op. cit. p. 263. ### [c33.40](https://indike.org/pramana/ins.general.tarn-1938.c33.40) ⁵ Ptol. vi, 16, 5 and 6. ### [c33.41](https://indike.org/pramana/ins.general.tarn-1938.c33.41) ⁶ Bailey op. cit. p. 888. ### [c33.42](https://indike.org/pramana/ins.general.tarn-1938.c33.42) ⁷ I suppose the occurrence together of the two names Θαγόρα (Tagora in some Latin versions) and Θοάνα in Ptol. vii, 2, 7, on the Great Gulf in Further India, near the equator, might be held to point to a connection. But what the names can mean here I have no idea, unless it is merely one of the usual cases where Ptolemy turns one place into two, see pp. 231 sq. ### [c33.43](https://indike.org/pramana/ins.general.tarn-1938.c33.43) I am not of course trying to prove anything in this Appendix; that is a matter for philologists. I am only pointing out that the Greek forms suggest that further investigation may be desirable before it be stated as an ascertained fact that the centum dialects A and B cannot represent the language of the historical Tochari. ## Linked texts (2) - [Naturalis historia](https://indike.org/pramana/inf.la.imperial.pliny-elder) — C. Plinius Secundus, Latin - [Historiae Philippicae](https://indike.org/pramana/inf.la.imperial.pompeius-trogus) — Pompeius Trogus, Latin --- Source: [https://indike.org/pramana/ins.general.tarn-1938.c33](https://indike.org/pramana/ins.general.tarn-1938.c33) · Pramāṇa